You are on page 1of 11

Journal of Experimental Social Psychology 49 (2013) 453463

Contents lists available at SciVerse ScienceDirect

Journal of Experimental Social Psychology


journal homepage: www.elsevier.com/locate/jesp

The consequences of faking anger in negotiations


Stphane Ct a,, Ivona Hideg b, Gerben A. van Kleef c
a
b
c

University of Toronto, Canada


Wilfrid Laurier University, Canada
University of Amsterdam, The Netherlands

H I G H L I G H T S
In two experiments, we examine the consequences of faking anger in negotiations
Faking anger increases the demands of negotiation counterparts
Faking anger has this effect because it erodes trust

a r t i c l e

i n f o

Article history:
Received 27 April 2012
Revised 27 November 2012
Available online 8 January 2013
Keywords:
Anger
Regulation
Surface acting
Deep acting
Authenticity
Negotiation

a b s t r a c t
Past research has found that showing anger induces cooperative behavior from counterparts in negotiations.
We build on and extend this research by examining the effects of faking anger by surface acting (i.e., showing
anger that is not truly felt inside) on the behavior of negotiation counterparts. We specically propose that
surface acting anger leads counterparts to be intransigent due to reduced trust. In Experiment 1, surface acting
anger increased demands in a face-to-face negotiation, relative to showing no emotion, and this effect was
mediated by (reduced) trust. In Experiment 2, surface acting anger increased demands in a video-mediated
negotiation, relative to showing no emotion, and this effect was explained by (reduced) trust, as in Experiment 1. By contrast, deep acting anger (i.e., showing anger that is truly felt inside) decreased demands, relative
to showing no emotion, and this effect was explained by (increased) perceptions of toughness, consistent with
prior research on the effects of showing anger in negotiations. The ndings show that a complete understanding of the role of anger in negotiations requires attention to how it is regulated. In addition, the results suggest
that faking emotions using surface acting strategies may generally be detrimental to conict resolution.
2013 Elsevier Inc. All rights reserved.

Introduction
Research on conict resolution has repeatedly found that negotiators cooperate when their opponents display anger. Across several
studies, negotiators made larger concessions to angry opponents
than to neutral or happy opponents (e.g., Sinaceur & Tiedens, 2006;
Sinaceur, Van Kleef, Neale, Adam, & Haag, 2011; Van Kleef, De Dreu,
& Manstead, 2004a). This research suggests that expressing anger
can be an effective way to evoke cooperative behavior in others and
get things done. But what happens when negotiators display anger
that they do not truly feel? Does faking anger also elicit cooperation?
Anecdotal evidence suggests that it may not. After the British
Petroleum oil spill in the Gulf of Mexico in the spring of 2010, U.S.
President Barack Obama was criticized for his calm response to the
incident (Pareene, 2010). He eventually showed anger about the incident on a television show, but his display further undermined his
support, because his display was deemed not to be genuine. Obama
drew considerable skepticism (Raum, 2010), and veteran newscaster
Corresponding author.
E-mail address: scote@rotman.utoronto.ca (S. Ct).
0022-1031/$ see front matter 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jesp.2012.12.015

Sam Donaldson commented that That's not him (Chalian, 2010).


This anecdote suggests that showing anger that is not truly felt may
not induce cooperation, and that it may even backre. To illuminate
this issue, in two experiments, we examined the effects of surface acting anger efforts to show anger that one does not feel on the behavior of negotiation counterparts.
Past ndings on the effects of expressing anger in negotiations
Past research has shown that displays of anger are interpreted as
signals of dominance and toughness (Knutson, 1996; Tiedens, 2001)
that intimidate others and persuade them to comply (Averill, 1983).
Studies of leadership, for instance, found that leaders' displays of
anger inuenced subordinates to exert more effort toward their
tasks (Sy, Ct, & Saavedra, 2005), especially when subordinates
paid close attention to the information contained in these displays
(Van Kleef et al., 2009).
Past research on negotiations has identied effects that are consistent with these ndings. Individuals who negotiated with angry opponents believed that these opponents were tough, had ambitious
goals, and were unlikely to make substantial concessions (Van Kleef

454

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

et al., 2004a). In turn, they tended to make large concessions to angry


opponents. In one study, participants made larger concessions to
angry opponents than to non-emotional or happy opponents because
participants believed that angry opponents had more ambitious goals
(Van Kleef et al., 2004a). Likewise, in another study, participants
perceived angry counterparts as tougher than non-emotional counterparts, and this led participants to concede more to angry counterparts
(Sinaceur & Tiedens, 2006). These ndings suggest that negotiators
who show anger elicit concessions from their counterparts because
these responses are perceived to be tough.
The previous studies, however, do not cover all of the potential effects of showing anger in negotiations. In previous studies, participants
had no reason to doubt that their counterparts did not truly feel the
anger that they displayed. For instance, in the studies conducted by
Van Kleef et al. (2004a), participants received a message from their
counterparts that they were angry (without actually seeing their counterparts and their facial displays of anger). In addition, participants
were led to believe that their counterparts did not know that their emotional reactions would be sent to the participants. Participants had an
explicit reason to believe that their counterparts' expressions were
real rather than faked for strategic reasons. Thus, in these studies,
there was nothing to indicate that the counterparts might not have actually felt angry. The past ndings may primarily concern the effects of
anger that is presumed to be truly felt, and not necessarily the effects of
showing anger that one does not truly feel. The effects may be different
when displayed anger is not truly felt, because the discrepancy between
the displayed and felt anger has consequences for the authenticity of
the expressions.
The consequences of surface acting anger for the authenticity of the
displays
Individuals often fake anger by showing anger that they do not
genuinely feel to achieve their personal goals and promote their
own interests (Tamir, Mitchell, & Gross, 2008). For instance, in one
study, participants acknowledged expressing more anger than they
felt to inuence another person to make better offers in an ultimatum
game (Andrade & Ho, 2009). In another study, managers indicated
that they sometimes deliberately showed anger that they did not
feel to inuence their subordinates (Fitness, 2000). And, one investigation showed that bill collectors displayed anger even when they felt
a different emotion, because this strategy was deemed most effective
by their organization to compel certain debtors to pay (Sutton, 1991).
Although this past research suggests that surface acting anger is
prevalent, its effects have so far been overlooked in the research on
negotiations.
One common way in which individuals fake their emotions is surface acting (Ct, 2005; Grandey, 2000; Gross, 1998). Surface acting is
an emotion regulation strategy in which individuals modify their external display of emotion, but they do not change their internal feelings (Ct, 2005; Grandey, 2003; Hochschild, 1983). When surface
acting anger, individuals put on a show by generating expressive
displays of anger that they do not actually feel (Gross, 1998). For example, negotiators may pretend to be angry externally about an offer
made by their counterparts even though they are not really angry inside. Surface acting anger thus creates a mismatch between publicly
displayed and privately felt emotions, because subjective experiences
of emotions are left intact (Ct, 2005).
The divergence between outwardly expressed and internally felt
emotions has important consequences. Past research has shown that
feigned displays of emotions appear different than spontaneously felt
emotions (Ekman, 2003; Ekman, Friesen, & O'Sullivan, 1988). Relative
to spontaneous displays, feigned displays were less symmetrical, so
that muscle movements that appeared on one side of the face were
less likely to be the same as those that appeared on the other side
(Ekman, Hager, & Friesen, 1981; Hager & Ekman, 1985; Skinner

& Mullen, 1991). Individuals who surface act emotions tend to be perceived as relatively inauthentic by observers because of these differences in the displays (Ct, 2005).
In one study, observers rated service providers who surface acted
emotions as less authentic than those who expressed genuine emotions (Grandey, Fisk, Mattila, Jansen, & Sideman, 2005). In another
study, surface acting emotions at work was negatively related to
coworker ratings of affective delivery, which encompasses sincerity,
warmth, courtesy, and friendliness (Grandey, 2003). And, across several studies, individuals who reported surface acting often tended to
feel less authentic (Brotheridge & Lee, 2002; Gross & John, 2003),
reported that they had less social support available to them
(Srivastava, Tamir, McGonigal, John, & Gross, 2009), and were perceived by their peers as having lower quality relationships (Gross &
John, 2003). These ndings suggest that during negotiations, surface
acting anger may lead to different behaviors from counterparts than
expressing anger that is truly felt or showing no emotions.
The consequences of surface acting anger in negotiations
We propose that in the context of negotiations, surface acting reduces trust. Trust is dened as expectations of goodwill, benevolence,
and integrity in the other party (Ferrin, Bligh, & Kohles, 2007; Kramer,
1999; Yamagishi & Yamagishi, 1994), and assurance that the other
party will not exploit them (Kelley & Thibaut, 1978). Negotiators
may infer that because the anger is not truly felt, their counterparts
do not have a valid reason for being angry. Negotiators may infer
that the counterparts have not been treated unfairly, otherwise they
would truly be angry, given that anger tends to arise when one has
been treated unfairly by others (Smith, Haynes, Lazarus, & Pope,
1993). Thus, negotiators may interpret displays by counterparts
who surface act anger as dishonest, opportunistic, and calculated attempts to control them (Rafaeli & Sutton, 1989). In turn, negotiators
may believe that they are at risk of being exploited when their counterparts pretend to be angrier than they really are as a result of surface acting.
Support for this reasoning is indirect yet suggestive. In previous
investigations, participants exhibited more trust, assigned more favorable traits (e.g., honesty), and cooperated more in a prisoner's dilemma and a trust game when others displayed authentic rather than
inauthentic smiles (Johnston, Miles, & Macrae, 2010; Krumhuber et
al., 2007). These ndings support the notion that faking anger will
erode the trust of negotiation counterparts.
In turn, past research suggests that negotiators who do not trust
each other protect themselves from potential harm in an effort to ensure that they are not exploited (Ferrin et al., 2007; Mayer, Davis, &
Schoorman, 1995). If negotiators perceive that their opponents lack
integrity, they may believe that their opponents may be willing to
take advantage of them. Negotiators who do not trust each other
are thus particularly likely to adopt a competitive stance.
In support of this proposition, across a series of studies, lack of trust
was associated with less exchange of information between negotiators
(De Dreu, Beersma, Stroebe, & Euwema, 2006; De Dreu, Giebels, & Van
de Vliert, 1998) and more retribution (Ross & LaCroix, 1996). Lack of
trust should thus lead negotiators to place high demands on their opponents. Surface acting anger and the ensuing erosion in trust should also
harm the relational outcomes of negotiations intangible assets such as
mutual liking and commitment to continuing the relationship (Gelfand,
Major, Raver, Nishii, & O'Brien, 2006) because negotiators should develop negative attitudes about those who lack integrity.
Overview of the research strategy
We conducted two experiments to examine the consequences of
surface acting anger in negotiations. In Experiment 1, we tested the
effects of surface acting anger, relative to a neutral control condition,

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

on the demands of negotiation counterparts and relational outcomes


in a face-to-face negotiation. We expected that compared to showing
neutral emotions, surface acting emotions would increase demands
and harm relational outcomes (Hypotheses 1a and 1b). We further hypothesized that these effects would be mediated by (reduced) trust
(Hypotheses 2a and 2b). In Experiment 2, we compared the effects
of surface acting anger to the effects of both deep acting anger
(a strategy that consists of creating both the feeling and the display
of anger) and showing neutral emotion in a video-mediated negotiation. We expected that compared to showing neutral emotions, surface acting anger would increase demands due to (reduced) trust
(Hypotheses 1a and 2a). We also expected that deep acting anger would
decrease demands (Hypothesis 3), and that this association would be mediated by (increased) perceptions of toughness (Hypothesis 4).
Experiment 1
We rst conducted an experiment in which participants
interacted face-to-face with a confederate in the role of negotiation
counterpart. We manipulated emotional expressions by training the
confederate to either surface act anger or show no emotion during
the negotiation. We examined the effects of surface acting anger on
the demands of the counterparts and relational outcomes assessed
by the counterparts. We tested Hypothesis 1a that compared to
showing no emotion, surface acting anger would increase demands,
and Hypothesis 2a that (reduced) trust would mediate the effect of
surface acting anger on demands. In addition, we tested Hypothesis
1b that surface acting anger would be associated with poorer relational outcomes and Hypothesis 2b that (reduced) trust would mediate the negative effect of surface acting anger on relational outcomes.
Method
Participants
Participants were 130 undergraduate students (75 women and 55
men) at a large university and individuals from the community between the ages of 19 and 50 (M = 22.19, SD= 4.29). They participated
in exchange for credit or monetary compensation of 10 dollars. 1 Participants were randomly assigned to a condition in which the counterpart
surface acted anger (n = 65) or a control condition in which the counterpart did not show emotion (n = 65). The dependent variables were
demands and relational outcomes.
Procedure
An experimenter welcomed participants to a laboratory room.
Participants sat on one of two chairs put in front of each other. They
were informed that they would be doing a two-person exercise that
involved negotiating the price of a used car. They were told that
they had been randomly assigned to the role of the seller, and that
the other participant, who was receiving instructions in the other
room, had been randomly assigned to the role of the buyer.
The experimenter explained the role of seller verbally to participants and then also gave them a written copy of these instructions.
We adapted the Blue Buggy negotiation exercise that involves selling or buying a used car (Paulson, 2006). We adapted this exercise
because it has been used for many years in negotiation research, ensuring face validity. Plus, it is relevant to a sample of students
whose nancial resources often limit them to buying used cars.
Participants read that they wanted to sell their used car. Until recently, they lived far from their university and used their car to
1
Participants who received credit (n = 72) did not differ from those who were paid
(n = 58) on the proposed mediator, trust, t(128) = .13, p= .90, the rst measure of demands, t(128) = 1.79, p = .08, the second measure of demands, t(128) = 1.30, p = .19,
and the measure of relational outcomes, t(127) = .30, p = .77. Thus, we combined the
participants in the analyses.

455

commute. Recently, they moved close to the university and no longer


needed the car because they could walk to campus. The car had become an unnecessary expense. They were presented with information about the car and the results of market research about the car's
value. They read that an expert had evaluated the car as worth between $2500 and $3500, and that they had advertised it for $3500.
They learned that they had already met a potential buyer who
seemed to like the car, but also had some reservations about it and
needed time to reect on the decision. Finally, they read that the
buyer called them and wanted to meet with them again to potentially
buy the car.
Participants were instructed to make an opening statement that
included some description of the car and a reminder of the asking
price of $3500. They were instructed to try to sell the car for as high
a price as possible, while keeping in mind that the car should be
sold within a limited period of time. They were also asked to nd
the experimenter outside of the laboratory room after every two
rounds of negotiation so that the experimenter could administer
questionnaires. A round was dened as an offer by the buyer and a
counteroffer by the seller (if the seller declined the offer). The experimenter veried that participants understood the instructions. The
experimenter then said that she would get the other participant
from the other room and that they would negotiate in the current
room because, in the scenario, the buyer visits the seller.
When the counterpart (a confederate) entered the room, he was introduced to the participant. We chose a confederate to regulate anger to
have greater control over the display of the counterpart. Past research
shows individual variation in the ability to regulate displays of emotions
in laboratory experiments (Gyurak et al., 2009; Schmeichel, Volokhov,
& Demaree, 2008), suggesting that some participants may not have
been able to follow instructions to surface act anger and, hence, introduced considerable noise in the data. Three separate individuals acted
as confederates in both conditions.2 The confederates followed scripted
protocols for the initial greetings, the rst offer, and the second offer.
We wrote the scripts so that they could realistically follow any statement by participants. The only variable was whether the confederates
surface acted anger or remained emotionally neutral throughout the
interaction.
The confederates were Caucasian male actors who were close in
age to the participants. We hired actors because reading a script
while modifying displays of emotions is challenging, and actors receive formal training for modifying emotional expressions. Actors
have been hired to create similar stimuli in past research (e.g., Bono
& Ilies, 2006; Grandey et al., 2005; Van Kleef et al., 2009), including
negotiation studies (e.g., Kopelman, Rosette, & Thompson, 2006).
We chose male actors because past research has suggested that
expressing anger is less appropriate for females (Kring, 2000;
Timmers, Fischer, & Manstead, 1998). The confederates did not
know the goals and the hypotheses of the study.
We trained the confederates to regulate their emotions as required
by the experimental conditions. We adapted procedures used by
Grandey et al. (2005) to train confederates to regulate emotions in simulated customer service interactions and procedures used by Levenson
and his colleagues (Gross & Levenson, 1997; Kunzmann, Kupperbusch,
& Levenson, 2005) to instruct participants to regulate emotions in the
laboratory. For the surface acting anger condition, we instructed the
confederates to remain emotionally neutral inside but express anger
in their faces. They followed these instructions by manipulating only
their facial muscles without modifying their thoughts, subjective experiences of emotions, or vocal cues such as vocal tone and intensity.

2
ANOVAs revealed no signicant differences across the three confederates on trust,
F(2, 127) = 1.12, p = .33, the rst measure of demands, F(2, 127) = 1.65, p = .20, and
the second measure of demands, F(2, 127) = 1.35, p = .26, and the measure of the relational outcomes, F(2, 126) = 3.06, p = .0502. Thus, we combined the results from the
three confederates in the analyses.

456

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

Confederates were further instructed to lower their eyebrows and draw


them together, and to ensure that their eyes are glaring and their jaw is
clenched. For the neutral condition, we instructed the confederates to
relax facial muscles so as to avoid showing emotion. We stressed the
importance of keeping all other aspects of behavior constant across
the conditions. The confederates obtained feedback from experimenters and students uninvolved with this project to help them calibrate their acting and expressions prior to the study.
Participants made the opening statement, as requested by the instructions. The counterpart then made a number of statements about
the car and listed a number of concerns about it. At the end of this
speech, the counterpart offered $2400 for the car. Participants decided to accept or reject the offer, and proposed a counteroffer. The
counterpart then made a number of other statements about the car
and repeated several concerns about it. This speech was designed so
that it could reasonably follow any statement and offer by participants. At the end of this speech, the counterpart offered $2550. Participants again decided to accept or reject that offer, and proposed a
counteroffer. After the end of this second round, the confederate got
up to nd the experimenter outside of the room. The experimenter
re-entered the room and assigned the participant and the confederate
to two computers that were separated by other computers so that
they could not see each other's responses. Participants lled out the
measures of trust, relational outcomes, and manipulation checks described below. The confederate stayed in the room to alleviate suspicion and lled out the questionnaires randomly. When participants
nished the questionnaires, the experimenter asked the confederate
if he was done, and he said yes. The experimenter announced that
the experiment was over. The experiment lasted approximately 1 h,
with the actual negotiation lasting approximately 10 min. There
was little variation in the duration of the actual negotiation because
the responses of the confederate were predetermined and largely
constant for all participants. At the end, the experimenter debriefed
participants and asked them not to discuss the study with anyone.
Measures
All measures in this article used a Likert-type response scale ranging
from 1 (strongly disagree) to 7 (strongly agree), unless otherwise noted.
Manipulation checks. To verify that we successfully manipulated
expressed anger, participants rated how much the counterpart displayed
anger and neutrality, using items from Van Kleef, De Dreu, Pietroni, and
Manstead (2006). All items followed the stem The buyer appeared
Three items assessed displays of anger (angry, irritated, and
aggravated; alpha=.91) and another three items assessed neutral displays (unemotional, dispassionate, and neutral; alpha=.72).
Past theory and research proposed that surface acting emotions is
associated with displays that are perceived as inauthentic (Ct,
2005; Grandey et al., 2005). To verify that displays of anger were perceived as inauthentic and, thus, that the confederates were surface
acting, we used six items assessing the authenticity of displays of
emotions adapted from Grandey et al. (2005) and Frank, Ekman,
and Friesen (1993). Sample items were The buyer was faking
how he felt (reversed scored) and The buyer genuinely expressed
emotions (alpha = .87).
Demands. There were two measures of demands. The rst measure
occurred after the rst offer of $2400 by the counterpart, and the second measure occurred after the second offer of $2550 by the counterpart. Participants proposed counteroffers by saying the amount that
they wanted for their car. The confederate reported the counteroffers
after each session ended. We veried that the counteroffers were correct using video-tapes of the sessions. The confederate was correct in
each case.

Relational outcomes. We used ve items from past research (Van Kleef


et al., 2004a) to create a composite measure of relational outcomes
(alpha = .88): I am satised with the course of the negotiation, I
would like to do business with the same buyer in the future, I
have a good feeling about the negotiation, I would like to avoid future negotiations with the same buyer, and I am pleased with this
negotiation. One participant skipped this measure.
Trust. We used three items to assess participants' trust in the buyer
(Van Kleef, De Dreu, & Manstead, 2006). Participants rated the degree
to which the buyer was trustworthy, unreliable (reverse scored),
and honest (alpha = .77).
Results
Manipulation checks
Displayed anger. As expected, participants perceived that their counterpart was more angry in the surface acting anger condition (M =
4.62, SD = 1.62) than in the neutral condition (M = 2.44, SD = 1.23),
t(128) = 8.64, p b .001 (d = 1.21). Participants also perceived the
counterpart to be more emotionally neutral in the neutral condition
(M = 4.01, SD = 1.30) than in the surface acting anger condition
(M = 3.26, SD = 1.42), t(128) = 3.12, p b .01 (d = .53). These results
indicate that the displays were manipulated as intended.
Authenticity. As expected, participants perceived the opponent to be
less authentic in the surface acted anger condition (M = 3.63, SD =
1.25) than in the neutral condition (M = 4.25, SD = 1.19), t(128) =
2.91, p b .01 (d = .50).
Focal analyses
Effect of surface acted anger on demands. Before conducting our main
analyses, we examined whether women and men behaved differently, given that gender differences in negotiation behavior were found
in some past research (e.g., Kray, Galinsky, & Thompson, 2002;
Small, Gelfand, Babcock, & Gettman, 2007). There were no gender differences on the rst counteroffer, t(128) = .41, p = .68 (d = .07), nor
the second counteroffer, t(128) = .29, p = .78 (d = .05). Thus, we
combined the data from male and female participants, and we did
not control for gender in the analyses.
Hypothesis 1a predicted that participants would be more intransigent with a counterpart who surface acts anger than with an emotionally neutral counterpart. We tested this hypothesis for each
round using t tests to compare demands across the conditions. Fig. 1
presents the level of demands in each condition by round. In Round
1, demands in the surface acting anger condition (M = 3109.23,
SD = 173.86) did not differ from demands in the neutral emotion condition (M = 3087.68, SD = 187.31), t(128) = .68, p = .50 (d = .12). In
Round 2, demands in the surface acting anger condition (M =
2943.08, SD = 196.63) were signicantly higher than demands in
the neutral condition (M = 2873.83, SD = 189.17), t(128) = 2.05,
p b .05 (d = .35). Thus, participants were more intransigent with opponents who surface acted anger, but only in the second round. Hypothesis 1a was partially supported. We further interpret this
nding in the discussion.
Effect of surface acting anger on trust. We expected that participants
would trust a counterpart who surface acts anger less than an emotionally neutral counterpart. As expected, trust was lower in the surface acting anger condition (M = 4.29, SD= .96) than in the neutral emotion
condition (M = 5.26, SD= .88), t(128)= 6.01, p b .001 (d = .94).
Mediation of the effect of surface acting anger (relative to neutral emotion)
on demands. Our mediational model predicted that higher demands in

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

Demands

457

surface acting anger on demands through trust (b = 36.83; 95% condence interval around b = .14 to 85.14; 5000 bootstrap resamples).
Thus, surface acting anger increased demands in Round 2, in part, via
reduced trust, in support of Hypothesis 2a.

3200

3100

Effect of surface acting anger on relational outcomes. Hypothesis 1b predicted that surface acting anger would hurt relational outcomes, relative to the neutral control condition. As expected, relational outcomes
in the surface acting anger condition (M = 3.47, SD = .99) were
poorer than in the neutral condition (M = 4.26, SD = 1.02), t(127) =
4.49, p b .001 (d = .74), in support of Hypothesis 1b.

3000
Surface Acting Anger
Neutral Emotion

2900

2800

2700
Round 1

Round 2

Fig. 1. Results from Experiment 1. The effect of surface acting anger (relative to showing no emotion) on the demands on the opponent.

the surface acting anger condition (relative to the control condition)


would be explained by (reduced) trust (Hypothesis 2a). We tested
this mediational model with the second counteroffer, because the second counteroffer in the surface acting anger condition was signicantly
higher than in the neutral condition. The results of this analysis are
displayed in the top panel of Fig. 2. When accounting for the associations between surface acting anger and trust, b = .97, t(128)= 6.01,
p b .001 (f2 = .28), and between trust and demands, b = 45.13,
t(128)= 2.81, p b .01 (f2 = .06), the original association between surface acting anger and demands was no longer signicant, b = 32.42,
t(127)= .86, p = .39 (f2 = .01). A bootstrapping procedure (Preacher
& Hayes, 2008; Shrout & Bolger, 2002) showed an indirect effect of

Mediation of the effect of surface acting anger (relative to neutral


emotion) on relational outcomes. We predicted in Hypothesis 2b that
the poorer relational outcomes in the surface acting anger condition
(relative to the control condition) would be explained by (reduced)
trust. The results of this analysis are displayed in the bottom panel
of Fig. 2. When accounting for the associations between surface acting
anger and trust and between trust and relational outcomes, b = .39,
t(127) = 4.53, p b .001 (f 2 = .16), the original association between surface acting anger and relational outcomes was reduced, b = .54,
t(126) = 2.77, p b .01 (f 2 = 06). A bootstrapping procedure showed
an indirect effect of surface acting anger on relational outcomes
through trust (b = .25; 95% condence interval around b = .48
to -.07; 5000 bootstrap resamples), in support of Hypothesis 2b.
Discussion
In Experiment 1, participants were relatively intransigent in the
second round of negotiations and reported poorer relational outcomes with counterparts who surface acted anger. These effects occurred, in part, because participants had little trust in counterparts
who surface acted anger. These ndings provide initial evidence

Trust of the
Opponent
-45.13**/ -37.79*

-.97***

Demands
of the
Opponent

Surface Acting
Anger (vs. Neutral
Emotion)

69.25* / 32.42

Trust of the
Opponent
.39***/ .27**

-.97***

Surface Acting
Anger (vs. Neutral
Emotion)

-.79*** / -.54**

Relational
Outcomes
Assessed by
the Opponent

Fig. 2. Results from Experiment 1. Top panel: Mediation of the effect of surface acting anger (relative to showing no emotion) on the demands of the opponent, via the trust of the
opponent. Bottom panel: Mediation of the effect of surface acting anger (relative to showing no emotion) on relational outcomes assessed by the opponent, via the trust of the opponent. Values are unstandardized coefcients. p b .05. p b .01. pb .001.

458

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

that surface acting anger does not have the favorable consequences
that were identied in past research on the effects of showing anger
and that surface acting anger instead carries costs for negotiators.
Surface acting anger increased demands on the second, but not on
the rst counteroffer. Past research on the effect of timing on behavior
in negotiations (e.g., Pruitt, 1981; Sinaceur & Neale, 2005; Sinaceur et
al., 2011) may help interpret this result. In multiple-round negotiations,
individuals know that more information about their partners may become available after the initial round. Negotiators may thus withhold
judgment, leading to no effect of displays in the early round(s) and increasingly important effects across rounds. In Experiment 1, participants may have waited until the second round, when they observed a
second instance of surface acting anger by their opponents, to decide
how much they trusted their opponents and to adjust their behavior
accordingly. Consistent with this interpretation, in prior studies
employing more than one round, the effects of displays of emotion on
negotiation behavior were not seen on the rst round, and the effects
became stronger after each round (e.g., Van Kleef & Ct, 2007; Van
Kleef, De Dreu, & Manstead, 2004b; Van Kleef et al., 2004a).
By contrast, in single-round negotiations, negotiation counterparts may draw immediate inferences about each other because
they do not have the luxury of obtaining any additional information
later on. Participants may know that they have garnered all of the information that they will obtain about their opponents after the rst
and only round. Participants may thus make immediate judgments
about how much they trust their opponents and adjust their behavior
accordingly by placing high demands on opponents who surface act
anger. Prior studies employing single-round scenario-based paradigms typically found effects of displays of emotions in the rst and
only round (Sinaceur & Tiedens, 2006; Van Kleef, De Dreu, et al.,
2006). We examine whether surface acting anger has similar immediate effects in single-round negotiations in Experiment 2.
In Experiment 1, we compared the effects of surface acting anger to a
neutral emotional display condition. A potential alternative explanation
of our results may be that participants demanded less and reported better relational outcomes with neutral counterparts than with counterparts who surface acted anger because they perceived the former to
be more poised and in control of the situation (Sinaceur et al., 2011).
To address this possibility, in Experiment 2, we compared the effects
of surface acting anger to the effects of deep acting anger, a strategy in
which individuals change both their external displays of emotions and
their internal feelings and, as such, create more authentic displays of
emotions.
Experiment 2
In Experiment 2, we examined the effects of surface acting anger
in a single-round negotiation to examine whether it will have an immediate effect on the behavior of counterparts. In addition, we compared the effects of surface acting anger to both a neutral emotion
condition, as in Experiment 1, and a condition in which negotiators
deep acted anger. Deep acting is a strategy for emotion regulation in
which individuals inuence both their subjective feelings and their
public displays of emotions (Ct, 2005; Grandey, 2003; Hochschild,
1983). With deep acting, subjectively felt emotions match external
displays of emotion. Because this form of regulation does elicit the
subjective experience of the emotion that is displayed, negotiators
should have little reason to doubt that the anger that is displayed as
a result of deep acting anger is truly felt (Ct, 2005; Grandey,
2003; Hochschild, 1983). Thus, the effects of deep acting anger should
be comparable to those of past research on displayed anger, where
opponents had little reason to doubt the authenticity of the displayed
anger. Indeed, past research has shown that the displays of emotions
generated by deep acting are perceived as authentic (e.g., Grandey,
2003; Hennig-Thurau, Groth, Paul, & Gremler, 2006). We thus suggest
that relative to the neutral condition, deep acting anger should

decrease counterparts' demands, consistent with past research on the


effects of anger on negotiation outcomes (e.g., Sinaceur & Tiedens,
2006; Van Kleef et al., 2004a, 2004b), and surface acting anger should increase demands, as in Experiment 1.
Moreover, these opposite effects should be carried by different
mechanisms. Surface acting anger should elicit higher demands than
neutral emotions because of (reduced) trust, as in Experiment 1. By
contrast, deep acting anger should elicit lower demands than neutral
emotions because of (increased) perceptions of toughness. This reasoning is consistent with past research that found that displays of anger
that is presumed to be truly felt are interpreted as signals of toughness
(Knutson, 1996; Tiedens, 2001) that lead negotiators to infer that opponents will not back away from their positions (Sinaceur & Tiedens,
2006; Van Kleef et al., 2004a).
To test these predictions, we conducted an experiment in which
participants interacted with a video-recorded counterpart. We manipulated the regulation of anger in the video-recording. The counterpart made an offer while surface acting anger, deep acting anger, or
showing no emotion. Because participants viewed a pre-recorded
message, and an actual interaction did not take place, we did not examine relational outcomes in Experiment 2.
We tested whether compared to neutral emotion, surface acting
anger would elicit higher demands (Hypothesis 1a) and deep acting
anger would elicit lower demands (Hypothesis 3). We expected that
two distinct processes would underlie these effects. We predicted
that relative to neutral emotion, surface acting anger would reduce
trust and that (reduced) trust would mediate the effect of faking
anger on demands (Hypothesis 2a). We also predicted that deep acting anger would create perceptions of toughness and that (increased)
toughness would mediate the effect of deep acting anger on demands
(Hypothesis 4).
Method
Participants
Participants were 140 undergraduate students (66 men and 74
women) at a large university between the ages of 18 and 28 (M =
20.87, SD = 1.73). They were compensated with credit that counted
toward their nal grade in one of their courses. Participants were randomly assigned to one of the three conditions: surface acting anger
(n = 46), deep acting anger (n = 47), and a control condition (n =
47) in which the counterpart did not show emotion. The dependent
variable was demands.
Procedure
Between one and six individuals participated in the study at the
same time. When they arrived at the laboratory, participants were
seated in front of a computer. The entire experiment was conducted
on the computer to standardize the instructions and stimuli. Participants wore headphones to listen to the video-recording so that the
other participants in the room were not distracted. Participants rst
read the instructions for the negotiation. They were provided with a
hard copy of these instructions so that they could refer to them
throughout the experiment. Participants were informed that the purpose of the study was to learn more about how people negotiate via
video-conferencing, because negotiations conducted via this communication medium were becoming increasingly frequent. Participants
were told that they would engage in a car selling negotiation and
given the details of the negotiation case. As in Experiment 1, we
adapted the Blue Buggy negotiation exercise (Paulson, 2006). The
information that participants were provided was the same as in
Experiment 1.
After reading the information, participants were asked to watch the
buyer's video-recorded offer, decide whether to accept or reject the
offer, and, if they rejected the offer, propose a counteroffer. They were
informed that the buyer's message had been video-recorded to make

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

the experience real and involving. They viewed a 90-second videorecording of the potential buyer making an offer. After we developed
this script, we asked two experts who teach undergraduate and graduate level courses in negotiations at different universities to review it.
Both experts rated the script as highly realistic and as neutral in emotional tone.
The potential buyer introduced himself and said that he was a university student who was hoping to buy a decent used car. He stated a
number of features of the car, such as its condition, size, and gas performance. He then mentioned a few concerns with the car, such as the
need for body work and new tires. He said that for these reasons, he
could not pay $3500, and offered $2400 instead. We chose the value
of $2400 because it was below the lower bound of the bargaining
zone according to the expert's evaluation of the car. Thus, we could
expect a large proportion of participants to reject the offer and then
propose a counteroffer, which was the dependent variable. In a
pre-test with a separate sample, 96% of individuals rejected this offer.
There were three versions of the video-recording, one for each experimental condition. The script was exactly the same across the conditions, but the displays varied. We hired three actors and lmed them in
the same room where the participants completed the experiment. The
actors were asked to memorize the script one week in advance. We
instructed the actors to deliver the script while deep acting anger, surface acting anger, or not expressing any emotions. Prior to lming
each version of the script, we gave the actors instructions on how to regulate their displays based on procedures used in past research (Grandey
et al., 2005; Gross & Levenson, 1997; Kunzmann et al., 2005) as in
Experiment 1. For the deep acting anger condition, the actors were
instructed to recall an event that had truly made them feel angry.
These instructions were designed to make the actors truly feel angry
again about this event and, thus, express anger that they truly feel
while reciting the script. For the surface acting condition, the actors
were asked to put on a show and pretend to be angry. As in Experiment 1, we instructed the confederates to lower their eyebrows and
draw them together and to ensure that their eyes are glaring and their
jaw is clenched. These instructions were designed to make them only
show anger in their face and, thus, express anger that they do not
truly feel. Prior to lming the neutral condition, the actors were
instructed to be business-like and to relax the muscles in their faces
so as to show no emotion. Each actor recited the script in each condition
several times in front of the video camera.
All of the video-recordings were then rated by each author independently and blind to condition on the intensity of anger and the authenticity of expression of emotion. The ratings of the authors
converged on the same actor as being the best. This actor was invited
to re-lm some of the video-recordings. We then selected the
video-recordings of this actor that best met the desired criteria (i.e.,
the anger expression conditions were rated equally high in anger,
and the surface acting anger condition was rated lower in authenticity than the deep acting anger condition).
Past research showed that inauthentic displays of emotions are less
symmetrical than authentic displays (Ekman et al., 1981; Hager &
Ekman, 1985). To verify that the authenticity of the displays was manipulated as intended, we assessed the symmetry of the displays of anger
in the video-recordings. We rst divided each video-recording into
5-second segments. Two raters then independently watched each segment and assigned a score of 1 if the display was symmetrical and 0 if
it was asymmetrical. We calculated intra-class correlations to estimate
the amount of variance observed within segments and across raters
(Bliese, 2000). ICCs were .73 for fake anger and .69 for genuine anger,
showing adequate inter-rater reliability. As expected, displays were
coded as more symmetrical in the deep acting anger (M = .81, SD=
.40) than in the surface acting anger condition (M = .47, SD= .52),
t(29) = 2.09, p b .05.
After viewing the video-recording, participants indicated whether
they accepted the buyer's offer. If they rejected the offer, they

459

proposed a counteroffer. These questions were followed by measures


of mechanisms, manipulation checks, and demographic questions.
We included the manipulation checks after the measures of the dependent variable to avoid priming participants to think about the displays and authenticity of emotions while deciding whether to accept
or reject the offer and what to counter-offer. Participants were
debriefed at the conclusion of the experiment, which lasted about
half an hour. Within this half hour, the negotiation lasted about
3 min. This included the 90-s video plus the time participants took
to respond to the video.
Measures
Manipulation checks. We used the same measures of perceived anger
(alpha = .88), neutral emotion (alpha = .70), and authenticity
(alpha = .82) as in Experiment 1.
Demands. The dependent variable was the demands of participants
following the counterpart's initial offer of $2400. As expected, almost
all participants (137 out of 140, or 98%) rejected this offer. One additional participant did not respond to this question. After rejecting the
offer, participants proposed a counteroffer by indicating the amount
that they wanted for their car. The four participants with missing
values for demands were removed from the analyses involving demands (but not the other analyses).
Trust. We used the same three items as in Experiment 1 (alpha = .82).
Perceived toughness
We used three items to assess participants' perceptions of the
buyer's toughness. Participants rated the degree to which the buyer
was tough, competitive, and ambitious (alpha = .62).
Results
Manipulation checks
Displayed anger. The ANOVA on ratings of anger was signicant, F(2,
137) = 51.48, p b .001 ( 2 = .43). Planned contrasts revealed that participants provided lower ratings of anger in the control condition
(M = 2.80, SD = 1.12) than in the deep acting anger condition (M =
5.20, SD = 1.29), t(137) = 8.90, p b .001 (d = 1.40) and the surface acting anger condition (M = 5.14, SD = 1.48), t(137) = 8.65, p b .001 (d =
1.37). Also, the surface acting anger condition and the deep acting
anger condition did not differ from each other in their ratings of
anger, t(137) = .20, p = .84 (d = .03). The similar level of intensity indicates that any difference in the effects of surface acting anger and
deep acting anger is not due to a difference in the intensity with
which anger is displayed.
The ANOVA on ratings of neutral emotion was also signicant, F(2,
137) = 36.18, p b .001 ( 2 = .35). As expected, participants perceived
the counterpart as more neutral in the neutral condition (M = 4.95,
SD = 1.16) than in the deep acting anger condition (M = 3.28, SD =
1.18), t(137) = 6.52, p b .001 (d = 1.10) and the surface acting anger
condition (M = 2.90, SD = 1.37), t(137) = 2.05, p b .001 (d = 1.35).
Also, the surface acting anger condition and the deep acting anger
condition did not differ from each other in their ratings of neutral
emotion, t(137) = 1.50, p = .14 (d = .25).
Authenticity. The ANOVA on ratings of anger was signicant, F(2,
137) = 7.81, p b .001 ( 2 = .10). Planned contrasts revealed that, as
expected, participants provided lower ratings of authenticity in the
surface acting anger condition (M = 3.43, SD = 1.22) than in the
deep acting anger condition (M = 3.95, SD = .89), t(137) = 2.35,
p b .05 (d = .46) and the control condition (M = 4.31, SD = 1.11),
t(137) = 3.93, p b .001 (d = .78). The deep acting anger condition

460

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

and the neutral emotion condition did not differ from each other in
their ratings of authenticity, t(137) = 1.59, p = .11 (d = .31).
Focal analyses
Effect of surface acting anger and deep acting anger on demands. As in
Experiment 1, we rst examined potential gender differences.
Women (M = 3060.42, SD = 215.25) placed lower demands than
men (M = 3131.23, SD = 189.49), t(134) = 2.03, p b .05 (d = .34). We
repeated all of the analyses using ANCOVA (controlling for gender),
and the conclusions were the same. For the sake of simplicity, we report the results of the analyses that do not control for gender.
We had predicted that surface acting anger (Hypotheses 1a) and
deep acting anger (Hypothesis 3) would inuence the demands of
the counterpart, relative to the neutral control condition. An ANOVA
on demands was signicant, F(2, 133)= 9.66, p b .001 (2 = .13). Participants in the surface acting anger condition exerted higher demands
(M = 3187.19, SD = 181.94) than those in the neutral condition,
(M = 3093.62, SD= 209.20), t(133) = 2.29, p b .05 (d = .45), in support
of Hypothesis 1a. Participants in the deep acting anger condition
exerted lower demands (M = 3006.52, SD= 188.15) than those in the
neutral condition, t(133) = 2.17, p b .05 (d = .42), in support of Hypothesis 3. Logically, the two anger conditions also differed from each other,
t(133)= 4.39, p b .001 (d = .88). These results indicate that participants
were relatively lenient with counterparts who deep acted anger, and
relatively intransigent with counterparts who surface acted anger (see
Fig. 3). Thus, Hypotheses 1a and 3 were supported.
Effect of surface acting anger and deep acting anger on trust. We
expected that participants would trust a counterpart who surface
acts anger less than an emotionally neutral counterpart. The ANOVA
on trust was signicant, F(2, 137) = 18.27, p b .001 ( 2 = .21). As
expected, trust was lower in the surface acting anger condition
(M = 3.59, SD = 1.22) than in the neutral condition (M = 4.98, SD =
1.06), t(137) = 6.00, p b .001 (d = 1.11).
In addition, trust was lower in the surface acting anger condition
than in the deep acting anger condition (M= 4.44, SD= 1.07),
t(137)= 3.68, p b .001 (d = .68), and lower in the deep acting anger condition than in the neutral condition, t(137) = 2.34, p b .05 (d = .43).
Effect of surface acting anger and deep acting anger on perceptions of
toughness. We expected that participants would perceive a counterpart who deep acts anger to be tougher than an emotionally neutral
counterpart. The ANOVA on perceived toughness was signicant,
F(2, 137)= 4.89, p b .01 (2 = .07). Perceived toughness was higher in
the deep acting anger condition (M = 5.31, SD= 1.18) than in the neutral condition (M = 4.72, SD= 1.04), t(137)= 2.73, p b .01 (d = .55).

Demands
3300

3100
Surface Acting Anger
Neutral Emotion

2900

Deep Acting Anger

2700
Fig. 3. Results from Experiment 2. The effects of surface acting anger and deep acting
anger (relative to showing no emotion) on the demands of the opponent.

In addition, perceived toughness was higher in the surface acting


anger condition (M = 5.30, SD = .89) than in the neutral condition,
t(137) = 2.68, p b .01 (d = .54). However, it was not higher in the
deep acting anger than in the surface acting anger condition,
t(137) = .04, p = .97 (d = .01). Participants perceived counterparts
who expressed any anger as tough. We report the results of a
follow-up experiment to illuminate this result below.

Mediation of the effect of surface acting anger (relative to neutral emotion) on demands. We predicted in Hypothesis 2a that higher demands in the surface acting anger condition (relative to the neutral
control condition) would be explained by low trust. The results of
this analysis are displayed in the top panel of Fig. 4. When accounting
for the associations between surface acting anger and trust, b = 1.39,
t(91) = 5.88, p b .001 (f 2 = .28), and between trust and demands,
b = 63.15, t(88) = 4.33, p b .001 (f 2 = .21), the original association
between surface acting anger and demands was no longer signicant,
b =9.28, t(87) = .20, p = .84 (f2 = .00). A bootstrapping procedure
showed an indirect effect of surface acting anger on demands through
trust (b = 84.29; 95% condence interval around b = 33.68 to 150.43;
5000 bootstrap resamples). These results indicate that surface acting
anger increased demands because of reduced trust, consistent with
the ndings of Experiment 1, and in support of Hypothesis 2a.

Mediation of the effect of deep acting anger (relative to neutral emotion)


on demands. In Hypothesis 4, we predicted that lower demands in the
deep acting anger condition (relative to the neutral control condition)
would be explained by perceptions of toughness. The results of this
analysis are displayed in the bottom panel of Fig. 4. When accounting
for the associations between deep acting anger and toughness, b =
.59, t(92) = 2.56, p b .05 (f 2 = .07), and between toughness and demands, b = 46.86, t(91) = 2.63, p b .01 (f 2 = .08), the original association between deep acting anger and demands was no longer
signicant, b = 63.74, t(90) = 1.52, p = .13 (f 2 = .03). Moreover, a
bootstrapping procedure showed an indirect effect of deep acting
anger on demands through toughness (b = 23.36; 95% condence
interval around b= 63.33 to 4.21; 5000 bootstrap resamples).
Deep acting anger increased demands due to higher perceptions of
toughness, consistent with the ndings of past research (e.g., Sinaceur
et al., 2011; Van Kleef et al., 2004a) and in support of Hypothesis 4.

Discussion
The ndings of Experiment 2 provide additional evidence that the
effects of surface acting anger are detrimental to negotiation outcomes, and are opposite to the benecial effects of showing anger in
negotiations found in previous studies. In particular, consistent with
the ndings of Experiment 1, Experiment 2 shows that surface acting
anger increases demands due to reduced trust. At the same time, deep
acting anger decreases demands because it increases perceptions of
toughness. What is common to the displays of anger examined in
past research (Sinaceur & Tiedens, 2006; Van Kleef et al., 2004a,
2004b) and the deep acting anger condition in Experiment 2 is that
the observers had no reason to doubt that the anger that is displayed
is not truly felt. The fact that we replicate past ndings in the deep
acting anger condition in Experiment 2 supports this notion. The results are different in the surface acting condition because observers
of anger trusted the expresser less. The same emotion, anger, had opposite effects on negotiation outcomes depending on how it was
regulated.
Experiment 2 also extended the previous ndings by showing that
surface acting anger can have immediate effects on the behavior of
negotiation counterparts if there is only one round of negotiations.

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

-1.39***

Surface Acting
Anger (vs. Neutral
Emotion)

Trust of the
Opponent

93.57* / 9.28

Perceived
Toughness
by the
Opponent

.59*

Deep Acting
Anger (vs. Neutral
Emotion)

-87.10* / -63.74

461

-63.15***/ -61.34***

Demands
of the
Opponent

-46.86**/ -39.72*

Demands
of the
Opponent

Fig. 4. Results from Experiment 2. Top panel: Mediation of the effect of surface acting anger (relative to showing no emotion) on the demands of the opponent, via the trust of the
opponent. Bottom panel: Mediation of the effect of deep acting anger (relative to showing no emotion) on the demands of the opponent, via perceived toughness of the opponent.
Values are unstandardized coefcients. p b .05. p b .01. p b .001.

Experiment 2 follow-up
Experiment 2 yielded an unexpected nding: both surface acting
anger and deep acting anger were perceived as tough. Surface acting
anger may be perceived as tough for a different reason than deep acting anger. Specically, surface acting may communicate that negotiators are willing to break rules, as they are showing emotions that they
do not truly feel, to gain what they want. In contrast, deep acting may
communicate that negotiators are ambitious. To shed light on this
issue, we collected additional data from 148 undergraduate students
who did not participate in the main experiments. We randomly
assigned them to view the surface acting anger, deep acting anger,
or neutral video-recording from Experiment 2. We administered the
three-item toughness scale ( = .70) along with an item to assess
ambition (He is determined to do well in this negotiation) and an
item to assess willingness to break rules (He is willing to be dishonest to gain something in this negotiation), rated on a 1 (strongly disagree) to 7 (strongly agree) scale. Again, both deep acting and surface
acting anger were perceived as tough (relative to neutral). There was
an indirect effect of surface acting anger (relative to neutral) on
toughness through willingness to break rules (b = .14; 95% C. I.
around b = .02 to .37), but not through ambition (b = .07; 95% C. I.
around b = .10 to .33). Conversely, there was an indirect effect of
deep acting anger (relative to neutral) on toughness through ambition (b = .40; 95% C. I. around b = .15 to .75), but not through willingness to break rules (b = .03; 95% C. I. around b = .01 to .19). These
tentative results suggest that negotiators who deep act anger and
who surface act anger are both perceived as tough, but for different
reasons.
General discussion
This research examined the effects of surface acting anger on the
behavior of counterparts in negotiations. Taken together, the ndings
show that individuals place particularly high demands, are relatively
dissatised, and have relatively little interest in negotiating again

with opponents who surface act anger, because they have little trust
in them. These ndings extend previous work suggesting that expressions of anger in negotiation are benecial (e.g., Sinaceur & Tiedens,
2006; Van Dijk, Van Kleef, Steinel, & Van Beest, 2008; Van Kleef, De
Dreu, et al., 2006; Van Kleef et al., 2004a, 2004b). Thus, the ndings
contribute to theory by showing that the same emotion anger
has opposite consequences on negotiation processes and outcomes
depending on whether and how it is regulated and, in turn, on how
authentic the display is perceived to be. Thus, adequate theories of
anger and conict resolution should incorporate the role of how displays of anger originate and, hence, the regulation of the displays of
anger, because the effects differ in meaningful ways.
In both Experiments, the effects of surface acting were mediated
by trust. These mediation ndings, along with the addition of the
deep acting anger condition in Experiment 2, alleviate concerns
about some alternative explanations of the results. First, neutral
counterparts may seem more poised than counterparts who show
anger (Sinaceur et al., 2011). The effect in Experiment 1 could have
been driven by the neutral condition rather than by the surface acting
anger condition, so that participants tended to pose lower demands
on counterparts who were neutral because they were more poised.
This concern is alleviated by the inclusion of the deep acting anger
condition in Experiment 2. Participants posed even lower demands
with counterparts who deep acted anger. An explanation based on
poise would have predicted instead that participants would demand
less from the neutral counterparts than from both counterparts who
surface acted anger and counterparts who deep acted anger.
A second concern could be that counterparts who surface acted
anger may have been perceived as particularly intense, and participants could have been reacting to the intensity of anger rather than
any doubts about the authenticity of the anger. Alleviating this concern, Experiment 2 showed that even though there were no differences in intensity perceptions between surface acting and deep
acting anger, the results were opposite, speaking against a role of perceived intensity of display in explaining the results. Third, although
counterparts who surface acted anger were unexpectedly found to

462

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463

be perceived as tough, the analyses in Experiment 2 showed that perceived toughness did not mediate the effect of surface acting anger on
demands. Finally, the mediations through trust and toughness guard
against alternative explanations based on the display and perception
of other emotions. For example, recalling anger-inducing event could
have induced nervousness in confederates who deep acted anger.
However, in past research, anxious negotiators made low initial offers
and ultimately did worse in negotiations than neutral negotiators
(Brooks & Schweitzer, 2011). Thus, it is unlikely that participants
would perceive nervous counterparts to be tough and, in turn, pose
low demands with these counterparts.
More broadly, our research has implications for social-functional
accounts of emotion (e.g., Keltner & Haidt, 1999; Morris & Keltner,
2000; Van Kleef, 2009). According to this perspective, emotions are
not merely internal states they are expressed and observed by
others, and they may inuence others' behavior, cognitions, and emotions. As such, emotions fulll an important coordinating function in
social life. Most theories of the social functions of emotions have not
considered whether genuine or deep acted displays of emotions
have the same effects as surface acted displays. Our ndings indicate
that whether and how emotions are regulated need to play more central roles in theoretical accounts of the social effects of emotions.
Our research also contributes to the literature on emotion regulation. Previous research has uncovered consequences of emotion
regulation for several intra-individual outcomes, such as well-being,
physiology, cognitive functioning, and job performance (e.g., Grandey,
2003; Gross & Levenson, 1997; Gyurak et al., 2009). This research has
focused principally on the consequences for the person regulating the
emotion; few studies have considered the social consequences of emotion regulation (for exceptions, see Butler, Lee, & Gross, 2007; Butler et
al., 2003; Srivastava et al., 2009). Research on the social effects of emotions suggests that how people regulate their emotions may have implications for how interaction partners act, think, and feel (Ct, 2005;
Keltner & Haidt, 1999; Van Kleef, 2009). Our research shows that the social effects of various emotion regulation strategies extend to negotiation behavior, indicating that one's choice of regulation strategy can
have important consequences. More broadly, our research indicates
that regulating emotions in a manner that leads observers to doubt
the authenticity of the emotions may be a maladaptive strategy during
negotiations that creates mistrust and escalates demands.
Caveats and future directions
One limitation of the experiments is that we only used male actors. Some past studies have found that displays of anger by women
and men are interpreted differently. These studies have shown that
people have more negative reactions when women display anger
than when men do (e.g., Gibson, Schweitzer, Callister, & Grey,
2009). This suggests that whether women's displays of anger originate in surface acting or deep acting may be less important for negotiation outcomes, because the reactions could be more consistently
negative. It is important in future research to examine whether reactions to surface acting displays of anger by women are the same ones
we found for men.
The use of video-recorded displays of anger in Experiment 2 constitutes another possible limitation. This feature of Experiment 2 provided us with experimental control that increased our condence
about the causal effects. This control, however, came at the expense
of mundane realism. We are reassured by the consistency of the results with the video-recording paradigm in Experiment 2 and the
face-to-face paradigm in Experiment 1. The ndings about surface
acting anger were consistent with Experiment 1 in which negotiations occurred face-to-face. The increasing popularity of information
technologies as communication media renders ndings about
computer-mediated negotiations theoretically and practically important (McKersie & Fonstad, 1997; Moore, Kurtzberg, Thompson, &

Morris, 1999). Even so, it is important to investigate the generalizability of the ndings to other communication media in future research.
In Experiment 2 in which participants reacted to previously
video-recorded counterparts, surface acting anger increased demands
on the rst (and only) counteroffer. In Experiment 1 in which negotiations took place face-to-face, however, surface acting anger increased
demands, relative to neutral displays of emotion, on the second, but
not on the rst counteroffer. The results for the initial round may have
been different because in multiple-round negotiations, such as in
Experiment 1, negotiators have the luxury of time to increase their condence in their impressions of their counterparts. Negotiators may thus
not act fully from these impressions or act on them until later rounds. In
single-round negotiations, such as in Experiment 2, negotiators may
need to form impressions and act on them immediately.
Another possibility is that negotiators are more focused on their
counterparts' displays of emotions in video-mediated negotiations,
as in Experiment 2, than in face-to-face negotiations, as in Experiment 1. In video-mediated negotiations, negotiators may be focused
primarily on the video, which would increase their likelihood of identifying and responding to how anger is regulated. This may explain
why we detected an effect of surface acting anger in the initial (and
only) round in Experiment 2. In face-to-face negotiations, there may
be more distractions and more stimuli that negotiators try to focus
on simultaneously. It would thus take longer to pick up on how
anger is regulated. This may explain why we did not nd an effect
in round 1 (of 2) in Experiment 1. These dynamics should be examined further in future research on how displays of anger shape negotiations over time and via different media.
Finally, in this investigation, we examined the effects of using surface acting to create a completely new display of anger that did not
previously exist. We did not examine the effects of using surface acting to exaggerate a display of anger that previously existed. It is possible that surface acting anger has different consequences when it is
used to exaggerate a relatively low level of anger that already exists
than when it is used to create a completely new display. This issue
should be examined in future research.
Conclusion
The present research contributes to the literature on anger in negotiations by demonstrating that the effects of faking anger are different
than the effects of showing genuine anger or showing no emotion. Genuine anger elicits concessions, but fake anger elicits demands in negotiation counterparts. Thus, the source and authenticity of the displays of
anger have to be considered to obtain a complete understanding of
the role of expressed anger in negotiations. The strategic management
of anger has important consequences for the behavior of parties during
negotiation, and, we suspect, for social behavior in general.
References
Andrade, E. B., & Ho, T. -H. (2009). Gaming emotions in social interactions. Journal of
Consumer Research, 36, 539551.
Averill, J. R. (1983). Studies on anger and aggression: Implications for theories of emotion. The American Psychologist, 38, 11451160.
Bliese, P. (2000). Within-group agreement, non-independence, and reliability. In K.
Klein, & S. Kozlowski (Eds.), Multi-level theory, research, and methods in organizations:
Foundations, extensions, and new directions (pp. 349381). San Francisco, CA:
Jossey-Bass.
Bono, J. E., & Ilies, R. (2006). Charisma, positive emotions and mood contagion. The
Leadership Quarterly, 17, 317334.
Brooks, A. W., & Schweitzer, M. E. (2011). Can Nervous Nelly negotiate? How anxiety
causes negotiators to make low rst offers, exit early, and earn less prot. Organizational Behavior and Human Decision Processes, 115, 4354.
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, 5767.
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, 4867.
Butler, E. A., Lee, T. L., & Gross, J. J. (2007). Emotion regulation and culture: Are the social consequences of emotion suppression culture-specic? Emotion, 7, 3048.

S. Ct et al. / Journal of Experimental Social Psychology 49 (2013) 453463


Chalian, D. (2010). David Chalian Says Farewell to Top Line. Available from http://
abcnews.go.com/Politics/video/david-chalian-farewell-top-line-10888445
Ct, S. (2005). A social interaction model of the effects of emotion regulation on work
strain. Academy of Management Review, 30, 509530.
De Dreu, C. K. W., Beersma, B., Stroebe, K., & Euwema, M. (2006). The interaction between social motives and epistemic motives in negotiation. Journal of Personality
and Social Psychology, 90, 927943.
De Dreu, C. K. W., Giebels, E., & Van de Vliert, E. (1998). Social motives and trust in integrative negotiation: The disruptive effects of punitive capability. The Journal of
Applied Psychology, 83, 408422.
Ekman, P. (2003). Emotions revealed. New York: Henry Holt.
Ekman, P., Friesen, W. V., & O'Sullivan, M. (1988). Smiles when lying. Journal of Personality and Social Psychology, 54, 414420.
Ekman, P., Hager, J. C., & Friesen, E. V. (1981). The symmetry of emotional and deliberate facial actions. Psychophysiology, 18, 101106.
Ferrin, D. L., Bligh, M. C., & Kohles, J. C. (2007). Can I trust you to trust me? A theory of
trust, monitoring, and cooperation in interpersonal and intergroup relationships.
Group & Organization Management, 32, 465499.
Fitness, J. (2000). Anger in the workplace: An emotion script approach to anger episodes between workers and their superiors, co-workers and subordinates. Journal
of Organizational Behavior, 21, 147162.
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, 8393.
Grandey, A. A. (2000). Emotional regulation in the workplace: A new way to conceptualize emotional labor. Journal of Occupational Health Psychology, 5, 95110.
Gelfand, M. J., Major, V. S., Raver, J. L., Nishii, L. H., & O'Brien, K. (2006). Negotiating
relationally: The dynamics of the relational self in negotiations. Academy of
Management Review, 31, 427451.
Gibson, D. E., Schweitzer, M., Callister, R. R., & Grey, B. (2009). The inuence of anger
expressions on outcomes in organizations. Negotiation and Conict Management
Research, 2, 236262.
Grandey, A. (2003). When the show must go on: Surface and deep acting as determinants
of emotional exhaustion and peer-rated service delivery. Academy of Management
Journal, 46, 8696.
Grandey, A., Fisk, G., Mattila, A., Jansen, K., & Sideman, L. (2005). Is service with a smile
enough? Authenticity of positive displays in service encounters. Organizational
Behavior and Human Decision Processes, 96, 3855.
Gross, J. J. (1998). Antecedent- and response-focused emotion regulation: Divergent
consequences for experience, expression, and physiology. Journal of Personality
and Social Psychology, 74, 224237.
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, 348362.
Gross, J. J., & Levenson, R. W. (1997). Hiding feelings: The acute effects of inhibiting
negative and positive emotion. Journal of Abnormal Psychology, 106, 95103.
Gyurak, A., Goodkind, M. S., Madan, A., Kramer, J. H., Miller, B. L., & Levenson, R. W.
(2009). Do tests of executive functioning predict ability to down-regulate emotions
spontaneously and when instructed to suppress? Cognitive, Affective, & Behavioral
Neuroscience, 9, 144152.
Hager, J. C., & Ekman, P. (1985). The asymmetry of facial actions is inconsistent with
models of hemispheric specialization. Psychophysiology, 22, 307318.
Hennig-Thurau, T., Groth, M., Paul, M., & Gremler, D. (2006). Are all smiles created
equal? How emotional contagion and emotional labor affect service relationships.
Journal of Marketing, 70, 5873.
Hochschild, A. R. (1983). The managed heart. Berkeley, CA: University of California Press.
Johnston, L., Miles, L., & Macrae, C. N. (2010). Why are you smiling at me? Social functions
of enjoyment and non-enjoyment smiles. The British Journal of Social Psychology, 49,
107127.
Kelley, H. H., & Thibaut, J. W. (1978). Interpersonal relations. New York: Wiley.
Keltner, D., & Haidt, J. (1999). Social functions of emotions at four levels of analysis.
Cognition and Emotion, 13, 505521.
Knutson, B. (1996). Facial expressions of emotion inuence interpersonal trait inferences. Journal of Nonverbal Behavior, 20, 165182.
Kopelman, S., Rosette, A. S., & Thompson, L. (2006). The three faces of eve: An examination of the strategic display of positive, negative, and neutral emotions in negotiations. Organizational Behavior and Human Decision Processes, 99, 81101.
Kramer, R. M. (1999). Trust and distrust in organizations: Emerging perspectives, enduring questions. Annual Review of Psychology, 50, 569598.
Kray, L. J., Galinsky, A. D., & Thompson, L. (2002). Reversing the gender gap in negotiations: An exploration of stereotype regeneration. Organizational Behavior and
Human Decision Processes, 87, 386409.
Kring, A. M. (2000). Gender and anger. In A. H. Fischer (Ed.), Gender and emotion: Social
psychological perspectives (pp. 211231). New York: Cambridge University Press.
Krumhuber, E., Manstead, A. S. R., Cosker, D., Marshall, D., Rosin, P. L., & Kappas, A.
(2007). Facial dynamics as indicators of trustworthiness and cooperative behavior.
Emotion, 7, 730735.
Kunzmann, U., Kupperbusch, C. S., & Levenson, R. W. (2005). Behavioral inhibition and
amplication during emotional arousal: A comparison of two age groups. Psychology
and Aging, 20, 144158.
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (1995). An integrative model of organizational trust. Academy of Management Review, 20, 709734.
McKersie, R. B., & Fonstad, N. O. (1997). Teaching negotiation theory and skills over the
Internet. Negotiation Journal, 13, 363368.
Moore, D. A., Kurtzberg, T. R., Thompson, L. L., & Morris, M. W. (1999). Long and short
routes to success in electronically mediated negotiations: Group afliations and
good vibrations. Organizational Behavior and Human Decision Processes, 77, 2243.

463

Morris, M. W., & Keltner, D. (2000). How emotions work: The social functions of emotional expression in negotiations. Research in Organizational Behavior, 22, 150.
Pareene, A. (2010, June 2). Why won't Obama just get even madder about this oil
spill. Salon.com. Retrieved from http://www.salon.com/2010/06/02/obama_anger_
pundits/singleton/.
Paulson, G. (2006). Blue Buggy. In J. M. Brett, & L. Thompson (Eds.), Negotiation, teamwork,
and decision making exercises. Evanston, IL: Dispute Resolution Research Center,
Northwestern University.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing
and comparing indirect effects in multiple mediator models. Behavior Research
Methods, 40, 879891.
Pruitt, D. G. (1981). Negotiation behavior. New York: Academic Press.
Rafaeli, A., & Sutton, R. I. (1989). The expression of emotion in organizational life.
Research in Organizational Behavior, 11, 142.
Raum, T. (2010). Obama struggles to show he's mad about the BP oil spill, but he's drawing
lots of skepticism. The Associated Press (Retrieved from http://www.news957.
com/news/world/article/64315obama-struggles-to-show-he-smad-about-the-bpoil-spill-but-he-s-drawing-lots-of-skepticism)
Ross, W., & LaCroix, J. (1996). Multiple meanings of trust in negotiation theory and
research: A literature review and integrative model. International Journal of Conict
Management, 7, 314360.
Schmeichel, B. J., Volokhov, R. N., & Demaree, H. A. (2008). Working memory capacity and
the self-regulation of emotional expression and experience. Journal of Personality and
Social Psychology, 95, 15261540.
Shrout, P. E., & Bolger, N. (2002). Mediation in experimental and nonexperimental studies:
New procedures and recommendations. Psychological Methods, 7, 422445.
Sinaceur, M., & Neale, M. (2005). Not all threats are created equal: How implicitness
and timing affect the effectiveness of threats in negotiations. Group Decision and
Negotiation, 14, 6385.
Sinaceur, M., & Tiedens, L. Z. (2006). Get mad and get more than even: When and why
anger expression is effective in negotiations. Journal of Experimental Social Psychology,
42, 314322.
Sinaceur, M., Van Kleef, G. A., Neale, M. A., Adam, H., & Haag, C. (2011). Hot or cold: Is
communicating anger or threats more effective in negotiation? The Journal of
Applied Psychology, 96, 10181032.
Skinner, M., & Mullen, B. (1991). Facial asymmetry in emotional expression: A
meta-analysis of research. The British Journal of Social Psychology, 30, 113124.
Small, D. A., Gelfand, M., Babcock, L., & Gettman, H. (2007). Who goes to the bargaining
table? The inuence of gender and framing on the initiation of negotiation. Journal
of Personality and Social Psychology, 93, 600613.
Smith, C. A., Haynes, K. N., Lazarus, R. S., & Pope, L. K. (1993). In search of the hot
cognitions: Attributions, appraisals, and their relation to emotion. Journal of
Personality and Social Psychology, 65, 916929.
Srivastava, S., Tamir, M., McGonigal, K. M., John, O. P., & Gross, J. J. (2009). The social
costs of emotional suppression: A prospective study of the transition to college.
Journal of Personality and Social Psychology, 96, 883897.
Sutton, R. I. (1991). Maintaining norms about expressed emotions: The case of bill collectors. Administrative Science Quarterly, 36(245), 268.
Sy, T., Ct, S., & Saavedra, R. (2005). The contagious leader: Impact of the leader's
mood on the mood of group members, group affective tone, and group processes.
The Journal of Applied Psychology, 90, 295305.
Tamir, M., Mitchell, C., & Gross, J. J. (2008). Hedonic and instrumental motives in anger
regulation. Psychological Science, 19, 324328.
Tiedens, L. Z. (2001). Anger and advancement versus sadness and subjugation: The effect
of negative emotion expressions on social status conferral. Journal of Personality and
Social Psychology, 80, 8694.
Timmers, M., Fischer, A. H., & Manstead, A. S. R. (1998). Gender differences in motives for regulating emotions. Personality and Social Psychology Bulletin, 24,
974985.
Van Dijk, E., Van Kleef, G. A., Steinel, W., & Van Beest, I. (2008). A social functional approach to emotions in bargaining: When communicating anger pays and when it
backres. Journal of Personality and Social Psychology, 94, 600614.
Van Kleef, G. A. (2009). How emotions regulate social life: The emotions as social information (EASI) model. Current Directions in Psychological Science, 18, 184188.
Van Kleef, G. A., & Ct, S. (2007). Expressing anger in conict: When it helps and when
it hurts. The Journal of Applied Psychology, 92, 15571569.
Van Kleef, G. A., De Dreu, C. K. W., & Manstead, A. S. R. (2004a). The interpersonal
effects of anger and happiness in negotiations. Journal of Personality and Social
Psychology, 86, 5776.
Van Kleef, G. A., De Dreu, C. K. W., & Manstead, A. S. R. (2004b). The interpersonal effects of emotions in negotiations: A motivated information processing approach.
Journal of Personality and Social Psychology, 87, 510528.
Van Kleef, G. A., De Dreu, C. K. W., & Manstead, A. S. R. (2006). Supplication and appeasement in conict and negotiation: The interpersonal effects of disappointment, worry, guilt, and regret. Journal of Personality and Social Psychology, 91,
124142.
Van Kleef, G. A., De Dreu, C. K. W., Pietroni, D., & Manstead, A. S. R. (2006). Power and
emotion in negotiations: Power moderates the interpersonal effects of anger and
happiness on concession making. European Journal of Social Psychology, 36,
557581.
Van Kleef, G. A., Homan, A. C., Beersma, B., van Knippenberg, D., van Knippenberg, B., &
Damen, F. (2009). Searing sentiment or cold calculation? The effects of leader emotional displays on team performance depend on follower epistemic motivation.
Academy of Management Journal, 52, 562580.
Yamagishi, T., & Yamagishi, M. (1994). Trust and commitment in the United States and
Japan. Motivation and Emotion, 1, 129166.