3.1 Precipitating the Turning Point
Turning points are typically preceded by a longer period in which negotiators make no progress towards a mutually beneficial agreement (Druckman
1986). A number of studies confirm that before starting to integrate and create value, negotiators engage in distributive value claiming—they assume that their goals are incompatible, make ambit claims, posture for position, and their argumentation focuses on their sources of power and perceived rights (Adair and Brett
2005; Lytle et al.
1999; Olekalns and Weingart
2003; Putnam
1990). When perpetuating this pattern of positional zero-sum bargaining, negotiators are likely to experience goal frustration and sense that an impasse is looming. Not moving closer to their desired objective, in turn, gives rise to negative emotions (Frijda
1986; Kumar
1997; Lazarus
1991). Prior research has shown that negotiators in a negative affect state have less mutual trust (Anderson and Thompson
2004), are less cooperative and have less concern for the counterpart (Forgas
1998; Yifeng et al.
2008).
We argue that the expression of negative emotions may also be beneficial as it serves as a precipitant to a turning point possibly steering the negotiation back to the right track. This view is supported by studies showing that the display of negative affect results in larger concessions by the counterpart (Van Kleef et al.
2004a,
b). Also, when perceived as justified, negative emotions do not necessarily lead to retaliation (Van Kleef and Côté
2007) and especially in competitive settings they may induce the counterpart to be more cooperative (Griessmair
2017; Van Kleef et al.
2010). Based on their function as negative reinforcer and feedback mechanism for counterproductive behavior (Cacioppo and Gardner
1999; Fischer and Roseman
2007; Morris and Keltner
2000), we propose that the turning points are initiated by a negative transformational sequence of goal-oriented emotions that destabilizes the ongoing negotiation pattern and signals that a behavioral change is required (see first interact in Fig.
2c, d). Thereby, it provides a stimulus for altering the dynamics of the negotiation and acts as a precipitant to the turning point.
The expression of negative emotions, however, does not only have informational and disincentive functions but can also have detrimental effects (Van Kleef and Côté
2007). Negotiators confronted with an opponent expressing negative affect are more likely to develop a negative impression of the counterpart (Van Kleef et al.
2004a,
b) and are less likely to reach an agreement (Friedman et al.
2004; Kopelman et al.
2006) or to be willing to engage in future negotiations with the counterpart (Kopelman et al.
2006). Furthermore, negative emotions are likely to be reciprocated by the counterpart (Barsade
2002; Hatfield et al.
1994), particularly if they appear unjustified and both actors are equal in power (Van Kleef and Côté
2007). For instance, Friedman et al. (
2004) show that the display of anger triggers an angry response by the counterpart, which, in turn, increases the likelihood of not reaching an agreement. This pattern of contentious reciprocation and its detrimental effects has been confirmed with regard to emotions in negotiations (Nielek et al.
2010) as well as with strategies and tactics (Brett et al.
1998; Olekalns and Smith
2000; Weingart et al.
1999,
1990).
Thus, introducing the departure via a negative transformational sequence of goal-oriented emotions may backfire. Prior research investigating when negative emotions have positive or negative effects has primarily focused on contextual variables such as justifiability, interdependence and power structure, and information processing needs (Van Kleef and Côté
2007). We propose that the potentially detrimental effects that may arise from introducing a departure via a negative transformational sequence of goal-oriented emotions cannot only be countered by contextual factors but also by coupling it with a positive transformational sequence of other-oriented emotions.
The suggested mechanism is akin to Fisher and Ury’s (
1981) classic advice to separate the people from the problem and Brett et al.’s (
1998) concept of mixed communication. Inspired by Fisher and Ury (
1981), Steinel et al. (
2008) show that negative behavior-oriented emotions—emotions directed towards “the problem” such as the counterpart’s offer—induced the counterpart to make concessions. Conversely, negative person-oriented emotions—emotions directed towards “the people”—had no such effect. Based on this study, Van Kleef and Côté (
2007) conclude that while expressing negative emotions about a person’s offer seems to pay, expressing negative emotions at them personally may backfire.
While Steinel et al. (
2008) have investigated person- and behavior-oriented emotions separately in an experimental setting, in actual negotiation interactions allowing for naturalistic communication negotiators can convey multiple ideas within a single speaking turn (Brett et al.
1998; Weingart and Olekalns
2004; Weingart et al.
2004). That is, they can pair negative goal-oriented emotions directed towards “the problem” with positive other-oriented emotions directed towards “the people”. Brett et al. (
1998) refer to this strategy of conveying complementary messages that serve two purposes in a single speech act as mixed-communication. In their study investigating conflict cycles, they show that reciprocating with a contentious strategy while at the same time communicating cooperation is more effective than a uniform tit-for-tat strategy. The underlying rationale is that the contentious reciprocation without cooperative signals helps secure individual gain but at the same time is likely to result in conflict spirals and potentially lead to a breakdown of the negotiation (Brett et al.
1998; Olekalns and Smith
2000; Weingart et al.
1999,
1990). Conversely, reciprocating cooperatively without a contentious component is likely to signal weakness, leaving the negotiator prone to exploitation (Brett et al.
1998). Successful negotiators use both: they engage in contentious reciprocation in order not to be exploited, but at the same time communicate cooperation in order to prevent an escalation of the conflict (Brett et al.
1998). Following a similar rationale, Gibbons et al. (
1992) advise negotiators to use ‘thromises’—a combination of threats and promises—rather than employing them individually. The effectiveness of mixed communication is also supported by controlled experimental evidence. Lindskold and Bennett (
1973) conducted a prisoner’s dilemma study in which participants could choose to send a contentious message (a threat), a cooperative message (a promise), or both. Results show that combining the threat with the promise not only resulted in a more favorable evaluation of the sender but also in more cooperation.
We expect to observe a similar mechanism discriminating between successful and unsuccessful dyads with regard to social precipitants of turning points. A negotiator’s message contains several emotional layers simultaneously; even though they are not necessarily expressed directly and may vary in strength (Griessmair
2017; Griessmair and Koeszegi
2009). Controlled experiments investigating the role of emotions in negotiations employed direct emotional expressions such as “I am angry” or “I am happy” for their manipulations (e.g., Van Kleef et al.
2004a,
b). Research investigating naturalistic negotiation interactions, however, found that emotions are also conveyed implicitly in natural negotiation communication. Conveying emotions does not necessarily require explicit emotional expressions, but negotiators associate distinct emotions with specific utterances and expressions (for a discussion and examples, see, Gibbons et al.
1992; Griessmair and Koeszegi
2009; Schroth et al.
2005).
For instance, consider the following interact before the departure (TP
t). Negotiator A sends the message “Your prior proposal is not what we are looking for and we need to work on an agreement.” (TP
t−2) to which negotiator B (TP
t−1) reacts with one of the following two alternatives: (1) “Your current proposal is unacceptable and as we are a respected company providing high quality products we expect a better offer.” (2) “Your current proposal is unacceptable but I’m confident that if we work together on the problem we will be able to reach an agreement that is beneficial for both of us.”. Although not expressed directly, both messages (1) and (2) convey negative goal-oriented emotions (“your current proposal is unacceptable”) directed towards negotiator A’s offer (“the problem”). They serve as feedback mechanism for counterproductive behavior and signal that behavioral change at the substantive level is required (Cacioppo and Gardner
1999; Fischer and Roseman
2007; Morris and Keltner
2000). Message (1) combines this negative transformational sequence of goal-oriented emotions with a negative transformational sequence of other-oriented emotions (“as we are a respected company providing high quality products we expect a better offer.”). Thereby, negotiator B signals not only dissatisfaction at the substantive level but also conveys feelings of superiority and dominance, disengagement from the personal relationship, and a disruption of interdependence (Markus and Kitayama
1991; Russell and Mehrabian
1977). Conversely, message (2) conveys mixed emotional signals by combining the negative transformational sequence of goal-oriented emotions with a positive transformational sequence of other-oriented emotions (“I’m confident that if we work together on the problem we will be able to reach an agreement that is beneficial for both of us.”). Additionally to signal dissatisfaction on the substantive level, negotiator B simultaneously gives a positive outlook on the relational level by conveying feelings of connection, affiliativeness, and promoting interdependence (Markus and Kitayama
1991; Russell and Mehrabian
1977).
Drawing on Brett et al.’s (
1998) model of conflict spirals, we propose that the successful dyads compensate the potential drawbacks of the negative transformational sequence of goal-oriented emotions by pairing it with a positive transformational sequence of other-oriented emotions. Whereas the former signals dissatisfaction with the current progress of the negotiation, the latter conveys affiliative intent and commitment for the relationship (Griessmair
2017; Kumar
1997; Markus and Kitayama
1991). Next to the substantive dimension, relationship building and establishing an appropriate working climate are cornerstones of successful negotiations (Moore et al.
1999; Poole et al.
1992). By using mixed transformational sequences as social signals, settlement dyads initiate change at the substantive level and at the same time maintain a favorable relationship. Conversely, we expect that dyads concluding the negotiation with a stalemate do not engage in regulatory behavior and use uniform social signals by pairing the negative transformational sequence of goal-oriented emotions with a negative transformational sequence of other-oriented emotions (see first interact in Fig.
2c, d). Rather than engaging in compensatory behavior by employing mixed social signals, they disrupt the process on a substantive as well as relational level, eventually increasing the likelihood of an impasse.
3.1.1 Concluding the Turning Point
According to our model, the social signal conveyed via transformational emotional sequences acts as a precipitant that leads to a departure from zero-sum bargaining after which the parties start making more economically favorable offers. Following Druckman’s (
2001) framework, these turning points have positive or negative consequences depending on whether the changes are incorporated or not. Thus, once a party has introduced a favorable turning point, it has to be confirmed by the counterpart as a new stable pattern and dominant dynamic for the subsequent negotiation interaction (Druckman et al.
2009). In order to do so, we expect that in successful negotiations the counterpart matches the offer from the negotiator that has introduced the turning point (see third interact in Fig.
2a). This reciprocal sequence acts as an immediate reinforcement of the positive departure (Brett et al.
2004) and communicates shared understanding about the newly established direction (Putnam
1990).
The progress the negotiators make towards a mutually beneficial agreement should also be reflected in their emotional experience and expressions. As the departure brings the negotiators closer to their goals, both on a substantial and on a relational level, positive emotions are likely to emerge (Frijda
1986; Lazarus
1991). Druckman et al. (
2009) also show that parties with aligned goals are more likely to create a positive climate. Furthermore, concluding the departure with a positive transformational sequence of goal- and other-oriented emotions conveys satisfaction with the current development (Schroth et al.
2005; Van Kleef et al.
2004a,
2010) and rewards the desired behavior shown at the turning point (Cacioppo and Gardner
1999; Fischer and Roseman
2007). Thus, we propose that in successful negotiations the departure is not only confirmed on a substantive level by matching the turning point offer, but also with social signals (see third interact in Fig.
2c).
Conversely, we expect that impasse dyads fail to reinforce the newly introduced dynamic on both an emotional and a substantive level. Previous research has found that successful negotiation dyads are more likely to use transformational sequences to steer the negotiation away from distributive bargaining (Olekalns and Smith
2000). However, transformational sequences do not only provide a stimulus for change (Brett et al.
1998; Olekalns and Smith
2000), but also signal divergent perspectives and fail to reinforce the preceding behavior (Brett et al.
2004). This kind of counterproductive mismatching occurs “when negotiators strategically utilize the information conveyed by the counterpart’s behavior to better serve their own interests” (Butt et al.
2005: 688). That is, they take advantage of a concessionary counterpart by hard value claiming and react to conciliatory behavior with high demands (Bateman
1980). By taking the counterpart’s coming closer behavior at the turning point as sign of weakness and countering it with mismatching, unsuccessful dyads disrupt the positive change resulting in negative consequences for the negotiation (cf., Druckman
2001) (see third interact in Fig.
2b).
According to our model, the turning point is initiated by a negative transformational sequence consisting of a sharp emotional decline. Successful negotiators conclude the turning point with a positive transformational sequence of goal- and other-oriented emotions. Not only does this provide a reinforcement of the turning point behavior and signals goal-congruence (Cacioppo and Gardner
1999; Morris and Keltner
2000), but also helps to re-direct the negotiation towards positive emotional grounds after the decline that introduced the turning point. Stalemate dyads, on the other hand, are unlikely to experience progress towards an agreement that gives rise to positive emotions (Frijda
1986; Lazarus
1991). Thus, we expect that unsuccessful negotiators reinforce the emotional climate of the turning point with a reciprocal sequence of goal- and other-oriented emotions rather than introducing emotional change via a positive transformational sequence (see third interact in Fig.
2d).
3.4 Emotional Content of Offers
In order to measure the emotional content of the messages negotiators sent along with the offers, we employed multidimensional scaling (MDS).
2 This procedure allows to capture the emotional layers expressed in messages and has already been used extensively in emotion (e.g., Feldman Barrett
2004; Russell
1980; White
2000) as well as negotiation research (e.g., Griessmair and Koeszegi
2009; Pinkley
1990; Pinkley et al.
1995). MDS is an attribute-free approach based on proximity judgements that uncovers the “hidden structure” of behavioral data and complex psychological phenomena by producing a spatial representation of the evaluated stimuli (Pinkley et al.
2005). Following the dimensional approach of emotions (Mauss and Robinson
2009), in MDS each message is located in a coordinate system of
n independent emotional dimensions. That is, rather than assigning a single discrete emotion, each message is characterized by both other- and goal-oriented emotions with the loading of the message on the dimensions reflecting the strength of the conveyed emotion.
To generate the distance matrix serving as input for MDS we employed the subjective clustering method (Gelfand et al.
2001). It requires the independent raters to sort the messages into homogenous groups so that the messages in one group are emotionally similar to each other and dissimilar to the other groups. The frequency with which messages are sorted into the same group reflects their emotional proximity. The so obtained proximity matrix has been analyzed using PROXSCAL.
Following the inductive component of MDS, we used several criteria suggested in literature for determining the number of dimensions and their interpretation. This included verbal characterizations and ratings by the individuals performing the similarity judgments in the subjective clustering method, systematic comparisons of the messages loading high on the opposing pole of the dimensions, and regressing the dimensions on the negotiation outcome (Kruskal and Wish
1977; Perkins and Reynolds
1995; Pinkley et al.
2005). In two separate subsamples with different raters, the logistic regressions point to a 3-dimensional solution (Nagelkerke
R2 = 0.37 and 0.40; Stress-I = 0.19 and 0.25). To assure that the findings are grounded in theory, the interpretation has been performed in light of existing emotional models (cf., Adair and Brett
2005).
3.4.1 Goal-Oriented Emotions (Pleasure–Displeasure)
The first dimension resembles pleasure–displeasure (Feldman Barrett
1998; Feldman
1995). This emotional dimension is related to the progress individuals make towards obtaining their objectives (Frijda
1986; Lazarus
1991). The messages that load high on the emotional positive pole have been characterized by the raters during the similarity judgements as ‘pleasant’, ‘content’, ‘happy’, and ‘excited’. Also the dominant theme of these messages reflects goal-congruency and the rate of progress made by the negotiators (e.g., “(…) we are extremely happy that you have responded very positively to our offer, the only difference of opinion between us, is in terms of price.”; “It seems we have reached a very good solution, now that I realize the importance of return policy to you.”) (Carver and Scheir
1990; Fischer and Van Kleef
2010; Kumar
1997; Van Kleef et al.
2010). Conversely, messages that load high on the negative pole of this dimensions have been described by the raters as ‘annoyed’, ‘frustrated’, ‘cold’, and ‘angry’. The messages express negative emotions as a result of not making progress towards the desired goal and holding the other party responsible for it (e.g., “You have gotten what you wanted on every issue. I can’t see where I have gained something.”; “I’m sorry but your offer is absolutely inacceptable. Please reconsider your offer.”) (Barclay et al.
2005).
3.4.2 Other- Versus Self-Oriented Emotion (Solidarity-Conflict and Submission-Dominance)
The MDS procedure resulted in two dimensions reflecting other- versus self-oriented emotions. Rather than being associated with goal achievement, both dimensions have the relationship with the counterpart as primary referent. They address success or failure in nurturing the relationship, the extent to which interdependence is disrupted or promoted, the willingness to make amends, and perceptions of the nature of the relationship (Griessmair
2017; Kumar
1997; Markus and Kitayama
1991).
The first other- versus self-oriented emotional dimension, solidarity-conflict or affiliativeness (Kitayama and Markus
1990; Markus and Kitayama
1991; White
2000), indicates the extent to which an individual is (dis)engaged from a personal relationship (Markus and Kitayama
1991). The messages loading high on the emotional positive pole of this dimension convey respect, feelings of connection, familiarity, and promote interdependence (Markus and Kitayama
1991) (e.g., “I am sure that our agreement is the best both of us could expect. It confirms my impression that our two companies are complementary and I have no doubt this opportunity is the beginning of a long collaboration”). Raters have characterized these messages as ‘understanding’, ‘optimistic’, ‘insightful’, or ‘compromising’. Conversely, messages on the opposing pole of this dimension have been denoted by the raters as ‘repellent’, ‘hostile’, ‘indifferent’, or ‘vengeful’ and express feelings that disrupt interdependence (e.g., “Our payment and return terms remain the same. My new offer is based in negotiations we are making with other potential suppliers.”).
The second dimension related to other- versus self-oriented emotions reflects the emotional dimension dominance versus submission or potency (Bush
1973; Neufeld
1975; Russell and Mehrabian
1977; White
2000). As opposed to solidarity-conflict, submission-dominance focuses on positioning the self in relation to the counterpart by conveying feelings that relate to the degree of superiority, control, and influence over the counterpart (Russell and Mehrabian
1977). Accordingly, messages loading high on the emotional negative pole of this dimension include a commanding tone, imposing conditions, and communicating one’s own expectations in a coercive manner (e.g., “We are providing you sophisticated and high precision metal parts (…) recognized corporation in this market for the last 50 years. (…) we’ll wait for an initial reasonable offer to start talking for real”). They have been characterized by the raters as ‘dominant’, ‘demanding’, ‘scornful’, and ‘disdainful’. Conversely, messages loading high on the positive pole have been described by the raters as ‘considerate’, ‘obliging’, and ‘agreeable’. They convey emotional repair work and signs of appeasement (Van Kleef et al.
2010) and express feelings of appreciation, willingness to take the other’s perspective, and making amends for a possible transgression (e.g., “We thought about your last offer and we will agree to the return condition because your objection concerning the high quality components is understandable-of course we are responsible for bad manufactured goods.”).