Why do negotiations so often fail even when there are possible resolutions that would serve disputants better than protracted struggle? And why, when resolutions are achieved, are they so often sub-optimal for the parties, or achieved only after heavy and avoidable costs? These questions have intrigued me for a number of years. Several years ago I wrote an article called Why Negotiations Fail: An Exploration of Barriers to the Resolution of Conflict1 and later, with the help of colleagues at the Stanford Center on Conflict Negotiation, I edited and wrote the introduction to a volume called Barriers To Conflict Resolutions.2 These works on barriers explored from a variety of disciplinary perspectives the strategic, psychological and institutional barriers that often led to bargaining failures.
Título original
Strategic Barriers to Dispute Resolution a Comparison of Bilateral and Multilate
Why do negotiations so often fail even when there are possible resolutions that would serve disputants better than protracted struggle? And why, when resolutions are achieved, are they so often sub-optimal for the parties, or achieved only after heavy and avoidable costs? These questions have intrigued me for a number of years. Several years ago I wrote an article called Why Negotiations Fail: An Exploration of Barriers to the Resolution of Conflict1 and later, with the help of colleagues at the Stanford Center on Conflict Negotiation, I edited and wrote the introduction to a volume called Barriers To Conflict Resolutions.2 These works on barriers explored from a variety of disciplinary perspectives the strategic, psychological and institutional barriers that often led to bargaining failures.
Why do negotiations so often fail even when there are possible resolutions that would serve disputants better than protracted struggle? And why, when resolutions are achieved, are they so often sub-optimal for the parties, or achieved only after heavy and avoidable costs? These questions have intrigued me for a number of years. Several years ago I wrote an article called Why Negotiations Fail: An Exploration of Barriers to the Resolution of Conflict1 and later, with the help of colleagues at the Stanford Center on Conflict Negotiation, I edited and wrote the introduction to a volume called Barriers To Conflict Resolutions.2 These works on barriers explored from a variety of disciplinary perspectives the strategic, psychological and institutional barriers that often led to bargaining failures.
8 Harv. Negot. L. Rev. 1 Harvard Negotiation Law Review Spring 2003 Commentary STRATEGIC BARRIERS TO DIS!TE RESOL!TION" A CO#ARISON O$ BILATERAL AND #!LTILATERAL NEGOTIATIONS Robert H. Mnookin d1 Copyright (c) 2003 Harvard Negotiation La Revie! Robert H. Mnookin I. Introduction "hy do negotiation# #o o$ten $ai% even hen there are po##ib%e re#o%&tion# that o&%d #erve di#p&tant# better than protracted #tr&gg%e' (nd hy) hen re#o%&tion# are achieved) are they #o o$ten #&b*optima% $or the partie#) or achieved on%y a$ter heavy and avoidab%e co#t#' +he#e ,&e#tion# have intrig&ed me $or a n&mber o$ year#. -evera% year# ago . rote an artic%e ca%%ed "hy Negotiation# /ai%0 (n 12p%oration o$ 3arrier# to the Re#o%&tion o$ Con$%ict 1 and %ater) ith the he%p o$ co%%eag&e# at the -tan$ord Center on Con$%ict Negotiation) . edited and rote the introd&ction to a vo%&me ca%%ed 3arrier# +o Con$%ict Re#o%&tion#.2 +he#e ork# on barrier# e2p%ored $rom a variety o$ di#cip%inary per#pective# the #trategic) p#ycho%ogica% and in#tit&tiona% barrier# that o$ten %ed to bargaining $ai%&re#. . have taken the opport&nity o$ preparing thi# artic%e to begin to think thro&gh and compare the barrier# to the negotiated re#o%&tion o$ con$%ict in bi%atera% and m&%ti%atera% negotiation#. /or thi# initia% $oray) . i%% primari%y $oc&# on hat . have previo&#%y ca%%ed 4#trategic barrier#5**tho#e that ari#e $rom the e$$ort# o$ 4rationa%5 bargainer# to ma2imi6e individ&a% ret&rn#) and may prec%&de the achievement o$ the greate#t po##ib%e 7gain# in trade7 at the %oe#t *2 co#t. .n other ord#) #trategic barrier# are tho#e that can ca&#e rationa%) #e%$*intere#ted di#p&tant# to act in a manner that prove# to be both individ&a%%y and co%%ective%y di#advantageo&#. . enter the m&%tiparty or%d o$ #trategic interaction ith #ome trepidation. +o begin ith) the mo#t con#pic&o&# body o$ re%evant theoretica% #cho%ar#hip i# game theory) hich e2p%ore# i##&e# o$ #trategic interaction beteen rationa%) #e%$*intere#ted actor#. "hi%e #&gge#tive) it i# a%#o %imiting. +he a2iomatic approach o$ game theoretica% ork on 4n*per#on game#5 i# mathematica%%y da&nting) ha# re#tricted de#criptive poer) and make# no c%aim# at o$$ering poer$&% pre#criptive advice $or negotiator#.3 .ndeed) a# (nato% Rapoport pointed o&t over thirty year# ago) game theoretica% ana%y#i# o$ 4n*per#on5 game# 4direct# it# main thr&#t5 to game# in 4charact eri#tic $&nction $orm)5 hich operate at a %eve% o$ ab#traction that doe# not con#ider the #trategie# avai%ab%e to the p%ayer#.8 Nor i# there very m&ch theoretica% ork to b&i%d on o&t#ide o$ game theory. .ndeed) in 1998) "i%%iam :artman c%aimed) 7;c<&rrent%y no concept&a% ork addre##e# the va#t area o$ m&%ti%atera% negotiation# . . . . .mp%icit%y or e2p%icit%y a%% negotiation theory addre##e# bi*%atera% negotiation#) b&t the comp%e2ity o$ m&%ti%atera% negotiation# remain# &ntreated.5= :artman># c%aim #trike# me a# an e2aggeration. .n $act) my on thinking in the#e matter# ha# been #&b#tantia%%y promoted by the important ork o$ Hoard Rai$$a and ?ame# -ebeni&#. +enty year# ago) in hi# #emina% book) +he (rt and -cience o$ Negotiation) Hoard Rai$$a arned) 4;t<here i# a va#t di$$erence beteen con$%ict# invo%ving to di#p&tant# and tho#e invo%ving more than to di#p&tant#.5@ (nd even -ebeni&#) ho ha# in my vie made the mo#t important contrib&tion# to o&r kno%edge o$ m&%tiparty negotiation) ackno%edge# that 4the c%eare#t and mo#t poer$&% advance# in theory have been ithin or main%y in#pired by the bi*%atera% or to*party ca#e.5A /ina%%y) hi%e m&ch o$ my on #cho%ar#hip ha# been centra%%y concerned ith negotiation# and di#p&te re#o%&tion) near%y a%% o$ it ha# $oc&#ed on to*party ca#e#. 1ar%y in my academic career) . became intere#ted in ho the $orma% %ega% #y#tem act# a# a backdrop $or *3 o&t*o$*co&rt negotiation#** hat . ca%%ed 7bargaining in the #hado o$ the %a.7 +he conte2t in hich . e2p%ored thi# i##&e a# divorce) hich invo%ve# bargaining beteen to #po&#e# ho may be repre#ented by %ayer#. B -ome 2011 Thomson Reuters. No claim to original U.S. Government Works. 1 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 year# %ater . rote) together ith 1%eanor Maccoby) a di#ting&i#hed deve%opmenta% p#ycho%ogi#t) a book abo&t ho divorcing parent# re#o%ved c&#tody i##&e#. Cividing the Chi%d9 i# a %ongit&dina%) empirica% #t&dy that e2p%ored ho #ome 1100 divorcing $ami%ie# re#o%ved c&#tody i##&e#. Dnce again) the $oc&# a# bi%atera% negotiation#. -imi%ar%y) my ork on barrier# primari%y addre##ed) imp%icit%y i$ not e2p%icit%y) bi%atera% negotiation#. My mo#t recent book) 3eyond "inning)10 a# aimed at he%ping %ayer# and their c%ient# negotiate both dea%# and di#p&te# more e$$ective%y. "ith the e2ception o$ one chapter dea%ing ith m&%tiparty %ega% di#p&te#) thi# book primari%y ana%y6e# %ega% negotiation# in hich to individ&a% c%ient# each hire an attorney) thereby creating a $o&r*per#on #y#tem ith %ayer# in the midd%e. +hi# $o&r*per#on #tr&ct&re a%%o# the ana%y#i# to be 4#imp%e and c%ear)5 b&t the book ackno%edge# that 4rea%ity i# rare%y #o kind5 and that many %ega% di#p&te# and dea%# invo%ve not to partie#) b&t #evera%**con#ider bankr&ptcy proceeding#) environmenta% di#p&te#) or many tort# ca#e# in hich a #ing%e p%ainti$$ may #&e #evera% de$endant#) or m&%tip%e p%ainti$$# bring #&it c%aiming that a prod&ct i# de$ective.11 +hi# artic%e i# organi6ed a# $o%%o#. Craing on my previo&# ork) Eart .. ana%y6e# #trategic barrier# in the conte2t o$ the to*party negotiation#. . begin ith the ca#e o$ a b&yer and #e%%er #imp%y negotiating over price** di#trib&tive bargaining. . ne2t t&rn to m&%ti*i##&e to*per#on bargaining #it&ation# here there are va%&e*creating po##ibi%itie#. Eart ... repre#ent# an initia% $oray into the m&%tiparty or%d. . #&gge#t that the Eareto*criterion may not provide an appropriate #tandard to eva%&ate i##&e# o$ e$$iciency in m&%tiparty bargaining. .n a to*party ca#e) any negotiated dea% pre#&mab%y better #erve# the partie# than doe# the #tat&# ,&o. +he #ame co&%d be #aid in a m&%tiparty negotiation) b&t on%y i$ the con#ent o$ every party *4 ere nece##ary. ( re,&irement o$ &nanimity in m&%ti%atera% negotiation) hoever) create# potentia% ho%do&t prob%em# that may po#e #evere #trategic barrier# to re#o%&tion. +he#e prob%em# can be mitigated i$ the con#ent o$ %e## than a%% the partie# can permit action. 3&t other prob%em# may ari#e. .$ coa%ition# o$ %e## than a%% are ab%e to change the #tat&# ,&o) thi# nece##ari%y mean# that a party %e$t o&t o$ a coa%ition may potentia%%y be made or#e o$$. ( variety o$ proced&ra% r&%e# may permit deci#ion*making itho&t &nanimity in m&%tiparty negotiation#. MaFority voting i# b&t one o$ many po##ib%e mechani#m# to a%%ocate deci#ion*making a&thority. +he o&tcome o$ any m&%ti%atera% negotiation can be pro$o&nd%y a$$ected by the#e proced&ra% r&%e# and vario&# deci#ion# concerning agenda. Eart .G brie$%y e2p%ore# the app%ication o$ an &n&#&a% proced&ra% r&%e**the 4#&$$icient con#en# #tandard**that a# emp%oyed in the m&%tiparty 4con#tit&tiona%5 negotiation# in -o&th ($rica and in Northern .re%and. II. Strategic Barriers in Two-Party Negotiations 3eyond "inning #&gge#t# that negotiation re,&ire# the management o$ three ten#ion# inherent in negotiation0 the ten#ion beteen creating and di#trib&ting va%&e! the ten#ion beteen empathy and a##ertivene##! and the ten#ion beteen principa%# and agent#.12 .n to*party negotiation#) the primary #trategic barrier re%ate# to the $ir#t ten#ion beteen the de#ire $or di#trib&tive gain**getting a bigger #%ice o$ the pie**and the opport&nity $or Foint gain#**$inding ay# to make the pie bigger. Hnder#tanding thi# ten#ion) hich a%#o e2i#t# in m&%ti%atera% negotiation#) i# a nece##ary $o&ndation $or a%% that $o%%o#. . begin ith the ca#e here there are no va%&e*creating opport&nitie# at a%%**a ca#e o$ p&re%y di#trib&tive bargaining beteen to partie# negotiating over a $i2ed pie. . then move on to to*party negotiation# invo%ving m&%tip%e i##&e# here there are va%&e*creating opport&nitie#. A. Distributive Bargaining: Two Parties !ne Issue D&r ana%y#i# o$ #trategic barrier# begin# ith a bi%atera% negotiation that e##entia%%y invo%ve# di#trib&ting va%&e. (##&me to individ&a%#**3&yer and -e%%er**m&#t negotiate the price o$ a &#ed a&tomobi%e. +o #imp%i$y the matter) %et &# a##&me the#e partie# are concerned abo&t on%y thi# dea%) and that there are no %inkage# to #imi%ar prob%em# or i##&e# o$ repetitive p%ay. *5 Negotiation ana%y#i# #&gge#t# that each party need# to determine hi# 3(+N( (43e#t (%ternative +o a Negotiated (greement5).13 +he 3(+N( indicate# hat that negotiator can do aay $rom the tab%e i$ no dea% i# reached. +he 3(+N( m&#t be tran#%ated into a re#ervation va%&e**the amo&nt at hich the bargainer i# indi$$erent beteen reaching a dea% and a%king aay to hi# 3(+N(. .n o&r e2amp%e) . i%% a##&me that -e%%er># 3(+N( i# to #e%% hi# car to a dea%er) ho ha# o$$ered him IA)000. . i%% a##&me that thi# i# -e%%er># re#ervation va%&e a# e%%. Let &# a##&me that 3&yer># 3(+N( i# to b&y a #omehat neer) &#ed car $rom a dea%er ith #%ight%y %oer mi%eage $or I11)=00. 3&yer tran#%ate# thi# into a re#ervation va%&e o$ I9)000 $or the car that -e%%er i# o$$ering. Dn the#e $act# the :DE( (4:one D$ Eo##ib%e (greement5)**the bargaining range created by the to re#ervation va%&e#**%ie# 2011 Thomson Reuters. No claim to original U.S. Government Works. 2 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 beteen IA)000 and I9)000. (t #take in thi# negotiation i# ho the #&rp%&# o$ I2)000 i%% be divided. 3oth 3&yer and -e%%er are made better o$$ by any tran#action ith a price anyhere in the bargaining range. (# #&ch) any dea% ithin the :DE( i# Eareto*#&perior to no dea% at a%%. ( #a%e $or any price beteen IA)000 and I9)000 i# Eareto*e$$icient) and other thing# being e,&a%) 3&yer o&%d obvio&#%y pre$er to pay %e## and -e%%er o&%d pre$er to receive more. Note that thi# i# not a 6ero*#&m game beca&#e both p%ayer# o&%d pre$er to negotiate a dea% ithin the :DE( than to have no dea% at a%%. 4;"<hen bargaining) b&yer and #e%%er may have divergent intere#t# ith re#pect to price) b&t both may pre$er to reach #ome agreement.518 +he de#ire $or di#trib&tive gain may %ead to a bargaining $ai%&re in hich the partie# do not reach a dea%) notith#tanding the :DE() beca&#e each i# #eeking to ma2imi6e hi# on ret&rn. +o $actor# he%p e2p%ain thi# po##ibi%ity0 in$ormation a#ymmetrie# and #trategic behavior. .n mo#t negotiation#) each party ha# at %ea#t #ome materia% in$ormation that the other party doe# not have. Dne potentia% in$orma* tion a#ymmetry i# the condition or the ,&a%ity o$ the good# to be traded. +he #e%%er typica%%y kno# $ar more abo&t the ,&a%ity o$ hat i# being #o%d than the b&yer. .t i# e%% e#tab%i#hed that in$ormation *6 a#ymmetrie# abo&t ,&a%ity) the 4%emon# prob%em)5 can a%#o %ead to non*Eareto o&tcome#.1= +he more common a#ymmetry concern# re#ervation va%&e#. 3&yer# and #e%%er# &#&a%%y do not kno each other># 4bottom %ine.5 3&yer) $or e2amp%e) may kno neither the amo&nt a dea%er o&%d pay $or the car) nor ho eager -e%%er i# to get rid o$ it. -e%%er may not kno 3&yer># a%ternative# or ho bad%y 3&yer ant# thi# partic&%ar car. Negotiator# rare%y revea% hone#t%y their re#ervation va%&e. .n$ormation a#ymmetrie# o$ thi# #ort open opport&nitie# $or #trategic opport&ni#m. .ndeed) the e##ence o$ di#trib&tive bargaining invo%ve# the attempt on the part o$ negotiator# to #hape each other># perception# o$ hat i# po##ib%e. "hen deciding hat action to take) each p%ayer m&#t con#ider the other># po##ib%e reaction) and vice ver#a. +hi# i# the e##ence o$ #trategic interdependence. 1ach negotiator i# con#tant%y a##e##ing hat the other #ide might event&a%%y be i%%ing to do and ho $ar they might go. Negotiator# emp%oy a variety o$ tactic# to in$%&ence the other #ide># perception# o$ hat i# po##ib%e**#ome mi#%eading) #ome o&tright di#hone#t. (# a con#e,&ence) the partie# might never di#cover that there i# a :DE( beca&#e) a#piring to drive a hard bargain) each might make e2treme o$$er# $rom hich he recede# very #%o%y. +ypica%%y) neither negotiator kno# ho $ar it might be po##ib%e to p&#h the other #ide. (# a con#e,&ence) the drive $or di#trib&tive advantage may at time# be a barrier to a dea% being made or a di#p&te being #ett%ed. 1ven i$ a #a%e i# con#&mmated (or the %a#&it #ett%ed)) the tran#action co#t# o$ reaching agreement might be m&ch higher than nece##ary. "hat i$ 3&yer and -e%%er #omeho kno each other># re#ervation va%&e' .n thi# ca#e) an obvio&# $oca% point $or a dea% o&%d be to divide the #&rp%&# in ha%$. "hi%e #p%itting the #&rp%&# may be a common re#&%t here there i# $&%% in$ormation) #trategic opport&ni#m and the drive $or di#trib&tive gain can %ead to negotiation $ai%&re#. 1ven i$ the partie# kno (or can g&e##) each other># re#ervation price) there #ti%% can be no dea% i$ they di#agree abo&t ho the #&rp%&# #ho&%d be divided. 1ach might attempt to #hape each other># perception# o$ ho $irm%y they i%% in#i#t on a partic&%ar #hare o$ a knon #&rp%&#. Dne or both might try to per#&ade the other that he i# committed to a%king aay $rom a bene$icia% dea% rather than accept %e## than the %ion># #hare o$ the #&rp%&#. H#ing a variety o$ 4commitment *7 #trategie#)5 a party can make a 4take*it*or*%eave*it o$$er5 that o&%d give that party the %ion># #hare o$ the #&rp%&#. 1@ -e%%er might go #o $ar) $or e2amp%e) a# to give a poer o$ attorney to an agent to #e%% on%y at a high price and then %eave ton to make that c%aim credib%e. +o might p%ay that game and a dead%ock co&%d re#&%t0 E%ayer# in a bargaining game are in an akard po#ition. +hey ant to make the mo#t $avorab%e agreement that they can) hi%e avoiding the ri#k o$ making no agreement at a%%! and) to certain e2tent) the#e goa%# are contradictory. .$ one party indicate# a i%%ingne## to #ett%e $or any term#) even i$ the gain i# on%y margina%) he or #he i%% %ike%y arrive at an agreement) b&t not a very attractive one. Dn the other hand) i$ he take# a hard po#ition and #tick# to it) he i# %ike%y to reach a $avorab%e agreement i$ he reache# any agreement at a%%**b&t he #tand# a good chance o$ being %e$t o&t in the co%d.1A .n #&m) hard*bargaining tactic# that may be rationa% $or #e%$*intere#ted partie# concerned ith ma2imi6ing the #i6e o$ their on #%ice o$ the pie can #ometime# %ead to ine$$icient o&tcome#. +ho#e #&bFected to #&ch tactic# o$ten re#pond in kind) and the net re#&%t i#) at be#t) additiona% co#t# o$ the di#p&te re#o%&tion proce## and) at or#t) $ai%&re to con#&mmate a m&t&a%%y bene$icia% agreement. B. An Aside on t"e Li#its o$ %a#e T"eoretica& Ana&ysis o$ Bargaining D&r car e2amp%e genera%i6e# to a very common type o$ economic tran#action) one that economi#t# have characteri6ed a# a 2011 Thomson Reuters. No claim to original U.S. Government Works. 3 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 bargaining game. +he b&yer o&%d pre$er to pay %e##! the #e%%er o&%d pre$er to *8 receive more! b&t both o&%d pre$er any dea% ithin a 6one created by the their no*dea% a%ternative# to no re#o%&tion at a%%. (ccording to one very prominent game theori#t) hoever) 7bargaining i# an e2treme%y di$$ic&%t topic beca&#e in many #etting# it r&n# right &p again#t the thing# game theory i# not #o good at.71B .t i# he%p$&% to e2p%ain hat the#e %imitation# are. /ir#t) beca&#e game theory i# a branch o$ mathematic#) it re,&ire# the de$inition o$ very preci#e protoco%# or r&%e# o$ the game. .n o&r e2amp%e) and in mo#t rea% or%d bargaining #it&ation#) there are no r&%e# #et o&t in advance #peci$ying ho m&#t make the $ir#t o$$er) hether o$$er# can be made #im&%taneo&#%y) ho many ro&nd# o$ bargaining are permitted) or hen the negotiation end#. .ndeed) game theoretica% ana%y#i# #ho# F&#t ho #en#itive the o&tcome o$ any bargaining game i# to vario&# protoco%#. Con#ider the $o%%oing very #imp%e bargaining game ith the $o%%oing protoco%. Dn%y one ro&nd o$ bargaining i# permitted. 3oth partie# kno each other># re#ervation va%&e. Dne p%ayer get# to make a #ing%e take*it*or*%eave*it o$$er. .$ the other p%ayer accept#) the tran#action i# con#&mmated at thi# price. .$ not) the game end# and each p%ayer receive# hi# no*dea% a%ternative. .$ the protoco% #peci$ie# the b&yer get# to make the #ing%e take*it*or*%eave*it o$$er) e o&%d e2pect the b&yer to capt&re a%mo#t the entire #&rp%&# by o$$ering on%y a tiny amo&nt above IA)000. Hnder the protoco%) #ince the #e%%er o&%d be better o$$ accepting thi# price than #e%%ing to a dea%er) the b&yer o&%d get a very good dea%. Dn the other hand) i$ the protoco% provided that the #e%%er co&%d make a take*it*or*%eave*it o$$er) the #e%%er might e#tab%i#h a price o$ near%y I9)000 and the #e%%er o&%d e##entia%%y capt&re the entire #&rp%&#. Negotiation over price in the rea% or%d typica%%y doe# not invo%ve #&ch #imp%e and e%%*de$ined proced&re# that are agreed &pon in advance. .n#tead) p%ayer# can &#&a%%y make #o many po##ib%e move# and co&nter*move# that the 4game5 i# m&ch too rich $or game theoretic techni,&e#) hich 4re,&ire c%ear and di#tinct 7r&%e# o$ the game.5>19 +hi# i# not to #ay that game theory provide# no &#e$&% in#ight# into bargaining. Dn the contrary) non*cooperative game theory permit# mode%ing and ana%y#i# o$ ho di$$erent #ort# o$ r&%e# o$ the game a$$ect #trategic interaction beteen rationa% actor#. ( #econd common prob%em o$ game theoretic ana%y#i# i# that there are o$ten many e,&i%ibria and no ay to choo#e among them. *9 +hi# prob%em can a%#o be i%%&#trated ith a #imp%e one*ro&nd bargaining mode% ith one ro&nd in hich the b&yer and #e%%er #im&%taneo&#%y name a price itho&t any prior kno%edge o$ the amo&nt #peci$ied by the other #ide. +he r&%e# o$ the game or protoco% $&rther provide that i$ the o$$er# match) the car i# #o%d at that price! i$ the b&yer># price i# higher than the #e%%er># price) the tran#action i%% occ&r at the average o$ the to price#! and i$ the #e%%er># price i# higher than the b&yer># price) there i%% be no dea%. .n thi# game) any price beteen IA)000 and I9)000 repre#ent# a Na#h 1,&i%ibri&m i$ named by both partie#. .$ the named price# are the #ame) neither p%ayer ha# any incentive to change hi# price. +o #ee thi#) a##&me both p%ayer# named IA)2=0. .$ the #e%%er kne thi#) he o&%d %o#e the #a%e ith a higher price and receive %e## ith a %oer named price. (na%ogo&#%y) the b&yer o&%d %o#e the tran#action ith a %oer price) and o&%d pay more ith a higher named price. .n #hort) once each named IA)2=0) neither o&%d have any incentive to change the named amo&nt. +he parado2 i# that thi# #ame rea#oning app%ie# to any price in the bargaining range i$ named by both p%ayer#. Jame theory 4i# o$ no he%p in #orting o&t hether any one ;o$ the#e Na#h e,&i%ibria< i# the 7#o%&tion> and) i$ one i#) hich one i#.520 .ndeed) a# Cavid Krep# ha# empha#i6ed) 4#im&%taneo&#*o$$er bargaining i# on%y one conte2t here there i# thi# prob%em o$ too many e,&i%ibria and too %itt%e g&idance $or choo#ing among them.521 +o e#cape the prob%em o$ m&%tip%e e,&i%ibri&m#) ?ohn Na#h approached the bargaining prob%em in an a2iomatic ay**i.e.) he #et o&t a #erie# o$ characteri#tic# that any #o%&tion #ho&%d have. 3a#ed on the#e a2iom# he #&gge#ted an 4(rbitration -cheme5 in hich partie# o&%d agree to have an arbitrator divide the #&rp%&# in a ay that ma2imi6ed the prod&ct o$ the p%ayer># &ti%itie#) taking the threat point# (the 3(+N(>#) into acco&nt.22 Need%e## to #ay) it i# not c%ear ho an arbitrator co&%d e2pect to %earn the p%ayer#> re#pective &ti%itie#) nor hy partie# to a negotiation o&%d decide to %eave their $ate to an arbitrator. (rie% R&bin#tein bri%%iant%y mode%ed a%ternating*o$$er bargaining in a ay that prod&ced a #ing%e e,&i%ibri&m.23 Hi# game ork# a# $o%%o#. +here are to p%ayer# and each can make one o$$er at a time. *10 E%ayer 3 ha# a choice o$ either accepting p%ayer (># o$$er or re#ponding ith an o$$er o$ hi# on. ( key e%ement o$ thi# game i# that the #&rp%&# that the p%ayer# are dividing dimini#he# over time and each p%ayer ha# a di#co&nt rate. +he o&tcome o$ thi# game i# that the $ir#t p%ayer on the $ir#t ro&nd o$$er# a tiny bit more than =0L and the #econd p%ayer agree# immediate%y. +he R&bin#tein re#&%t ho%d# i$ in#tead o$ &#ing di#co&nt rate#) each party bear# tran#action co#t# each time he make# an o$$er. ( nice $eat&re o$ thi# mode% i# that i$ the p%ayer# have the #ame time pre$erence# or the #ame tran#action co#t#) they i%% e##entia%%y divide the #&rp%&# even%y**an o&tcome that matche# o&r int&ition. Dn the other hand) i$ there are #&b#tantia% di$$erence#) it can have very dramatic e$$ect# 2011 Thomson Reuters. No claim to original U.S. Government Works. 4 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 on the e,&i%ibri&m**e$$ect# ith m&ch %e## int&itive appea%.28 More $&ndamenta%%y) the e2ten#ion o$ the R&bin#tein mode% to three or more p%ayer# no %onger ha# a #ing%e) &ni,&e e,&i%ibri&m.2= ( third %imitation o$ game theoretica% ana%y#i# #ho&%d a%#o be empha#i6ed. Jame theory provide# no theory abo&t here the r&%e# o$ the game come $rom. .n partic&%ar) a# . i%% deve%op in the ne2t #ection) beca&#e an important e%ement o$ negotiation o$ten invo%ve# changing the game) by adding partie#) by adding i##&e#) or by #trategica%%y changing proced&re#) $orma% game theoretica% ana%y#i# cannot he%p &# &nder#tand hy a party change# the r&%e# or p%ayer# or i##&e# hen he did. .ndeed) an important part o$ many m&%tiparty negotiation# concern# negotiating the r&%e# and proced&re#. (# a $ina% point) . o&%d %ike to #&gge#t that game theory i# not e#pecia%%y poer$&% either a# a de#criptive theory (predicting ho peop%e act&a%%y behave) or a# a pre#criptive theory (#&gge#ting ho rationa% peop%e #ho&%d be#t negotiate in a or%d here many peop%e do not behave in ay# con#i#tent ith the theory). +he#e point# are be#t i%%&#trated by con#idering the va#t e2perimenta% %iterat&re re%ating to the 4&%timat&m game)5 the #tr&ct&re o$ hich i# akin to the #e,&entia% one*ro&nd bargaining game de#cribed above. +he#e e2periment# #&gge#t that) economic theory notith#tanding) a party ith $&%% in$ormation i%% o$ten re$&#e to accept a $ina% o$$er ithin the :DE( i$ the re#&%ting divi#ion o$ the #&rp%&# o$$end# hi# #en#e o$ $airne##. *11 .n the 7&%timat&m game7 one p%ayer i# given the right to divide a $i2ed #&m o$ money) #ay I10) and then give an &%timat&m to the #econd p%ayer0 either accept the #hare o$$ered) in hich ca#e the I10 i# divided according to the o$$er) or reFect the divi#ion) in hich ca#e neither p%ayer receive# anything. +hi# i# a p&re%y di#trib&tive game) here a #&rp%&# o$ I10 i# being divided into to knon #hare# #et by the $ir#t p%ayer. Jame theory #&gge#t# that a rationa% o$$eree o&%d pre$er a pittance**#ay a penny**to receiving nothing) even i$ the o$$eror i# getting I9.99. .n other ord#) the o$$eree># re#ervation va%&e #ho&%d be 4ep#i%on5**a #ma%% amo&nt greater than 6ero. (nd yet the evidence #&gge#t# that hen the $ir#t p%ayer divide# the #&rp%&# in a very one*#ided ay) many o$$eree# i%% reFect the o$$er and take nothing in order to deprive the o$$eror o$ the %ion># #hare. "here there i# $&%% in$ormation and the i##&e o$ di#trib&tion %oom# %arge) a party># #en#e o$ $airne## can %ead to hat economi#t# o&%d ca%% Eareto*ine$$icient o&tcome#. .n #hort) a# a de#criptive matter) the game theoretica% mode% o$ thi# #imp%e game doe# not capt&re the behavior o$ many individ&a%#.2@ .$ thi# i# #o) there are important imp%ication# $or pre#criptive theory) or hat Hoard Rai$$a #&gge#t# i# the appropriate 4de#criptive*pre#criptive5 #tance o$ #omeone giving advice to rea% or%d negotiator#. Rai$$a #&gge#t# that a good pre#criptive theory #ho&%d be ba#ed on ho one be%ieve# the negotiating co&nterpart act&a%%y behave#. +he be#t pre#criptive advice to #omeone ho get# to divide the pie in the &%timat&m game i# that he i%% probab%y ma2imi6e hi# ret&rn by dividing the I10 to o$$er the other #ide #omething in the I8 range**$ar more than a penny. "hi%e #ome may t&rn don anything other than an even #p%it) mo#t i%% accept an &neven #p%it #o %ong a# it# not too e2treme.2A '. Two Parties Severa& Issues: T"e Tension Between 'reating and Distributing (a&ue Mo#t negotiation# invo%ve more than dividing a $i2ed pie**they a%#o o$$er opport&nitie# $or creating va%&e and thereby e2panding the pie. "hat do . mean by creating va%&e' . think it i# be#t de$ined in term# o$ Eareto*e$$iciency. (n o&tcome i# #aid to be Eareto*e$$icient or Eareto*optima% i$ one party can be made better o$$ on%y by making *12 the other party or#e o$$. 1conomic# teache# that there i# a Eareto $rontier con#i#ting o$ vario&# Eareto*optima% o&tcome# that have di$$erent di#trib&tive con#e,&ence# $or the partie#.2B 3y de$inition) henever there i# a negotiated agreement in a to*party negotiation both partie# m&#t be%ieve that a negotiated o&tcome %eave# them at %ea#t a# e%% o$$ a# they o&%d have been i$ there ere no agreement. +here$ore) in a $i2ed) po#itive #&m game) any re#o%&tion i# Eareto*e$$icient in compari#on to no dea% at a%%. Dn the other hand) in negotiation# here it i# po##ib%e to create va%&e) partie# may reach negotiated o&tcome# that are not Eareto*e$$icient i$ there i# #ome other negotiated o&tcome avai%ab%e that might make one party better o$$ itho&t harming the other. +hi# notion o$ creating va%&e b&i%d# on a %ong*e#tab%i#hed tradition that negotiation %iterat&re re$er# to a# the 7integrative7 po##ibi%itie# pre#ent in #ome negotiation#.29 3e$ore de#cribing hy #trategic interaction may prevent negotiator# $rom reaching re#o%&tion# that are Eareto*e$$icient) one #ho&%d $ir#t &nder#tand the #o&rce# o$ va%&e and their economic &nderpinning#. .n my vie) there are $o&r #o&rce# o$ va%&e thro&gh negotiation0 di$$erence# beteen the partie#! non*competitive #imi%aritie#! economie# o$ #ca%e and #cope! and opport&nitie# $or red&cing tran#action co#t# and dampening economic opport&ni#m. +he notion that di$$erence# can create va%&e i# co&nterint&itive to many negotiator#) ho be%ieve they can reach agreement on%y by $inding common gro&nd. 3&t the tr&th i# that di$$erence# are o$ten more &#e$&% than #imi%aritie# in he%ping partie# reach a dea%. Ci$$erence# #et the #tage $or po##ib%e gain# thro&gh trade) and it i# thro&gh trade that va%&e i# mo#t common%y 2011 Thomson Reuters. No claim to original U.S. Government Works. 5 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 created. +here are $ive type# o$ di$$erence# that are a%% potentia% #o&rce# o$ va%&e creation0 di$$erent re#o&rce#! di$$erent re%ative va%&ation#! di$$erent $oreca#t#! di$$erent ri#k pre$erence#! and di$$erent time pre$erence#. +hro&gh the negotiation proce##) to the e2tent that partie# di#cover that #&ch di$$erence# e2i#t) they can #tr&ct&re tran#action# that create va%&e by trading on hat amo&nt# to di$$erence# in re%ative va%&ation. *13 ( #econd #o&rce o$ va%&e i# non*competitive #imi%aritie#. .n the#e in#tance#) one per#on># gain doe# not mean the other per#on># %o##. /or e2amp%e) negotiator# o$ten have a #hared intere#t in a prod&ctive) cordia% orking re%ation#hip. 3oth gain to the e2tent that they can improve their re%ation#hip. Likei#e) parent# #hare an intere#t in the e%%*being o$ their chi%dren. .$ a chi%d $%o&ri#he#) both parent# derive #ati#$action. +h&#) even $or divorcing parent#) arrangement# that bene$it a chi%d create Foint gain# $or both ad&%t#. ( third #o&rce o$ va%&e re%ate# to economie# o$ #ca%e and #cope. +o $irm# that each have a #ma%% p%ant may be ab%e to red&ce the &nit co#t o$ prod&ction by having a Foint vent&re that b&i%d# one %arge prod&ction $aci%ity. Creating or pre#erving #ca%e economie# i# a rich #o&rce o$ va%&e creation. -imi%ar%y) economie# o$ #cope can a%#o create va%&e. +he#e ari#e hen more than one good or #ervice can be prod&ced &#ing the #ame ba#ic re#o&rce#) th&# red&cing the co#t# o$ each. Negotiator# can a%#o create va%&e by red&cing the tran#action co#t# o$ reaching an agreement and by dampening #trategic opport&ni#m. +hi# can occ&r in #evera% ay#0 by making the proce## o$ negotiation %e## time*con#&ming and co#t%y! by red&cing the ri#k# that the partie# i%% deceive each other! and by better a%igning $&t&re incentive#. "hat) then) i# the ten#ion beteen creating and di#trib&ting va%&e in a negotiation' .t i# a ten#ion that) i$ managed bad%y) i%% %ead to o&tcome# that $ai% to $&%%y e2p%oit the va%&e creating po##ibi%itie#. .n$ormation drive# thi# ten#ion. "itho&t #haring in$ormation) negotiator# $ind it di$$ic&%t to identi$y trade# that might create va%&e and potentia%%y make both partie# better o$$. Hnreciprocated openne##) hoever) can be e2p%oited. Ci#c%o#ing one># pre$erence#) re#o&rce#) intere#t#) and a%ternative# can he%p to create va%&e) b&t can po#e a grave ri#k ith re#pect to di#trib&tive i##&e#. Negotiator# are con#tant%y ca&ght beteen the#e competing #trategic demand#. .n the end) an individ&a% negotiator i# typica%%y concerned $ir#t ith the #i6e o$ her #%ice and on%y #econdari%y ith the #i6e o$ the pie a# a ho%e. .ndeed) a negotiator ho can ea#i%y c%aim a %arge #hare o$ a #ma%% pie may ind &p ith more to eat than one ho bake# a m&ch bigger pie b&t end# &p ith on%y a #%iver. ( #ki%%$&% negotiator move# nimb%y beteen imaginative #trategie# to en%arge the pie and con#ervative #trategie# to #ec&re an amp%e #%ice no matter hat #i6e the $ina% pie t&rn# o&t to be. Ci$$ic&%tie# ari#e) hoever) beca&#e many negotiator# do not manage thi# ten#ion e%% and) a# a con#e,&ence o$ *14 $oc&#ing primari%y on the di#trib&tive dimen#ion o$ bargaining) $ai% to #ei6e va%&e*creating opport&nitie#. .n #hort) acc&rate in$ormation abo&t each negotiator># goa%#) prioritie#) pre$erence#) re#o&rce#) and opport&nitie# i# e##entia% to reach agreement# that o$$er optima% 7gain $rom trade7**agreement# tai%ored to take $&%% advantage o$ the #o&rce# o$ va%&e o&t%ined above. (t the #ame time) hoever) partie# have a c%ear incentive to concea% their intere#t# and prioritie#) or even mi#%ead the other #ide abo&t them. 3y $eigning attachment to hatever re#o&rce# they are ready to give &p in trade and $eigning re%ative indi$$erence to hatever re#o&rce# they #eek to gain hi%e concea%ing opport&nitie# $or &ti%i6ation o$ the#e re#o&rce#) each party #eek# to in the be#t po##ib%e term# o$ trade $or it#e%$. .n other ord#) tota% $rankne## and 7$&%% di#c%o#&re7**or #imp%y greater $rankne## and $&%%er di#c%o#&re than practiced by the other #ide o$ the negotiation**%eave one #ide v&%nerab%e in the di#trib&tive a#pect o$ bargaining. (ccording%y) the #harp bargainer i# tempted) and may rationa%%y deem it advantageo&#) to practice #ecrecy and deception. Dnce again) hoever) the#e tactic# can %ead to &nnece##ary dead%ock# and co#t%y de%ay# or) more $&ndamenta%%y) $ai%&re# to di#cover the mo#t e$$icient trade# or o&tcome#.30 ) ) ) ) +hi# #hort #&mmary #&gge#t# the &nder%ying rea#on# that #trategic interaction**the #e%$*con#cio&# rationa% behavior o$ individ&a% negotiator#** may act a# a barrier to Eareto*e$$icient o&tcome#. .n #imp%e negotiation# that on%y imp%icate di#trib&tive i##&e#) the re#&%t may be no agreement &nder circ&m#tance# here both partie# co&%d be made better o$$ by one or more po##ib%e agreement#. "here a negotiation invo%ve# #evera% dimen#ion#) there are o$ten integrative po##ibi%itie#** opport&nitie# to make the pie bigger. Here) the prob%em can be to*$o%d. .n #ome in#tance#) the $oc&# on di#trib&tive gain may re#&%t in no dea% or re#o%&tion at a%%. More common%y) it may %ead to a negotiated o&tcome that i# not Eareto*optima%. III. *u&ti+arty Negotiations My ana%y#i# o$ bi%atera% negotiation# e2p%ain# hy a negotiator># p&r#&it o$ di#trib&tive gain can o$ten %ead to negotiation 2011 Thomson Reuters. No claim to original U.S. Government Works. 6 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 $ai%&re#. 1ven tho&gh there are po##ib%e agreement# that o&%d better #erve the partie#) agreement may not be reached at a%%) or even i$ re#o%&tion *15 i# achieved) may not #ati#$y the economi#t># Eareto*criterion o$ e$$iciency. +he#e #ame #trategic prob%em#**and the &nder%ying ten#ion beteen creating and c%aiming va%&e**e2i#t in m&%tiparty negotiation# a# e%%. 3&t in thi# #ection) . o&%d %ike to e2p%ore #evera% additiona% #trategic comp%ication# that may act a# barrier# in the m&%tiparty conte2t a%one. +he $ir#t prob%em concern# the meaning o$ creating va%&e and e$$iciency in the m&%tiparty conte2t. Like other negotiation ana%y#t#) in bi%atera% negotiation# . de$ine creating va%&e in term# o$ Eareto*e$$iciency. +hi# mea#&re doe# not re,&ire inter* per#ona% &ti%ity compari#on#. .t #imp%y #&gge#t# that e$$iciency i# improved henever there i# an a%ternative re#o%&tion that either make# both partie# better o$$ or make# one party better o$$ hen the other party i# no or#e o$$. (n important $eat&re o$ bi%atera% negotiation i# that henever the partie# have agreed to an o&tcome) that o&tcome can be pre#&med to better #erve the intere#t# o$ each party than the #tat&# ,&o**otheri#e there o&%d have been no agreement. Mea#&ring va%&e i# m&ch %e## #traight$orard in a m&%tiparty conte2t. Earado2ica%%y) the re,&irement o$ Eareto*e$$iciency may be very ine$$icient. +he Eareto*criterion re,&ire# that every party to a negotiation have veto poer over the dea%. .$ there are ten partie# to a negotiation) the on%y ay e can kno that a negotiated agreement i# Eareto*#&perior to the #tat&# ,&o i# i$ a%% ten partie# agree to make the dea%. .n other ord#) Eareto*e$$iciency in a m&%tiparty negotiation re,&ire# &nanimity. Hn$ort&nate%y) a &nanimity r&%e create# the #trategic ri#k o$ ho%do&t prob%em#. (##&me) $or the moment) that the :DE( in a m&%tiparty conte2t inc%&de# many po##ib%e dea%# ith di$$erent di#trib&tive imp%ication#) a%% o$ hich make a%% ten negotiator# better o$$ than the #tat&# ,&o. "ith a &nanimity r&%e) the di#trib&tive a#pect# o$ m&%tiparty bargaining can become e2treme%y di$$ic&%t and time con#&ming to manage. "henever nine o&t o$ ten negotiator# have agreed to a partic&%ar dea%) the tenth per#on may credib%y threaten to veto that dea% beca&#e he ant# a dea% that give# him a %arger #hare o$ the pie. Dbvio&#%y more than one negotiator can p%ay thi# game) and a# a con#e,&ence) the tran#action co#t# o$ attempting to achieve &nanimo&# agreement can be very high and negotiation# can break don. ( $e year# back) . #a an e2amp%e o$ thi#. ( Cambridge) Ma##ach&#ett# neighborhood a# given the opport&nity to have it# e%ectrica% ire# p&t &ndergro&nd. +he city o&%d provide a partia% #&b#idy) b&t the ten neighbor# a%% had to con#ent to an a%%ocation o$ *16 the remaining co#t# $or in#ta%%ing the ne main ire. "hatever the a%%ocation) each neighbor o&%d a%#o need to pay hi# on hook*&p co#t# $or bringing the poer $rom the ne &ndergro&nd ire to hi# ho&#e. Mo#t neighbor# ,&ick%y agreed to a%%ocate the co#t o$ the main ire ba#ed on $rontage $eet o$ each %ot. 3&t one neighbor) 4Mr. ?one#)5 ho#e ho&#e a# #et $arther back $rom the #treet) obFected. He arg&ed that #ince hi# connection co#t to a ne &ndergro&nd ire o&%d be higher) he #ho&%d pay a #ma%%er #hare o$ the Foint co#t#. +o other neighbor# #a ?one#> #&gge#tion a# #trategic**a# ho%ding o&t to get more o$ the #&rp%&#** and a# a con#e,&ence) no dea% co&%d be made at a%%0 the e%ectrica% ire# ere not p%aced &ndergro&nd and the #&b#idy a# %o#t. My on int&ition i# that the probabi%ity o$ a bargaining $ai%&re) even tho&gh there i# a :DE() i%% increa#e a# the n&mber o$ the partie# to a negotiation goe# &p) i$ each ha# a veto. (nd even i$ a dea% i# made) . o&%d a%#o predict that the tran#action co#t# o$ reaching agreement are %ike%y to be m&ch higher than ith a %e## #trict deci#ion r&%e. E&tting the ho%do&t i##&e a#ide) there i# an even more $&ndamenta% i##&e. (# partie# are added) i$ each ha# a veto) it o&%d #tand to rea#on that the ri#k o$ there being no :DE( increa#e#. +hi# i# not a prob%em i$ one be%ieve# a negotiated dea% #ho&%d not be po##ib%e i$ it ha# even #%ight%y &n$avorab%e di#trib&tiona% con#e,&ence# $or even one party. 3&t in the m&%tiparty conte2t) there o$ten may be #it&ation# here the gain# $or the inner# $rom a #ocia% per#pective are tho&ght to #&b#tantia%%y o&teigh the %o##e# o$ the %o#er#. "hat i$ a po##ib%e agreement bene$it# near%y a%% o$ the partie# a great dea%) b&t impo#e# #%ight %o##e# on a $e' 3eca&#e the Eareto*criterion impo#e# a di#trib&tion re,&irement that privi%ege# the #tat&# ,&o) the %o#er# might be e2pected to e2erci#e their veto. (nd yet in many conte2t#) e#pecia%%y i$ #ide*payment# are not po##ib%e) one might conc%&de that at %ea#t in term# o$ ea%th ma2imi6ation and ordinary notion# o$ e$$iciency) the comm&nity o&%d be better o$$ ith the dea% even tho&gh it doe# not meet the re,&irement# o$ Eareto*e$$iciency. .ntere#ting%y) &#ing the 7Ka%dor*Hick#7 compen#ation criterion a# a ay o$ de$ining economic e$$iciency avoid# the con#traint# o$ the c%a##ica% Eareto*criterion. Ka%dor*Hick# a#k# hether it i# po##ib%e $or the inner# to make #ide payment# to the %o#er# #o that the %o#er# o&%d be indi$$erent beteen the ne arrangement and the #tat&# ,&o. No #&ch #ide payment i# act&a%%y re,&ired to be paid) hoever. .$) a# a theoretica% propo#ition) the gain# are #&$$icient to permit #&ch payment#) that i# eno&gh to conc%&de that the propo#ed arrangement *17 i# more e$$icient that the #tat&# ,&o. .n Ka%dor># ord#) hether the %o#er# 4#ho&%d be given compen#ation . . . i# a po%itica% ,&e#tion on hich an economi#t co&%d hard%y prono&nce an 2011 Thomson Reuters. No claim to original U.S. Government Works. 7 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 opinion.531 3eca&#e the Ka%dor*Hick# te#t a#k# hether it might be po##ib%e to make a compen#atory payment) not hether in $act one i# going to be made) it mitigate# the ho%do&t prob%em. -ome partie# might be made or#e o$$) b&t the ne arrangement o&%d nonethe%e## be vieed a# more e$$icient than the o%d #tat&# ,&o. 3&t Ka%dor*Hick# i# hard%y a comp%ete #o%&tion to determining economic e$$iciency in a m&%tiparty negotiation. (# a matter o$ economic theory) the criterion i# not itho&t it# prob%em#) hether beca&#e o$ potentia% change# in re%ative price# or income e$$ect#.32 "hi%e app%ied economi#t# and po%icy maker# may &#e co#t*bene$it ana%y#i# or 4ea%th ma2imi6ation5 princip%e# to eva%&ate economic e$$iciency) they are imp%icit%y making 4a %eap o$ $aith)5 and 4eigh;ing< together 1ve># %o##e# and (dam># gain#.533 More $&ndamenta%%y) in a negotiation) no #ocia% p%anner i# trying to eigh overa%% co#t# and bene$it# to a##e## hether a #&$$icient #ide*payment i# theoretica%%y po##ib%e. .$ %e## than a%% can impo#e co#t# on a minority itho&t the need to make #ide*payment#) ho can e be con$ident that the ne arrangement improve# #ocia% e%$are in compari#on ith the #tat&# ,&o' Notith#tanding the#e concept&a% di$$ic&%tie#) in many conte2t# &nanimity re,&irement# are #imp%y bypa##ed beca&#e #&bgro&p# can $orm coa%ition# and make their on agreement#) &ncon#trained by a veto o$ tho#e ho might be di#advantaged. 4Dnce three or more con$%icting partie# are invo%ved) coa%ition# may $orm and act in concert again#t the other di#p&tant#.538 +ho#e %e$t o&t o$ a coa%ition may o$ten be or#e o$$ than they ere be$ore negotiation# began. Hn%ike in bi%atera% negotiation#) here typica%%y one party cannot a$$ect the other party># 3(+N() in m&%tiparty negotiation#) tho#e %e$t o&t o$ dea% may o$ten $ind them#e%ve# or#e o$$ i$ no dea% ere made. Con#ider) $or e2amp%e) merger negotiation# in an ind&#try ith three competing companie#. (%tho&gh a 4grand coa%ition5 invo%ving a merger o$ a%% three partie# might be po##ib%e) #o too o&%d be merger# *18 beteen any to o$ the $irm#. Moreover) a merger beteen any to $irm# might make the company %e$t o&t or#e o$$ than i$ no dea% ere made.3= "henever coordinated action# or deci#ion# by %e## than a%% the partie# can change the #tat&# ,&o) the potentia% $or coa%itiona% dynamic# become# part and parce% o$ m&%tiparty negotiation. "itho&t attempting to p%&mb $&%%y the concept&a% comp%e2itie# o$ coa%ition#) it i# #a$e to a##ert that a variety o$ #trategic barrier# can ari#e that do not e2i#t in bi%atera% negotiation#. N*per#on game theory demon#trate#) $or e2amp%e) that ith re#pect to game# that have no 4core)5 the prob%em i# not too many e,&i%ibri&m#) b&t none. 4.$ a game ha# no core) it i# &n#tab%e in the #en#e that hatever the payo$$) #ome coa%ition ha# the poer and motivation to break &p the imp&tation and go o$$ on it# on.53@ Coa%itiona% in#tabi%ity may %ead to an obvio&# #trategic barrier in #&ch in#tance#. I(. Decision-*a,ing Procedures in *u&ti+arty Negotiation: T"e -.a#+&e o$ /Su$$icient 'onsensus0 M&%ti%atera% negotiation# o$ten take p%ace ithin in#tit&tion# that have created proced&ra% r&%e# a%%oing %e## than a%% the partie# to change the #tat&# ,&o) notith#tanding the obFection# o$ #ome.3A (n intere#ting ,&e#tion i# ho #&ch proced&re# might be &#ed to mitigate the ho%do&t prob%em and a%%o a negotiated re#o%&tion that i# ide%y accepted a# %egitimate. Recent e2amp%e# o$ comp%e2 m&%tiparty po%itica% negotiation# took p%ace in -o&th ($rica and Northern .re%and. .n -o&th ($rica) the goa%# o$ the negotiation# ere to create an interim con#tit&tion) di#mant%e apartheid) and prepare $or the co&ntry># $ir#t open e%ection#. .n Northern .re%and) the goa% a# to create a $rameork $or #e%$*government. . am e#pecia%%y intrig&ed by the &#e o$ a 4#&$$icient con#en# criterion $or deci#ion*making in the comp%e2 m&%tiparty po%itica% negotiation# that recent%y took p%ace in -o&th ($rica and Northern .re%and. .n both in#tance#) thi# #tandard mitigated the ho%do&t prob%em and enco&raged the $ormation o$ a broad coa%ition o$ moderate centri#t partie# that had previo&#%y been in deep con$%ict. .n each ca#e) *19 m&%tiparty negotiation# ere beg&n in the hope# o$ tran#$orming ba#ic po%itica% in#tit&tion#. +he p&rpo#e o$ the 4CDC1-(53B negotiation# and the M&%ti*Earty Negotiating Eroce## (4MENE5) in -o&th ($rica a# to create an interim con#tit&tion that di#mant%ed apartheid and #et gro&nd r&%e# $or %egi#%ative e%ection# in hich a%% -o&th ($rican# eighteen year# or above o&%d vote. .n Northern .re%and) here the 4a%% party peace ta%k#5 invo%ved at %ea#t eight partie#) Jeorge Mitche%% &#ed the #&$$icient con#en#&# mechani#m to orche#trate the e2tended negotiation proce## that %ed to the 199B Jood /riday (ccord#) hich created a $rameork $or #e%$*government and ere #&b#e,&ent%y rati$ied in a re$erend&m. +he#e negotiation# invo%ved repre#entative# o$ many di$$erent po%itica% partie#**a# many a# tenty*#i2 partie# in -o&th ($rica) and at %ea#t eight in Northern .re%and. .n both #it&ation#) partie# repre#enting a #ing%e race (i.e.) b%ack# in -o&th ($rica) or a #ing%e re%igion (i.e.) Erote#tant# in Northern .re%and) might have c%aimed to #peak $or a maFority o$ the tota% pop&%ation. 3&t 2011 Thomson Reuters. No claim to original U.S. Government Works. 8 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 deci#ion*making by #&ch a maFority co&%d never have re#o%ved either con$%ict. +he cha%%enge in both negotiation# a# to b&i%d a coa%ition o$ moderate centri#t partie# acro## racia% or re%igio&# %ine# and prevent more e2treme partie#**b%ack or hite) Catho%ic or Erote#tant**$rom b%ocking progre##. -&$$icient con#en#&# a# a ay in both in#tance# to maintain a #pirit o$ inc%&#ivene## hi%e avoiding the #trict&re# o$ &nanimity. A. De$ining Su$$icient 'onsensus "hat a# the 4#&$$icient con#en# criterion' .t di$$ered #omehat in the to ca#e#. .n -o&th ($rica) there a# no attempt to de$ine the preci#e ,&antitative re,&irement) b&t a r&%e did enco&rage a proce## o$ broad participation and #&pport. .n Northern .re%and) the $orma% te#t a# more comp%icated and had #evera% component#. 1. T"e Test o$ Su$$icient 'onsensus in Sout" A$rica (ccording to the -tanding R&%e# o$ the M&%ti*Earty Negotiating Eroce##) #&$$icient con#en#&# orked a# $o%%o#0 "hen the chair o$ any o$ the n&mero&# #pecia%i6ed committee# charged ith partic&%ar i##&e# determined that &nanimity) or 4genera% con#en#&#)5 co&%d not be initia%%y achieved on an i##&e) the committee a# in#tr&cted to &#e *20 a variety o$ mechani#m# in order to reach the ide#t po##ib%e con#en#&#. .n partic&%ar #&ch mechani#m ;#ic< #ha%% inc%&de adFo&rnment# to enab%e in$orma% di#c&##ion# beteen participant#) #etting &p technica% committee# compo#ed a# the meeting deem# appropriate $or the partic&%ar matter &nder con#ideration) a# e%% a# a%%oing participant# to con#&%t their principa%#.39 .$ &nanimity #ti%% co&%d not be reached) the chair co&%d then decide 7at hi#Mher di#cretion7 hether #&$$icient con#en#&# e2i#ted. Eartie# #ti%% obFecting might di#p&te the deci#ion) b&t the intention a# that 4in the #pirit o$ cooperation ;they o&%d< not hinder the proce## $rom going $orard.580 -&$$icient con#en#&# a# &#ed a# a deci#ion*making mechani#m many time# in the 1991*1998 period o$ -o&th ($rica># tran#ition) b&t mo#t notab%y to #et the date $or e%ection# to a $ina% con#tit&tion*making body. Dn ?&ne 3) 1993) the ($rican Nationa% Congre##) the moderate b%ack organi6ation %ed by Ne%#on Mande%a) propo#ed (pri% 2A) 1998 a# the e%ection date a$ter a $%&rry o$ ta%k# and conce##ion# to #ma%%er b%ack partie#. (# part o$ the propo#a%) the (NC a%#o #tip&%ated that $ina%i6ation o$ the deci#ion be po#tponed &nti% ?&ne 1=) in order to 4ma2imi6e con#en#&# on thi# matter.5 81 Dn ?&ne 1=) a #ma%%er gro&p) the .nkatha /reedom Earty (4./E5)) p&t $orard a motion o$ it# on) that 4any $ina% determination o$ the e%ection date be po#tponed5 &nti% #evera% condition# had been met.82 /i$teen o$ the gro&p# pre#ent voted again#t the ./E># motion and eight voted $or it. M.?. Mah%ang&) then the chair o$ the negotiating co&nci% in hich the e%ection date a# to be determined) r&%ed that there a# not #&$$icient con#en#&# $or the ./E motion. +hi# meant) in e$$ect) that there a# #&$$icient con#en#&# $or the (pri% 2A) 1998 e%ection date.83 2. T"e Test o$ Su$$icient 'onsensus in Nort"ern Ire&and .n Northern .re%and) #&$$icient con#en#&# a# #aid to e2i#t $or an i##&e here there a# agreement that a# #&pported by0 (1) partie# *21 repre#enting a maFority o$ the Catho%ic comm&nity! (2) partie# repre#enting a maFority o$ the Erote#tant comm&nity! (3) a n&merica% maFority o$ a%% the partie# pre#ent at the ta%k#! and (8) the government# o$ the Rep&b%ic o$ .re%and and the Hnited Kingdom. "herea# the MENE -tanding R&%e# #tip&%ated that #&$$icient con#en#&# be &#ed 4$or re#o%ving #&b#tantia% i##&e# and not $or $orma% and admini#trative deci#ion#)588 #&$$icient con#en#&# in Northern .re%and enFoyed a m&ch broader app%ication. Jeorge Mitche%%) one o$ the ta%k#> chairmen) &#ed it to make many proced&ra% and admini#trative deci#ion#. /re,&ent%y) imp%ementing #&$$icient con#en#&# made the negotiation# protracted and $r&#trating) b&t it a%#o a%%oed Mitche%% to gain tr&#t and e#tab%i#h a&thority ith the negotiator#. .n Northern .re%and) #&$$icient con#en#&# a# mo#t memorab%y emp%oyed to adopt the very 7R&%e# o$ Eroced&re7 that introd&ced it. .n %ate ?&%y 199@) it %ooked a# tho&gh the negotiation participant# o&%d not be ab%e to agree on a #et o$ gro&nd r&%e# be$ore the (&g&#t rece##. /ina%%y) Eeter Robin#on) a de%egate $rom the Cemocratic Hnioni#t Earty (4CHE5)) #&gge#ted that the de%egate# vote on each individ&a% #ection o$ the propo#ed r&%e# and on the r&%e# a# a ho%e. +ho#e ith obFection# co&%d vote) i$ they #o de#ired) again#t each o$ the individ&a% r&%e#) b&t then vote $or the r&%e# a# a ho%e. +hat ay) minority partie# concerned ith maintaining their rep&tation# ith their con#tit&encie# co&%d appear a# having voted overhe%ming%y again#t the r&%e#) hi%e #ti%% enab%ing them to pa##.8= B. T"e Su$$icient 'onsensus 'riterion -#+owered t"e *a3or *oderate Parties 3oth -o&th ($rica and Northern .re%and># #&$$icient con#en#&# had the e$$ect o$ creating a poer$&% centri#t coa%ition re#pon#ib%e $or mo#t o$ the progre## in the ta%k#. 2011 Thomson Reuters. No claim to original U.S. Government Works. 9 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 1. Sout" A$rica .n the -o&th ($rican negotiation#) #&$$icient con#en#&# a# #aid to e2i#t hen there a# agreement beteen the Nationa% Earty (4NE5)) /.". de K%erk># party hich repre#ented a maFority o$ hite#) and the ($rican Nationa% Congre## (4(NC5)) Ne%#on Mande%a># party *22 and by $ar the %arge#t b%ack party. +ho#e ho di#agreed ith a po##ib%e deci#ion ere enco&raged to voice their obFection# and tho#e obFection# o&%d be d&%y con#idered and taken #erio&#%y. 3&t hen nece##ary $or progre##) a deci#ion by the NE and (NC o&%d genera%%y be eno&gh $or #&$$icient con#en#&#) even i$ the concern# o$ a minority co&%d not be #ati#$ied. 4(ccording to one acco&nt)5 rite# the Centre $or Eo%icy -t&die# in it# 1993 report on the negotiation#) 4thi# magica% princip%e a# anno&nced to a #enior Code#a participant the night be$ore the convention began by a key NE negotiator and hi# (NC co&nterpart. 7"hat doe# 4#&$$icient con#en# mean'> the participant a#ked. 7"hatever e ;the (NC and NE< decide it mean#)> a# the rep%y.58@ +he 4Record o$ Hnder#tanding)5 $or e2amp%e**a doc&ment created by the (NC and the NE in -eptember 1992 that dea%# ith the character o$ po#t*apartheid in#tit&tion#**i# evidence o$ a #trong moderate b%oc in the -o&th ($rica negotiation#. 3y o$$icia%%y reconci%ing the vie# o$ the (NC and the NE on maFor i##&e#) the Record o$ Hnder#tanding en#&red #moother $&t&re negotiation# and prevented ta%k# $rom becoming mired in the obFection# o$ e2treme gro&p#! the (NC and NE o&%d agree on an i##&e $ir#t) and then bring it be$ore the re#t o$ the partie# $or con#ideration. Dther partie#) #&ch a# the ./E and E(C (4Ean*($ricani#t Congre##5)) and %eader# #&ch a# D&pa J,o6o o$ Ci#kei and L&ca# Mangope o$ 3oph&that#ana) acc&#ed the (NC and NE o$ co%%&#ion) b&t the $act remain# that the (NC*NE coa%ition provided the $rameork $or near%y a%% the negotiation# o$ -o&th ($rica># tran#ition. 2. Nort"ern Ire&and .n Northern .re%and) #&$$icient con#en#&# enab%ed the -ocia% Cemocratic Labo&r Earty (4-CLE5) and the H%#ter Hnioni#t Earty (4HHE5)) the to %arge#t and mo#t moderate po%itica% gro&p#) to $orm a centri#t coa%ition ith con#iderab%e poer. +he term# o$ the deci#ion*making mechani#m ere #&ch that i$ either party did not agree to a propo#a%) that propo#a% co&%d not pa##. 7.t e##entia%%y provided that the o&tcome o$ the ta%k# co&%d not re$%ect a one*#ided victory by one overa%% $action or another. (# #&ch) any re#&%t o&%d need the #&pport o$ a centri#t coa%ition.78A 3y giving the -CLE and HHE the *23 abi%ity to veto propo#a%#**the #o&rce o$ a rea% #en#e o$ poer** #&$$icient con#en#&# a$$orded the to %arge) moderate partie# a 7#a$er7 environment in hich to cooperate and make conce##ion#.8B '. Su$$icient 'onsensus Had t"e -$$ect o$ Si#+&i$ying t"e *u&ti+arty Negotiations .n the negotiation# o$ -o&th ($rica and Northern .re%and) mo#t o$ the #&b#tantive progre## a# made in #ma%%) &#&a%%y bi%atera% or tri%atera% ta%k#. .n -o&th ($rica) thi# a# the ca#e $rom the very beginning. .n Northern .re%and) the chairmen made ear%y attempt# to cond&ct negotiation# a# a%%*party ta%k#) b&t %ater opted $or #ma%%er gro&p#. +hi# #ty%e o$ proced&re a# adopted o&t o$ con#ideration# o$ e$$iciency and made po##ib%e by #&$$icient con#en#&#. -&$$icient con#en#&# co&%d o$ten red&ce a cacophono&# and ine$$ective m&%tiparty negotiation into a to*party negotiation beteen the organi6ation# o$ the moderate coa%ition and their a$$i%iate#. 1. Sout" A$rica 1ven be$ore CDC1-( a# convened $or the $ir#t time) the (NC and NE had been engaged in bi%atera% ta%k#) and thi# coa%ition g&ided the progre## o$ a%% o$ the con#tit&tiona% negotiation#. +he $ir#t CDC1-( a# cond&cted on%y in p%enary ta%k#) one re#&%t o$ hich a# the Cec%aration o$ .ntent) 4a #et o$ princip%e# and p&rpo#e# to g&ide CDC1-(.589 +he co%%ective #igning o$ thi# doc&ment gave the impre##ion o$ a tr&e a%%*party negotiation. Hoever) 4Code#a 1 a# in the main a rit&a%i#tic event) apparent%y de#igned more to impre## a dome#tic p&b%ic and $oreign dignitarie#.5=0 +he con$erence># organi6er# began thi# $ormat ith the pre## and the p&b%ic in mind! the p%enary meeting# ere %arge%y ceremonia%. +he rea% negotiation# o&%d take p%ace not in p%enarie#) b&t in the $ive #ma%%er 4"orking Jro&p#5 that the Cec%aration o$ .ntent de%ineated. +hey had a #imi%ar compo#ition0 a p%enary that met on%y #poradica%%y! three #ma%%er gro&p#! and #even technica% committee#. "rite# -i#k) the 4centra% deci#ion*making #tr&ct&re o$ the 20B*member) tenty*#i2 party) par%iamentary*#ty%e a##emb%y a# the ten*per#on Negotiating Co&nci%.5=1 Hoever) not even the#e e%even #ma%%er gro&p# ere re#pon#ib%e $or a%% o$ the negotiation0 7;M<&ch pre%iminary bargaining a# cond&cted *24 in a #erie# o$ behind*the*#cene# bi%atera% ta%k# thro&gho&t 1993.5=2 .n the end) the #tr&ct&re o$ -o&th ($rica># ta%k# a# m&ch the #ame a# Northern .re%and>#0 the (NC and the NE ere a$$orded the abi%ity to behave a# tho&gh they ere participating in to*party negotiation# beca&#e the #&$$icient con#en#&# criterion re,&ired that on%y they agree $or a deci#ion to be taken. 2011 Thomson Reuters. No claim to original U.S. Government Works. 10 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 2. Nort"ern Ire&and Hnti% ear%y ?&%y 199@) p%enary meeting# o$ the Northern .re%and peace proce## ere he%d in a room ith #i2ty*eight peop%e aro&nd a %arge) #,&are) ring*#haped tab%e) a #it&ation 4more cond&cive to attack than to compromi#e.5=3 +he#e p%enary meeting# ere %arge to the point o$ being &nie%dy) and deci#ion*making) even by #&$$icient con#en#&#) a# %aborio&#. 3eca&#e the#e meeting# a%#o took p%ace aro&nd 7marching #ea#on7 and ere not making progre##) Mitche%% cance%%ed the p%enary #e##ion# $or the ne2t to eek# and in#tead #ched&%ed bi%atera% meeting#. +he#e ta%k# made more progre## than the a%%*party one# had) and $rom then on) Mitche%% made &#e o$ hat Mark C&rkan o$ the -CLE de%egation ca%%ed 7variab%e geometry>**the concept o$ $%e2ibi%ity in the #hape and #i6e o$ meeting#. 3i%atera% di#c&##ion# beteen the HHE and the -CLE and %ater tri%atera% di#c&##ion# invo%ving the (%%iance managed to prod&ce a orking paper on decommi##ioning on hich there a#) even according to HHE #o&rce#) ninety per cent agreement. (t the #ame time) bi%atera% di#c&##ion# had prod&ced agreement beteen the -CLE and the HHE on a po##ib%e Comprehen#ive (genda $or the #&b#tantive negotiation# hich a# de#igned to rea##&re each party abo&t the intention# o$ the other.=8 +he #&cce## o$ variab%e geometry can be attrib&ted to #&$$icient con#en#&#. +he bi%atera% ta%k# beteen the -CLE and the HHE ere e2treme%y important beca&#e they ere the gro&p# ho#e approva% con#tit&ted #&$$icient con#en#&#. "hen the#e to moderate partie# co&%d agree on #omething) #&$$icient con#en#&# a# con#idered to be attained. +he to moderate partie# had the privi%ege o$ behaving a# *25 tho&gh they ere part o$ a to*party negotiation rather than a m&%tiparty one. D. T"e -$$ects o$ t"e Su$$icient 'onsensus 'riterion on t"e *inor and *ore -.tre#e Parties /or minor partie# that ere not e2treme) the #&$$icient con#en#&# criterion enco&raged them to Foin the maFor party in $orming a centri#t coa%ition. +he more intere#ting e$$ect a# on the o&t%ier#. .n both -o&th ($rica and Northern .re%and) the &#e o$ #&$$icient con#en#&# #&bFected more e2treme partie# to very di$$ic&%t choice#. "hen a deci#ion ent again#t an o&t%ier># intere#t#) one option a# to bo%t $rom the negotiation# and ork to &ndermine the proce## $rom the o&t#ide. +he a%ternative a# to remain in the negotiation# ith the hope o$ con#training $&t&re action#. 1. Sout" A$rica .n -o&th ($rica) the ten#ion beteen anting to take part in the negotiation# and anting to prevent the $ormation o$ an agreement i# be#t e2emp%i$ied by the .nkatha /reedom Earty. +he ./E participated in the $ir#t CDC1-() b&t beca&#e o$ the po#ition it took on #ome i##&e#) it 4began to be #een not on%y a# an o&tbidder on the government># right b&t a# an in#ec&re po%itica% $oe o$ the (NC) th&# #etting it#e%$ &p a# a potentia% #poi%er.5== .t a# at the $ir#t CDC1-( that the ./E 4revea%ed it# on &nder#tanding o$ being a minority) regiona%%y ba#ed party5!=@ it a# not one o$ the gro&p# ho#e approva% a# needed $or #&$$icient con#en#&#. +he ./E a%#o participated in MENE) hich took p%ace beginning in (pri% 1993) b&t a%ked o&t o$ the negotiation# that ?&ne a$ter rea%i6ing that #&$$icient con#en#&# o&%d a%%o the centri#t (NC*NE coa%ition to #et the e%ection date regard%e## o$ it# obFection. +he ./E then #hi$ted gear# $rom participating in negotiation# to trying to prevent an agreement $rom taking p%ace. .t bro&ght a #&it cha%%enging #&$$icient con#en#&# be$ore the -o&th ($rican -&preme Co&rt in -eptember o$ that year) b&t the ca#e a# &%timate%y di#mi##ed.=A *26 2. Nort"ern Ire&and .n Northern .re%and) the Cemocratic Hnioni#t Earty (4CHE5)) a more e2treme Erote#tant Earty) $o&nd it#e%$ in a #imi%ar #it&ation. "hen -inn /Nin a# admitted to ta%k# in ?&%y 199A a# a re#&%t o$ a deci#ion ba#ed on #&$$icient con#en#&#) both the CHE and another %oya%i#t party) the Hnited Kingdom Hnioni#t Earty (4HKHE5)) #tormed o&t o$ ta%k# and never ret&rned. +hey had been participating in ta%k#) b&t their avoed aim 4a# they repeated%y in#i#ted) ;a#< to end thi# proce##.5=B Hoever) the CHE and HKHE %earned $ir#thand the di%emma o$ the di##enting minor party in m&%tiparty ta%k# ith a #&$$icient con#en#&# criterion. "rite# Mitche%%) 7their a%ko&t a# a $ate$&% error . . . . +heir ab#ence $reed the HHE $rom dai%y attack# at the negotiating tab%e) and gave the party room to negotiate that it might not otheri#e have had.5=9 +he deci#ion o$ the#e %oya%i#t partie# to ,&it negotiation# co&%d not prevent an agreement $rom taking p%ace. .n thi# ca#e) it may have act&a%%y $aci%itated the proce##) %eaving the CHE and HKHE in a or#e #it&ation than they had been hi%e participating in the ta%k#) beca&#e no they had no in$%&ence hat#oever over an agreement. ($ter %eaving the ta%k#) .an Eai#%ey) the CHE %eader) made a %a#t*ditch e$$ort to prevent an agreement. .n ear%y (pri% 199B) hen the -CLE and the HHE anno&nced agreement on #&b#tantia% i##&e#) Eai#%ey #hoed &p at the ta%k#> ven&e ith an ento&rage0 2011 Thomson Reuters. No claim to original U.S. Government Works. 11 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 /or month#) Eai#%ey and hi# co%%eag&e# had $o&nd them#e%ve# practica%%y ignored by the media a$ter #ta%king o&t o$ negotiation#! no they ere determined to make their pre#ence $e%t. Eai#%ey to%d an e2cited gathering o$ $%ag*aving %oya%i#t# c%o#e to the ta%k#0 7Oo& have got the right to come &p here tonight and demon#trate that it># No -&rrenderP>@0 Eai#%ey># e$$ort# ere %arge%y #corned by tho#e pre#ent) b&t the important a#pect o$ thi# anecdote i# Eai#%ey># #hi$t $rom a participant in negotiation# ith #ome abi%ity to in$%&ence their o&tcome) to an o&t#ider itho&t in$%&ence) trying de#perate%y to end them. (. 'onc&usion ( n&mber o$ year# ago) . #&gge#ted that an important inte%%ect&a% ta#k re,&ire# the e2p%oration o$ hy) &nder circ&m#tance# hen *27 partie# might make them#e%ve# better o$$ thro&gh a negotiated re#o%&tion o$ con$%ict) negotiation# o$ten $ai%ed. Negotiation obvio&#%y invo%ve# #trategic interaction**i.e.) Foint deci#ion*making in hich each party># choice# m&#t nece##ari%y be in$%&enced by hat he think# other partie# are %ike%y to do. +hi# artic%e ha# #o&ght to e2p%ore #trategic barrier#) both in the bi%atera% and m&%ti%atera% conte2t. "hat conc%&#ion# might be dran $rom thi# mode#t $oray' /ir#t) negotiation ana%y#i# p%ain%y indicate# that #trategic barrier# o$ten do e2i#t. .n #ome circ&m#tance#) the rationa% p&r#&it o$ di#trib&tive gain in negotiation can %ead to ine$$icient o&tcome#. . have #&gge#ted hy thi# i# #o in bi%atera% negotiation#) and my on be%ie$ i# that it i# even more the ca#e in m&%ti%atera% negotiation#. -econd) game theoretica% ana%y#i# o$$er# he%p$&% in#ight# into #trategic interaction) b&t in many circ&m#tance#) it i# not a &#e$&% #o&rce $or pre#criptive advice to tho#e act&a%%y engaged in negotiation#. 3eca&#e game theory re,&ire# $orma% ana%y#i# in mathematica% term#) the r&%e# o$ the game m&#t be e%% #peci$ied in advance. (# my brie$ e2p%oration# o$ the m&%tiparty negotiation# in -o&th ($rica and Northern .re%and demon#trate) deci#ion r&%e# can have a poer$&% impact on the po##ibi%ity that inning coa%ition# can be $ormed and the comparative capacity o$ di$$erent partie# to in$%&ence the $ina% o&tcome. .n negotiation# genera%%y and e#pecia%%y in m&%tiparty negotiation#) the $orm o$ #trategic interaction o$ten invo%ve# adding or #&btracting i##&e#) adding or #&btracting partie# to the negotiation) negotiating abo&t the content o$ deci#ion r&%e#) and attempting to manip&%ate the agenda. -&ch comp%e2itie# typica%%y cannot be mathematica%%y mode%ed. D&r &nder#tanding o$ con$%ict re#o%&tion o&%d #&re%y be enriched by care$&% e2p%oration o$ barrier# $rom the per#pective# o$ a variety o$ the #ocia% #cience#) certain%y game theory and the economic# o$ bargaining have m&ch to contrib&te. 3&t my on be%ie$ i# that no theoretica% per#pective) and no #ing%e di#cip%ine) ha# a monopo%y on &#e$&% in#ight# concerning the barrier# to the $air and e$$icient re#o%&tion o$ con$%ict. .ndeed) . #&#pect that progre## i%% t&rn very $&ndamenta%%y on the abi%ity o$ peop%e $rom di$$erent di#cip%ine# to %earn $rom one another and to ork together to improve both theory and practice. D&r goa% #ho&%d &%timate%y be to go beyond #imp%y &nder#tanding hy negotiation# #ometime# $ai% and #ometime# #&cceed. .t #ho&%d be to he%p &# overcome the barrier# and achieve more con#i#tent #&cce## in the negotiated re#o%&tion o$ con$%ict. /ootnote# d1 . o&%d %ike to grate$&%%y ackno%edge the va%&ab%e re#earch a##i#tance o$ Karen +enenba&m and L&ka#6 Ro6deic6er. (n ear%ier ver#ion o$ thi# artic%e a# pre#ented at a con$erence on Ca&#e# and Management o$ Con$%ict# in "oer%it6) Jermany in ?&ne) 2002) and a# previo&#%y p&b%i#hed a# part o$ the con$erence proceeding# in 1=9 ?. .n#tit&tiona% and +heoretica% 1con. 199 (2003). .t i# p&b%i#hed here ith the permi##ion o$ that Fo&rna%. 1 Robert H. Mnookin) "hy Negotiation# /ai%0 (n 12p%oration o$ 3arrier# to the Re#o%&tion o$ Con$%ict) B Dhio -t. ?. on Ci#p. Re#o%. 23= (1993). 2 Robert H. Mnookin Q Lee Ro##) 3arrier# to Con$%ict Re#o%&tion (Kenneth (rro Q Robert H. Mnookin et a%. ed#.) EDN 3ook# 1999). 3 -ee Cavid M. Krep#) Jame +heory and 1conomic Mode%ing (1990). 8 (nato% Rapoport) N*Eer#on Jame +heory @A (19A0). = .nternationa% M&%ti%atera% Negotiation (pproache# to the Management o$ Comp%e2ity 2i ("i%%iam :artman ed.) 1998). 2011 Thomson Reuters. No claim to original U.S. Government Works. 12 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 @ Hoard Rai$$a) +he (rt and -cience o$ Negotiation 11 (19B2). A Cavid (. La2 Q ?ame# K. -ebeni&#) +hinking Coa%itiona%%y0 Earty (rithmetic) Eroce## Dpport&ni#m) and -trategic -e,&encing) in Negotiation (na%y#i# 1=3 (H. Eeyton Oo&ng ed.) 1991) ;hereina$ter La2 Q -ebeni&#) +hinking Coa%itiona%%y<. B Robert Mnookin Q Lei# Kornha&#er) 3argaining in the -hado o$ La0 the Ca#e o$ Civorce) BB Oa%e L.?. 9=0 (19A9). 9 1%eanor 1. Maccoby Q Robert H. Mnookin) Cividing the Chi%d0 -ocia% and Lega% Ci%emma# o$ C&#tody (1992). 10 Robert Mnookin) 3eyond "inning0 Negotiating to Create Ga%&e in Cea%# and Ci#p&te# (2000). 11 -ee id. at 29=*318. 12 -ee id. at 9*10. 13 +hi# phra#e) hich ha# no entered the #tandard negotiation vocab&%ary) a# #&gge#ted $ir#t by /i#her and Hry. -ee Roger /i#her et a%.) Jetting to Oe#0 Negotiating (greement "itho&t Jiving .n 9A (19B3). 18 Morton C. Cavi#) Jame +heory 8= (19A0). 1= -ee Jeorge (. (ker%o$) +he Market $or 4Lemon#50 R&a%ity Hncertainty and the Market Mechani#m) B8 R.?. 1con. 8BB (19A0). 1@ Commitment #trategie# invo%ve per#&ading the other #ide that yo& have no $reedom o$ choice. .n a c%a##ic e2amp%e o$ten attrib&ted to +homa# -che%%ing) to car# begin to enter an inter#ection at the #ame time. 3oth driver# ant to get thro&gh the inter#ection $ir#t b&t neither ant# a co%%i#ion. 3&t i$ one driver co&%d detach hi# #teering hee% and con#pic&o&#%y thro it o&t o$ the car) the other driver o&%d have to permit him to pa## thro&gh. "hy' 3eca&#e the $ir#t driver ha# changed the #econd driver># option#. He m&#t no de$er or ca&#e a co%%i#ion) herea# be$ore there a# #ome chance that the other driver o&%d #%o don $ir#t. Commitment #trategie# come in many $orm# b&t to be e$$ective) a commitment m&#t #eem binding. Creating a commitment that i# credib%y irrever#ib%e i# not ea#y) beca&#e o$ten a party i%% be perceived a# having the poer to change co&r#e. -ee genera%%y +homa# C. -che%%ing) +he -trategy o$ Con$%ict (19@0). -ometime# commitment #trategie# invo%ve making promi#e# to third partie#. /or e2amp%e) a b&yer (ho a# a member o$ the Cemocratic Earty) might #omeho commit to paying I2)000 to a right* ing charity i$ he pay# a #e%%er more than IA)000 $or a good. +hi# commitment) i$ credib%e) might e%% per#&ade a re%&ctant #e%%er to accept the b&yer># 4$ina%5 take*it*or*%eave*it o$$er o$ IA)000. 1A Cavi#) #&pra note 18) at 119*20. 1B Krep#) #&pra note 3) at 92. 19 .d. at 98. 20 .d. at 9A. 21 .d. at 9A. 22 -ee ?ohn /. Na#h) ?r.) +he 3argaining Erob%em) 1B 1conometrica 1== (19=0). 23 -ee (rie% R&bin#tein) Eer$ect 1,&i%ibri&m .n a 3argaining Mode%) =0 1conometrica 9A (19=0). 2011 Thomson Reuters. No claim to original U.S. Government Works. 13 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 28 Krep#) $or e2amp%e) #&gge#t# that i$ ( can make o$$er# a$ter = #econd# hi%e it take# 3 1= #econd#) the R&bin#tein e,&i%ibri&m i%% give 3M8 o$ the #&rp%&# to p%ayer ( and on%y 1M8 to 3. -ee Krep#) #&pra note 3) at 12@*2A. 2= -ee Cre /&ndenberg Q ?ean +iro%e) Jame +heory 181 (1991). 2@ 1mpirica% re#earch a%#o #ho# that R&bin#tein># a%ternating o$$er mode% doe# not acc&rate%y de#cribe bargaining behavior in e2perimenta% #it&ation# &#ing R&bin#tein># protoco%#. -ee (%vin 1. Roth) +he Handbook o$ 12perimenta% 1conomic# 2=3*38B (?ohn H. Kage% Q (%vin 1. Roth ed#.) 199=). 2A .d. 2B -ee Ha% R. Garian) .ntermediate Microeconomic#0 ( Modern (pproach 1= (1999). 29 -ee Cynamic (dmini#tration0 +he Co%%ected Eaper# o$ Mary Earker /o%%ett (Henry C. Metca%$ Q L. Hrick ed#.) 1982)! Richard 1. "a%ton Q Robert 3. McKer#ie) ( 3ehaviora% +heory o$ Labor Negotiation#0 (n (na%y#i# o$ a -ocia% .nteraction -y#tem (19@=)! #ee a%#o Cean J. Er&itt) Negotiation 3ehavior (19B1)! Hoard Rai$$a) +he (rt and -cience o$ Negotiation 11 (19B2)! Cavid (. La2 Q ?ame# K. -ebeni&#) +he Manager a# Negotiator0 3argaining $or Cooperation and Competitive Jain (19B@)! /i#her et a%.) #&pra note 13. 30 +he i##&e# are di#c&##ed at %ength in Mnookin) 3eyond "inning) #&pra note 10) and in Mnookin Q Ro##) #&pra note 2) at 3*28. 31 Nicho%a# Ka%dor) "e%$are Eropo#ition# o$ 1conomic# and .nterper#ona% Compari#on# o$ Hti%ity) 89 1con. ?. =89 (1939). 32 -citov#ky and -am&e%#on #&b#e,&ent%y revea%ed tro&b%ing prob%em# ith Ka%dor*Hick# hen one compared circ&m#tance# here the to #tate# o$ the or%d invo%ved di$$erent combination# o$ good#. -ee (. /riedman) Ka%dor*Hick# Compen#ation) in +he Ne Ea%grave Cictionary o$ 1conomic# and La (E. Neman ed.) 199B). -ee genera%%y La2 Q -ebeni&#) +hinking Coa%itiona%%y) #&pra note A. 33 La2 Q -ebeni&#) +hinking Coa%itiona%%y) #&pra note A) at 821. 38 Rai$$a) #&pra note @) at 11. 3= -ee id. at 2=A*@2. 3@ Morton C. Cavi#) Jame +heory0 ( Nontechnica% .ntrod&ction 1B8 (19B3). 3A -ee genera%%y La2 Q -ebeni&#) +hinking Coa%itiona%%y) #&pra note A) at 1=3 (noting that #cho%ar# have #t&died the $ormation o$ coa%ition# ithin the#e in#tit&tion# and the e$$ect# o$ voting proced&re# and agenda #etting on negotiation# ithin %egi#%at&re#) internationa% a%%iance#) and in treaty making). 3B CDC1-( #tand# $or the Convention $or a Cemocratic -o&th ($rica) hich met in Cecember 1991. ( #econd CDC1-( took p%ace in May 1992. +he M&%ti*Earty Negotiating Eroce##) hich $ir#t convened in (pri% 1993) a# CDC1-(># #&cce##or. 39 Jov>t o$ the -e%$*Joverning +erritory o$ Ka:&%& v. Mah%ang&) 1998 (1) -( @2@) @30 (+EC). 80 .d. at @29. 81 .d. at @31. 2011 Thomson Reuters. No claim to original U.S. Government Works. 14 WILLIAM GOLDMAN 10/3/2011 For Educational Use Only !"A!EGI# $A""IE" !O DI%U!E "EOLU!ION& A'''( ) *ar+' Ne,ot' L' "e+' 1 82 .d. 83 +he ./E di#p&ted Mah%ang&># deci#ion and cha%%enged the &#e o$ #&$$icient con#en#&# in a -eptember 1993 -o&th ($rican -&preme Co&rt ca#e. +he Co&rt di#mi##ed the ca#e) citing the app%icant># inabi%ity to demon#trate that deci#ion# taken by #&$$icient con#en#&# ere %ega%%y binding. 88 Ka:&%&) 1998 (1) -( at @30. 8= Canie% /. C&rran and ?ame# K. -ebeni&#) +he Mediator a# Coa%ition 3&i%der0 Jeorge Mitche%% in Northern .re%and) B ?. .nt>% Negot. ($orthcoming 2003). 8@ +he Long ?o&rney0 -o&th ($rica># R&e#t $or a Negotiated -ett%ement 28*2= (-teven /riedman ed.) 1993). 8A C&rran Q -ebeni&#) #&pra note 8=. 8B .d. 89 +imothy C. -i#k) Cemocrati6ation in -o&th ($rica 20= (199=). =0 +he Long ?o&rney0 -o&th ($rica># R&e#t $or a Negotiated -ett%ement) #&pra note 8=) at 23. =1 -i#k) #&pra note 89) at 22@. =2 .d. =3 Harvard 3&#ine## -choo% Ca#e 9*B01*393) +o He%% "ith the /&t&re) Let># Jet Dn "ith the Ea#t0 Jeorge Mitche%% .n Northern .re%and 13 (Canie% /. C&rran and ?ame# K. -ebeni&# ed#.) 2001). =8 .d. at 1@. == -i#k) #&pra note 89) at 20A. =@ .d. =A .n it# deci#ion to di#mi## the ca#e) the -&preme Co&rt cited the ./E># inabi%ity to demon#trate that deci#ion# taken by #&$$icient con#en#&# ere binding. (# evidence) it pointed to the ./E># on a%ko&t) hich) it arg&ed) o&%d not have been po##ib%e i$ #&$$icient con#en#&# entai%ed a %ega%%y binding re%ation#hip. Ka:&%&) 1998 (1) -( at @30. =B Jeorge ?. Mitche%%) Making Eeace 110 (1999). =9 .d. @0 1amonn Ma%%ie Q Cavid McKittrick) 1ndgame in .re%and 23=*3@ (2001). -nd o$ Docu#ent S 2011 +hom#on Re&ter#. No c%aim to origina% H.-. Jovernment "ork#. 2011 Thomson Reuters. No claim to original U.S. Government Works. 15