2010 - thompson et.al - negotiation

ANRV398-PS61-19 ARI 28 October 2009 11:59

NegotiationLeigh L. Thompson, Jiunwen Wang,and Brian C. GuniaKellogg School of Management, Northwestern University, Evanston, Illinois 60208;email: [email protected]

Annu. Rev. Psychol. 2010. 61:491515

First published online as a Review in Advance onOctober 7, 2009

The Annual Review of Psychology is online atpsych.annualreviews.org

This articles doi:10.1146/annurev.psych.093008.100458

Copyright c 2010 by Annual Reviews.All rights reserved

0066-4308/10/0110-0491$20.00

Key Words

mixed-motive interaction, decision making, bargaining, valuecreation, value claiming

AbstractNegotiation occurs whenever people cannot achieve their own goalswithout the cooperation of others. Our review highlights recent em-pirical research that investigates this ubiquitous social activity. We se-lectively review descriptive research emerging from social psychologyand organizational behavior. This research examines negotiation behav-ior and outcomes at ve levels of analysis: intrapersonal, interpersonal,group, organizational, and virtual. At each level, we review research onnegotiation processes and outcomes, and we discuss the implications ofvarious processes and outcomes for the two functions of negotiation:value creation (integrative negotiation) and value claiming (distributivenegotiation).

491

Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


Contents

INTRODUCTION . . . . . . . . . . . . . . . . . . 492The Intrapersonal, Interpersonal,

Group, Organizational,and Virtual Systems . . . . . . . . . . . . . 493

Integrative and DistributiveNegotiation . . . . . . . . . . . . . . . . . . . . 493

INTRAPERSONAL LEVEL . . . . . . . . . 494Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 494Gender . . . . . . . . . . . . . . . . . . . . . . . . . . . . 496Affect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497

INTERPERSONAL LEVEL . . . . . . . . . 499Economic and Social Psychological

Foundations . . . . . . . . . . . . . . . . . . . . 499Interpersonal Effects of Emotions

in Negotiation . . . . . . . . . . . . . . . . . . 499Interpersonal Improvisation

in Negotiation . . . . . . . . . . . . . . . . . . 500Subjective Value in Negotiation. . . . . 500Trust and Tactics . . . . . . . . . . . . . . . . . . . 501Relationships and Negotiations . . . . . 502

GROUP LEVEL . . . . . . . . . . . . . . . . . . . . . 502Social and Group Identity . . . . . . . . . . 502Relational and Collective Identity . . . 503Culture . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504Groups and the Discontinuity

Effect . . . . . . . . . . . . . . . . . . . . . . . . . . 504ORGANIZATIONAL LEVEL. . . . . . . . 505

Choice of Negotiation Partner. . . . . . 505Reputation and Negotiation

Through Time . . . . . . . . . . . . . . . . . 506VIRTUAL LEVEL . . . . . . . . . . . . . . . . . . . 507

Moderators and Mediators . . . . . . . . . 507CONCLUSION . . . . . . . . . . . . . . . . . . . . . 508

INTRODUCTION

Anytime people cannot achieve their goalswithout the cooperation of others, they are ne-gotiating. By this denition, negotiation is aubiquitous social activity. Research on negoti-ation has been inuenced by a wide variety ofelds, including mathematics, management, or-ganizational behavior, social psychology, cog-nitive psychology, economics, communicationstudies, sociology, and political science. The

products of this multidisciplinary approachhave been intense theoretical development andan impressive body of empirical ndings.

Negotiation research has undergone severalphases, characterized by different paradigmsof thought. For example, during the 1980s,negotiation research was heavily inuenced bygame theory and behavioral decision theory.During the 1990s, negotiation research wasstrongly inuenced by social psychology. At theturn of the millennium, negotiation researchhas become decidedly cognitive in avor. Eachgeneration of research has provided scholarswith a new vantage point from which to exam-ine the complex dance of negotiation.

One of the most important theoretical dis-tinctions in negotiation scholarship is the onedening normative and descriptive research(Raiffa 1982). Normative research, largely de-rived from game theory, economics, and math-ematics, proposes optimal models of the ne-gotiation problem and prescribes what peoplewould do if they were wise and all-knowing(cf. Luce & Raiffa 1957, Nash 1951). In thisreview, we focus on descriptive research, whichrecognizes that negotiators do not always be-have in a game-theoretic, optimal fashion. Theway negotiators actually behave usually departssignicantly from normative, economic mod-els (but not necessarily from behavioral eco-nomic models; Camerer 2003). For example,whereas normative models predict that peo-ple will/should almost always defect in a pris-oners dilemma or social dilemma, actual de-fection rates are dramatically lower than 100%(Camerer 2003, Komorita & Parks 1995).Moreover, normative models of negotiationdictate that parties should reach Pareto-optimalsettlements, dened as agreements that cannotbe improved upon without hurting one or bothof the parties outcomes. However, very few ne-gotiators reach Pareto-optimal outcomes on aregular basis (Thompson 2009, Thompson &Hastie 1990).

Our focus is limited to descriptive researchinuenced by social psychology and its closecousin, organizational behaviorboth of whichhave strongly inuenced negotiation research

492 Thompson Wang Gunia

Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


since 1980. We focus on empirical studies thatexamine the individual negotiator within one ormore of ve systemsintrapersonal, interper-sonal, group, organizational, and virtual. Weuse these systems as a guide for organizingour review. Within each system, we focus ontwo overarching themes: integrative negotia-tion and distributive negotiation, described fur-ther below.

The Intrapersonal, Interpersonal,Group, Organizational,and Virtual Systems

We use the term intrapersonal system tosignify the ways that negotiation behavior andoutcomes depend upon the perceptions andinner experiences of the negotiator. For ex-ample, the intrapersonal system might includeresearch on how an individuals sense of powerinuences his or her negotiation behavior,satisfaction, and outcomes. The interpersonalsystem refers to the ways that negotiators be-havior and outcomes depend upon the presenceof the other party or partiesnegotiations inthe context of others, and the dyadic aspects ofnegotiation behavior. Investigations of how anegotiators mood inuences the other partysbehavior and the ultimate negotiation outcomeexemplify this system. The group system en-compasses social dynamics that extend beyonda single dyadfor example, group identity,cultural identity, coalitions, and conformity.The organizational system represents a higherlevel of analysis and examines the negotiatoras embedded in a larger network or market-place. For example, some studies at this levelinvestigate how negotiators choose optimalcounterparties in a marketplace of negotiators.Finally, the virtual system focuses on hownegotiators medium of interactionsuchas face-to-face, phone, or emailaffects thenature and quality of negotiation processes andoutcomes. Several studies have investigatedwhether negotiators are more likely to discovermutual value when negotiating face-to-face orvia computer (cf. Morris et al. 2002, Naquin &Paulson, 2003, Purdy et al. 2000).

Negotiation:an interpersonaldecision-makingprocess necessarywhenever we cannotachieve our objectivessingle-handedly.Negotiations includenot only the one-on-one business meetings,but also multiparty,multicompany, andmultimillion-dollardeals. People negotiatein their personal lives(e.g., with theirspouses, children,schoolteachers,neighbors) as well as intheir business lives

Pareto-optimal:Pareto optimality, orPareto efciency, is animportant concept ineconomics with broadapplications in gametheory, engineering,and the social sciences.The term is namedafter Vilfredo Pareto,an Italian economistwho used the conceptin his studies ofeconomic efciencyand incomedistribution.Informally, Pareto-optimal situations arethose in which anychange to make anyperson better offwould make someoneworse off

Integrative:negotiations areintegrative when allcreative opportunitiesare leveraged and noresources are left onthe table

Integrative and DistributiveNegotiation

Whereas the independent variables or causalfactors underlying negotiation have been highlyeclectic and strongly inuenced by the con-temporary theoretical milieu, the dependentvariables under investigation have remainedconsistent across several decades. The mainreason for this consistency is the inuence ofeconomics on negotiation research. Within ne-gotiation research, the two dependent variablesthat appear in virtually every published studyof negotiation are negotiation processes andoutcomes.

Negotiation processes include negotiatorsbehaviors, cognitions, emotions, and motiva-tions. For example, much social psychologicalresearch has focused on negotiator satisfactionand the perceived relationship between the par-ties (see Curhan et al. 2006 for a review). Ne-gotiation outcomes include the integrative anddistributive features of the agreement. By in-tegrative, we mean the extent to which the ne-gotiated outcome satises the interests of bothparties in a way that implies the outcome can-not be improved upon without hurting one ormore of the parties involved (i.e., Pareto op-timality) (Pareto 1935). A classic example ofPareto optimality is the story of the two sis-ters who quarreled bitterly over a single orange(Fisher & Ury 1981). The sisters resolved thedispute by cutting the orange in half, such thateach sister received exactly 50%. Later, the sis-ters discovered that one only needed the juicewhereas the other only needed the rind; un-fortunately they had failed to realize this dur-ing the negotiation itself. Cutting the orangein half was not an integrative outcome, becauseanother feasible solution would have simulta-neously improved both sisters outcomesonesister could have received all of the juice and theother all of the rind. This solution would havefully maximized both parties interests. The factthat another feasible solution would have beenbetter for both parties suggests that the actualoutcome was suboptimal or Pareto inefcient,as opposed to integrative.

www.annualreviews.org Negotiation 493

Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


Distributive:a negotiation isdistributive whennegotiators are mainlyconcerned about theirown economicoutcomes and not thejoint outcomes of allnegotiating parties

Best alternativeto a negotiatedagreement(BATNA):determines the point atwhich a negotiator isprepared to walk awayfrom the negotiationtable. In practice, itmeans that thenegotiators should bewilling to accept anyset of terms superior totheir BATNA and toreject outcomes thatare worse than theirBATNA

The distributive aspect of negotiation refersto how negotiators divide or apportion scarceresources among themselves. For example, inthe classic ultimatum game (Guth et al. 1982,Ochs & Roth 1989), one person (player 1)receives a xed amount of money (say $100) todivide with another person. Player 1 proposes asplit of the $100; if player 2 agrees, the proposedsplit takes effect. If player 2 rejects the proposal,each party gets $0. The split that Player 1 pro-poses can be perceived to be fair or acceptableto player 2, leading player 2 to accept the offer.In this case, the distributive aspect of the ne-gotiation is the proportion of the original $100that each negotiator receives.

Recently, the initial focus on the economicoutcomes of negotiation has widened to in-clude investigations of subjective outcomes.Whereas rational behavior in negotiation isusually equated with the maximization of eco-nomic gain, joint or individual, some have ar-gued that it is equally appropriate to con-sider social-psychological outcomes, such as thequality of the relationship, the degree of trustbetween parties, each negotiators satisfaction,and each persons willingness to negotiate withthe other in the future. In an attempt to measuresubjective concerns, Curhan and his colleaguessurveyed people on what they value in negotia-tion (Curhan et al. 2006). Four distinct consid-erations emerged: feelings about instrumentaloutcomes (i.e., how much money they made),feelings about themselves (e.g., how competentthey were in the negotiation), feelings aboutthe process (e.g., whether the conversation wasconstructive) and feelings about the relation-ship (i.e., whether the negotiation preserved orstrengthened it).

INTRAPERSONAL LEVEL

Negotiation research at the intrapersonal levelof analysis clearly recognizes the multipartynature of negotiation, but it emphasizes howthe inner experience of the negotiator im-pacts negotiation processes and outcomes, andvice-versa. We focus on three interrelatedintrapersonal constructs that have received

signicant research attention in recent yearspower, gender, and affect. Many studies ofpower, gender, and affect in negotiations fol-low from research stimulated by the work ofSteele (Steele & Aronson 1995), Banaji (Blair& Banaji 1996), Greenwald (Greenwald et al.1996), Bargh (Bargh & Pietromonaco 1982),and others on the behavioral effects of uncon-scious priming. This research examines howsubtle, below-threshold activation of conceptsinuences above-threshold behaviors. In ne-gotiations, above-threshold behaviors substan-tially impact negotiation processes and out-comes, which may unconsciously activate othercognitions and behaviors.

Power

Power refers to an individuals relative abilityto alter other peoples outcomes (Keltner et al.2003). Several studies examine psychologicalpower as a state, operationalized through prim-ing, but others examine power as a trait or in-dividual difference. Although negotiators mayhave several sources of structural power (French& Raven 1959), the most commonly investi-gated source of power is the negotiators bestalternative to a negotiated agreement (BATNA;Fisher & Ury 1981).

A negotiators BATNA has become the pri-mary indicator of a negotiators relative powerin negotiation. The BATNA concept was for-mally introduced by Fisher and Ury in 1981;however, the concept actually traces back tothe social exchange theory of Thibaut & Kelley(1959). Exchange theory cites rewards (bor-rowed from psychology) and resources (bor-rowed from economics) as the foundation ofinterpersonal exchanges. Rewards refer to thebenets a person enjoys from participating in arelationship (Thibaut & Kelley 1959), whereasresources are any commodities, material orsymbolic, that can be transmitted through in-terpersonal behavior (Foa & Foa 1975) andgive one person the capacity to reward another(Emerson 1976). Satisfaction with an exchangerelationship is derived in part from the evalua-tion of the outcomes available in a relationship.


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


Outcomes are equal to the rewards obtainedfrom a relationship minus the costs incurred.

People in social exchanges compare the out-comes of the current exchange with the out-comes they could achieve in an alternativeexchangethese alternative outcomes are op-erationalized as the comparison level of alter-natives, or CLalt. When the CLalt exceedsthe outcomes available in a current relation-ship, the person is more likely to leave therelationship. The concept of CLalt is paral-lel to BATNA. When ones BATNA is bet-ter than an agreement one can reach with aparticular negotiation counterpart, one shouldchoose to not agree and exercise the BATNAinstead.

Negotiators BATNAs are strongly relatedto their reservation point (RP). RPs are thequantication of a negotiators BATNA (Raiffa1982). According to Raiffa (1982), a negotia-tors RP is the point at which a negotiator isindifferent between reaching a deal with partyA or walking away from the table and exercisinghis/her BATNA. For a seller, prices exceedingreservation points are acceptable; for a buyer,prices less than reservation points are accept-able. RPs are generally operationalized as thevalue attached to a negotiators BATNA, plusor minus the value of any idiosyncratic prefer-ences they attach to reaching agreement versusexercising the BATNA.

Just as BATNA traces to Thibaut & Kelleys(1959) earlier concept of CLalt, reservationprice traces to Walton & McKersies (1965)concept of resistance point, described in theirbook A Behavioral Theory of Labor Relations.Resistance point is a negotiators subjectivelydetermined bottom linethe point at whichnegotiators are indifferent between reachingagreement and walking away, in the midst of thenegotiation. Walton & McKersie (1965) postu-lated that negotiators who had more attractiveresistance points were in a more powerful posi-tion because they could simply offer the otherparty just enough to meet their resistance pointand claim the rest (the surplus) for themselves.Although the concept of reservation price haslargely displaced the concept of resistance point

Reservation point:determined not bywhat the negotiatorwishes and hopes for,but rather by what herBATNA represents. Areservation point is aquantication of anegotiators BATNAwith respect to otheralternatives

Bargaining zone:the range betweennegotiatorsreservation points.Between a buyer and aseller, the bargainingzone will be betweenthe highest price abuyer is willing to payand the lowest price aseller is willing to sellfor

in recent academic research, resistance pointsprovided an important theoretical step towardspecifying the concept of bargaining zone. Bar-gaining zone is basically the overlap betweentwo negotiators RPsthe buyers RP minusthe sellers RP. If this number is positive, a zoneof possible agreement (ZOPA) is said to exist; ifit is negative, no ZOPA exists.

Research studying the effects of power havedocumented that there is a strong, causal rela-tionship between the attractiveness of a nego-tiators BATNA and the negotiators ability toclaim resources in a given negotiation (Galinsky& Mussweiler 2001, Magee et al. 2007,Mussweiler & Strack 1999). Negotiators withattractive BATNAs are considered powerful;these negotiators are decidedly more assertivein negotiations. For example, powerful peoplemove rst, both by initiating negotiations andby making the rst offer (Magee et al. 2007).When power is primed (by instructing peopleto write about a time when they felt powerfulor to perform a word-completion task involv-ing words about power), these individuals of-ten make the rst offer in negotiations. If theconcept of BATNA is a measure of structuralpower, then chronic tendencies to dominateothers in social relationships reect personalpower. Both structural and personal power canimprove negotiators outcomes by leading themto make the rst offer (Galinsky & Mussweiler2001, Magee et al. 2007, Mussweiler & Strack1999).

Although having power may increase a ne-gotiators propensity to make a rst offer, thismay depend on the nature of the negotiation.Specically, it is reasonable to assume that ifboth negotiators have attractive BATNAs, theirmotivation to reach mutual agreement is not ashigh as that of two negotiators with very pooralternatives. Thus, the effects of ones power ina negotiation may depend on the size of the bar-gaining zone. Given that BATNAs establish theminimum level of benets one would receive,irrespective of what occurs in the negotiation,their inuence quickly diminishes once bene-ts equivalent to the BATNA value have beenattained.


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


In one study, strong BATNAs improved ne-gotiators outcomes more when the bargain-ing zone was small rather than large (Kim &Fragale 2005). When the bargaining zone waslarge, power tended to derive more from a ne-gotiators contribution to the negotiation. Inthis case, contribution refers to the benetsthat a negotiator contributes beyond the valueof the counterpartys BATNA. For instance, ifthe counterparty is selling a house and has aBATNA (e.g., another buyer offering $200Kfor the house) and the negotiator offers $210Kfor the house, the difference, or $10K, is thecontribution.

Once an offer equaled the value of onesBATNA in Kim and Fragales research, out-comes depended more on the extent to whichthe counterparty could contribute value be-yond the BATNA. Contributions thus exertedan important inuence on negotiation out-comes, especially as the potential agreementbecame more valuable (relative to negotiatorsBATNAs).

Gender

Power is manifested and expressed by nego-tiators in many ways. For example, power candepend upon structural factors (e.g., BATNA)or on personal characteristics. A negotiatorsstructural power can change when environ-mental conditions change, but personal poweris, for the most part, xed. For example, anegotiator who is selling her house and has anattractive offer from a very motivated buyer hasa lot of structural power; however, if the buyersuddenly withdraws the offer on the house(perhaps due to a failed home inspection), thenegotiators power plummets. Conversely, anegotiator who is a vice president of a majorcompany and has a lot of personal charm alsoholds high power, which is more resilient totemporary uctuations of the market (exceptin the case of losing her job!). One importantsource of personal power is gender.

To exert inuence in a negotiation, gen-der must be activated or made salient (Kray &Thompson 2005). In a series of investigations

modeled after Steele & Aronson (1995), Krayet al. (2001) did just this. Specically, they inves-tigated whether the mere activation of gender(and its accompanying stereotypes) impacts ne-gotiation performance. The prevailing stereo-type is that women are less assertive and agenticthan men. Because many people see negotiationas a situation requiring assertive and agentic be-havior, stereotypically female traits may seeminconsistent with negotiation once the connec-tion is made salient. For these reasons, the meremention of negotiation might create an internalconict within women: On one hand, they maybelieve that performing well requires them toengage in counterstereotypical behaviors. Onthe other hand, they may believe that othersexpect them to behave in an accommodating,nonassertive fashion.

Kray et al. (2001) hypothesized that themention of gender might operate much likestereotype threat (Steele & Aronson 1995).Gender salience might thus operate like a low-power state, preventing women from acting as-sertively. In their study, women did, in fact,get worse outcomes than did men in mixed-gender negotiations, when an implicit genderstereotype was subtly activated. However, it wasreasoned that explicit activation of the genderstereotype may allow women to counteract it.As predicted, explicitly activated gender stereo-types led to a stereotype-reactance effect, inwhich women actually outperformed men byclaiming more resources (presumably in an at-tempt to defy the stereotype). Women effec-tively said, Well, unassertive behavior and ac-commodation may be the cultural stereotype ofwomen, but it is surely not me!

In another series of studies, Kray and col-leagues (2004) reasoned that negotiation, likeother social activities, can be construed as ei-ther a masculine or feminine activity. The mas-culine construal of negotiation involves agencyand assertiveness. It is also possible to construesuccessful negotiation as understanding humanbehavior, perceiving nonverbal cues, and build-ing trust. Arguably, these skills are more consis-tent with the classic female stereotype. Indeed,women outperformed men when traditionally


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


feminine traits were linked with negotiationsuccess, and each gender outperformed theother when the other gender was linked withnegotiation ineffectiveness (Kray et al. 2001).

The implications of stereotype activationmay also depend on whether negotiators havehigh or low power (Kray et al. 2004). Speci-cally, activation of an explicit male stereotypeled to negotiated outcomes that favored thehigh-power negotiator, whereas activation of anexplicit female stereotype led to more integra-tive, win-win outcomes that were benecial forboth parties, much like the sisters who discov-ered the juice-and-rind tradeoff.

Other studies (Small et al. 2007) examinegender differences in the willingness to initiatenegotiation (Babcock et al. 2006, Bowles et al.2007; but see Gerhart & Rynes 1991), tracingthese differences to power differences. Becausewomen traditionally have less power than domen in U.S. society (Eagly & Wood 1982), theyinitiate negotiations less often; however, thisdifference is attributable to the fact that situa-tions framed as negotiation conict with po-liteness norms that prevail in low-power groups(Babcock et al. 2006). Consistent with this rea-soning and the links between gender and power,framing negotiations as opportunities to askeliminated gender differences in negotiationinitiation, as did priming psychological power(Kray et al. 2001). Along similar lines, Bowleset al. (2007) traced differences in the initiationof negotiation behavior to observers reactions.Both male and female observers penalized fe-male job candidates for initiating negotiations.Consistent with Small et al.s (2007) politenessargument, participants rated women who ini-tiated negotiations as less nice and more de-manding. Moreover, women were less likelythan were men to initiate negotiations with amale (but not a female) evaluator.

Another stream of gender research exam-ines what happens when women do, in fact,initiate negotiation. Although gender differ-ences in actual negotiation behavior have re-ceived exhaustive research attention (e.g., Deal2000, Gerhart & Rynes 1991, Major et al.1984, Stevens et al. 1993, Watson 1994), recent

meta-analyses (Stuhlmacher & Walters 1999,Walters et al. 1998) characterize such differ-ences as modest and context dependent. Ac-cording to these meta-analyses, women nego-tiate slightly more cooperatively than do men,but situational factors such as relative power ofthe negotiator, integrative potential of the task,and mode of communication often override thiseffect. In addition, other individual differences(i.e., social motives) explain cooperation in ne-gotiation more readily than gender does. Forexample, negotiators with a prosocial motivebehave more cooperatively (and achieve betteroutcomes) than do those with an egoistic motive(De Dreu et al. 2000). Gender differences seemto explain relatively little variance by compari-son, and it is possible that the variance they doexplain reects underlying gender differencesin social motives.

There are behavioral implications ofgender-dependent power. For example, doesmaintaining steady eye contact have differ-ent power implications for male and femalenegotiators (Swaab & Swaab 2009)? Whennegotiators made eye contact (and when visualaccess was possible), agreement quality wasmaximized for women but minimized for men(Swaab & Swaab 2009). Apparently, womenand men had different affective experiencesduring negotiation. When men made eyecontact, perceived power differences wereexacerbated, creating a sense of discomfortthat undermined agreement quality.

Affect

Forgass (1995) affect infusion model consid-ers the impact of mood on cognitive process-ing, identifying two overarching conditions un-der which mood is likely to affect informationprocessing. The rst condition is when situa-tions require cognition about difcult, periph-eral subjects; the second is when situations re-quire judgment of obscure, atypical subjects(Forgas 1995). According to the affect infusionmodel, the adoption of information processingstyle also depends on a combination of factorssuch as the novelty, complexity, and salience of


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


the task, and the personality, motivation, affec-tive state, and cognitive capacity of the personinvolved in the judgment process.

The implication for negotiation processesand outcomes is that feeling good or feelingbad should have important consequences fornegotiator cognition and strategies (Lanzetta1989). In one study, positive mood generatedsuperior individual outcomes in negotiationswith integrative potential characterized by co-operative negotiation strategies (Forgas 1998).In another study, positive mood decreased eva-sive and equivocal communications, especiallyin high-conict negotiations (Forgas & Cromer2004).

One line of research qualies these ndingsby demonstrating that the impact of affectdepends on power. For example, agreementquality was better predicted by the chronic,positive affect of high-power negotiators thanthat of low-power negotiators (Anderson &Thompson 2004). Apparently, the more power-ful negotiators emotions were more inuentialthan the less powerful negotiators emotions.Furthermore, trait-positive affect, com-bined with high structural power (i.e., a strongBATNA), helped negotiation dyads reach moreintegrative agreements without harming eithernegotiators individual outcomes (Anderson &Thompson 2004). Recently, investigations ofnegative affect such as anger expressions havealso been examined (Sinaceur & Tiedens 2006).Anger expressions produced concessions fromnegotiators with a poor BATNA, presumablybecause the angry negotiator communicatedtoughness. This nding contrasts somewhatwith earlier investigations in which feelingsof high anger and low compassion producedlower joint outcomes, but not lower individualoutcomes (Allred et al. 1997). Similarly, whennegotiators expressed positive affect, negativeaffect, or neutral affect in a take-it-or-leave-itultimatum, positive-affect negotiators weremost likely to have their ultimatum accepted.Negative-affect negotiators were the leastsuccessful (Kopelman et al. 2000).

Other research examined the relation-ship between economic outcomes, negotiator

behavior, and satisfaction (an affective responseto negotiation). For example, a negotiators fo-cus on RP or aspirations inuences feelings ofsuccess in a negotiation (Thompson 1995). Ne-gotiators with low RPs felt more successful thandid those with high RPs, even though their nalsettlements were identical. Furthermore, nego-tiators with low aspirations felt more successfulthan did negotiators with high aspirations, eventhough the nal settlement was identical. Aspi-rations inuenced negotiators perceptions ofsuccess more than did RPs. In general, aspira-tions, relative to RPs, exerted a more powerfulinuence on the demands people made to oth-ers in negotiations and how successful they feltabout negotiated outcomes.

Along similar lines, negotiators might feeldissatised when the counterparty accepts theirrst offer (Galinsky et al. 2002). Apparently,when the counterparty immediately acceptsones rst offer, a counterfactual thoughtprocess is produced (e.g., Oh no, I shouldhave asked for more!). This counterfactualthought process results in dissatisfaction, evenwhen negotiators outcomes were objectivelysuperior to agreements reached later in negoti-ations. Thoughts about how much better theycould have done overwhelmed negotiators ob-jective outcomes. These ndings are consistentwith studies demonstrating that negotiatorssatisfaction depends heavily on the compar-ison value on which they focus attention:Negotiators who focused on their target priceconsistently achieved better outcomes but wereless satised than those who focused on theirBATNA (Galinsky et al. 2002). Yet, focusingon the target price during a negotiation andthe BATNA after a negotiation allowed nego-tiators to achieve superior outcomes withoutthe accompanying dissatisfaction.

An array of negative cognitions and emo-tions confront negotiators who fail to reachdeals (OConnor & Arnold 2001). For exam-ple, negotiators who failed to reach agree-ment (i.e., impassed) found themselves caughtin a distributive spiral such that they inter-preted their performance as unsuccessful, ex-perienced negative emotions, and developed


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


negative perceptions of their counterpart andthe process. Moreover, they were less willingto work with their counterpart in the future,planned to share less information and behaveless cooperatively, and lost faith in negotia-tion as an effective means of managing conicts(OConnor & Arnold 2001).

INTERPERSONAL LEVEL

Economic and SocialPsychological Foundations

Traditionally, negotiation at the interper-sonal level has been viewed via the lens ofmixed-motive interaction. The concept ofmixed-motive interaction was rst introducedby economist Thomas Schelling (1960) to referto situations where two or more parties facea conict between two motives: cooperation(the integrative aspects of negotiation) andcompetition (the distributive aspects). Innegotiations, individuals must cooperate toavoid impasse and reach mutual agreement, butcompete to gain sufcient resources for them-selves. Two-person bargaining is thus a classicexample of a mixed-motive interaction. Indeed,Lax & Sebenius (1986b) emphasize that allnegotiators must balance the twin tasks ofnegotiation: creating value and claiming value.

The interpersonal system in negotiationwas also richly stimulated by basic researchin the areas of emotional contagion, mimicry,and behavioral synchrony (Chartrand & Bargh1999). One nding in these areas, for exam-ple, is that people tend to engage in facerubbing, foot shaking, and smiling more inthe presence of someone who engages inthat behavior (Chartrand & Bargh 1999). An-other is that behavioral mimicry increases lik-ing and rapport between interaction partners(Tiedens & Fragale 2003). Beyond behavioralmimicry, more complex interpersonal mimicrysuch as mood contagion (see Neumann &Strack 2000) and dominance complementarity(see Tiedens & Fragale 2003) have also beendocumented. Mood contagion effects demon-strate that people easily assume the moods of

others. Dominance complementarity ndingsdemonstrate that people respond to othersdominant behavior with a submissive stance,and vice versa. Furthermore, they demonstratethat when one party complements dominantbehavior with submissive behavior, this facili-tates interpersonal liking (Tiedens & Fragale2003).

Interpersonal Effects of Emotionsin Negotiation

Emotions inuence negotiations at the inter-personal as well as intrapersonal level. In fact,research on emotions in negotiation bridges theintrapersonal and interpersonal level. Two spe-cic emotions, anger and happiness, have re-ceived particular attention from negotiation re-searchers (Van Kleef et al. 2004a). Participantsin one study received information about theemotional state (anger, happiness, or none) oftheir opponent (Van Kleef et al. 2004a). Con-sistent with the research noted above, partic-ipants conceded more to an angry opponentthan to a happy one. Apparently, people usedemotion information to infer the others limit(i.e., their RP), and they adjusted their demandsaccordingly. However, this effect was absentwhen the other party made large concessions.Angry communications (unlike happy ones) in-duced fear and thereby mitigated the effect ofthe opponents experienced emotion. Negotia-tors were especially inuenced by their oppo-nents emotions when they were motivated toconsider them (Van Kleef et al. 2004b).

The processes and mediators behind the in-terpersonal effects of emotions may be inu-enced by the extent to which individuals aremotivated to process information systemati-cally and deeply (De Dreu & Carnevale 2003,Van Kleef et al. 2004b). For instance, partici-pants in one study (Van Kleef et al. 2004b) re-ceived information about the opponents emo-tion (anger, happiness, or none). Those in theangry condition received a message saying thisoffer makes me really angry, whereas those inthe happy condition received a message say-ing I am happy with this offer. As predicted,


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


negotiators conceded more to an angry oppo-nent than to a happy one, but only when theyhad low (rather than high) need for cognitiveclosurea measure of their chronic motivationto process information systematically. Also, par-ticipants were only affected by the others emo-tion under low rather than high time pressure,because time pressure reduced their capacityfor information processing. Finally, negotiatorswere only inuenced by their opponents emo-tions when they had low (rather than high)power, presumably because high-power nego-tiators had less need and were less motivated toprocess this information. These results supportthe motivated information-processing model,which argues that negotiators are only affectedby their opponents emotions if they are moti-vated to consider them.

Interpersonal Improvisationin Negotiation

Other research has utilized a more qualita-tive approach to unpack interpersonal pro-cesses in negotiations. Beyond the focus oneconomic outcomes in negotiations, negotia-tors may sometimes also be focused on rela-tionship processes and outcomes (McGinn &Keros 2002).

Specically, McGinn & Keros (2002) high-light the improvisation and the logic of ex-change in socially embedded transactions. So-cially embedded transactions take into accountthe fact that negotiators can have deep socialties or share mutual social ties with one another.This is in contrast to the arms length trans-action between individuals, in which individu-als share little familiarity or affect and no pro-longed past or expected future ties (Granovetter1973, Podolny & Baron 1997, Uzzi 1997).

By improvisation, McGinn & Keros (2002)conjecture that most people at the outset of anegotiation do not construe it as such. Thisis because, whereas arms length transactionsare often guided by a logic of prot maximiza-tion, embedded transactions (such as betweenfriends) go beyond the focus on outcomes alone;

they tend to focus on rules of friendships as op-posed to rules of the market.

In a qualitative fashion, McGinn & Keros(2002) used a sense-making lens to illuminatemicroprocesses underlying socially embeddedtransactions, investigating how social networksaffect the logic of exchange governing the trans-action. Transcript analysis of two-party negoti-ations revealed that most pairs of negotiatorsquickly coordinated a shared logic of exchangeand improvised in accord with its implied rulesthroughout their interaction. The improvisa-tion took the form of opening up, workingtogether, or haggling. Negotiators used threedynamic processestrust testing, process clar-ication, and emotional punctuationwhenthey had difculty moving the interaction to-ward a coherent, mutually agreed-upon pattern.Social embeddedness, or the extent to which anindividual shares other social connections withanother individual (Granovetter 1973), easescoordination within negotiation (McGinn &Keros 2002).

Subjective Value in Negotiation

As noted above, negotiators have noneconomic,relational concerns as well as economic ones.Besides their concern with economic gains,negotiators are also concerned about their feel-ings about the self, the negotiation process, andthe relationship (Curhan et al. 2006). More-over, the subjective value accrued from thesecomponents of negotiation have long-lastingimpact (Curhan et al. 2009). For example,the subjective value that actual managersderived from job offer negotiations predictedtheir subsequent job attitudes and turnoverintentions better than the economic value theyachieved: Subjective value measured at theoutset of a negotiation predicted managers jobsatisfaction and likelihood of quitting a full oneyear later. Curiously, negotiators economicoutcomes (i.e., their actual salaries) did notpredict satisfaction or turnover. Arguably, thesubjective value gained from a negotiation mayhave more long-lasting impact than the actualeconomic gains from the negotiation.


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


However, one potentially important con-sideration is whether subjective value conictswith economic value in negotiations. To exam-ine this, negotiators who held relational goalswere compared with negotiators who held eco-nomic goals. If relational goals hinder economicgain, then it would be reasonable to expectnegotiators to underperform relative to eco-nomically motivated negotiators (Curhan et al.2008). Indeed, negotiators in egalitarian or-ganizations reached less-efcient (i.e., worse)economic outcomes but had higher relationalcapital than did those who negotiated in hier-archical organizations. By directly pitting eco-nomic gain against relational considerations,this study showed how the structure of ones en-vironment (egalitarian versus hierarchical) caninuence ones own goals and therefore nego-tiation outcomes.

Trust and Tactics

Mutual trust is an essential ingredient in effec-tive organizations (see Dirks & Ferrin 2001)and negotiations (Kimmel et al. 1980). Trust,dened as the intention to accept vulnerabilitybased upon positive expectations of the coun-terparts behavior and intentions (Rousseauet al. 1998), allows negotiators to exchangethe information necessary for integrative agree-ments. Distrusting negotiators are reluctant toshare information or ask questions, believingthat their counterparts will take advantage ofshared information and respond to their ques-tions dishonestly. Conversely, trusting negotia-tors believe their counterparts will use informa-tion to identify integrative agreements. Theyalso tend to believe information that the coun-terpart shares, accepting it as sincere and ac-curate (Parks et al. 1996). As a result, trustingnegotiators exchange more information aboutpreferences and priorities and achieve more in-tegrative outcomes (Butler 1995, Kimmel et al.1980, Pruitt & Kimmel 1977, Weingart et al.1993).

Despite the importance of trust, viola-tions of trust are common (see Elangovan &Shapiro 1998 for a review), jeopardizing the

integrativeness of negotiation outcomes. Giventhe mixed-motive nature of negotiation, it istempting for negotiators to use deception tomaximize their personal gain. Yet, deception islikely to compromise trust. Thus, an impor-tant question is when people will lie in ne-gotiations. People tend to lie when the luresof temptation and uncertainty align with pow-erless and anonymous victims (Tenbrunsel &Diekmann 2007). The more negotiators standto gain economically, the more likely they areto lie (Bazerman et al. 1998). Moreover, themore uncertainty negotiators have about ma-terial factors, the more likely they are to lie. Ofcourse, liars often garner a reputation as such,making it more difcult for them to win coun-terparts trust in the future (Glick & Croson2001).

Given that negotiators may sometimes re-sort to deceptive tactics in negotiations, anotherimportant consideration is how interpersonaltrust broken by deceptive behavior can be re-stored. One theory holds that broken trust cannever be fully restored, even if the trust breakerperforms a series of consistently trustworthy ac-tions (Schweitzer et al. 2006), such as fullledpromises, apologies, and consistently reliablebehavior. A promise to change behavior cansignicantly speed the trust recovery process,but prior deception harms the effectiveness of apromise in accelerating trust recovery. Anotherperspective holds that apologies can effectivelyrestore trust when the trust violation concerns amatter of competence, but not when it concernsa matter of integrity (Kim et al. 2004).

In a given negotiation, tactics such as threats,bluffs, and disclaimers can affect negotiatorsrelationships and the grounds for their trust.For example, a buyer-seller simulation with twonegotiation periods examined the behavioraland attitudinal consequences of threats, bluffs,and disclaimers (Shapiro & Bies 1994). Somenegotiators received a threat stated as a dis-claimer, whereas others did not. Changes in ne-gotiators evaluations of their partner and ne-gotiation outcomes were examined after somewere led to believe their partner had stateda false threat (a bluff ). Negotiators who used


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


threats were perceived as more powerful, butthey were also perceived as less cooperativeand achieved lower integrative agreements thanthose who did not use threats.

Relationships and Negotiations

Perhaps the most straightforward question onecould investigate about the interpersonal as-pects of negotiation is whether people involvedin a relationship can fashion integrative agree-ments better than strangers can. Kelley (1982)studied how couples negotiate problems of in-terdependence. Yet, the rst study that trulyexamined how people in relationships, ver-sus strangers, negotiate was Fry et al.s (1983)study of dating couples. Paradoxically, strangerswere more likely to reach win-win (mutuallybenecial agreements) than were dating cou-ples, although the effect did not reach conven-tional levels of signicance. The authors rea-soning was that couples (and perhaps friends)are uncomfortable asserting their own needsand therefore are more willing to settle for sub-optimal agreements.

The orientation that friends bring to anegotiation seems to dictate the outcomes theyachieve. Pairs of friends who are similar in com-munal orientation are most likely to capitalizeon joint interests (Thompson & DeHarpport1998). However, when friends are dissimilarin communal orientation, their ability to iden-tify compatible issues declines precipitously.Friends who are high in communal orientationare more likely to allocate resources equallythan are friends low in communal orientation.The existence of friendships also has signicantimplications for ones negotiation outcomes(Seidel et al. 2000). Seidel and colleagues ana-lyzed more than 3000 actual salary negotiationsand found that having friends in high placeswithin the relevant organization improvedsalary negotiation outcomes.

Whereas the studies reviewed above tendto focus on economic outcomes, negotiationsalso involve symbolic resources such as identityand legitimacy. Glynn (2000) studied identityand legitimacy during a musicians strike at the

Atlanta Symphony Orchestra. Glynn analyzedthe musicians and administrators as compet-ing parties vying for the legitimacy to denethe core identity of the orchestra. Embeddedwithin the multilayered negotiation, Glynn re-ports, were conicts over status and power and,implicitly, control over the resources that wouldconfer such status and power (p. 291). Thisstudy illustrates that relationships not only in-uence negotiations, but negotiations can re-constitute and reshape relationships.

GROUP LEVEL

The group system focuses on how group dy-namics inuence negotiation processes and out-comes. In this section, we selectively focus onfour major streams of research at the grouplevel: social and group identity, relational andcollective identity, group culture, and teams andthe discontinuity effect. Some of this researchuses paradigms derived from game theory (e.g.,social dilemmas), but we include it in this reviewbecause it speaks directly to descriptive negoti-ation research.

Social and Group Identity

According to the group identity perspective,which is part of a larger social identity tradition(e.g., Tajfel et al. 1971), the stronger an individ-uals group identity, the less sharply he or shedistinguishes between self-interest and collec-tive interest. For negotiation, this implies thatdistributive (personal gains) are less focal thanintegrative (mutual gains) for negotiators whoconsider counterparts members of their group.This conjecture has been examined most di-rectly in the social dilemma literature, whichexamines situations where individual and col-lective interest are largely opposed.

There are two perspectives concerningchoice in a social dilemma situation. From apurely economic point of view, the rationalchoice is to defect because it yields greateroutcomes. Of course, if everyone defects, thenthe collective welfare of the group suffers. Thesocial psychological viewpoint is that defection


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


is undesirable and people are best served wheneveryone puts self-interest aside and chooses tomaximize group interests. Kramer & Brewer(1984) pioneered the study of group identityin social dilemma and negotiation research. Byemphasizing the common fate among groupmembers and the salience of a superordinategroup identity, they showed that the degreeof cooperation in social dilemmas increases(Brewer & Kramer 1986, Kramer & Brewer1984). Another way of inducing group identityis to extend the length of time a person expectsto be part of a group. In one study (Mannix& Loewenstein 1993), people who expected tobe part of a group for a long time were moreconcerned with the welfare of the group thanwere people who anticipated a eeting inter-action. Moreover, negotiators who perceivedthat other group members would leave coop-erated less than did those who expected thegroup to remain intact (Mannix & Loewenstein1993).

These studies suggest that making groupidentity salient tends to activate different ne-gotiation processes, producing different out-comes. Yet, the importance of group identity inmixed-motive interactions such as negotiationhas not gone unchallenged. Kerr & Kaufman-Gilliland (1994) examined the impact of socialidentity on cooperation in social dilemmas. In acarefully constructed set of studies, they foundstrong support for the idea that it is negotiatorsverbal promises that increase cooperation in so-cial dilemmas, not simply the extent to whichnegotiators feel identied with their group.

Relational and Collective Identity

Recently, work on identity has moved from theextent to which individuals feel they are a partof their group to the nature of the identity. Forexample, Markus & Kitayama (1991) focusedon whether people hold independent or inter-dependent identities, or self-construals. A per-son who holds an independent self-construaldenes himself or herself in terms of the at-tributes, preferences, and traits that make him

or her unique and autonomous. In contrast, aperson with an interdependent self-construal ismore likely to dene himself or herself in termsof his or her social and group relationships(Gardner et al. 1999, Markus & Kitayama1991). In a one-on-one, dispute-negotiationcontext, Seeley et al. (2007) primed indepen-dent versus interdependent self-construals andfound that negotiators with interdependentself-construals were more generous than wereindependent negotiators. However, this effectcompletely reversed in a team-on-team con-text, such that teams with independent self-construals (i.e., highly dened by their ownattributes) were more generous than teamswith interdependent self-construals (i.e., de-ned with reference to the other team). All ofthese effects held primarily for high-power ne-gotiators. The implication is that interdepen-dent self-construals seem to evoke a benevolentuse of power in dyadic contexts but a more ex-ploitative use of power in intergroup contexts.

Very little research has examined the pos-sibility of reverse causality between negotia-tion and social identitythat the negotiationprocess itself could inuence peoples identity.Thompson (1993) examined how negotiationaffects intergroup relations. People who nego-tiated with an out-group member developedmore favorable evaluations of the out-group,whereas people who negotiated with an in-group member were more likely to show in-group favoritism. However, when the negoti-ation situation dictated that negotiators couldnot reach a mutually benecial agreement,the positive effects of interpersonal negotia-tion disappeared. Thus, negotiation with out-group members improves intergroup relationsin negotiations with integrative potential. Fur-thermore, outcomes are comparable regard-less of the counterparts group membership.Whereas individuals expecting to negotiatewith out-group members thought they wouldobtain lower outcomes than those expect-ing to negotiate with in-group members, thevalue of the actual outcomes achieved did notdiffer.


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


Culture

One important aspect of group identity is cul-ture, or the distinctive characteristics of a par-ticular social group (Lytle et al. 1995). Culture ismanifest in a groups values, beliefs, norms, andbehavioral patterns. An underlying feature ofWestern cultures is the use of formal logic andavoidance of contradiction (Nisbett et al. 2001).In contrast, in non-Western cultures, cogni-tion is characterized by a holistic system ofthought. Individuals view themselves as embed-ded and interdependent with a larger social con-text. They also tend to focus their cognitive at-tention on relationships and context (Peng &Nisbett 1999).

One result of this difference in systems ofthought is that negotiators from different cul-tures make more or less use of emotional ap-peals. Emotional appeals are relatively inconsis-tent with formal logic. Thus, negotiators fromnon-Western cultures tend to make more emo-tional appeals than do U.S. negotiators (Drake1995). For instance, Taiwanese negotiators usedmore normative statements referring to socialroles and relationships than did U.S. negotia-tors (Drake 1995). Conversely, U.S. negotia-tors used more statements emphasizing logicand reasoning than did Taiwanese negotiators.

Another important cultural difference be-tween Western negotiators and non-Westernnegotiators is the motivation that they bringto the negotiating table. Motivation is the fo-cused and persistent energy that drives cog-nition and behavior (Mook 2000). Motivationimpacts how negotiators approach negotiationsand evaluate outcomes. In Western cultures,negotiators tend to judge negotiation outcomesby the joint prot that accrues and the valuethat they themselves claim (Lax & Sebenius1986c, Neale & Bazerman 1992). However,in non-Western cultures, negotiators may caremore about relational capitalthe mutual trust,knowledge, and commitment that can accruefrom negotiatingmore than economic out-comes (Gelfand et al. 2006).

For example, Japanese negotiators place ahigh value on relational capital: They prefer

and even insist on negotiating with peoplewith whom they have a relationship or so-cial network, even if it means forgoing poten-tial economic benets (Graham & Sano 1989,Yamagishi & Yamagishi 1994). Indian man-agers, on the other hand, may assume lowerrelational capital in the form of mutual trustthan do American managers, and negotiationsmay serve to reafrm their assumptions (Guniaet al. 2009). In two studies, Indian managerslower level of trust led to low joint gains rela-tive to the gains of American managers.

Culture also has important effects on howindividuals perceive causality. Psychologicalresearch has demonstrated that members ofWestern cultures tend to make the fundamentalattribution error more often than do membersof non-Western cultures (Nisbett et al. 2001,Peng & Nisbett 1999). That is, they underesti-mate the impact of situational factors and over-estimate the impact of others dispositional fac-tors in causing events (Ross 1977). The resultfor negotiation is that U.S. negotiators tend tomake dispositional attributions for their coun-terparts behaviors and discount potential situa-tional attributions (Morris et al. 1999). Disposi-tional attributions for negative behaviors lead tonegative consequences in negotiations. Specif-ically, dispositional attributions led to compet-itive perceptions of the situation and counter-part, resulting in a preference for adversarialinstead of collaborative procedures.

Groups and the Discontinuity Effect

A central question in group research is whethertwo heads are better than one (Insko et al.1987, 1988, 1990; Schopler et al. 1991, 1993).This question was rst addressed using a sim-ple prisoners dilemma game in which ne-gotiators were offered a cooperative (trust-ing) choice or a defecting (self-interested,exploitive) choice. Overwhelmingly, one-on-one negotiators made more cooperative choicesthan did group-on-group negotiators, underidentical payoffs. Insko et al. (1987) coinedthe term discontinuity effect to describe


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


the empirical nding that one-on-one nego-tiation behavior cannot be simply extrapo-lated to group-on-group negotiation behavior.Schopler & Insko (1992) argued that the dis-continuity effect was driven by group membersfear of being exploited by the out-group as wellas their greed for additional payoffs.

Thompson et al. (1996) examined thediscontinuity effect in a markedly different ne-gotiation paradigm, in which parties interestswere not completely opposed and a mutuallyattractive, optimal outcome existed but wasnot apparent to negotiators. This paradigmwas similar to the sisters-and-orange parablein the introduction. In terms of integrativeoutcomes, group-on-group congurationsproduced more integrative agreements thandid solo-on-solo or solo-on-group. In termsof distributive outcomes, groups earned morethan solos. The authors reasoned that insuch a negotiation, information processingis paramount; indeed, groups asked morerelevant questions, shared more information,and formed more accurate judgments than didsolos (see also Peterson & Thompson 1997).The group-on-group conguration apparentlyallowed negotiators to seek and process moreof the relevant information.

Morgan & Tindale (2002) attempted to re-solve the disparate ndings between Insko et al.(1987) and Thompson et al. (1996). Morganand Tindales insight was that the disparate-appearing ndings were based upon dramati-cally different negotiation tasks: Insko and col-leagues used a prisoners dilemma task, whereasThompson and colleagues used an integrativebargaining task; the tasks differ in many im-portant ways (see Thompson 2009 for a re-view of the differences). In Morgan & Tin-dales (2002) study, negotiators were allowed toreach an agreement on either a cooperative orcompetitive integrative bargaining task in oneof three formats (group versus group, groupversus single, or one-on-one). Next, negotia-tors were asked to choose between maintainingthe agreed-upon settlement or defecting withina prisoners dilemma payoff structure. Groupscontinued to show the discontinuity effect, such

that they opted to defect. This was true evenwhen they had performed better than the solonegotiator with whom they had just negotiated.Groups shared motives for defection that differdepending upon the nature of the task and op-ponent (Morgan & Tindale 2002).

ORGANIZATIONAL LEVEL

The organizational system represents a higherlevel of analysis than the previous levels; it ex-amines the negotiator as embedded in a largernetwork or marketplace. This level of analy-sis is crucial because in organizations and inmarkets, dyads rarely operate in isolation fromtheir social context. Instead, each negotiatortypically participates in multiple dyadic rela-tionships, and these dyadic relationships aggre-gate to form a complex social structure that sur-rounds each dyad and inuences trust, expecta-tions, and interpersonal perceptions.

Heider (1958) documented that two peo-ple can be connected by a third party, whostrengthens or disturbs the relationship amongthe two. Contemporary sociologists have alsodocumented how dyadic relationships and in-terpersonal behavior may be inuenced by theoverall network structure in which the dyad isembedded (e.g., Burt & Knez 1996, Coleman1990, Granovetter 1985). Despite these foun-dations, relatively little research has examinedhow negotiation dyads operate within theirlarger social context. In this section, we re-view three streams of negotiation research atthe organizational level. The rst two exam-ine how interpersonal connections (choosingnegotiation partners and reputations) inuencenegotiation processes. The third looks at howorganizational or institutional forces impactnegotiations.

Choice of Negotiation Partner

A critical issue facing employees and employers,buyers and suppliers, and joint venture partnersis whom to select as a negotiation partner. Thevast majority of studies in the existing negotia-tion literature have simply assigned negotiation


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


partners (Tenbrunsel et al. 1999). One of theearliest studies that examined this problem ofsearch and deliberation in partner choice wasSondak & Bazermans (1989) study of match-ing in quasi-markets. In this paradigm, a largemarket of buyers and sellers was created and ne-gotiators were told to partner with whomeverthey pleased, to make a deal. The main ndingwas that substantial economic suboptimality ex-ists as the result of selection mismatches. Peo-ple may choose to negotiate with their friends,even though the integrative potential of ne-gotiating with a stranger may be higher (seealso Northcraft et al. 1998). Similarly, whenpeople had the option to choose their friendas negotiation partner in a simulated hous-ing market, they often stopped searching andreached a deal with the friendoverlookingother, potentially fruitful negotiation relation-ships. Ultimately, this led to market inefcien-cies (Tenbrunsel et al. 1999).

Reputation and NegotiationThrough Time

One consideration that inuences the inte-grative and distributive outcomes negotiatorsachieve in organizational systems is their rep-utation. Much sociological and macro organi-zational research has documented the impor-tance of reputation in markets (e.g., Raub &Weesie 1990). In one investigation (Glick &Croson 2001), the impact of reputations amongmanagement students in a semester-long ne-gotiation course was examined. Students ratedone another on the basis of rsthand experi-ence, from least cooperative to the most co-operative. Four reputational proles emerged:the liar-manipulator (who will do anythingto gain advantage), tough-but-honest (verytough negotiator who makes few concessionsbut will not lie), nice-and-reasonable (makesconcessions), and cream puff (makes conces-sions and is conciliatory regardless of what theother does). Once reputations spread throughthe market, behavior changed. People actedmuch tougher when dealing with perceivedliar-manipulators, for example. Furthermore,

people used tough or manipulative tactics ina defensive fashion with liar-manipulators andtough-but-honest negotiators, but used themin an opportunistic fashion with cream puffs(Glick & Croson 2001).

Other research examined how reputation isrelated to history of negotiation behavior, alsoin an MBA class (Anderson & Shirako 2008).The development of reputations was trackedamong individuals who engaged in multiple ne-gotiation tasks across several weeks. Reputa-tions were only mildly related to the actual his-tory of behavior. However, the link betweenreputation and behavior was much strongerfor some individuals than others. The link wasstrongest for those who were well known andreceived the most social attention. In contrast,behavior had little impact on the reputations oflesser-known individuals.

Another, similar perspective suggests thatdyadic negotiation is not an isolated event, butrather inuences subsequent dyadic negotia-tions (OConnor et al. 2005). Specically, thequality of the deals negotiators reached at anypoint in time were strongly inuenced by theirprevious bargaining experiences. Negotiatorswho reached an impasse in a prior negotiationwere more likely either to impasse in their nextnegotiation or to reach deals of low joint valuerelative to those who had reached an initialagreement. Notably, the impact of past perfor-mance on subsequent deals was just as strong fornegotiators who changed partners on the sec-ond occasion. These results highlight the role ofbargaining history as a predictor of negotiationbehavior. Moreover, they suggest that, at leastin some cases, negotiations should be concep-tualized as interrelated exchanges rather thandiscrete incidents.

Organizations also impact negotiations viainstitutional forces. One controversial per-spective argues that organizations or institu-tions may serve as barriers to negotiations(Wade-Benzoni et al. 2002). Specically,normative factors (obligations, operating pro-cedures), cognitive factors (cultural values, cog-nitive frameworks), and regulatory factors (reg-ulations and laws) may impede negotiations.


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


For example, organizations with cultures em-phasizing strict adherence to procedure maydiscourage negotiation by explicitly prohibitingit (normative factor) or by preventing employ-ees from even perceiving it as a viable alternative(cognitive factor). The value-laden lens that or-ganizationally embedded actors bring may alsolead to impasse or prevent people from reachingeconomically efcient outcomes.

VIRTUAL LEVEL

Given the ubiquity of computer-mediated com-munication technology in business commu-nications, consumer transactions, and inter-personal relationships, virtual negotiation iscurrently a fertile ground for research (Nadler& Shestowsky 2006).

A straightforward question one might askis whether negotiation is best conducted face-to-face or via computer-mediated communica-tion technology. Answers to this question aresurprisingly mixed (see Nadler & Shestowsky2006 for a review). In some cases, nego-tiators who interact via computer-mediatedtechnology are less likely to reach integra-tive outcomes than are negotiators who inter-act face-to-face (Arunachalam & Dilla 1995,Barefoot & Strickland 1982) or via paper andpencil (Grifth & Northcraft 1994). On theother hand, some studies report no reliable ef-fect of communication medium (Morris et al.2002, Naquin & Paulson 2003, Purdy et al.2000).

With regard to condence and satisfaction,parties who negotiate face-to-face feel morecondent in their performance and satisedwith their negotiation outcome than do thosewho negotiate via computer (Naquin & Paulson2003, Purdy et al. 2000, Thompson & Coovert2003). Moreover, compared to parties who ne-gotiate face-to-face, parties who negotiate viaemail desire less future interaction with theircounterpart (Naquin & Paulson 2003). De-spite these differences in subjective outcomes,studies that examined the emotional content ofmessages in email and face-to-face negotiations

found no differences between the two mediums(Morris et al. 2002).

Moderators and Mediators

Though the effects of information technologyon interpersonal outcomes in negotiation maycurrently seem inconclusive, some studies haveidentied important mediators that may help toexplain the effects of technology on negotiationin the future. For instance, negotiators behavemore honestly when negotiating face-to-facethan via writing (Valley et al. 1998). The com-munication medium in which bargaining takesplace also affects the efciency and distributionof outcomes (Valley et al. 1998). Face-to-face communication may facilitate moretruth-telling and trust than communication viawriting, thus inuencing negotiation outcomes.

However, negotiators may sometimes be-have less cooperatively when they have visualaccess to one another than when they do not(Carnevale & Isen 1986, Carnevale et al. 1981).In one investigation, researchers examined theinuence of positive affect and visual accesson the process and outcome of negotiation inan integrative bargaining task (Carnevale &Isen 1986). Only when negotiators were face-to-face and not in a positive affective statewere there heavy use of contentious tactics, re-duced tradeoffs, and few integrative solutions.In other words, when negotiators had visual ac-cess and were potentially experiencing negativeaffect, they were more likely to use contentioustactics.

Other research has examined contexts inwhich email negotiations may fail or succeed.For instance, Moore et al. (1999) proposed thatthere were long and short routes to successin electronically mediated negotiations. A longroute to success would involve many of the as-pects of deliberate cognitive processing; a shortroute would involve more heuristic, supercialprocessing of information (Fiske 1988, Sloman2002).

To understand why email negotiations of-ten fail, another study (Moore et al. 1999) ex-amined two distinct elements of negotiators


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


relationships: shared membership in a socialgroup and mutual self-disclosure. Some par-ticipants negotiated with a member of an out-group (a student at a competitor university),whereas others negotiated with a member ofan in-group (a student at the same univer-sity). In addition, some negotiators exchangedpersonal information with their counterparts,such as their hometown and hobbies, whereasothers did not. When neither common in-group status nor a personalized relationship ex-isted between negotiators, email negotiationswere more likely to end in impasse. These re-sults were attributable to the positive inuenceof mutual self-disclosure and common groupmembership on negotiation processes and rap-port between negotiators, especially in a rela-tively impersonal context like email.

CONCLUSION

Our review has focused on a subset of re-search ndings that have strongly impactedthe study and practice of negotiation. The re-search ndings span several decades, but theinvestigations meaningfully build upon one an-other because the key criteria by which schol-ars evaluate the quality of negotiation has re-mained essentially unchanged since the dawn ofnegotiation research. Modern negotiation re-search has greatly benetted from its economicroots, which have provided rigorous methodsby which to measure the mutual value createdby two or more parties, each motivated to pur-sue their own interests. The robust empiricalfact that most negotiators fail to fully maximizetheir own gains (as well as mutual gains) whenseated at the bargaining table has greatly fueledthe res of negotiation research.

Our focus on intrapersonal, interpersonal,group, organizational and virtual systems hasallowed us to examine the wide lens throughwhich the apparently simple task of negotia-tion may be meaningfully studied. The intra-personal system provides the most close-upview of negotiation, taking us into the mindand heart of the negotiator, who is either an-ticipating or engaging in a negotiation. The

interpersonal system is particularly meaning-ful in negotiation research because the dyadicprocess allows us to examine the presence orabsence of interpersonal phenomena such as be-havioral synchrony and mutual gaze, which can-not be reduced to the intrapersonal level. Thegroup and organizational systems have been in-uenced by rich social psychological, as well associological and organizational, traditions. Ne-gotiation research, like the universe, appears tobe expanding rather than contracting. Indeed,the virtual level has allowed globally dispersedresearchers themselves to collaborate while in-vestigating negotiation at a virtual level. Ratherthan reporting to a physical laboratory, todaysresearch participants often negotiate via com-puter with people they will never meet.

It is curious how some research topics withinthe domain of social and organizational psy-chology sustain themselves over time, whereasothers are mere ashes in the pan. Negotiationand bargaining research, by nearly any stan-dard, has withstood the test of time. There areseveral reasons for its longevity. First, the mul-tidisciplinary nature of negotiation has broughtscholars together, especially from social psy-chology and organizational behavior and alsofrom game theory and economics. These mul-tidisciplinary collaborations have created a richnetwork of negotiation scholars that lead toshared volumes, conferences, and even jobsand research positions, thereby ensuring thelongevity of the eld. Nearly every businessschool offers a course in negotiation that manyMBA students take, requiring a cadre of trainedfaculty members. The faculties often receivetheir training in PhD programs or in post-doctoral programs that focus primarily on ne-gotiation. Graduate students are attracted tosuch positions and develop research ideas thatare relevant to the broad array of negotiationtheory.

A second factor that has contributed to thecontinued popularity of negotiation research isthe fact that it is considered an essential busi-ness, if not a life, skill. The demand for ne-gotiation skills spurs the development of ne-gotiation books, courses, seminars, cases, and


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


teaching materials that require theoretical rigorand background. The existence of a normativetheory by which to evaluate the performance ofnegotiators provides a foundation for meaning-ful research and theory. The existence of de-scriptive theory provides meaningful insightsinto negotiations as they typically unfold.

If there is a downside to negotiation researchit might be that negotiation has done more tak-ing than giving, meaning that often the negotia-tion scholarship is essentially about social or or-ganizational phenomena that could frankly bestudied as easily in other contexts. For example,

one might study behavioral synchrony or mir-roring in negotiation, but it is equally plausibleto study these same phenomena in other con-texts, like small, collaborative teams or job in-terviews. Similarly, more than two decades ofresearch have focused on extending Kahnemanet al.s (1982) research on judgment biases (e.g.,framing, anchoring, overcondence) to two-party negotiations (for a review, see Neale &Bazerman 1994). Despite this prodigious bor-rowing, our review suggests that negotiation re-search has yielded many insights of its own andis poised to yield many more in the future.

SUMMARY POINTS

1. Intrapersonal processes such as ones psychological power and mood impact negotiationprocesses and outcomes.

2. Interpersonal processes such as display of emotions also impact negotiation processesand outcomes.

3. When negotiation takes place not between individuals but rather between groups, groupidentity, culture, and structure of negotiation will affect whether groups (teams of nego-tiators) do better than solo negotiators.

4. The social context and network in which one is embedded also inuences negotiations,through choice of negotiation partner and formation of reputation.

5. When negotiations are not face-to-face but rather are computer-mediated, many vari-ables come into play in determining whether computer-mediated negotiations harm orfacilitate negotiations.

DISCLOSURE STATEMENT

The authors are not aware of any biases that might be perceived as affecting the objectivity of thisreview.

LITERATURE CITED

Allred KG, Mallozzi JS, Matsui F, Raia CP. 1997. The inuence of anger and compassion on negotiationperformance. Organ. Behav. Hum. Decis. Process. 70(3):17587

Anderson C, Shirako A. 2008. Are individuals reputations related to their history of behavior? J. Personal. Soc.Psychol. 94(2):32033

Anderson C, Thompson LL. 2004. Affect from the top down: how powerful individuals positive affect shapesnegotiations. Organ. Behav. Hum. Decis. Process. 95(2):12539

Arunachalam V, Dilla WN. 1995. Judgment accuracy and outcomes in negotiationa causal-modeling analysisof decision-aiding effects. Organ. Behav. Hum. Decis. Process. 61(3):289304

Babcock L, Gelfand M, Small D, Stayhn H. 2006. Gender differences in the propensity to initiate negotiations.In Social Psychology and Economics, ed. D De Cremer, M Zeelenberg, JK Murnighan, pp. 23962. Mahwah,NJ: Erlbaum


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


Barefoot JC, Strickland LH. 1982. Conict and dominance in television-mediated interactions. Hum. Relat.35(7):55966

Bargh JA, Pietromonaco P. 1982. Automatic information processing and social perception: the inuence oftrait information presented outside of conscious awareness on impression formation. J. Personal. Soc.Psychol. 43(3):43749

Bazerman MH, Tenbrunsel AE, Wade-Benzoni K. 1998. Negotiating with yourself and losing: making deci-sions with competing internal preferences. Acad. Manag. Rev. 23(2):22541

Berderskey C, McGinn KL. 2009. Open to negotiation: phenomenological assumptions and knowledge dis-semination. Organ. Sci. In press

Blair IV, Banaji MR. 1996. Automatic and controlled processes in stereotype priming. J. Personal. Soc. Psychol.70(6):114263

Bowles HR, Babcock L, Lai L. 2007. Social incentives for gender differences in the propensity to initiatenegotiations: Sometimes it does hurt to ask. Organ. Behav. Hum. Decis. Process. 103(1):84103

Brewer MB, Kramer RM. 1986. Choice behavior in social dilemmas: effects of social identity, group size, anddecision framing. J. Personal. Soc. Psychol. 50(3):54349

Butler JK. 1995. Behaviors, trust, and goal achievement in a win-win negotiating role play. Group Organ.Manag. 20(4):486501

Burt RS, Knez M. 1996. A further note on the network structure of trust: reply to Krackhardt. Rationality Soc.8(1):11720

Camerer C. 2003. Behavioral Game Theory: Experiments in Strategic Interaction. New York: Russell Sage Found./Princeton, NJ: Princeton Univ. Press

Carnevale PJD, Isen AM. 1986. The inuence of positive affect and visual access on the discovery of integrativesolutions in bilateral negotiation. Organ. Behav. Hum. Decis. Process. 37(1):113

Carnevale PJD, Pruitt DG, Seilheimer SD. 1981. Looking and competing: accountability and visual access inintegrative bargaining. J. Personal. Soc. Psychol. 40(1):11120

Chartrand TL, Bargh JA. 1999. The Chameleon effect: the perception-behavior link and social interaction.J. Personal. Soc. Psychol. 76(6):893910

Coleman JS. 1990. Foundations of Social Theory. Cambridge, MA: Harvard Univ. PressCurhan JR, Elfenbein HA, Xu H. 2006. What do people value when they negotiate? Mapping the domain of

subjective value in negotiation. J. Personal. Soc. Psychol. 91(3):493512Curhan JR, Neale MA, Ross L, Rosencranz-Engelmann J. 2008. Relational accommodation in negotiation:

effects of egalitarianism and gender on economic efciency and relational capital. Organ. Behav. Hum.Decis. Process. 107(2):192205

Deal JJ. 2000. Gender differences in the intentional use of information in competitive negotiations. SmallGroup Res. 31(6):70223

De Dreu CKW, Carnevale PJ. 2003. Motivational bases of information processing and strategy in conict andnegotiation. In Advances in Experimental Social Psychology, ed. MP Zanna, 35:23591. New York: Academic

De Dreu CKW, Weingart LR, Kwon S. 2000. Inuence of social motives on integrative negotiation: a meta-analytic review and test of two theories. J. Personal. Soc. Psychol. 78(5):889905

Dirks KT, Ferrin DL. 2001. The role of trust in organizational settings. Organ. Sci. 12(4):45067Drake LE. 1995. Negotiation styles in intercultural communication. Int. J. Con. Manag. 6(1):7290Eagly AH, Wood W. 1982. Inferred sex differences in status as a determinant of gender stereotypes about

social inuence. J. Personal. Soc. Psychol. 43(5):91528Elangovan AR, Shapiro DI. 1998. Betrayal of trust in organizations. Acad. Manag. Rev. 23(3):54766Emerson RM. 1972. Exchange theory: Part II. Exchange relations in networks. In Sociological Theories in

Progress, ed. J Berger, M Zelditch, B Anderson, 2:5887. Boston, MA: Houghton MifinFisher R, Ury W. 1981. Getting to Yes: Negotiating Agreement Without Giving In. Boston, MA: Houghton MifinFiske ST. 1988. Compare and contrast: Brewers dual-process model and Fiske et al.s continuum model.

In Advances in Social Cognition, Vol. 1: A Dual Model of Impression Formation, ed. TK Srull, RS Wyer,pp. 6576. Hillsdale, NJ: Erlbaum

Foa UG, Foa EB. 1975. Resource Theory of Social Exchange. Morristown, NJ: General Learning PressForgas JP. 1995. Mood and judgment: the affect infusion model (AIM). Psychol. Bull. 117(1):3966


Ann

u. R

ev. P

sych

ol. 2

010.

61:4

91-5

15. D

ownl

oade

d fro

m ar

journa

ls.an

nualr

eview

s.org

by U

nive

rsity

of P

enns

ylva

nia

on 0

7/31

/10.

For

per

sona

l use

onl

y.


Forgas JP. 1998. On feeling good and getting your way. J. Personal. Soc. Psychol. 74(3):56577Forgas JP, Cromer M. 2004. On being sad and evasive: affective inuences on verbal commu

2010 - thompson et.al - negotiation

Documents