Common Identity Mediation in Intergroup Disputes

Abstract

The present research aims to explore the effectiveness of Common Identity Mediation (cf. Gaertner, Mann, Murell & Dovidio, 1989), a third-party intervention procedure especially conducive to facilitating intergroup dispute resolution. Therefore, Common Identity Mediation was compared to another third-party procedure (i.e. Arbitration-Mediation; cf. McGillicuddy et al. 1987; Ross und Conlon 2000) and an intergroup control condition without third-party assistance. We used a face-to-face, distributive intergroup dispute paradigm, including circumstances that elevate the risk for failed third-party interventions (e.g. high emotional involvement of disputants); with the results indicating that Common Identity Mediation is as effective as Arbitration-Mediation in reducing the risk of partial impasses and improving economic outcomes. In addition, Common Identity Mediation shows a slight advantage in boosting psychological outcomes such as feelings of connectivity, shared identity, judgments of counterparts and satisfaction with the arranged settlement. Findings are discussed in relation to social identity theory, the Common Ingroup Identity Model and the Ingroup Projection Model, and with respect to their implications for practitioners.

Keywords: inter-group conflict, negotiation, dispute resolution procedure, Common Identity

Mediation, Arbitration-Mediation, Common Ingroup Identity Model, Ingroup Projection Model

Common Identity Mediation in Intergroup Disputes:

A Superordinate Identity outperforms Arbitration-Mediation on psychological outcomes.

Our daily life is filled with disputes, whereby one party or person makes a claim (i.e. the demand for compensation to another party who is thought to have caused the relevant damage) that the other party rejects (Felstiner, Abel & Sarat, 1981). Indeed, the current rate dispute between the Independent Flight Attendants Organisation (UFO) and Lufthansa illustrates how expensive and intense such disputes can become. In simplified terms, the UFO has rejected Lufthansa corporate management’s demand to implement extensive economies on wages and leave days. The UFO has consequently gone on strike, repeatedly costing the Lufthansa millions, and not to mention consequential charges for customers being trapped at the airport and the goods traffic industry suffering severe delays. Unremarkably, any form of strike victims may become upset with the UFO, while others show sympathy. In addition, disputes on wages always involve diverging and typically conflicting values and social perspectives, thus boosting the emotional involvement of the disputing parties. Based upon this, and the fact that costs are increasing with every day of the unresolved conflict, there is a strong need for effective dispute resolution procedures. Therefore, disputants often rely on third-party interventions designed to facilitate dispute resolution (Conlon & Sullivan, 1999; Kressel & Pruitt, 1989a), with formal third-party procedures used for example in labor-management disputes (e.g. between Lufthansa and the UFO) where it is essential to reach an agreement and avoid strikes in order to avoid extensive costs. Other contexts involving third-party interventions include inter- and intra- organizational disputes as well as consumer complaint disputes (cf. Kressel & Pruitt, 1989a). Given that all third-party interventions aim to mediate the dispute by helping the disputants reach to an agreement (McGillicuddy, Welton & Pruitt, 1987), they share many similarities regardless of their changing contexts and variable formal procedures. Indeed, this focus on similar strategies and modes of action of third-party interventions is referred to as the "generic view" of conflict resolution (Pruitt & Carnevale, 1993), and constitutes the foundation for research on third-party procedures. Nevertheless, it should be noted that third-party interventions can rely on different formal attributes (e.g. the extent of process- or decision control) and also produce diverging outcomes (e.g. quality of joint outcomes). It is thus necessary for dispute resolution system designers and other practitioners to know which third-party intervention represents the most promising approach for respective dispute contents and contexts. Previous research has primarily focused on the most common third-party interventions: arbitration and mediation (cf. Conlon, Moon & Ng, 2002) and their combination in hybrid procedures, namely Arbitration- Mediation and Mediation-Arbitration (hybrid procedures: cf. Conlon et al., 2002, Loschelder & Trötschel, 2010, McGillicuddy et al., 1987).

However, less attention has been paid to Common Identity Mediation, a method incorporating a theoretical background based on findings of the social identity theory (social identity theory: cf. Tajfel & Turner, 1979), which states that the categorization of people into distinct groups as in dispute resolution or negotiation settings arouses intergroup bias (cf. Brewer, 1979; Tajfel, 1978). Accordingly, this intergroup bias poses a serious threat to third- party intervention success, given that it elevates the disputants' perception of competitiveness and competitive behavior, and can thereby influence the negotiation outcome (cf. Brewer & Brown, 1998; Sherif, 1966; Trötschel, Hüffmeier, Loschelder, Schwarz & Gollwitzer, 2010). Furthermore, members of divergent groups who function as group representatives in negotiations are likely to use more competitive strategies (Pruitt et al., 1978). Therefore, the intended integration of distinct group identities into a shared superordinate identity by Common Identity

Mediation should decrease the prevalence of intergroup bias (Gaertner & Mann, 1989) and lower the risk for conflict resolution in disputes. The present paper aims to further this mediation approach through comparison with the well-established third-party intervention of Arbitration- Mediation.

Previous research evaluating third party procedures has mainly focused on measures such as the quality of joint outcomes and the frequency of negotiation impasses (i.e. non-agreements; cf. Conlon et al., 2002; Gaertner, Höhne, Loschelder, Trötschel & Swaab, 2011; McGillicuddy et al., 1987). Given that such measures pose the most direct and objectively measurable outcomes that can easily be transferred into financial costs or earnings, they will be referred to as economic outcomes, which can be used to describe and compare the basic functionality of third-party interventions in terms of their ability to foster agreements and raise disputants' benefits (cf. McGillicuddy et al., 1987; Trötschel, 2001). However, as dispute resolution procedures aim to influence human behavior, their consequences for disputants' perception, experience and evaluation represent a further important source of information, addressing: disputants' satisfaction with the negotiation outcome (Thompson, 1995); the evaluation of respective counterparts during the dispute (Gaertner & Mann, 1989); the extent of experienced shared identity with the counterpart (cf. Swaab, Postmes, van Beest & Spears, 2007); the will to work together with the counterpart in the future (Gaertner et al., 2011); and the satisfaction with the third-party performance (Gaertner et al., 2011). These measures differ from economic outcomes in that they can hardly be transferred into financial costs or earnings, and will be distinguished accordingly as psychological outcomes. Indeed, psychological outcomes can be used to further evaluate (and possibly differentiate) procedures that produce similar economic outcomes (Gaertner et al., 2011), and can also provide vital information to practitioners through indicating how disputants judge the provided service (i.e. the third-party intervention). Therefore, supplementary to economic outcomes, the present research also aims to shed light on the evaluation of third-party procedures using psychological outcomes.

The present study first provides an overview of classic third-party procedures, including their combination in hybrid third-party interventions. Second, Common Identity Mediation is presented as an alternative third-party intervention. Third, possibilities to evaluate the efficiency of third-party interventions are discussed. Fourth, different threats to mediation success involving practical implications for the conducting of this study are explained. Fifth, the present study is summarized, with the goals and hypotheses are outlined. Sixth, methods and results are presented and explained, and finally, seventh, the results are discussed in light of social theory, the Common Ingroup Identity Model, the Ingroup Projection Model and Elangovan’s propositions for practitioners.

Classic Third-party Interventions

This section focuses on the two predominant third-party intervention procedures, namely arbitration and mediation (Pruitt, 1981), and their combination in hybrid procedures (cf. Pruitt & Carnevale, 1993). Arbitration and mediation differ in terms of the third party's decision control (Thibaut & Walker, 1975). In fact, while mediators cannot determine a binding settlement, an arbitrator can force the negotiating parties to agree on a proposed deal or dictate a binding settlement itself (Kressel & Pruitt, 1989a). In other words, if the disputing parties fail to reach agreement in a straight mediation, the third-party service ends, whereas if disputants in an arbitration procedure fail to reach agreement by the end of a pre-determined deadline, the third party announces a binding settlement (McGillicuddy et al., 1987). Therefore, arbitration guarantees a settlement (Ross & Conlon, 2000) and often motivates disputants to settle voluntarily owing to fearing the loss of decision control inherent to arbitration (Thibaut & Walker, 1975). However, arbitration also encourages arbitrators to use heavy tactics (e.g. threatening to terminate the negotiation or arguing for one specific proposal) during the negotiation phase to reach an agreement, which bears the risk for disputants making premature concessions (McGillicuddy et al., 1987). Furthermore, compliance with the arranged settlement might be in danger because the settlement is forced as opposed to voluntarily crafted (McGillicuddy et al., 1987).

On the other hand, mediation offers the possibility of engaging in integrative agreements (McEwen & Maiman, 1984), and moreover it usually leads to higher voluntary settlement rates and high compliance (Kressel & Pruitt, 1989b). Commonly, mediation is rated a fair third-party procedure (Ross & Conlon, 2000), but it cannot guarantee an agreement.

In recognition of the unique strengths of both third-party intervention procedures, previous research has examined their combination in hybrid procedures (Conlon et al., 2002; Pruitt & Carnevale, 1993). Indeed, both possible combinations, namely Mediation-Arbitration and Arbitration-Mediation, have different effects on the dispute resolution process and its outcomes.

Mediation-Arbitration consists of two phases: the mediation phase and arbitration phase (Conlon et al., 2002). In the mediation phase, both parties are supported in finding a voluntary settlement by the third party, and the negotiation only enters into an arbitration phase if the mediation phase reaches a predetermined deadline without agreement. During the arbitration phase, the third party holds an arbitration hearing and subsequently imposes a binding dispute settlement, in a procedure consistent with the common proposition that dispute resolution should follow a "low-to-high-cost sequence" for the participants (Ury, Brett & Goldberg, 1988, pp. 62- 63). In contrast to Arbitration-Mediation, Mediation-Arbitration is less expensive and time consuming (Ross & Conlon, 2000). If arbitration ensues, Mediation-Arbitration also produces more compliance with the arbitrated settlement, while disputants usually perceive the whole procedure as fairer than they would in Arbitration-Mediation concerning process and outcome distribution (Ross & Conlon, 2000). In summary, Mediation-Arbitration offers a slight advantage over Arbitration-Mediation concerning effects that either appear some time after the dispute situation or those lasting for a longer period of time during the third-party intervention (longterm effects: cf. Kressel & Pruitt, 1989b; Ross & Conlon, 2000).

Arbitration-Mediation includes three phases: an arbitration phase, a mediation phase and a ruling phase (Conlon et al., 2002). During the arbitration phase, the third party meets each disputing party in private to familiarize itself with the respective interests and claims. At the end of this phase, the arbitrator decides over a binding settlement, which is not revealed to the disputants, but is rather written down and sealed in an envelope. In the mediation phase, both parties have time to negotiate while the arbitrator acts as a mediator facilitating dispute resolution. The parties will only enter the ruling phase if they fail to reach a voluntary agreement during the mediation phase. During this final phase, the arbitrator opens the sealed envelope and pronounces the previously-determined settlement, which thus becomes binding for the disputants. The advantages of Arbitration-Mediation primarily lie in the ability to boost the disputant's motivation to reach agreement during the mediation phase (Ross & Conlon, 2000), which occurs due to the disputants’ fear of losing their decision control at the end of the mediation phase (theory of procedure: cf. Thibaut & Walker, 1975). Consequently, they try to settle the dispute themselves and thus use more cooperative behaviors during the mediation phase (cf. Curry & Pecorino, 1993; Pruitt, 1981), including increased concession making and enhanced revealing of confidential information, which is associated with a higher probability of integrative settlements (Pruitt, 1981; Thompson, 1991). In addition, disputants show a greater approval of third-party settlement suggestions, which derives from the salient threat of losing during the arbitration phase (loss aversion: cf. DeDreu & van Lange, 1995; Trötschel, 2001). Moreover, disputants also lower their outcome expectations (Ross & Conlon, 2000). Overall, these combined effects prompt Arbitration-Mediation to produce a higher rate of voluntary settlements than Mediation-Arbitration (Ross & Conlon, 2000), and also higher joint outcomes (Conlon et al., 2002). In summary, it can be stated that Arbitration-Mediation offers a slight advantage over Mediation-Arbitration concerning effects that appear temporally close to the dispute situation (short-term effects: cf. Ross & Conlon, 2000).

Similar to the unique strengths of each procedure, both hybrid third-party interventions differ in terms of their advantages and disadvantages. Accordingly, dispute resolution system designers and other practitioners are strongly advised to follow a contingency approach to procedure selection (Ross & Conlon, 2000). Based on the theory of procedure (Thibaut & Walker, 1975), Elangovan (1995, 1999) has proposed six questions that can be used to ascertain the adequate intervention form: (1) How important is the dispute?; (2) How important is it to resolve the conflict quickly?; (3) Does the dispute concern the application of existing rules, or does it concern the alteration of existing rules?; (4) Do the parties expect frequent work- related interaction?; (5)Would the parties be committed to a settlement imposed by a third party?; (6) What is the probability that the parties would negotiate a settlement compatible with the larger organization or system? Ross and Conlon (2000) answer these questions for the decision between Arbitration-Mediation and Mediation-Arbitration as follows: (1) Third parties should retain decision control for important disputes (Elangovan, 1995). While both procedures fulfill this demand, Arbitration-Mediation offers a slight advantage concerning the salience of the upcoming loss of decision control and the acceptance of third-party suggestions during the negotiation phase. (2) Mediation-Arbitration produces faster settlements and hence should be used if brevity is important. (3) Both procedures can alter existing or create new rules. (4) Mediation-Arbitration produces better long-term outcomes and should thus be preferred over Arbitration-Mediation if the disputants will have to interact in the future. (5) Mediation- Arbitration produces stronger compliance with the settlement. (6) Mediation-Arbitration grants more process control to the disputants and hence should be used when the parties can find a solution in accordance with the larger system's goal (Elangovan, 1995). However, such a goal is more abstract than a specific dispute resolution outcome. For example, in the present study, the third party has to respect the interests of the over-arching company before each party's individual outcome, and a larger system goal thus often sets an important frame for the conductance of third-party interventions.

Following Elangovan's propositions, one could initially argue that the dispute setting in the present study might be handled using Mediation-Arbitration, with reasons lying in the fictive need for the participants to quickly settle the dispute (Elangovan Question 3) and continue their cooperation in the future (Elangovan Question 4). However, the importance of the conflicting issues (Elangovan Question 1) and high tension between the disputants, causing them to forget the interests of their common enterprise (Elangovan Question 6), would advise the usage of Arbitration-Mediation in granting the third party more process and outcome control. In addition, Arbitration-Mediation produces a higher voluntary settlement rate, which is one of the most important criteria for evaluating dispute resolution procedures (Sheppard, 1984, Thomas, Jamieson & Moore, 1978). Therefore, we decided to use Arbitration-Mediation as a comparison for the effectiveness of Common Identity Mediation.

Common Identity Mediation

Common Identity Mediation is based on the Common Ingroup Identity Model (Gaertner, Mann, Murell & Dovidio, 1989), and aims to change disputants’ mental representation from perceived membership of different groups to perceived membership of the same group. In fact, these divergent cognitive representations represent different social identities (cf. Turner, Hogg, Oakes, Reicher S. D & Wetherell, 1987) that can be organized hierarchically, with a higher-level identity including multiple lower-level identities (circles of inclusion: cf. Allport, 1954). This mechanism causes Common Identity Mediation to change the disputants’ level of social identity, in order that lower-level, subgroup identities merge in one higher-level, superordinate identity (Gaertner, Dovidio, Anastasio, Bachmann & Rust, 1993). Consequently, former outgroup members are included within the ingroup, so that individuals develop a common ingroup identity on a superordinate level (Gaertner, Rust, Dovidio, Bachmann & Anastasio, 1996). This can be achieved in two ways: first, the former (sub-)group identity can be eliminated (Hornsey und Hogg 2000); or second, the superordinate identity is established whilst maintaining the distinct group identities (Gaertner & Dovidio, 2000). Existing research has shown that group members may be reluctant to abandon their (sub-)group identity, especially when they feel attached to it (Wenzel et al. 2007). Therefore, group members resist the intervention procedures aiming at reducing the distinct group identities, and only try harder to maintain their ingroup-outgroup categorization (Hornsey und Hogg 2000). In addition, if the ingroup-outgroup distinction is underplayed in concrete situation, positive experiences in the contact with members of the other group will hardly be generalized (Wenzel et al. 2007). Accordingly, Hewstone and Brown (1986) argued that contact partners should explicitly act as members of their own (sub-)group, with group members able to derive a positive social identity from their ingroup’s positive attributes while accrediting positive attributes of the outgroup (Hewstone und Brown 1986). In addition, research has shown that recategorization does not necessarily force disputants to forsake their former subgroup identity. Instead, they maintain a dual identity, with both the former subgroup identity and the newly-acquired superordinate identity simultaneously salient (shared identity: cf. Gaertner & Dovidio, 2000; Gaertner et al., 1993). Allport (1954) supports this notion by suggesting that "Concentric loyalties need not clash. To be devoted to a large circle does not imply the destruction of one's attachment to a smaller circle." (p. 44).

We argue that the advantages of Common Identity Mediation lie in its good structural fit to the classic dispute resolution composition, given that conflicts typically emerge between two (or more) parties or their representatives and revolve around their contrasting interests (Gaertner et al., 2011). In addition, Common Identity Mediation directly addresses the social identity of disputants, which plays an important role concerning the extent of competition and trust in intergroup disputes (cf. Trötschel, Hüffmeier & Loschelder, 2010). Furthermore, Common Identity Mediation has been shown to boost psychological outcomes (e.g. intergroup relations between group representatives; cf. Gaertner et al., 2011).

Finally, Common Identity Mediation doesn't pose the threat of a forced settlement via arbitration. This could naturally also be disadvantageous, as Common Identity Mediation does not guarantee a settlement at the end of the dispute resolution process. Overall, Common Identity Mediation should represent an effective dispute resolution procedure with respect to economic outcomes, also showing potential advantages over Arbitration-Mediation on psychological outcome measures.

Evaluation of Third-party Performance

Voluntary settlements (i.e. rate of total impasses), joint benefits or costs and settlement speed represent the most important outcomes for dispute resolution procedures (cf. Sheppard, 1984, Thomas et al., 1978). While the frequency of voluntary settlements is negatively reflected in the rate of total impasses, voluntariness in evaluation studies is assured via the experimental design (e.g. in this study, arbitrated settlements are coded as impasses because they were forced). Concerning impasses, previous research has mainly focused on how negotiators fail to reach an integrative agreement (for a review, see Bazerman, Moore & Gillespie, 1999; DeDreu, Weingart & Kwon, 2000), while few empirical studies have examined how disputants fail to reach agreement and leave issues unresolved on the bargaining table (Gaertner et al., 2011). Emphasizing the importance of impasses for negotiation research, Tripp and Sondak (1992) state that "(...) impasse rates across conditions may be one of the most significant findings" and that "(...) what leads to impasse is an important question in its own right" (p.279). To measure impasses, research differentiates between total and partial impasses, with the former occurring when the complete dispute resolution has failed and there is not one agreement on a single issue (e.g. Ben-Yoav & Pruitt, 1984). On the other hand, a partial impasse occurs if the dispute resolution was not successful for all conflicting issues yet managed to produce a settlement for part of the issues (e.g. Neal & Bazerman, 1985; Trötschel & Gollwitzer, 2007). The measurement of partial impasses allows for a differentiated perspective on dispute resolutions, whereby it is possible to measure and analyze the course of partial impasses over time, which can provide additional insight within the dispute resolution process.

The differentiation between total and partial impasses leads to another vital distinction for dispute resolution settings. For example, the classic negotiation paradigm used by Pruitt and Lewis (1975) was designed to force its participants to agree on all three conflicting issues (i.e. iron, sulfur, coal) or break up the negotiation without result. Accordingly, it was not possible to reach an agreement on one issue (e.g. coal) while leaving the other two issues unresolved (e.g. sulfur and iron). In this way, issues are interlinked. The aforementioned dispute between Lufthansa and the UFO includes such interlinked issues, with wages and leave days equally influencing the working conditions of flight attendants. Therefore, both issues cannot be settled independently as they address the same underlying interest for the disputing parties. By contrast, issues can also be non-linked, whereby disputants can at least partially settle their conflict while leaving some conflicting issues unresolved. For example, a car manufacturer negotiates the price for the deliverance of wheels and windows with his component supplier, where they could settle on a good price for the wheels yet find no solution for the windows, which the manufacturer thus decides to buy from another supplier. Note that issues need to be non-linked to produce partial impasses, and that dispute settings with interlinked issues can only produce complete solutions or total impasses.

Joint outcomes are commonly assessed by adding both parties' individual benefits or losses at the end of a negotiation or dispute, and hence the measurement of joint outcomes cannot be used to evaluate the value distribution between disputants. However, it provides important information concerning the general quality of an agreement. First, in the case of a distributive dispute without the possibility of accepting partial impasses, the joint outcome will always be zero as one party's gain is the other party's loss (cf. Trötschel, 2001; Walton & McKersie, 1965). Second, if partial impasses are allowed and all issues are of same value, the joint outcome indicates how many issues were left on the bargaining table at the end of negotiation. In this case, the analysis of partial impasses can provide additional insight into the negotiation process yet generates no additional information concerning the negotiation outcome, as joint outcomes and partial impasses are 100% correlated. Third, if the conflicting issues are of different value (e.g. the ceiling costs 49.000€ whereas the sauna only costs 12.100€), the same joint outcome could be produced by different constellations of partial impasses, or vice versa. In this case, partial impasses and joint outcomes are no longer correlated, and it is important to analyze both variables as they provide complementary information concerning the achieved settlement. Fourth, if the disputants have different preferences for the conflicting issues (integrative potential: cf. Walton & McKersie, 1965), the joint outcome can indicate the extent to which both parties were able to establish a win-win solution, as they allow for higher joint outcomes than compromises (win-win solution: cf. Raiffa, 1982; Trötschel, 2001).

The tempo in which a settlement is agreed upon is usually measured by the elapsed time (cf. Conlon et al., 2002) or number of negotiation rounds over a predefined time until the disputants agree on a binding settlement (cf. Trötschel, 2001). Indeed, the speed of settlement represents one important criterion in choosing between different dispute resolution procedures, particularly if a prolonged dispute produces great financial or psychological costs for disputants and third parties (Conlon et al., 2002).

[...]