THE ROLE OF MEDIATION IN CONFLICT MANAGEMENT CONDI

THEROLEOFMEDIATIONINCONFLICTMANAGEMENT:CONDITIONSFORSUCCESSFULRESOLUTIONJonathanWilkenfeld,UniversityofMarylandSaritKraus,BarIlanUniversityandUniversityofMarylandTaraE.Santmire,UniversityofMarylandChristopherFrain,UniversityofMarylandFebruary10,1999PreparedforpresentationattheAnnualMeetingoftheInternationalStudiesAssociation,February1999,WashingtonDC.1BackgroundThispapergrowsoutofaconferenceonMethodologyinInternationalRelationsconvenedatRiceUniversityJune26-28,1998.TheconferencewasorganizedbytheDivisionofSocialSciencesandtheJamesA.BakerIIIInstituteforPublicPolicyatRice,andthePrograminForeignPolicyDecisionMakingatTexasA&M.Theobjectiveoftheconferencewastoexplorewaysinwhichvariousmethodologicaltraditionsaddressasubstantivetopicininternationalrelationsresearch.ResearchdesignsweretobedevelopedanddiscussedattheRiceconference,followedbyfull-blownpaperspreparedforpresentationattheFebruary1999InternationalStudiesAssociationMeet-ingsinWashingtonDC.Ultimately,aneditedvolumecontainingthesepapersistobepublished.Fivemethodologicaltraditionsarerepresentedamongthepapers:rationalchoice/gametheory;dynamicmodeling;quantitativeapproaches;quasi-experimentalandsimulationapproaches;andqualitative/casestudyapproaches.Theunifyingfocusofthepapersisoninternationalconictmanagementandinternationalcon-ictresolutionresearch.Thepapersweretoaddresssomeorallofthefollowingquestions:(1)whatarethedierencesbetweenapproaches/procedures/methods/conceptionsofconictmanage-mentandconictresolution?Howcanwesystematicallydistinguishbetweenthetwoconcepts?(2)Howcanwetheoreticallyandempiricallydetectwhichprocessisunderway?(3)Whataretheconditions/circumstances/strategiesformovingfromconictmanagementtoconictresolution?(4)Whatistherelationshipbetweentheconditions/circumstances/strategiesoftransitionfrommanagementtoresolutionofinternationalconictsandtheoutcomesofresolutioneorts?Theoriginalplanforthepresentpapergrewoutofinitialdiscussionswithinthegroupchargedwithdevelopingresearchdesignsutilizingquasi-experimentalandsimulationapproachestothestudyofconictmanagementandconictresolution.Earlyinthesediscussions,itwasdecidedthatallofthepaperswouldfocusspecicallyontheroleofmediationintheprocessofmovingfromconictmanagementtoconictresolution.Infact,averyelaborateresearchdesignwasdevelopedwhichinvolvedinputfromallthosepresent-PhilSchrodt,JonWilkenfeld,CharlesTaber,NehemiaThismaterialisbaseduponworksupportedinpartbytheNationalScienceFoundationunderGrantNo.IRI-9423967.1Geva,andAlexMintz-whichwouldbringtogetherarticialintelligence,automatedeventsdata,simulation,andexperimentalapproachesinasinglepaper.Whilethissoonprovedunfeasibleforthisroundofpapers,thereremainssomecollaborationamongthefourresearchers,withtheintentthattheywilleventuallyjoinforcesinamulti-methodassaultonthisbroadresearchquestion.Thatis,WilkenfeldandGevawilldesignexperimentalsimulationenvironmentsinwhichtheroleofmediatorswillbeexaminedinseveralconictenvironments.Schrodtwillprovideempiricalinputfromreal-worldcasesofmediationbeginningwiththeArab-IsraeliconictandYugoslavia,andatest-bedforcomparingexperimentalwithempiricalresults,whileTaberwilleventuallydevelopanautomatednegotiationagentcapableofparticipationintheexperiments.Awordaboutthespecicpapertobepresentedbelow.Itispreliminary.Experimentalworkhassomeuniquerequirementswhichmakeitdiculttobringtoearlyclosure.Threeinparticularaectedthecurrenteort.First,whiletheresearchteamweassembledhadworkedtogetherprevi-ouslyonaseriesofexperimentalstudieshavingtodowithcrisissituations,anewconictscenariowasdevelopedspecicallyforthenewlineofresearchfocusingontheroleofmediation.Inaddition,theprojectwasintransitionfromtheuseofaspecicdecisionsupportsystemfocusingexplicitlyonahostagecrisissimulation,toageneralizeddecisionsupportsystemcapableofincorporatinganyconict/crisisscenarioanditsattendantparametersandutilityfunctions.Theseexperimentsweretobethersttoutilizethisnewsetofsoftwarepackages.Finally,wehaveonlyoneshoteachsemesteratagroupofstudentsattheUnivesityofMarylandwhowouldmakegoodsubjectsforourexperiments,andhencewewererestrictedinthedegreetowhichwecoulddevelopanelabo-rateexperimentaldesign.Forallofthesereasons(andseveralothersuniquetotheexperimentalenvironmentweused),theresultsoftheexperimentsruninfall1998,whileprovidingagooddealofinsightintotheroleofmediationinconictresolution,shouldbetakenasextremelypreliminary.2IntroductionDecisionmakersarefrequentlyoverwhelmedbythevastamountsofinformationwhichtheymustconsider.Often,theyareforcedtomakepartiallyinformeddecisionswhichignorecriticalissuesbecauseofthecomplexityofthesituationbeinganalyzed.Thistendencybecomesevenmorepronouncedininternationalcrisissituations,characterizedbythreattobasicvalues,nite(oftenshort)timeforresponse,andheightenedprobabilityofinvolvementinmilitaryhostilities[BrecherandWilkenfeld1997].Theseelementsallcontributetothepossibilitythatindividualdecisionmakerswillbeoverwhelmedwithdetailandpossiblyunabletoidentifyoptimaloutcomes,eitherindividuallyorcollectively.Thevastliteratureoncrisi