Gender Negotiation

Organizat iona iona l B ehavior ehavior a nd H uma n D ecisi ecisio on P roc rocesses esses Vol. 87, No. 2, March, pp. 386– 409, 2002 doi:10.100 doi:10.1006/ 6/obhd.2001.297 obhd.2001.2979, 9, a va ila ble onlin e a t ht t p://w w w.idea libra ry.com ry.com on

R ever ev erss ing the G end er G a p in Neg otia otia tions : An Expl Explo o ra tio n o f S te re o type Reg ene ra tio n Laura J . Kra y E l l er er C ol ol l eg e of B u s i n es ess a n d P u b l i c A d m i n i s t r a t i on , U n i v er er s i t y o f A r i z on on a  

Ad a m D . G a l in in s k y U n i v er er s i t y of of U t a h  

a nd Leigh Thompson Kell Kell ogg Grad uat e Schoo Schooll of M anagem anagem ent, N or thw ester ter n U ni ver ver si ty 

We exam xamine how gender nder stereo tereotyp type es affe affect ct performa performance nce in mixed-g mixed-gender negotiati negotiations. ons. We We extend extend recent cent work demo demonstranstrating tha that stere tereo oty type pe activa ctivatio tion n lea leads to a male adv dva anta ntage and a com complem plementary ntary fem female disadva disadvanta ntag ge at the barga bargaiining ning table table (Kray, (Kray, T hom hompso pson, & Galinsky, Galinsky, 2001).I ). I n the pres present inves investigat tigation, ion, we rege regenerate nerate the stereo tereotyp type e of effective ffective neg negot otiiators ators by ass associ ociatating stereotypically feminine skills with negotiation success. In E xper xperiment iment 1, wo wom men performe performed d bette better r in mixedxed-g gender nder neg negotiations tions when when stereo tereoty typica picall lly y fem feminine ini netraits traitswe were re linke li nked d to tos succes uccessful neg negotiating, tiating, but but not not when when gender-ne nder-neutral utral traits were were linke li nked d to neg negotiation tiation succes uccess. Gende Gender r differen difference ces s were were mediate diated by the performa performance ncee expecta xpectations tionsand andg goals oals set by neg negot otiiato ators.I rs. I n E xperi xperi-ment 2, we rege regenerate nerated d the stereo tereotyp type e of effective ffective neg negot otiiato ators by linking li nking stereo tereoty typica picall lly y masculine or fem feminine ini netraits traitswith with neg negotiation ineffe neffective ctivene nes ss. Wome omen outpe outperforme rformed d men in mixedxed-g gender nder neg negotiations tiations when when ste stereo reoty typica picall lly y mas masculine culine traits were were linke li nked d to poo poor neg negotiation tiation perform performa ance, nce, but but men outpe utperforme rformed wo wom men The dat a reported were col collec lected ted w hile th e second second a ut hor wa s a post post doct doct oral fello fellow w a t North w estern U niversity. niversity. This This resea resea rch wa s support support ed by tw o grant s from from t he Nat iona iona l Science Science Founda tion (SES-0073682 awarded to Laura Kray and SES-9870892 awarded to Leigh Thompson). Address co correspondence rrespondence a nd reprint reprint requ ests to La ura K ra y, Depa rt ment of Ma na gement &P olic olicyy, U nivers ity of Arizona Arizona , P.O. B ox 21010 210108, 8, Tucson, Tucson, AZ 8572 857211-010 0108. 8. E -ma il: kra y@elle [email protected] r.ar izona .edu. Fax: (520) 621-4171. 0749-5978/02 $35.00  2002 Elsevier Science (USA) All right s reserved. reserved.

386 38 6

R E VE V E R S I N G TH E G E N D E R G AP

when when stereo tereotyp typically ically fem feminine ini ne traits were were linke li nked d to topo poo or neg negotiation perform performa ance. nce. I mplications plications fo for r stereo tereoty type pe threat threat theo theory and and  2002 E lsevier lsevier Science Science (USA) neg negot otiiati ations are discuss discussed.   200

387

R E VE V E R S I N G TH E G E N D E R G AP

387

when when stereo tereotyp typically ically fem feminine ini ne traits were were linke li nked d to topo poo or neg negotiation perform performa ance. nce. I mplications plications fo for r stereo tereoty type pe threat threat theo theory and and  2002 E lsevier lsevier Science Science (USA) neg negot otiiati ations are discuss discussed.   200

One widely held stereotype about women is that they are less effective at negotiating compared to men. For example, a typical negotiation scenario inv ol ol ve ve s b u y in i n g a n e w ca r . P op u la l a r w i sd s d om om s u g ge g e st s t s t h a t w om en e n br b r in in g a m a n w ith t hem to the dea de a lership lership so th a t t hey a re “ta ken ken serio seri ously” usl y” a nd given given a fair s h a k ee.. I n d ee e e d , on e a u d it it of n ew e w ca r d ea ea l er er s hip h i p s r ev e v ea ea l ed ed t ha h a t s a l es e s pe p e op l e quoted w omen signific significa nt ly higher prices prices tha n t hey did men w ho used used exa exa ct ly the same scripted bargaining strategies as the wome n (Ayres & Siegelman, 1995). B ut over ve r a nd a bove bove any bia bia s on the pa rt of th e dea de a ler ler, w omen ca c a rry a n a dditiona dditiona l burden burde n w ith th em into the dea de a lership lership,, w hich hich is the possi po ssibi bili lity ty th a t a n y t hi n g t he y s a y o r d o w i l l be i n t er p r e t e d i n l i gh t o f t h e st e r e o t y p e a b o u t w omen’s men’s inferio inferiorr negotia negotia ting a bility bility (Kr a y, Thompso hompson, n, & G a linsky linsky,, 2001). This bur den is t erm ed st ereoty pe t hr ea t (S (S t eele, eele, 199 1997; 7; S t eele & Aronson, 1995 1995)) a nd severa l investiga investiga t ions ions h a ve fo found th a t peo people’ ple’ss beha vior vior is a ffec ffect ed by th e mere activation of a stereotype or simply making a stereotype-relevant task dia gnostic of of a bility bility.. The q uestio ues tion n w e a ddress ddres s in th is a rt icle icle is is w hether th is burden burde n be lif l ifted, ted, or or even even t ra nsferred nsferred t o men, to improve improve w omen’s men’s perfo performa rma nc ncee a t the bar ga ining ining ta ble? ble? In a nsw ering ering t his question question w e extend extend theory theory a nd resea rese a rch rch in a number numbe r of importa importa nt w a ys. First , w e provide provide a s t rong test of S t eele eele’’s th eory eory of st ereot ereot ype t h r ea t b y d e m o n s t r at i n g t h a t m e r el y l i n k i n g t r ai t s t h a t a r e s te r e o t y p i c o f a group t o performa performa nce out out comes ca n produc producee st ereo ereoty pe t hrea t effe effeccts. We do t h i s b y m a n i p u l a ti n g w he t h e r p o s i t i v e o r n e g a t i ve o u t co m e s ar e g en e r a l l y conn ect ect ed to st ereot ereot ypica ypica lly feminin feminin e a nd ma sculine sculine tr a its. S eco econd, w e provide provide th e first first clea lea r evidence evidence th a t stereoty stereoty pe th rea t a ffec ffects perfo performa rma nce expe expeccta tions a nd t hese perfo perform rm a nce expec expectt a t ions ions media t e the observed observed decrement decrement s in perfo perforrma nce. nce. Fina lly, lly, we ha ve improved improved on on ea rlier resear ch t ha t confounded confounded positivit positivit y w ith mut ua lit lit y (a (a social social ca ca tegory tegory sha red by by ma le count erpar ts) w hen exp explo loring ring w h et e th h e r p os it it i v e s t e r eo eoty t y p es es ca n i mpr m pr ov ov e t h e n eg e g ot ot i a t i on on p e r for f orm m a nce n ce of women (Kray et al., 2001).

STEREOTYPES AND PERFORMANCE

A wide body of evidence suggests that stereotypes have a pernicious effect on the behavior and performance of the stereotyped. Even the United States Waterr hous ho use  e   recognized that stereotyping Supreme Court in H opki ns v. Pr i ce Wate can result in unequal outcomes in organizational se ttings (Fiske, 1993). In th a t par ticula ticula r ca ca se, a w oma n wa s denied denied par tn ership in a prestigious prestigious acco account ing firm, despite evidence of superior job performance, because her behavior did n ot conform nform t o gendergende r-ba ba sed expe expeccta tions. The The S upreme Court noted t ha t stereoty stereoty pes pes a re pa pa rt icula icula rly influentia influentia l wh en dec decisio isions a re based on on q ua lit lit a tive, idio idiosyncra tic dimensio dimensions ns (i. (i.e. e.,, interpersona interpersona l skill skill)) ra th er t ha n on qu a nt ita tive

388

KR AY, G ALI NS KY, AND TH OMP S ON

feat ures (i.e., a mount of business brought in t o a firm). Despit e t he court ’s ruling th a t negat ive stereoty pes ha d impa cted the t reat ment of an individua l within t his pa rt icula r orga niza t ion, overcoming t he impa ct of st ereot ypes is a formida ble ta sk. In a n interview cont ext , Word, Za nn a , a nd Cooper (1974) found th a t White interview ers tr eat ed Bla ck int erviewees w ith less immedia te nonverba l

388

KR AY, G ALI NS KY, AND TH OMP S ON

feat ures (i.e., a mount of business brought in t o a firm). Despit e t he court ’s ruling th a t negat ive stereoty pes ha d impa cted the t reat ment of an individua l within t his pa rt icula r orga niza t ion, overcoming t he impa ct of st ereot ypes is a formida ble ta sk. In a n interview cont ext , Word, Za nn a , a nd Cooper (1974) found th a t White interview ers tr eat ed Bla ck int erviewees w ith less immedia te nonverba l beha viors (e.g., less eye cont a ct a nd furt her int erpersona l dista nce) w hile also const ructing few er gra mma tically correct q uestions compa red to the tr eat ment of White interview ees. Ra th er t ha n overcoming t he impact of th is tr eat ment, B la ck a pplica nt s succumbed to it by responding w ith m ore gra mma t ica l errors and less eloquent and confident speech. Through interaction, stereotypic expect a ncies of perceivers can lea d t o stereot ypic responses of t a rget s. D evine (1989) has even suggested that, in the presence of a member of a stereotyped group, stereotype a ctivat ion is inevita ble a nd una voidable. St er eotype T h r eat 

The mere knowledge that negative stereotypes exist about one’s group can impair performance on job-relevant tasks. The performance of members of negatively stereotyped groups suffers when the task they are performing is relevant to the stereotype and believed to be diagn ostic of ability (Steele & Ar onson , 1995). S t eele (1997) h a s coined t h e phen omen on ster eotyp e th r eat , a n d it describes t he concern a per son feels a bout confirm ing , a s self-cha ra ct erist ic, a nega t ive stereot ype a bout one’s gr oup. St eele a nd Aronson describe stereoty pe t hr ea t a s th e follow ing predica ment : “The existence of such a st ereot ype mea ns t h a t a n y t h in g o n e d o e s o r a n y o f o n e ’s f e a tu r e s t ha t c o nf o r m t o i t m a ke t h e stereoty pe more pla usible a s a self-cha ra cteriza tion in t he eyes of oth ers, a nd perh a ps even in one’s ow n eyes.” Alth ough n ot empirica lly demonst ra t ed, St eele argued that the threat becomes reality because conc ern over confirming the stereotype produces anxiety, lowers expectations, and reduces performance a nd, th us, u nw itt ingly confirms t he st ereot ype (S teele, 1997). U nlike Word et a l. (1974), S t eele an d collea gues ha ve shown t ha t st ereot ypeconsist ent beha vior does not req uire a beha viora l-confirm a t ion process wh erein a perceiver elicits behavior from a ta rget. In fact, it is not n ecessa ry th a t th e ta rget person believe t he stereotype for his or her beha vior to be negat ively a ffected; t he most ca pable members of st ereot yped gr oups t end t o be the most a dversely a ffect ed in t heir perform a nce by st ereot ype th rea t (S t eele, 1997). All th a t a ppears to be necessa ry for stereoty pe th reat to emerge is the knowledge t h a t a s t e r eot y p e ex is t s a n d t h e ex pl ici t a r t i cu la t i on t h a t a p a r t icu la r t a s k i s d i a g n os t i c o f ab i l i t y . In m o s t c as e s , t a r ge t s m a y v e h e m e n t ly d e n y t h at t h e stereotype applies to them. The implications of this research span important social classifications, such as race and gender, and apply to a wide range of ability domains, such as individua l test s of ma t hema t ics (S pencer, S t eele, & Quinn , 1999), genera l int ellectual performance (Steele & Aronson, 1995), and athletics (Stone, Lynch, Sjomeling, & Darley, 1999). In addition to affecting these individual tasks, negative stereotypes also influence competitive, interactive negotiation tasks

R E VE R S I N G TH E G E N D E R G AP

389

(Kray et al., 2001). Kray et al. (2001) found that, in mixed-gender negotiating dya ds, simply la beling t he negotia t ion a s dia gnostic of a bility improved men’s a bility t o negotia te but hindered women’s performa nce a t t he bar ga ining ta ble. Extensions of this research on negatively stereotyped groups revealed that th e effect of stereotype a ctiva tion depends on th e va lence of th e stereotype,

R E VE R S I N G TH E G E N D E R G AP

389

(Kray et al., 2001). Kray et al. (2001) found that, in mixed-gender negotiating dya ds, simply la beling t he negotia t ion a s dia gnostic of a bility improved men’s a bility t o negotia te but hindered women’s performa nce a t t he bar ga ining ta ble. Extensions of this research on negatively stereotyped groups revealed that th e effect of stereotype a ctiva tion depends on th e va lence of th e stereotype, th e social gr oup to w hom it is a pplied, an d w heth er t he stereoty pe is relevan t to a pa rt icula r ta sk. Negat ive stereoty pes ca n a ffect members of nonst igma t ized gr oups (Arons on et a l., 1999; B row n & J osephs , 1999). Leyen s, D esert , Cr oizet , a nd D a rcis (2000) crea t ed a st ereot ypic expect a ncy by in st ructin g some people t ha t men (a gr oup t ypica lly exempt from stigm a t iza t ion) performed more poorly on a ffective t a sks. Whereas “th rea tened m en” did not perform w orse on oth er ta sks, they ma de significa nt ly more errors on t he a ffect ive discrimina tion ta sk. Consistent with Steele’s (1997) analysis of the causes of stereotype threat, L ey en s et a l . f ou n d t h a t en d or s em en t of t h e s t e r eot y p e d i d n o t m ed ia t e t h e effects, but th a t ident ifica tion w ith t he performa nce doma in under th reat w a s an important predictor of threat-induced decrements in performance. The framing of a task as relevant to positive versu s negative stereotypes differen t ia lly impa cts performa nce (S t one et a l., 1999). For exa mple, one st ereotype of White athletes is that they are high in “sp orts intelligence,” whereas A f r i c an - A m e r i c an a th l e t e s a r e g e n e ra l l y r e g a r d ed t o b e h i g h er in “ na t u r a l a t hletic ability.” The a ct ivat ion of t hese stereoty pes d ifferent ially a ffected th e performance of White and African-American athletes: when a golf task was deemed t o be dia gnostic of sport s int elligence, Whites did bett er t ha n w hen it w a s fra med a s dia gnostic of na tura l a th letic ability. The opposite w a s t rue of Africa n-America n a t hletes, w ho performed bett er wh en th e ta sk wa s diagn ostic of na tur a l a th letic ability th a n sports intelligence. Whereas stigmatized groups experience threat when t heir performance is interpreted a s diagn ostic of a n a bility th a t is releva nt to th e stereotype of th e stigmatized group, advantaged groups experience a performance boost. For inst a nce, Spencer et a l. (1999) found t ha t m en out performed w omen on a m a t h test only w hen t est-ta kers w ere told t ha t gender differences a ctua lly existed (w ith no mention of gender stereotypes). Moreover, men w ho w ere told th a t gender differences exist performed better than men who were told no gender differences exist. Our own investigations suggest that members of positively stereotyped groups (i.e., men in negotiations)experience improved performance when traditional stereotypes have been activated (K ray et al., 2001). In one experiment, activating stereotypes relevant to negotiation performance had a st ronger positive effect on men’s negot ia t ing performa nce t ha n it ha d a nega t ive effect on fema le performa nce. I n our empirica l extension of st ereoty pe th rea t (Kra y et a l., 2001), w e a lso demonstr a ted tha t performa nce expecta tions a nd goa ls mediat e t he effect of st ereoty pe a ctiva t ion on performa nce for positively s t e r eot y p ed g r ou p m em b er s (i .e. , m en ) a s w e ll a s n e ga t i v el y s t e r eot y p ed group m embers. Kr a y et a l. (2001) a lso found t ha t a ct ivat ing a positive ident ity t ha t is mut ua lly sha red by ma le a nd fema le negot iat ors essent ially levels t he play ing field. In a mixed-gender negotiation, linking negotiation performance to a shared

390

KR AY, G ALI NS KY, AND TH OMP S ON

ident ity (stud ent), ra th er t ha n t o a divisive ident ity (gender), equa lized performance across genders. Although this finding suggests positive identities can have a potent effect on both negotiators, Kray et al.’s investigation confounded pos it i v it y w i t h m u t u a l it y. Th u s t h e ir d a t a l ea v e u n cl ea r w h et h e r a pos it i v e stereotype improves performance for women when thei r male counterparts

390

KR AY, G ALI NS KY, AND TH OMP S ON

ident ity (stud ent), ra th er t ha n t o a divisive ident ity (gender), equa lized performance across genders. Although this finding suggests positive identities can have a potent effect on both negotiators, Kray et al.’s investigation confounded pos it i v it y w i t h m u t u a l it y. Th u s t h e ir d a t a l ea v e u n cl ea r w h et h e r a pos it i v e stereotype improves performance for women when thei r male counterparts d o not s h a r e i t .

St er eot ype Regen er at i on 

I n th i s i n v es t i g a t i o n, w e e x am i n e t h e e f f e c t t h a t r eg e n e r a t ed s t e r e o ty p e s h a v e on g r ou ps t h a t a r e t r a d it i on a l ly s t i gm a t i z ed a n d , h e n ce, d i sa d v a n t a g ed . B y st ereoty pe regenera tion, w e refer t o a process by w hich beha viors a nd t ra its associated with a group are modified or redefined. Specifically, stereotypical beha viors a nd tra its tha t a re assumed to be lia bilities a t a ta sk are tra nsformed into a ssets. For ma ny cultur a l groups, stereoty pe regenera tion is not possible. However, for many roles, such as manager or negotia tor, stereotypical traits a re often regenera ted. For exa mple, there is a tr end tow a rd t he “feminiza t ion” of management (Rudman & Glick, 1999), in which stereotypically feminine tr a its such a s inclusivity, emotiona l sensitivity a nd expression, a nd support iveness a re legitimized a nd va lued. This feminiza tion of ma na gement , par t icularly a t th e middle ma na gement level, could lea d females w ho conform to t he stereotype to have an advantage over their male cou nterparts. Given that the conditions t ha t produce t hrea t for one gr oup (e.g. w omen a nd Africa n America ns) ca n produce performa nce enha ncement for t he corresponding a dva nt a ged group (e.g., men and Caucasians), improving perform ance for women at the ba rga ining ta ble might simply be a ma t ter of str ength ening the perceived link betw een stereotypica lly feminine tra its a nd negotia tion success in th e minds of n egotia tors. Although ma ny individuals hold na ¨ ı ve theories tha t link stereoty pica lly ma sc u l i ne t r a i ts t o ne g o t i a t i o n s u c c es s ( K r ay e t al . , 2 00 1 ) , m an y o f t h e t ra i t s regarded by experts to be critica l t o negotia tion success a re in fact feminine in na ture. F or exa mple, th e t ra its tha t Ra iffa (1982) used to define effective negotia t ors a re composed of bot h st ereot ypica lly ma sculine a nd st ereot ypica lly feminine tr a its. In par ticula r, verba l a bility —effective commun ica t ion a nd listening skills—are equated with negotiation success (Raiffa, 1982). Likewise, being insightful an d emot iona lly expressive ca n be a dva nt a geous to th e negotia tor. We predict th a t a ctiva ting th ese t ra its w ill improve t he performa nce of w omen a t t he ba rga ining t a ble. More specifica lly, if cult ura l st ereot ypes cont ribu t e t o t h e ge n d e r g ap w i t n e s se d w h e n m e n a n d w o m e n n e g ot i a t e w i th e a c h oth er un der norma l circumsta nces (Kra y et a l., 2001), t hen regenera ting the link betw een stereotypica l t ra its a nd successful negotia ting should elimina te (or possibly r everse) t he gender ga p. Armed w ith a stereoty pe tha t implicitly portrays women as more advantaged than men in negotiations, we expect women to outperform men at the bargaining table.

R E VE R S I N G TH E G E N D E R G AP

391

H ow does ster eotype r egen er at i on i m pact per form an ce?  In t he origina l empir-

ica l demonstr a t ion of stereot ype t hr ea t , St eele a nd Aronson (1995) demonstr a t ed that conditions that produce stereotype threat effects also produce increased self-doubt. In a more recent investigation of the effect of racial stereotypes on athletic performance, Stone et al. (1999) determined that stereotypes impact

R E VE R S I N G TH E G E N D E R G AP

391

H ow does ster eotype r egen er at i on i m pact per form an ce?  In t he origina l empir-

ica l demonstr a t ion of stereot ype t hr ea t , St eele a nd Aronson (1995) demonstr a t ed that conditions that produce stereotype threat effects also produce increased self-doubt. In a more recent investigation of the effect of racial stereotypes on athletic performance, Stone et al. (1999) determined that stereotypes impact performance expectations as well as performance. Prior to engaging in each hole of a golf task, participants estimated how many strokes they would require to complete the hole. These expectations mirrored the outcome measures described a bove, a lthough St one et a l. noted th a t these expecta tions w ere likely a ffected by a ctua l performa nce in a n ea rlier sta ge of t he multista ged ta sk. C onsistent w ith t hese tw o progra ms of resear ch on st ereoty pe th rea t, w e predict t ha t simply being r eminded of a st ereot ype releva nt t o one’s socia l group will a ffect how w ell an individual expects to do even before the task begins. We expect stereotype activation to influence expectations, which in turn will impact performance. One reason why we expect the activation of stereotypes concerning an import a nt socia l ident ity, such a s gender, t o impa ct perform a nce is tha t st ereot ypes can alter one’s sense of self-worth and the goals an individual sets (Crocker & Ma jor, 1989; S t one et a l., 1999). Women gen era lly set low er goa ls for t hem selves than men in negotiations, and controlling for diffe rences in goals has been shown to elimina te negotia tion performa nce differences betw een men a nd women (St evens, B a vetta , & G ist, 1993). We expect stereoty pe regenera tion to impact th e performa nce goa ls tha t men a nd w omen set for th emselves in negotia tions such that women will expect more than they otherwise would and men would expect less than they otherwise would when the trad itional stereotype of a successful negotiator has been redefined. Negotiators with high outcome goals a re genera lly more persistent th a n negotia tors w ith low outcome goa ls a nd t his persistence translates into better performance (Bazerman, Magliozzi, & Neale, 1985; Huber & Neale, 1987; Neale & Bazerman, 1985). Thus, we expect the effect of stereot ype activa t ion on performa nce t o be media t ed by negot ia t or goals. To test t he hy pot heses out lined a bove, we first condu ct ed a pretest t o confirm that several of the traits regarded by experts to b e indicative of negotiation success ar e stereot ypica lly feminine. From th is pret est w e crea t ed a list of st ereotypica lly feminine t ra its a nd a list of gender-neutra l tr a its. We th en a ctivat ed these traits for men and women before engaging them in a negotiation task. After exposing them to the traits, participants ind icated their goals for the negotiation and then engaged in a mixed-gender negotiation. PRETEST

To det ermine th e extent to w hich key n egotia tion skills a re a ssociat ed w ith gender, we conducted a pretest t ha t a ssessed men a nd w omen’s na ¨ ıve t heories about the relationship between gender and negotiation skills. A total of 24 s t u d en t s on a col leg e ca m pu s (12 m a l es a n d 12 f em a l es ) w er e s ol ici t ed t o com pl et e a s h or t q u e st i on n a i r e a s ki n g t h e m t o a s s es s 13 t r a i t s d et e r m in ed by negotia tion expert s to be critica l for negotia tion success (Ra iffa, 1982). P a rt icipa nt s w ere a sked to eva luat e whether t hese tr a its w ere stereotypica lly

392

KR AY, G ALI NS KY, AND TH OMP S ON

feminine, stereoty pica lly ma sculine, or gender-neutr a l. S pecifica lly, par ticipa n t s r ea d , “ P l ea s e t a k e a cou ple of m in u t es t o t h in k a b ou t t h e cu lt u r a l stereotypes of men and women with regard to negotia tion skills. Please rate th e follow ing descriptions for how w ell t hey fit th e cultur a l stereot ypes a bout m e n a n d w o m e n . N o t e th a t th e s e c h a r a c t er i s t i c s m a y n o t re f l e c t y o u r o w n

392

KR AY, G ALI NS KY, AND TH OMP S ON

feminine, stereoty pica lly ma sculine, or gender-neutr a l. S pecifica lly, par ticipa n t s r ea d , “ P l ea s e t a k e a cou ple of m in u t es t o t h in k a b ou t t h e cu lt u r a l stereotypes of men and women with regard to negotia tion skills. Please rate th e follow ing descriptions for how w ell t hey fit th e cultur a l stereot ypes a bout m e n a n d w o m e n . N o t e th a t th e s e c h a r a c t er i s t i c s m a y n o t re f l e c t y o u r o w n persona l beliefs. Ra te t he a djectives a ccording t o how w ell th ey fit th e cultur a l stereotype whether or not you believe the stereotype to be true.” Participants then rated each trait on a 9-point scale, with higher numbers indicating more st ereot ypica lly ma sculine tr a its. As show n in Ta ble 1, key negot ia t or skills a re of t en l in k ed t o a s peci fi c gen d er. We i n t e r pr e t t r a i t s t h a t w er e r a t e d i n t h e upper t hird of th e sca le (5. 9) t o b e m a s cu lin e i n n a t u r e a n d t r a i t s t h a t w er e r a t ed i n t h e l o w er t h i r d o f t he s c a l e (3.1) to be feminine in na ture. Tra its t ha t fell in t he midr a nge of th e sca le (3.1– 5.9) w ere considered gend er-neut ra l. We then conducted paired sample t   t e s t s t o d e t e r m i n e t h a t t h i s m e t h o d o f dividing the categories resulted in distinct groupings. As expected, all three groups differed fr om ea ch oth er [ma sculine vs feminine, t (23)  9.79, p   .001; masculine vs neutral, t (23)  5.76, p   .001; feminine vs neut ra l, t (23)  7.53, p     .001]. We also examined whether the gender of the respondent affected r a t i ng s w i t h i n e a c h c a t e go r y . A s ex p e c t e d , t h e r a ti n g s b y m a l e a n d f e m a l e respondent s did n ot significa nt ly differ, F   1, n s . EXPERIMENT 1

H a ving esta blished th e exist ence of gender-specific a nd gend er-neut ra l tra its th a t a pply to negotia t ion success, w e were ready to proceed w ith our investigat ion of st ereot ype regenera t ion in a ct ua l fa ce-t o-fa ce negot ia t ions. We hypoth esized t ha t w omen w ould experience a performa nce boost relat ive to men w hen TABLE 1 Means and Standard Deviations of Evaluations of Successful Negotiator Traits According to Gender Stereotypes Negot ia t or t ra it

Mea n

S t a nda r d D evia t ion

Assert ive G ood pr oblem solver H igh rega rd for ow n int er est s K n ow ledgea ble R a t ion a l G ood judgm en t S en se of hum or P a t ien t P repa red Ver ba lly expressive G ood list enin g skills I n sight ful E m ot ion a l

7.3 7.0 6.3 6.0 5.6 5.3 5.0 4.0 3.7 2.8 2.5 2.3 1.9

1.6 1.5 1.5 1.7 1.8 1.9 .80 2.3 1.7 1.8 1.4 1.3 .90

Note.  Tra its w ere presented on a 9-point sca le, w ith higher va lues indica ting a str onger a ssocia-

tion w ith the ma le stereotype.

R E VE R S I N G TH E G E N D E R G AP

393

p o s i t iv e n e go t i a t i o n t r a i t s t h a t a r e p ar t o f t h e tr a d i t i o n a l f e m a l e s t e r e ot y p e (insightful, good listener, and verbally communicative) were associated with t h e n e g o t i a t o r s t e r e o t yp e . B e c au s e t h e n e g o t i a t i on w a s r e ga r d e d a s h i g h l y diagnostic of important abilities, men were expected to suffer from “stereotype t h r ea t ” a f t e r p o s i ti v e f e m a l e t r ai t s h a v e b e en ac t i v a t e d , me a n i n g t h a t th e i r

R E VE R S I N G TH E G E N D E R G AP

393

p o s i t iv e n e go t i a t i o n t r a i t s t h a t a r e p ar t o f t h e tr a d i t i o n a l f e m a l e s t e r e ot y p e (insightful, good listener, and verbally communicative) were associated with t h e n e g o t i a t o r s t e r e o t yp e . B e c au s e t h e n e g o t i a t i on w a s r e ga r d e d a s h i g h l y diagnostic of important abilities, men were expected to suffer from “stereotype t h r ea t ” a f t e r p o s i ti v e f e m a l e t r ai t s h a v e b e en ac t i v a t e d , me a n i n g t h a t th e i r performa nce w ould suffer. As a cont rol, w e included a condit ion in w hich tr a its relat ed to negotia tion performa nce, but una ssociat ed with ma sculine a nd feminine st ereot ypes, w ere activa t ed. We expect ed men to preva il und er th is genderneutral condition because describing a task as diagnostic of ability leads to stereotype threat for a negatively stereotyped group even when stereotypical tr a its ha ve not been a ctiva ted (Kr a y et a l., 2001; S teele & Aronson, 1995). With respect t o the un derlying process guiding t hese performa nce effect s, w e expect ed the a ct iva t ion of gender-relevan t versu s gender-neut ra l st ereot ypes to affect negotiators’ expectations and aspirations. I n particular, we expected women to set higher goals for themselves and to exp ect to do better when a f e m a l e s t er e o t y p e h a d b e e n a c ti v a t e d a n d l i n k e d t o s u c c e s sf u l n e g o t i a t i ng compa red t o w hen gend er-neut ra l tr a its w ere activa t ed. We expect ed th e opposite to occur for males; their own performance expe ctations and aspirations were hypothesized to suffer when stereotypically feminine traits were associated with negotiation success compared to gender-neutral traits. To test our st ereot ype regenera t ion h ypotheses, w e exa mined m ixed-gender dyads in which both negotiators experienced the same manipulation. In our first set of studies (Kray et al., 2001), we established that the effects of gender stereotype activation on negotiation performance ar e limited to the mixedgender ca se a nd most pronoun ced wh en bot h negotia t ors experience t he ma nipula t ion. Tha t is, exposing only one negot ia t or t o t he experiment a l ma nipula t ion impacted negotiation agreements, but the impact of stereotype activation was grea test w hen both negotia t ors experienced the ma nipula t ion. For this rea son, w e chose t o focus on t he set of circumst a nces in w hich ster eot ype regenera t ion processes a re most likely t o occur. M eth od 

The experiment involved tw o conditions, w ith fema le positive stereot ype a ctiva t ion a nd gender-neutr a l stereoty pe a ct ivat ion a s the tw o levels of the between groups factor. In both conditions, each negotiator in a dyad received th e sa me ma nipula t ion. Role a ssignments (buyer vs seller) w ere counterbalanced. Overview.

P a rt icipa nt s were 122 full-time a nd evening M.B .A. stud ents at a business school enrolled in a course in negotiations. Sixty-one dyads were formed, each w ith one female a nd one ma le pa rt icipa nt . The experiment w a s conducted during th e first w eek of a 10-w eek a ca demic term . Participants.

The experiment a l procedure consist ed of four pha ses. Firs t , pa rticipa nt s w ere each given a confidential packet of m a terials describing t he genera l na t ure of the negotia tion an d w ha t r ole th ey would pla y in th e exercise. Procedure.

394

KR AY, G ALI NS KY, AND TH OMP S ON

They w ere informed it w a s a n “honor code” violat ion t o excha nge a ny physica l r o l e i n f o r m a t i on w i t h an y o n e e l se , a l t h o ug h t h e y w e r e f re e t o s ay a n y t h i ng th ey w ished during t he negotia tion. The pa cket of ma teria ls a lso cont a ined th e key experiment a l ma nipula tions. B a sing our ma nipula tions on th ose of S teele a nd Aronson (1995) a nd Kr a y et

394

KR AY, G ALI NS KY, AND TH OMP S ON

They w ere informed it w a s a n “honor code” violat ion t o excha nge a ny physica l r o l e i n f o r m a t i on w i t h an y o n e e l se , a l t h o ug h t h e y w e r e f re e t o s ay a n y t h i ng th ey w ished during t he negotia tion. The pa cket of ma teria ls a lso cont a ined th e key experiment a l ma nipula tions. B a sing our ma nipula tions on th ose of S teele a nd Aronson (1995) a nd Kr a y et a l. (2001), pa rt icipa nt s in bot h condit ions w ere t old th a t t he negotia t ion exercise w a s highly diagn ostic of import a nt , ma na gerial n egotia tion a bilities based on individua l bar ga ining sty les. The tr a its t ha t w ere linked t o successful negotia tion performance varied across conditions. Based on results from our pretest, pa rt icipan t s in the fema le positive condit ion w ere t old, “Highly skilled negot ia t ors ha ve: (1) a keen a bilit y t o express th eir though t s verba lly; (2) good list ening skills; an d (3) insight int o t he ot her negotia t or ’s feelings.” In t he gender-neut ra l condition, participants were told, “Highly skilled negotiators are: (1) wellprepar ed; (2) a ble to ma int a in a sense of hu mor; a nd (3) open-mind ed.” In bot h conditions, part icipa nts w ere urged t o put fort h a strong effort on th e ta sk. After reading their role materials, participants completed a prenegotiation self-a ssessment of t heir expect a t ions for t he u pcoming negotia t ion. We expected, to th e extent th a t st ereoty pe a ctiva t ion a ffects relat ive performa nce expecta tions of individua ls, tw o compa rison groups w ould be relevan t —those par ticipa nt s in t he sa me role a nd one’s negotia ting part ner. To a ddress t hese tw o groups, pa rt icipan t s in a ll condit ions w ere a sked to indica t e how confident t hey w ere th a t t hey w ould perform w ell in t he negotia t ion, relat ive to oth er student s i n t h e cl a s s w i t h t h e s a m e r ol e, a n d w h a t por t i on of t h e “ pi e” t h e y ex pect e d to negotiate relative to their negotiating partner. For each of the preceding q uest ions , responses w ere on a n 11-point sca le, ra ng ing fr om 0 t o 100%. H igh er va lues indica te more confidence a nd bett er persona l performa nce. We a lso included a measure of expectations that was not directly tied to individuals’ relat ive performa nce: P a rt icipa nt s indica ted th eir goa l/ta rget level (i.e., sa le price) for t he single-issue n egot ia t ion. The next sta ge w a s the a ctua l fa ce-to-face negotia tion. We used th e sa me negotiation exercise used by Kray et al. (2001, Experiments 1 and 3), which w a s a st a nda rd negotia tion th a t involved th e potentia l purcha se of a pha rma ceuticals pla nt betw een a buyer a nd a seller. The t a sk a llow ed for a qu a nt ita t ive assessment of negotiation performance, as determine d by the sole issue of selling price. The bargaining zone spanned from $17.5 to $26 million. Partici p a n ts w e r e i n f o r m ed i m m e d i a te l y p r io r t o t h e ne g o t i a t i o n w h o th e y w o u ld negotia te w ith a nd t hen a ssigned a privat e meeting place to conduct th e negot ia t i on , w h i ch cou ld l a s t u p t o 30 m i n . Th e n e got i a t i on i n st r u ct i on s cl ea r l y indica ted t ha t th e objective of par ticipa nts w a s t o ma ximize their ow n profit, w heth er buyer or seller. The negot iat ion concluded w hen both par t ies cam e t o a mutual agreement on price or when time was called . U pon completion of th e t a sk, par ticipa nt s completed a n individua l a ssessment of the negotiation process and outcome. Participants were asked to estima te t heir performa nce, once a ga in, rela t ive to ot her stu dents in t he sa me role a nd t heir negot ia t ing par t ner, w ith t he sa me 11-point sca le from t he prenegotia t ion a ssessment (endpoint s: 0 a nd 100). Also on a n 11-point sca le, pa rt icipan t s

R E VE R S I N G TH E G E N D E R G AP

395

indica ted h ow equally ba lanced th e tw o roles w ere in terms of pow er (higher va lues indica ting great er persona l pow er relat ive t o par t ner). P a rt icipa nt s w ere a lso a sked to indica te h ow prepared a nd know ledgea ble th ey persona lly w ere regarding the negotiation and then assessed their n egotiating partner on the sa me m ea sur es. These items w ere presented on 7-point scales (endpoint s: not

395

R E VE R S I N G TH E G E N D E R G AP

indica ted h ow equally ba lanced th e tw o roles w ere in terms of pow er (higher va lues indica ting great er persona l pow er relat ive t o par t ner). P a rt icipa nt s w ere a lso a sked to indica te h ow prepared a nd know ledgea ble th ey persona lly w ere regarding the negotiation and then assessed their n egotiating partner on the sa me m ea sur es. These items w ere presented on 7-point scales (endpoint s: not at all, “extremely”). Results 

To exam ine th e effect of stereotype a ct iva t ion on individual expectations of negotiation performance, we conducted an ANOVA on each of the negotiation premeasures, with stereotype activation condition a nd n egot ia t or gender a s betw een-pa rt icipan t fa ct ors (see Ta ble 2 for a corr elation ma tr ix tha t includes a ll varia bles in t he study). Follow ing t he procedure developed by Kray et al. (2001), we controlled for role assignment by creating a s t a n d a r d iz ed Z   score, w ith higher values indica ting more a mbitious goa ls. The unit of a na lysis w a s th e individua l a s th ese mea sures w ere collect ed before a ny intera ction w ith one’s negotia ting part ner. We predicted t ha t th e ta rgets set by negotia tors w ould vary a s a function of both stereotype a ctiva tion a nd individua l gender. Specifica lly, w e predict ed th a t men w ould set h igher t a rgets in the gender-neutral condition than in the female positive condition, but the r e ver s e w ou ld b e t r u e f or w o m en . As s h ow n i n F i g. 1, t h e p a t t e r n of m ea n s confirmed our predictions: Wherea s t he sta nda rdized goa l w a s higher for men in t he gender-neut ra l condit ion (M   .43) t ha n in t he fema le positive condit ion (M   .17), w omen set h igher g oa ls for t hemselves in t he fema le positive condition (M   .42) th a n in t he gender-neutr a l condition (M   .17). This pattern was statistically confirmed with a significant Stereotype Activation  Gender interaction, [F (1, 121)    5.91, p     .05]. Relative performance expectations compa red to ot hers in the sa me role (M  g r a n d  54.96) or their negot ia t ing pa rt ner (M  g r a n d  53.88) were n ot a ffect ed by stereoty pe a ctiva tion (F s  1, n s ). Pr en egoti at i on m easu r es.

TABLE 2 Study 1: Correlations between Variables Va ria ble 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

P recon fidence Prepie Goal Sa le price P ostconfidence Postpie P ow er P repa red self P repa red opponen t Knowledge self K now ledge opponen t

1

2

3

— .70 .02 .02 .36 .23 .10 .06 .04 .33 .00

— .04 .05 .33 .28 .21 .05 .03 .32 .05

— .37 .06 .02 .15 .02 .05 .01 .03

4

5

6

— .07 — .02 .66 — .12   .45 .42 .07 .25   . 12 .01 .02 .06 .02 .22   .12 .07 .07 .13

Note.   Significa nt correlat ions ( p     .05) are in bold.

396

KR AY, G ALI NS KY, AND TH OMP S ON

7

8

— .15 .09 .08 .16

— .79 .45 .03

9

10

11

— .43 — .04 .09

—

396

KR AY, G ALI NS KY, AND TH OMP S ON

FIG. 1.   Experiment 1: Standardized negotiation goals by stereotype activation and gender. St a nda rdized goal is computed such tha t higher values indica te a more a mbitious goal, collapsing a cross role a ssignment.

To a n a l y ze per for m a n ce, w e f ir s t exa m in ed w hether role a ssignments (fema le buyer/ma le seller vs fema le seller/ma le buyer) influenced performa nce by condu ct ing a n a na lysis of va ria nce (ANOVA) on sa le price, w ith role as a betw een-groups fa ct or (see Ta ble 3 for descriptive data across experimental conditions). As expected, the effect of role on sale price w a s not st a tist ica lly significa nt [F (1, 59)  .62, n s ], so w e colla psed a cross role a ssignments by tr a nsforming negotia tion a greements into a sta nda rdized Z   score, with higher values indicating better perfor mance for men relative to w omen. To test our hypoth esis t ha t gender stereotype a ctiva tion a ffects negotiation outcomes, we then examined the standardized performance score across stereotype activation conditions with an ANOVA. In support of our hypothesis, men performed better in the gender-neutral condition (M    .46), N egoti ati on per for m ance.

TABLE 3 Study 1: Means and Standard Deviations of Dependent Variables by Stereotype Activation Condition and Role Assignments F em a le posit ive Va r i a b l e P r econ fiden ce P repie G oa l S a le pr ice P os tcon fiden ce P ost pie P ow er P repa red self P repa red opponent K n ow ledge self K now ledge oppon ent Note.  F



 fema le; M

F s el ler /M b u y er 55.56 51.48 22.75 22.19 52.80 55.20 47.41 4.33 4.78 3.30 3.67 

(14.23) (15.86) (3.00) (1.73) (19.04) (15.84) (15.59) (1.18) (1.09) (1.44) (1.49)

G en der-n eut ra l

F b u y er /M s el ler 55.16 55.67 21.95 21.33 56.21 54.48 50.81 4.65 4.87 4.19 4.42

(14.35) (13.57) (2.60) (1.70) (13.47) (14.29) (17.42) (1.38) (1.20) (1.42) (1.15)

F s el le r/M b u y er 53.41 52.00 21.09 20.80 55.77 53.17 51.22 4.39 4.44 3.07 5.51

(15.91) (13.63) (2.90) (1.59) (15.50) (15.72) (11.66) (1.26) (1.16) (1.56) (9.75)

F b u y er /M s el le r 50.87 52.73 22.78 22.57 54.29 52.86 49.13 4.74 4.78 3.39 3.86

(16.49) (11.62) (2.86) (2.03) (17.20) (19.78) (13.11) (1.32) (1.31) (1.75) (1.49)

 ma le.

R E VE R S I N G TH E G E N D E R G AP

397

R E VE R S I N G TH E G E N D E R G AP

397

FIG.2. Experiment 1: Sta nda rdized negotia tion performa nce by stereotype activat ion. Relat ive performance is computed such that higher values indicate a better outcome for men relative to women.

a nd w omen perform ed bet t er in th e fema le positive condit ion (M   .30), [F (1, 59)    10.38, p     .01]1 (see Fig. 2). To exa mine more closely th e source of t h i s e f f e c t , w e e xa m i n e d wh e t h e r t h e d i f fe r e n c e b e t w e e n m a l e a n d f e m al e performa nce w ithin each condition differed from zero w ith a one-sa mple t  test . I n th e gender-neutr a l condition, w e replica ted our previously esta blished finding t ha t ma les outperform fema les w hen t he negotia tion is considered t o be diagnostic of performa nce [t (31)   2.79, p     .01]. In the female positive condition, women performed significa nt ly bett er th a n t heir ma le count erpar ts, [t (28)  1.79, p   .05]. 2 Rel ati onshi p betw een pr enegoti ati on aspi r at i ons an d n egoti ati on per form ance.

We next sought to determine bett er t he rela tionship betw een perform a nce goa ls constructed before the negotiation and subsequent negotiation performance. Overall, aspirations and performance were significa ntly correlated (r     .37, p     .001), with higher aspirations associated with better performance. The rela tionship betw een t a rget a nd outcome w a s sta t istica lly significa nt for ma les (r   .55, p   .001) but not fema les (r   .26, n s ) in t he gender-neut ra l condit ion; th e opposite w a s tr ue in t he fema le positive condit ion —ta rget va lues of fema les (r   .45, p   .01) w ere significa nt ly r ela ted t o out comes, but not t hose of ma les (r   .16, n s ). We next explored w het her performa nce goa ls media t ed th e effect of ster eot ype a ctiva t ion on n egot ia t ion performa nce. To do so, w e computed a difference score of th e prenegotia tion a spira tion m ea sure so tha t it r epresent ed the goal differences betw een m en a nd w omen in t he dya d, w ith th e woma n’s score subtra cted from the man’s scores. According to Baron and Kenny (1986), the mediating va ria ble (perform a nce expect a t ions) mu st st ill predict t he out come w hile cont rolling for experimental condition and, finally, the effect of experimental condition must be reduced when controlling for the effect of the mediator. When both 1

The 2  2 ANOVA, w ith role a ssignment a nd st ereotype activa tion a s betw een-group factors, is a lso stat istically significa nt, [F (1, 61)   10.06, p    .002]. 2 Because we had directional hypotheses all of our one-sample t   t est w ere one-ta iled.

398

KR AY, G ALI NS KY, AND TH OMP S ON

perform a nce expecta tions a nd experiment a l condit ion w ere simult a neously entered into the regression equation, performance expectations continued to predict negotiated outcomes,      .39 [t (57)    4.70, p     .001], but the effect of experimenta l condition w a s reduced t o ma rgina l significa nce,    .43 [t (57)  1.96, p   .06]. U sing t he corr ect ed procedure origin a lly specified in K enny,

398

KR AY, G ALI NS KY, AND TH OMP S ON

perform a nce expecta tions a nd experiment a l condit ion w ere simult a neously entered into the regression equation, performance expectations continued to predict negotiated outcomes,      .39 [t (57)    4.70, p     .001], but the effect of experimenta l condition w a s reduced t o ma rgina l significa nce,    .43 [t (57)  1.96, p   .06]. U sing t he corr ect ed procedure origin a lly specified in K enny, Ka shy, a nd B olger (1998), w e next t ested w het her t his reduction w a s sufficient ly large to be significant, and indeed it was, (Z    2.10, p     .05). The impact of stereotype activation and diagnosticity on the agre ements reached between men a nd women w a s mediat ed by the negotia tors’ performa nce expecta tions. St ereoty pes can empow er a nd disempow er individua ls in t he n egotia tion context, depending on w hich st ereot ypica l tr a its a re linked t o positive performa nce. We crea ted a difference score represent ing t he ma le response minus t he fema le response for ea ch postn egotia tion mea sure t o exa mine how gender affected self-a ssessment s a nd negot iat ing par tn er a ssessm e n t s a c r o s s ex p e r i m e n ta l c o n d i ti o n s . T h e d y a d w a s t he u n i t of a n a l y s i s. Because the difference score was constructed such t hat female assessments w ere subt ra ct ed from ma le a ssessment s, eva lua t ions concern ing one’s negotia t i n g p a r t n er w er e e xpect e d t o b e p os it i v e t o t h e e xt e n t t h a t m a l es ev a l u at ed t h e i r fe m a l e p a r t ne r m o r e p o s i t i ve l y a n d n e g a t i ve t o t h e e x t e n t th a t m a l e s eva luat ed th eir fema le par tn er less positively. We h ypothesized tha t w omen would be evaluated more positively in the female positive condition compared to t he gender-neutra l condition, but th a t men w ould be evalua ted more positively in th e gender-neutra l condition. We tested our hypoth eses using the difference score a s th e dependent va ria ble in separ a t e ANOVAs, w ith st ereot ype activation condition as the between-groups factor. How prepared participants perceived their negotiating partner to have been in the negotiation depended on t heir gender a nd st ereoty pe activa tion condition [F (1, 59)  8.29, p   .01]. Whereas the difference in perceptions in the gender-neutral condition favored ma les (M   .69), th e difference in perceptions w ith in a dya d fa vored fema les in th e fema le positive condition (M   .34). We examined the simple effects of this analysis with one-sample t  t ests. The difference bet w een ma le an d fema le p e r ce p t i o n s w as g r e a te r t h a n z e r o i n t h e g en d e r - n e u t ra l c o n d i ti o n [t (31)  2.69, p     . 0 5] ; t h e d i f f e r e n ce i n p e rc e p t i o n s w a s m a r gi n a l i n t h e f e m al e positive condition [t (28)  1.38, p   .10]. No ot her post negotia t ion a ssessment effect s w ere significa nt . Postn egoti at i on m easur es.

Discussion 

The negotia t ion process ca n be cha ra ct erized a s a t hr ee-st a ged process: preparation, actual negotiation, and performance assessment (Thompson, 2001). The results of t his st udy suggest stereotype a ctiva tion a nd stereotype t hreat ma tt er in each sta ge of the negot iat ion process. First , the goa ls tha t individual negotia tors set w ere affect ed by t he a ctiva tion of stereoty pica lly feminine tr a its th a t ha d been linked t o positive negotia ting performa nce. Empow ered by t he stereot ype regenerat ion process, such t ha t stereot ypica lly feminine tra its w ere

R E VE R S I N G TH E G E N D E R G AP

399

positively rela t ed to negotia t ion performa nce, women set high er goa ls for t hems e l v es ; t h e o p p o s i t e wa s t r u e o f m e n , w h o s e g o a ls w e r e lo w e r e d d u e t o t h e linking of countermasculine traits to negotiation success. Consistent with the goals th ey set for t hemselves, w omen’s perform a nce w a s bet t er wh en stereot ypically feminine traits were activated relative to ge nder-neutral traits. Men’s

R E VE R S I N G TH E G E N D E R G AP

399

positively rela t ed to negotia t ion performa nce, women set high er goa ls for t hems e l v es ; t h e o p p o s i t e wa s t r u e o f m e n , w h o s e g o a ls w e r e lo w e r e d d u e t o t h e linking of countermasculine traits to negotiation success. Consistent with the goals th ey set for t hemselves, w omen’s perform a nce w a s bet t er wh en stereot ypically feminine traits were activated relative to ge nder-neutral traits. Men’s performa nce, in cont ra st, w orsened a fter t he fema le stereotype w a s a ctivat ed compa red t o t he gender-neutr a l tr a its. In a ddition, cont rolling for th e negotia t ors’ goals significa nt ly reduced t he effect of st ereot ype a ct iva t ion on negot ia t ed out comes. Follow ing th e negotia t ion, st ereot ype a ct iva t ion a ffect ed how n egot iator’s evaluated their negotiating partner. That is , men perceived women to b e b e t t e r p r e p a re d w h e n s t e r e ot y p i c a l l y f e m i n i n e t r a i ts h a d b e e n l i n k e d to negotia tion success tha n w hen gender-neutr a l tra its w ere linked to negotia tion success. Although it is un clea r w heth er t hese postn egotia t ion perceptions res u l t ed f r o m t h e s t er e o t y p e a c t i va t i o n i n t h e m i n d o f m en o r t he b eh a v i o r s o f w o m en , i t i s c l e a r th a t t h e a c t i v a t i on o f s t e re o t y p i c t ra i t s w i t h p o s i t i v e connota tions for w omen is import a nt for both sides of t he ba rga ining ta ble. This study expands our knowledge of how stereotypes influence behavior by demonstrating that even nonstigmatized groups can experience stereotype t h r ea t . M e n, w h o a r e tr a d i t i o n a ll y s e e n a s mo r e a d v a n t a g e d i n n eg o t i a t i o n s (Kr a y et a l., 2001), performed w orse in t he fema le positive condition t ha n th e g en d er -n e ut r a l b a s e lin e . J u s t li ke f r a m i ng a n a t h l et i c t a s k a s d ia g n os t i c of “ s p o r ts i n t e l li g e n c e ” v e r s u s “n a t u r a l a t h l e t i c a b i l i t y ” i m p a c t e d w h e t h e r t h e performance of White or Black athletes suffered (Stone et al., 1999), performa nce in a mixed-gender negotia tion cont ext depends on w hich stereot ypic tr a its a re seen a s relevan t a t t he time in wh ich th e negotia tion occurs. P erha ps most unique to this study is the evidence th a t even members of disa dva nt a ged groups can benefit from the activation of a regenerated stereotype—women’s performa nce improved th rough th e a ct ivat ion of stereot ypica lly feminine tra its with their explicit connection to positive negotiating performance. Multiple stereoty pic a tt ributes a re releva nt to most social groups a nd focusing on th e positives wh ile ignoring th e nega t ives a ppear s to be a n a dva nt a geous stra tegy. When stereotypically masculine traits are linked to successful negotiation outcomes, m en outper form w omen (K ra y et a l., 2001). The resu lt s of th e presen t i n ves t i ga t i on r ev ea l t h a t ev en w h en t h e t r a i t s t h a t a r e l in k ed t o n eg ot i a t i on p r o w e ss a r e g en d e r - n e u tr a l , m e n h a v e a n a d v a n t ag e o v e r w o m e n . A t fi r s t gla nce, t his finding a ppea rs cont ra dict ory t o our model of stereot ype a ct iva t ion, w hich w ould seem to predict t ha t describing gender-neut ra l t ra its a s predict ive of negotiation success levels the playing field. This might be true if not for the supposed diag nosticit y of th e negot ia t ion a cross bot h experiment a l condit ions — describing a t a sk a s dia gnostic of ability is enough t o produce st ereot ype t hr ea t effect s in disa dva nt a ged groups (Kr a y et a l., 2001; S teele & Aronson, 1995). Despite th e know ledge tha t gender-neutr a l tr a its a re import a nt for negot iat ion success, the knowledge that the task was diagnostic of ability might have dominated the cognitions of women, resulting in their relatively poor performa nce. On a positive note, th e current stud y demonst ra t es th a t linking stereotypically feminine traits to negotiation success, e ven in a highly diagnostic

400

KR AY, G ALI NS KY, AND TH OMP S ON

environment , is enough t o count era ct t he deleterious effect s of st ereot ype t hr ea t for t he disa dvan ta ged group. Stereotype activation affects negotiation performance through the goals negotiators set prior to arriving at the bargaining t able. Goals are predictive of performance across a wide variety of tasks (Locke, Frederick, Lee, & Bobko,

400

KR AY, G ALI NS KY, AND TH OMP S ON

environment , is enough t o count era ct t he deleterious effect s of st ereot ype t hr ea t for t he disa dvan ta ged group. Stereotype activation affects negotiation performance through the goals negotiators set prior to arriving at the bargaining t able. Goals are predictive of performance across a wide variety of tasks (Locke, Frederick, Lee, & Bobko, 1984; Locke & Latham, 1990), including negotiations (Stevens et al., 1993). This study exa mines one fa ctor t ha t a ffects th e goa ls tha t negotia tors set for th emselves—t he a ctiva t ion of stereoty pes releva nt t o th e ta sk. Although S tone et a l. (1999) showed evidence tha t th e a ctiva t ion of r a cial stereoty pes a ffect ed performance expectations, they also noted that expe ctations at one point in time w ere likely a ffected by a ct ua l performa nce on ea rlier holes of a multihole golf ga me. Thus, t he current study is th e first to our know ledge tha t directly mea sures performa nce expecta t ions in th e cont ext of st ereoty pe thr eat before comm encement of t he ta sk a nd t hen document s th e rela t ionsh ip betw een expectations and performance, uncontaminated by prior performance.

EXPERIMENT 2

T h e n a t u re o f s t e re o t y p e r e g e ne r a t i o n d e t e rm i n e s w h i c h g e n d e r h as t he a d v a n t a g e a t t h e b a r g a i ni n g t a b le. H ow ev er, s t e reot y p e t hr e a t i s m or e of t en linked to poor ra th er th a n successful performa nce expecta tions for a highly stereotyped group. African Americans are regarded as less intelligent than C a u ca s ia n s a r e t hou g h t t o b e s m a r t ; w o m en a r e t hou g h t t o b e w or s e at m a t h t h a n m en a r e t h ou g ht t o b e g ood a t m a t h . Th u s t h e r a c e a n d g en d er of t h e s t e r eo t y p e d gr o u p s ar e t h o ug h t t o b e mo r e d i a gn o s t i c t h a n th e y a re f o r t h e a dva nt a ged gr oups. I t is m ore su rprising (Miller, Ta ylor, & B uck, 1991) t o see a f e m a l e m a t h g e ni u s th a n t o se e a m a le i nc a p a b l e o f s o l v i n g r ud i m e n t a r y p r o b le m s . I n t h e n e x t s t u d y w e w a n t e d t o e xa m i n e ho w t h e r e g e n er a t i o n o f stereotypes associated with   poor   performance affects skill. Another goal of this experiment was to examine stereotype regeneration in th e cont ext of a multi-issue negotia tion w ith integrat ive potentia l. Although compet it ive skills a re requ ired for ma ximizin g one’s out come in a pur ely dist ributive ta sk, a s in E xperiment 1, a coopera tive problem solving a pproa ch th a t emphasizes information sharing and trade-offs tends to lead to higher joint ga in (Thompson, 2001). It is un clea r from previous resea rch w het her st ereoty pe a c t i va t i o n a f fe c t s t h e a b i l i t y t o c r e at e v a l u e a t t h e n e g o t i at i n g t a bl e . O n a n individual level, it is possible that stereotype regeneration would affect performance differently for purely distributive issues ve rsus integrative issues. If so, th e effect of stereoty pe regenera tion might not be evident in a n a ggrega t ed a greement from a m ult i-issue negot ia t ion. Anoth er benefit of exam ining n egot ia t i on p er f or m a n ce in a t a s k w i t h a g r eem en t v a l u es t h a t a r e n ot f ix ed i s t h a t it becomes possible t o exam ine th e perform a nce of ma le a nd fema le negotia t ors separ a t ely, t hu s cla rifying t he impa ct of st ereot ype regenera t ion on ea ch gender. Our fina l goa l for conducting t his experiment w a s t o provide a str onger t est for s ome of th e core a ssert ions of t he t heory of ster eot ype t hr ea t (S t eele, 1997)

R E VE R S I N G TH E G E N D E R G AP

401

and its empirical extensions (Kray et al., 2001; Stone et al., 1999). Namely we w a n t e d t o t e st t h e pr opos it i on t h a t t h e con t e xt a n d m a n n e r i n w h i ch g en d er is emphasized and the valence attached to each gender will determine performa nce. S tu dents just beginning a course in negotia tions comprise th e perfect sa mple to test th is theory. Although Kr a y et a l. found t ha t students a ssociat e

R E VE R S I N G TH E G E N D E R G AP

401

and its empirical extensions (Kray et al., 2001; Stone et al., 1999). Namely we w a n t e d t o t e st t h e pr opos it i on t h a t t h e con t e xt a n d m a n n e r i n w h i ch g en d er is emphasized and the valence attached to each gender will determine performa nce. S tu dents just beginning a course in negotia tions comprise th e perfect sa mple to test th is theory. Although Kr a y et a l. found t ha t students a ssociat e stereot ypica lly m a sculine t ra its w ith successful negotia ting performa nce, most s t u d en t s a r e t a k in g t h e cl a s s b eca u s e t h e y w a n t t o lea r n a b ou t t h e a t t r i bu t e s t h a t w i ll a l low t h e m t o b e s u cces s fu l r a t h e r t h a n i n ef fect i v e n eg ot i a t or s . I n fact, many students report on the first day of class great uncertainty as to w ha t tr a its cha ra cterize effective a nd ineffective negotia tors, w hich suggests the association between gender-linked traits and successful negotiating perform a n ce i s p a r t icu la r l y m a l lea b le a t t h i s p oi n t i n t i m e. I f S t e ele, S t on e e t a l ., a nd Kra y et a l. a re correct, th en relia ble effects should emerge rega rdless of whether stereotypic traits are linked to effective versus ineffective negotiators. We sought to test the stereotype regeneration proposition by linking stereoty pica lly ma sculine a nd feminine t ra its t o ineffect ive negotia t ing. We hypoth esized th a t w hen ma sculine t ra its a re linked t o ineffective negotia ting the outcom es of m en w o uld s u ff er in m ixed -g en d er d y a d s . I n con t r a s t , lin k in g st ereot ypica lly feminine t ra its to ineffect ive negotia t ing should result in w omen s t u m bli n g at t h e ba r g a i ni n g t a b le.

M eth od 

The experiment involved tw o condit ions, wit h ma le nega t ive st ereotype a ctiva tion a nd fema le nega tive stereotype a ctivat ion a s t he tw o levels of the between-groups factor. In both conditions, each negotiator in a dyad received the same manipulation. Role assignments were counterbalanced. Overview.

Our sa mple included 21 mixed-gender negotia t ing dya ds, for a t ot a l of 42 pa r t icipa n t s . P a r t icipa n t s w e r e M .B .A s t u d en t s en r olled in a negotiations course. Participants.

To extend our invest iga t ion t o a more complex negotia t ion, w e used a negot iat ion t a sk tha t concerned a n employment negotia tion in w hich a job ca ndidat e a nd a recruiter a tt empted t o negotia te severa l issues relevan t to both parties (i.e., salary, benefits, vacation time, and region of placement; Neale, 1997). The negotiation included eight issues in total. Preferences were induced in negotiators by assigning points to issues (greater points equaled more preferred). Negotia tors could earn betw een 8,400 points t o 13,200 points. Tw o issues w ere purely distr ibutive, mea ning th a t th e par ties’ preferences w ere in complete opposition. Tw o issues w ere compa t ible, mea ning t ha t the parties’ preferences were identical. The remaining issues formed two pairs of issues with integrat ive potentia l, mea ning th a t one part y cared more about issue A a nd t he oth er par t y car ed more a bout issue B . If both par ties conceded on th e issue they cared less a bout, both part ies could benefit in t erms of the number of points th ey earn ed. N egoti ati on task.

402

KR AY, G ALI NS KY, AND TH OMP S ON

As in E xperiment 1, pa rt icipa nt s in both conditions were told t ha t t he negotia tion exercise wa s highly diagn ostic of import a nt , ma na gerial negotia t ion a bilities ba sed on individual ba rga ining sty le. We var i e d w hi c h t r ai t s w e r e l i n k e d to   unsuccessful   negotiation performance across conditions. All par ticipan ts w ere t old t he follow ing: E x p er i m en t a l m a n i p u l a t i o n .

402

KR AY, G ALI NS KY, AND TH OMP S ON

As in E xperiment 1, pa rt icipa nt s in both conditions were told t ha t t he negotia tion exercise wa s highly diagn ostic of import a nt , ma na gerial negotia t ion a bilities ba sed on individual ba rga ining sty le. We var i e d w hi c h t r ai t s w e r e l i n k e d to   unsuccessful   negotiation performance across conditions. All par ticipan ts w ere t old t he follow ing: E x p er i m en t a l m a n i p u l a t i o n .

In preparing for this negotia tion, st udents a re often curious a bout wh at cha ra cteristics predict success and failure in complex negotiations that involve multiple issues. A recent series of studies examined the relationship between bargaining style and negotiation performance in multi-issue negotiations. It was determined that negotiators who display t he follow ing beha viors tend t o perform  worse (get poorer outcomes) th a n t hose who do not.

For pa rt icipan t s in th e ma le nega t ive st ereot ype condit ion, th e t ra its included (a ) High rega rd for persona l int erests, (b) Dependence on a ssertive beha viors to m ove negotia tion forw a rd, (c) Relia nce on ra tiona l a na lysis to understa nd the other negotiator’s preferences, and (d) Limited displays of emotion. For participants in the female negative stereotype cond ition, the traits included (a) Passive and reactive in expressing personal interests, (b) Dependence on ow n listening skills t o move negotia tion forwa rd, (c) Relian ce on int uitions to understand the other negotiator’s preferences, and (d) Clear displays of emotion. Results 

B efore a na lyzing th e negot iat ion performa nce of men a nd w omen a cross experiment a l conditions, we first determined w heth er it w a s a ppropria te to colla pse a cross r ole a ssignment by conducting a mixed ANOVA on perform a nce, including r ole as a w ith in-dya d fa ct or a nd r ole a ssignment (ma le seller/fema le buyer a nd ma le buyer/fema le seller) as a betw eendya d fa ctor. B eca use performa nce w a s not impa ct ed by role or r ole assignment (F s  1, n s ), w e proceeded to colla pse a cross role a ssignment by creat ing tw o va riables represent ing t he performa nce of w omen a nd men. To test our hypothesis that gender stereotype activation affects negotiation outcomes, we examined the performance of men and women across stereotype conditions w ith a mixed ANOVA. In support of our h ypothesis, w omen’s perform a nce w a s b et t er i n t h e m a l e n eg a t i ve con d it i on (M     6334) than the female negative cond it ion (M   4290), an d men’s performa nce w a s bett er in th e fema le nega t ive cond it ion (M   6330) th a n t he ma le nega tive condition (M   3864). (S ee F ig . 3). This pa t tern w a s confirmed with a sta tist ica lly significa nt t w o-w a y intera ction [F (1, 19)  7.84, p   .01]. Looking a t t he effect s w ith in ea ch gend er, st ereot ype regeneration affected both women [F (1, 21)    7.16, p     .05] and men [F (1, 21)  4.66, p   .05]. To det ermin e th e source of t his st ereot ype regenera t ion effect , w e next exa mined ea ch ty pe of negot ia t ion issu e sepa ra t ely. To do so, w e comput ed t w o scores for ma le a nd fema le negotia t ors: point s ea rn ed on distr ibut ive issues a nd point s ea rn ed on int egra t ive issues (point s w ere a lw a ys equ iva lent for bot h negot ia t ors on compa t ible issue). We t hen cond ucted a repea t ed-mea sur es ANOVA, including negotia tor gender a s a w ithin-group factor a nd stereoty pe condition a s a N egoti ati on per for m ance.

R E VE R S I N G TH E G E N D E R G AP

403

R E VE R S I N G TH E G E N D E R G AP

403

FIG. 3. Experiment 2: Negotia tion performa nce by negotia tor gender a nd st ereoty pe condition.

bet w een-gr oup fa ct or for ea ch of th ese ty pes of issues. The expected Negotia t or Gender  S tereot ype Condition intera ction wa s sta tist ica lly significa nt for the distr ibutive issues only, suggesting th a t it is in th e division of a fixed-pie th a t stereoty pe regenera tion h a s its great est effect [F (1, 19)  4.98, p   .05]. Although a similar pat tern emerged for integrat ive issues, t he effect w a s n ot statistically significant [F (1, 19)  2.72, p   .12]. This result suggests that the effect of stereoty pe regenerat ion is st ronger for distr ibutive issues th a n it is for issues w ith integrat ive potentia l. B e c a us e t h is n e g o ti a t i o n ha d i n t eg r a t i v e p o t e n t ia l , w e e x a m i n ed w h e t he r stereot ype regenera tion impacted joint ga in, or t he sum t ota l of points crea ted by th e t w o negot iat ors. To do so, w e conducted a n ANOVA on joint ga in, w ith stereotype condition and gender composition as between-groups factors. No significant effects emerged (F s  1, n s ) . Stereotype regeneration appears to a f f e ct t h e d i s t ri b u t i o n o f re s o u r c e s , b u t n o t t h e c r e a t i o n o f r e s o u r c e a t t h e b a r ga i n in g t a b le. DISCUSSION

The results of this study suggest that stereotype regeneration can occur by redefining traits that are normally associated with negotiation effectiveness such that an association is drawn between these traits and negotiation ineffectiveness. This study a lso extends our understa nding of stereotype t hreat a nd stereotype regeneration, suggesting that their grea test influence is with the claiming of distributive resources. That being said , this study also suggests th a t, a t lea st in negotia tions in w hich h ighly va lued issues a re distr ibutive in na tu re, th e benefit or detriment of st ereoty pe regenera tion is not can celed out by the inclusion of other issues in the negotiation. One result of the document ed stereoty pe regenera tion process is th a t members of groups who traditionally do not fare particularly well on the task (e.g., fema les) subsequ ent ly excel in th eir perform a nce. P resum a bly, stereot ype t h r ea t is a llev ia t e d b y r em in d in g m em b er s of t r a d it i on a l ly d is a d v a n t a g ed groups tha t tr a its th ey might norma lly a ssociat e with negotia tion success a nd

404

KR AY, G ALI NS KY, AND TH OMP S ON

which, according to stereotypes, they do not posses s, also have a downside. With out t he th rea t of confirm ing a n ega t ive st ereot ype a bout t heir socia l group, women negotiated more assertively than when stereotypically feminine skills w ere associat ed wit h n egot iat ion fa ilure. Likew ise, men’s a ssertiveness in t his d is t r i bu t i ve b a r ga i n in g s it u a t i on w a s p r es u m a bly h i n der e d w h e n m a s cu li n e

404

KR AY, G ALI NS KY, AND TH OMP S ON

which, according to stereotypes, they do not posses s, also have a downside. With out t he th rea t of confirm ing a n ega t ive st ereot ype a bout t heir socia l group, women negotiated more assertively than when stereotypically feminine skills w ere associat ed wit h n egot iat ion fa ilure. Likew ise, men’s a ssertiveness in t his d is t r i bu t i ve b a r ga i n in g s it u a t i on w a s p r es u m a bly h i n der e d w h e n m a s cu li n e tr a its w ere linked t o ineffective n egotia tors a s opposed to th e m ore common perception that feminine traits hinder negotiating ability (Kray et al., 2001). The reverse pattern occurred when stereotypically f eminine traits, the very ones th a t increa sed fema le perform a nce in Experiment 1, w ere linked t o ineffective, ra th er tha n effect ive, negotia t ing. For both men a nd w omen, linking tra its that were stereotypic of their gender to ineffectiv e negotiating performance hindered performance. We mentioned at the outset that we considered our participant population of M.B.A. student s embar king on a course in negotia t ions a s a par ticula rly a pt sample to examine the processes of stereotype regen eration. Many of these students presumably held naı¨ ve stereotypes about the traits associated with negotiation effectiveness, but these stereotypes were not so deeply ingrained and firmly held that they were not susceptible to change. The findings of this study ma ke a st rong case for t he a ssert ion t ha t th e negotia tion cont ext a ffects h ow a s s er t i v el y in d iv id u a l s b a r ga i n i n n eg ot i a t i on s . D e pen d in g on w h i ch th ought s a nd st ereoty pes (a nd t he va lence of th e stereoty pes) a re preva lent a t th e time tw o negotia tors a rrive a t t he ba rga ining ta ble, th e rela tive a dvan ta ge of men a nd w omen is determined. Al t h ou g h w e h a v e p r op os ed a pr oces s w h er e by b ot h m en a n d w om e n a r e a f f e ct e d b y t h e r e g e ne r a t i o n o f s t e re o t y p e s , i t i s i m po r t a n t t o n o t e t ha t w e ca nn ot r ea ch th is conclusion from th e curr ent st udy a lone. B eca use bot h negotia t ors experienced t he regenera t ion of stereot ypes before negot ia t ing, it rema ins unclea r w hether th e regenerat ion impa cted both genders or just one gender. To a ddr ess th is point , w e refer t o result s from Experiment 1 rega rdin g prenegot ia t ion goa ls t ha t w ere set prior to t he comm encement of t he negot ia t ion, w hich suggest stereotype regeneration affected both men a nd women. Drawing on t h e s e fi n di n gs , i t s eem s r ea s on a b ly cer t a i n t h a t t h e per f or m a n ce d a t a i n t h e current study are derived from similar processes in terms of goal setting and a ssert iveness a s inter vening forces. The pict ure t ha t emerg es from Experiment 1, the data from Spencer et al. (1999), and the data from Kray et al. (2001) in w hich only one m ember of m ixed-gender dya ds ha d stereotypes a ctiva ted suggests that stereotype threat produces dual effec ts of burden and benefit. The group with the purported advantage is benefited by stereotype activation a n d t h e g r ou p w i t h t h e pu r por t e d d is a d v a n t a g e is b ur d en ed b y s t er eot y p e activation. In competitive social interactions, the simultaneous burden and benefit of stereotype activation can produce surprisingly strong group differences on outcomes. GENERAL DISCUSSION

We emba rked upon t his resea rch a s a w a y of exploring a disconcert ing rea lity: Women tend to underperform in importa nt negotia tions compared to men

R E VE R S I N G TH E G E N D E R G AP

405

(S t uh lma cher & Wa lt ers, 1999; Wa lt ers, S t uh lma cher, &Meyer, 1998). B uildin g on current social psychologica l th eory rela ted to stereotype th reat (St eele, 1997), we reasoned that the association of traditional cultural gender stereoty pes w ith negotia tion prow ess m ight be a cont ributing fa ctor t o this gender ga p. We hypoth esized tha t the best w a y to countera ct the impact of nega tive

R E VE R S I N G TH E G E N D E R G AP

405

(S t uh lma cher & Wa lt ers, 1999; Wa lt ers, S t uh lma cher, &Meyer, 1998). B uildin g on current social psychologica l th eory rela ted to stereotype th reat (St eele, 1997), we reasoned that the association of traditional cultural gender stereoty pes w ith negotia tion prow ess m ight be a cont ributing fa ctor t o this gender ga p. We hypoth esized tha t the best w a y to countera ct the impact of nega tive st ereot ypes on women wa s t o “fight fire w ith fire,” or ma ke sa lient t hose negotia tion skills that are stereotypically associated with women. The results reported in this ar ticle suggest tha t, indeed, a rming disadva nt a ged group members w ith a set of cognitions that encourages them to set challenging goals is an effective tool for promoting success at the bargaining table. I n E x p e ri m e n t 1 , w e a t t e m p t e d t o r e g e ne r a t e t h e s t e r eo t y p e o f ef f e c t i v e negotiators such that it included traits that are feminine in nature. In so doing, we witnessed the improved performance of wom en relative to men in distributive negotiations. Armed with the cognition s that some of the traits stereotypically associated with women are predictive of negotiation success, women set higher goals for themselves and negotiate d more forcefully than w hen t he tr a its a ssociat ed wit h negotiat ion success were free of gender a ssocia t ions. As in previous resea rch (Kr a y et a l., 2001), simply ma king a negotia t ion dia gnostic of a bilit y, independent of stereoty pe a ct iva t ion, increa sed th e performa nce expect a t ions a nd out comes of men a nd d ecrea sed th e perform a nce expect a t ions a nd out comes of w omen. P erceiving a n egot ia t ion a s diag nostic of a bilit y ha s t hese effects beca use both men a nd w omen believe (1) t ha t ma les ha ve an a d v a n t a g e a t t h e ba r g a i ni n g t a bl e a n d (2) t h a t t h i s a d v a n t a g e s t e m s fr om t h e traits that males are presumed to possess (Kray et al., 2001). The fact that men out performed w omen even wh en the a ctiva t ed tr a its w ere gender-neutr a l suggests th a t w omen n eed to be part icularly mindful of t he positive qua lities th ey purport edly (a nd stereoty pica lly) possess if t hey a re t o claim th e grea test possible am oun t of resources. To impa ct t he goa ls a nd performa nce of w omen, it is not enough to regenerate the stereotype of negotiators by rendering gender irreleva nt . In E xperiment 2, w e explored a noth er ma nner in w hich stereoty pes can be regenerated. Instead of focusing on how stereotypic ally feminine traits are r e l a te d t o n e g o t ia t i o n e f f e c t i v en e s s , a s i n E x p er i m e n t 1 , w e e x a mi n e d th e link between stereotypically masculine and feminine traits and negotiation ineffectiveness. When stereoty pica lly feminine t ra its w ere a ssocia ted w ith negotia t ion failure, men out performed their fema le count erpa rt s; w hen stereotypically masculine traits were associated with negoti ation failure, women outperformed their male counterparts. The findings of this study suggest that stereot ypes ca n be effect ively regenera ted by reminding negotia tors th a t t ra its often thought to be predictive of success can in fact be detrimental to performa nce. In combina tion wit h oth er studies in th is st rea m of resea rch, it a ppear s that linking stereotypical traits to both positive consequences and negative consequences a ffect s goa l set ting a nd ba rga ining beha vior in a simila r fa shion. Note th a t n o ment ion of gender w a s ma de in the experiment a l ma nipula t ions of t h e t w o ex per i m en t s . Th e m a p pi n g of t h e t r a i t s t o g en d er w a s d o n e a t a n implicit level (G reenw a ld &B a na ji, 1995). Tha t is, t ra its t ha t w ere st ereot ypica l

406

KR AY, G ALI NS KY, AND TH OMP S ON

of men a nd w omen w ere linked t o effective a nd ineffective negot iat ors, but th e link to gender w a s done implicitly a s no informa tion a bout social ca tegories w a s p r o v i d ed . Pr e v i o u s r e s e ar c h s u g g e s t s t h at th e e f f e c t s o b s e r ve d i n t h is a rt icle depend on th e implicit na t ure of the a ctiva t ion of t he st ereot ypes. The implicit prim ing of know ledge str uctur es produces a ssimila t ion effect s, or judg-

406

KR AY, G ALI NS KY, AND TH OMP S ON

of men a nd w omen w ere linked t o effective a nd ineffective negot iat ors, but th e link to gender w a s done implicitly a s no informa tion a bout social ca tegories w a s p r o v i d ed . Pr e v i o u s r e s e ar c h s u g g e s t s t h at th e e f f e c t s o b s e r ve d i n t h is a rt icle depend on th e implicit na t ure of the a ctiva t ion of t he st ereot ypes. The implicit prim ing of know ledge str uctur es produces a ssimila t ion effect s, or judgments a nd behaviors, t ha t become more consistent w ith th e a ctiva ted know ledge structure th a n th ey w ould oth erwise (Higgins, Rholes, & J ones, 1977; Kr a y et a l., 2001; Moskow itz & S kurn ik, 1999). When people a re bla t a nt ly a nd explicitly primed, however, they perceive the primed construct to be a biasing influence, and to correct for this undesired influe nce on thought and deed, t h e y o f t e n r e ac t a g a i ns t t h i s b i a s i n g i n f l u e n c e ( M a r t i n , 1 9 8 6 ) . K r a y e t a l . ma nipula t ed wh ether linking stereoty pica lly m a sculine tra its t o effective negot ia t ing performa nce w a s done implicitly (w ith out m ent ioning gender) or explicitly (mentioning th a t th e t ra its differ by gender). When explicitly primed, w omen rea cted a ga inst t his imposed constr a int a nd outperformed t heir ma le count erpar ts. When w omen w ere explicitly t old t ha t a social cat egory to w hich they belong would hinder their ability to succeed, they dissociated from the tr a ditiona l fema le stereoty pe a nd enga ged in count erstereot ypic beha viors t ha t d e f i e d t h e st e r e o t y p e. F r o m o u r r es e a r c h w e h a ve f o u n d t h a t th e e f f ec t o f stereotypes on negotiated outcomes depends on four different variables: (1) whether the negotiation is diagnostic of ability, ( 2) the traits (masculine or f em i n in e) t h a t a r e a ct i v a te d , (3) t h e li n ki n g of t h e s e t r a i t s t o ef fect i v e a n d ineffective negotia t ion, a nd (4) w heth er th e tra its a re a ctiva ted implicitly or explicitly. Our research extends theory and research in a numbe r of important ways. Firs t , we provide a st rong t est of S t eele’s th eory of st ereot ype thr ea t by demonstr a ting tha t merely linking tr a its tha t ar e stereoty pic of a group to performa nce out comes can produce stereoty pe th rea t effect s. We do th is by ma nipula t ing whether positive or negative outcomes are generally connected to stereotypica lly feminine an d ma sculine tra its. We demonstr a te t he ma lleability of stereoty pes a nd th e ea se w ith w hich a stereoty pe a dvan ta ge ca n become a disadva nt a ge, dependin g on how it is fra med. Second, w e provide the first clea r evidence that this linkage affects performance expectations, ones unfettered by prior performa nce, a nd th a t these performa nce expecta tions media te th e observed decrements in performa nce. Fina lly, in demonst ra ting th a t connecting stereotypically feminine traits to negotiator success imp roves the performance of w omen, w e ha ve improved on ea rlier r esea rch th a t confounded positivity w ith mut ua lity in a ssessing th e impa ct of positive stereotypes on negotia t ion performance (Kray et al., 2001). Our resear ch suggest s severa l directions for fut ure research. First , th e ma in focus of this investigation has been on negotiation outcomes. We examined negotia tor goa ls a s one wa y of understa nding how a greement s differ a ccording to the activation of various stereotypes. More work is needed in this regard th ough in t erms of obta ining precise beha viora l measur es ba sed on a negotia t ion process a na lysis (Moore, Ku rt zberg , Thompson, & Morris , 1999). For exa mple, documenting the frequency of key negotiation actions, such as offers,

R E VE R S I N G TH E G E N D E R G AP

407

persua sion a tt empts, mention of goa ls, informa tion sha ring, informa t ion seeking, references to the relationship between negotiators, and procedural statement s, w ill increase our und erst a ndin g of how precisely th e gender ga p is ma de t o a ppea r a n d d is a p pea r t h r ou g h t h e a ct i va t ion of s t er eot y pes a n d t h e ir regeneration.

R E VE R S I N G TH E G E N D E R G AP

407

persua sion a tt empts, mention of goa ls, informa tion sha ring, informa t ion seeking, references to the relationship between negotiators, and procedural statement s, w ill increase our und erst a ndin g of how precisely th e gender ga p is ma de t o a ppea r a n d d is a p pea r t h r ou g h t h e a ct i va t ion of s t er eot y pes a n d t h e ir regeneration. Our r esea rch ha s focused on expect a t ions a nd goa ls as int ervening processes a ffecting th e link betw een stereotype a ctivat ion a nd performa nce. We h a ve documented that the goals of negatively stereotyped groups are less assertive th a n t he goa ls of positively st ereoty ped groups. Fut ure resea rch th a t exa mines oth er mediat ing factors might be worth w hile, a nd understa nding w ha t psychological factors influence goals is a worthwhile endeavor. More specifically, the extent to which the activation of stereotypes affec ts self-doubt, confidence, identification with the task, and perceived power i s undetermined. Each of these factors might mediate the effect of stereotyp e activation on goals and opening offers, w hich subsequent ly a ffect performa nce. B y broa dening t he scope of the investigation, a richer understanding of the impact of stereotypes on behavior will be obtained. E xa mining the effect s of stereoty pes tha t a re mutua lly sha red by negotia tors is also a w ort hw hile endea vor. As ment ioned a bove, one motiva t ion for exam ining st ereot ype regenera t ion w a s to disent a ngle th e effect of positive stereotypes from the effect of mutually shared stereotypes on n egotiation performance (Kray et al., 2001). When male and female negotiators were reminded that individua ls in competit ive, a ca demic environment s tend t o perform very w ell in negotia tions—rega rdless of gender —the gender ga p wa s significa nt ly reduced compared to a baseline condition. E xploring how mutua l stereotypes, both negative and positive ones, affect mixed-gender negotiations will allow us to a ddress issues surrounding negot iat or perceptions relat ed to pow er a nd st a tu s differentials that impact behavior at the bargainin g table. Finally, another important future direction for this research concerns exp a n d i ng t h e t y p e s o f s t er e o t y p e s t h a t a r e e xa m i n e d . O u r re s e a r c h s u g g e s ts that stereotypes about ascribed and thus stable social categories (i.e., gender and race) are somewhat mutable and, depending on how a gender-linked trait is interpreted, it can be perceived a s a help or a hindra nce to t he individual. Wha t our resea rch ha s n ot yet exa mined is w hether t he effects of st ereoty pe a ct iva t ion on chosen a nd t hus less st a ble socia l ca t egories, such a s one’s occupa tion or role in a negotiation, operate in a similar manner. It may be the case th a t disidentifying from a nega tive st ereot ype a bout one’s chosen socia l group is accomplished with little effort, and so these negative stereotypes have less impact on beha vior. Although it rema ins a n empirica l quest ion, w e specula te, on t he basis of th e wide ra nge of cont exts in w hich st ereoty pe th rea t ha s been shown to occur, that any situational variable that lowers one’s expectations and creates self-doubt—regardless of the stability of the category membership—should produce performa nce effects th a t mirror th ose a lrea dy documented in th e cont ext of st ereotype t hreat resea rch. To begin t o a nsw er t his qu estion, w e a re curr ent ly exa minin g how t he implicit a ct iva t ion of st ereot ypes

408

KR AY, G ALI NS KY, AND TH OMP S ON

a b ou t pow er b et w een b u y er s a n d s el ler s a f fect s n eg ot i a t i on per f or m a n ce (G a linsky, Thompson, & K ra y, 2001). Conclusion 

In the current studies we have provided a strong te st of Steele’s theory of

408

KR AY, G ALI NS KY, AND TH OMP S ON

a b ou t pow er b et w een b u y er s a n d s el ler s a f fect s n eg ot i a t i on per f or m a n ce (G a linsky, Thompson, & K ra y, 2001). Conclusion 

In the current studies we have provided a strong te st of Steele’s theory of s t e r eo t y p e t h r e a t a n d w e ha v e b e g u n t o an s w e r t h e q u e s t i o n o f wh a t ca u s e s gender differences in negotia t ions. I n bot h experiment s, th e cont ext a nd m a nner in w hich gender w a s empha sized and th e valence a tt a ched to ea ch gender determined performance in mixed-gender dyads. The s ituation, rather than the person, appears to be the primary determinant of negotiator performance. This r esearch suggests t ha t more t ha n one plausible stereoty pe exists a bout effective negotia tors, a nd the sha ping of th is st ereoty pe ha s pow erful effects on mixed-gender negotiations. Stereotypically masculine and feminine traits can be revalued and the stereotype of an effective negotiator regenerated. It seems clear that, regardless of gender, recognizing one’s strengths and the w e a k n es s e s o f o n e ’s n e go t i a t i n g p a r t n er i s a n i m p o r t a n t s t e p i n a c h i e vi n g success in the negotia tion a rena. REFERENCES Aronson, J ., Lu stina , M. J ., G ood, C ., Keough, K ., S teele, C . M., & B rown, J . (1999). When wh ite men ca n’t do ma th : Necessa ry a nd sufficient fa ctors in stereoty pe th reat .   Journal of Experimental  Social Psychology, 35, 29– 46. Ayres, I., & Siegelma n, P. (1995). Ra ce a nd gender discriminat ion in bar ga ining for a new ca r. T h e  Am er i can Economi c Review, 85, 304– 321. B a ron, R. M., & Kenn y, D . A. (1986). The modera tor-media t or va ria ble distinction in socia l psychological research: Conceptual, strategic, and statis tical considerations.   J o u r n a l o f P e r s o n a l i t y   and Social Psychology, 51, 1173– 1182. B a zerma n, M. H., Ma gliozzi, T., & Neale, M. A. (1985). I nt egrat ive bar ga ining in a competitive m a r k e t .   Or gani zati onal Behavi or and H um an Decision Pr ocesses, 35, 294– 343. B rown , R. P., &J osephs, R. A. (1999). A burden of proof : St ereoty pe releva nce a nd gend er differences in ma th performa nce.   J our nal of Per sonal it y & Soci al Psychology, 76, 246– 257. Crocker, J ., & Major, B. (1989). Social stigma and self-esteem: The self-protective properties of s t i g m a .   Psychological Review, 96, 608– 630. Devine, P. G . (1989). St ereotypes a nd prejudice: Their a utomat ic a nd controlled components. Journal of Personality and Social Psychology, 56, 5– 18. Fiske, S. (1993). Controlling other people.   American Psychologist, 48, 621– 628. G a linsky, A., Thompson, L., & Kr a y, L. J . (2001). Ta king st ereoty pes out of st ereoty pe th rea t: The threat of power derived from negotiator roles. Unpublished manuscript. Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Attitudes, self-esteem, and stereotypes.   Psychological Review,  102, 4– 27. H iggin s, E . T., Rholes, W. S., &J ones, C. R. (1977). Ca t egory a ccessibilit y a nd impr ession forma t ion. Journal of Experimental Social Psychology, 13, 141– 154. Huber, V. L., & Neale, M. A. (1987). Effects of self- and competitor goals on performance in an interdependent ba rgaining t a sk.   Journal of Applied Psychology, 72, 197– 203. Kenny, D. A., Ka shy, D . A., & Bolger, N. (1998). D a ta a na lysis in socia l psychology. I n D. G ilbert, S. Fiske, & G . Lindzey (E ds.),   The handbook of social psychology   (Vol. 1, 4th ed., pp. 233– 265). B oston, MA: McGra w – Hill.

R E VE R S I N G TH E G E N D E R G AP

409

Kr a y, L. J ., Thompson, L., &G a linsky, A. (in press). B a tt le of th e sexes: Gend er st ereot ype activa t ion in negotiations.   Journal of Personality and Social Psychology,  2001 80, 942– 958. Leyens, J . P., D esert, M., Cr oizet, J . C., & D a rcis, C. (2000). St ereoty pe t hrea t: Are low er sta t us a nd history of stigmatization preconditions of stereotype threat?   Personality and Social Psychology  B u l l e t i n , 26, 1189– 1199. Locke, E. A., Frederick, E., Lee, C., & Bobko, P. (1984). Effect of self-efficacy, goals, and task

R E VE R S I N G TH E G E N D E R G AP

409

Kr a y, L. J ., Thompson, L., &G a linsky, A. (in press). B a tt le of th e sexes: Gend er st ereot ype activa t ion in negotiations.   Journal of Personality and Social Psychology,  2001 80, 942– 958. Leyens, J . P., D esert, M., Cr oizet, J . C., & D a rcis, C. (2000). St ereoty pe t hrea t: Are low er sta t us a nd history of stigmatization preconditions of stereotype threat?   Personality and Social Psychology  B u l l e t i n , 26, 1189– 1199. Locke, E. A., Frederick, E., Lee, C., & Bobko, P. (1984). Effect of self-efficacy, goals, and task stra tegies on ta sk performa nce.   Journal of Applied Psychology, 69, 241– 251. Locke, E . A., & La th a m, G . P. (1990).   A theory of goal setting and task performance. E nglewood Cliffs, NJ : Prentice Hall. Ma rt in, L. L . (1986). Set /reset : U se/disu se of concepts in impr ession forma t ion.  J our nal of Per sonal -  ity and Social Psychology, 51, 493– 504. Miller, D . T., Ta ylor, B ., & Bu ck, M. (1991). G ender ga ps: Who needs t o be explain ed?   J o u r n a l of     Personality and Social Psychology, 61, 5– 12. Moore, D . A., Ku rt zberg, T. R ., Thompson, L . L ., & Morris, M. W. (1999). L ong a nd short rout es t o success in electronically mediat ed negotia tions: G roup a ffiliat ions an d good vibra t ions.  O r g a n i -   zational Behavior and Human Decision Processes, 77, 22– 43. Moskowitz, G. B., &Skurnik, I. (1999). The differing nature of contrast effects following exemplar versus trait primes. Standard of comparison versus motivated correction processes.   J o u r n a l of     Personality and Social Psychology, 76, 911– 927. Nea le, M. A. (1997). New recruit. In   Dispute Resolution Research Center Exercises . E v a n s t on , I L : Northwestern University. Neale, M. A., & B a zerma n, M. H. (1985). The effect of externa lly set goals on rea ching int egrat ive agreements in competitive markets.   Journal of Occupational Behavior, 6, 19– 32. Raiffa, H. (1982).   T he ar t and science of negoti ati on . C a mbridge, MA: B elknap. Rudman, L. A., & Glick, P. (1999). Feminized management and backlash toward agentic women: The hidden cost s to women of a kinder, gent ler ima ge of middle ma na gers.   Journal of Personality  and Social Psychology, 77, 1004– 1010. Spencer, S . J ., St eele, C . M., & Quinn, D . M. (1999). S tereoty pe thr eat a nd w omen’s m a th performance.   Journal of Experimental Social Psychology, 35, 4– 28. Steele, C. M. (1997). A threat in the air: How stereotypes shape intellectual identity and performance.   American Psychologist, 52, 613– 629. St eele, C . M., & Aronson, J . (1995). S tereoty pe th reat a nd th e int ellectu a l t est performan ce of African Americans.   Journal of Personality and Social Psychology, 69, 797– 811. St evens, C. K ., Ba vetta , A. G ., & G ist, M. E . (1993). G ender differences in th e acquisition of sa lar y negotiation skills: The role of goals, self-efficacy, and perceived control.   J ou r n a l of A p p l i ed   Psychology, 78, 723– 735. St one, J ., Lynch, C ., S jomeling, M. & Da rley, J . M. (1999). S tereoty pe th reat effects on B lack a nd White a th letic performa nce.   Journal of Personality and Social Psychology, 77, 1213– 1227. St uhlm a cher, A. F., & Wa lters, A. E . (1999). G ender differences in negotia tion out come: A meta a n a l y s i s .   Personnel Psychology, 52, 653– 677. Thompson, L. (2001).   T h e m i n d a n d h e a r t o f t h e n e g o t i a t o r   (2nd ed.). Upper Saddle River, NJ : Prentice Hall. Wa lters , A. E ., St uhlm a cher, A. F., & Meyer, L. L. (1998). G ender a nd n egot ia t or competitiveness: A meta -an a lysis.   Or gani zati onal Behavi or and H um an Decision Pr ocesses, 76, 1– 29. Word, C. O., Zanna, M. P, &Cooper, J . (1974). The nonverbal mediation of self-fulfilling prophecies in interracia l int eraction.   Journal of Experimental Social Psychology, 10, 109– 120. Received December 28, 2000