WewouldliketothankDanArielyNittaiBergmanBrunoBiaisArthurCampbellErnstFehrBobGibbonsLorenzGotteBengtHolmstromJonathanLevinStevenLippmanRobertMarquezMurielNiederle 1IntroductionPreviousres ID: 254242
Download Pdf The PPT/PDF document "IsPay-for-PerformanceDetrimentaltoInnova..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
IsPay-for-PerformanceDetrimentaltoInnovation?FlorianEdereryandGustavoMansozJuly14,2012AbstractPreviousresearchineconomicsshowsthatcompensationbasedonthepay-for-performanceprincipleise ectiveininducinghigherlevels WewouldliketothankDanAriely,NittaiBergman,BrunoBiais,ArthurCampbell,ErnstFehr,BobGib-bons,LorenzGotte,BengtHolmstrom,JonathanLevin,StevenLippman,RobertMarquez,MurielNiederle, 1IntroductionPreviousresearchonthee ectivenessofincentivepayarguesthatpayinganagentbasedonhisperformanceinducestheagenttoexertmoree orttherebyimprovingproductivity.Thereisampleevidencesupportingthisthesisindi erenttypesofstudies,rangingfromquasi-experimentsinempiricaldatatoeconomic eldexperimentsandlaboratorystudies.Forexample,Lazear(2000)showsthattheproductivityofwindshieldinstallersinSafeliteGlassCorporationincreasedwhenmanagementchangedtheircompensationfrom xedwagestopiece-ratepay.Shearer(2004) ndssimilarevidenceinarandomized eldexperimentwithCanadiantreeplanters.Dickinson(1999)showsthatsubjectsinalaboratoryexperimenttypemoreletterswhentheircompensationismoresensitivetoperformance.Asintheaboveexam-ples,mostoftheexistingevidenceofthee ectof nancialincentivesonperformancecomesfromstudyingsimpleroutinetasks,inwhiche ortisthemaindeterminantofproductivity.1Incontrast,asubstantialbodyofexperimentaland eldresearchinpsychologyprovidesevidencethat,intasksthatrequireexplorationandcreativity,pay-for-performancecanactu-allyundermineperformance.McGraw(1978),McCullers(1978),Kohn(1993),andAmabile(1996)summarizethe ndingsofthislineofresearchbystatingthatpay-for-performanceencouragestherepetitionofwhathasworkedinthepast,butnottheexplorationofnewuntestedapproaches.Thesestudiesconcludethatintasksthatinvolvecreativityandinno-vation,monetaryincentivesshouldnotbeusedatalltomotivateagents.Takentogethertheavailableevidenceseemstosuggestthatstandardperformancepayworkswellforcertaintypesoftasks(physicale ort)butnotforothers(creativity).Inthispaperwepresentevi-dencethatisfullyinlinewiththisview,but,moreimportantly,wealsoshowthatcorrectlystructuredperformance-based nancialincentivescanmotivateinnovationincreativetasksandcanleadtosuperiorresultsthanthoseachievedintheabsenceof nancialincentives.Our ndingsprovidesupportfortheideathatforanytask\thee ectsofincentivesdependonhowtheyaredesigned"(p.206)asGneezy,Meier,andRey-Biel(2011)suggestintheconclusionsoftheircomprehensivereviewarticleonincentivee ects.Inacontrolledexperimentalsettinginwhichsubjectsperformataskthatinvolvesatrade-o betweenexplorationandexploitation,weprovideevidencethatincentiveplansthattolerateearlyfailureandrewardlong-termsuccessleadtomoreinnovationandbetterper-formancethan xedwagesorstandardpay-for-performanceincentiveschemes.Wealso ndevidencethatthethreatofterminationcanundermineincentivesforinnovationwhilegoldenparachutescanalleviatetheseinnovation-reducinge ects.Finally,weshowthatrisk-aversion 1Detrimentalincentivee ectsone orthavealsobeenstudiedineconomicswithmorenuancedconclusionsastowhenmonetaryrewardsleadtobetterperformance(GneezyandRustichini2000a,GneezyandRustichini2000b,FreyandJegen2001,Ariely,Gneezy,Loewenstein,andMazar2009).1 isanimportantfactorinexplainingtheseresults.Ourresultshaveimplicationsforvarioussituationsinwhichenablingentrepreneurshipisanimportantconcern.Infact,\stimulatinginnovation,creativityandenablingentrepreneurship"isatoppriorityformanagementandwidelyregardedasthe\greatesthumanresourcechallenge"facingorganizationsaccordingtoCEOsurveys.2Thesesituationsrangefromthedesignofcompensationpackagesoftopexecu-tivesandmiddlemanagersinlargecorporationstostructuringcompensationforentrepreneursinstart-upcompanies.Forexample,someofthemostinnovation-andcreativity-driven rmsincludingpharmaceuticalandadvertisingcompanieshaveadoptedrewardsystemsthattol-erateorevenrewardthefailureofemployees.3Inthecontextofexecutivecompensation,toleranceforearlyfailurecanbeachievedthroughcommonlyusedpracticessuchasmanage-rialentrenchment,goldenparachutes,oroptionrepricing.Thesepracticesareoftencriticizedbecausetheyprotectorevenrewardthemanagerafterpoorperformanceand,therefore,caninducethemanagertoshirkordivertfundsfromthecorporation.Therehavebeensev-eralproposalstorestricttheuseofsomeofthesepractices.4Ourresultssuggestthatwhencombinedwithappropriatelong-termincentives,thesepracticescanbee ectiveinmotivatinginnovation.Regulationsthatrestricttheirusecouldthushaveanadversee ectoninnovation.Innovationisoftende nedastheproductionofknowledgethroughexperimentation(Arrow1969,Weitzman1979).AspointedoutbyMarch(1991),thecentralconcernthatariseswhenlearningthroughexperimentationisthetensionbetweentheexplorationofnewuntestedapproachesandtheexploitationofwell-knownapproaches.Asubstantialmanagementlit-eraturefollowingMarch(1991)hasinvestigatedtheorganizationalfactorsthatfavoreitherofthetwoformsofinnovationandhowtooptimallybalancethetensionbetweentheminambidextrousorganizations.5Thisliteraturehasabstractedawayfromhowdi erentincen-tivesystemsmighta ectthetrade-o ,anissuewhichwespeci callyaddressinthepresentpaper.Weconductaseriesoflarge-scaleeconomiclaboratoryexperimentswith379subjectsinwhichsubjectsfacethetensionbetweenexplorationandexploitation.Subjectscontroltheoperationsofacomputerizedlemonadestandandmustchoosebetween ne-tuningtheprod-uctchoicedecisionsgiventothembythepreviousmanagerorchoosingadi erentlocationandradicallyalteringtheproductmixtodiscoverabetterstrategyTostudytheimpactofdi erentincentiveschemesonproductivityandinnovationinthe 2In\CEOChallenge2004:PerspectivesandAnalysis,"TheConferenceBoard,Report1353,theauthorsreportthatthehighestresponse(31%)amongsurveyedCEOsisthatstimulatinginnovationisofthegreatestconcerntotheircompany.3Foradiscussionofthisissueinthepopularpresssee\BetterIdeasThroughFailure:CompaniesRewardEmployeeMistakesToSpurInnovation,GetBackTheirEdge",WallStreetJournal,September27,2011.4See,forexample,BebchukandFried(2004)and\RewardsforFailure,"BritishDTIconsultation,June2003.5AcomprehensiveoverviewofthisliteraturecanbefoundinthespecialissueonorganizationallearninginManagementScience(Argote,McEvily,andReagans2003).2 abovetasks,we rstconsiderthreedi erentbaselinetreatmentgroups.Theonlydi erencebetweenthesetreatmentgroupsisthecompensationo eredtosubjects.Subjectsinthe rstgroup( xedwage)receivea xedwageineachperiodoftheexperiment.Subjectsinthesecondtreatmentgroup(pay-for-performance)aregivenastandardpay-for-performance(orpro tsharing)contract,receivinga xedpercentageofthepro tsproducedduringtheexperiment.Subjectsinthethirdtreatmentgroup(exploration)areallocatedacontractthatistailoredtomotivateexploration.Theircompensationisa xedpercentageofthepro tsproducedduringthesecondhalfofeachofthetwoexperiments.6Ourmainhypothesisisthatsubjectsundertheexplorationcontractexploremoreandaremorelikelyto ndasuperiorstrategythansubjectsunderthe xed-wageorstandardpay-for-performancecontracts.Twofeaturesoftheexplorationcontractencouragesubjectstoexplore.First,toleranceforearlyfailurepermitssubjectstofailatnocostinthe rsthalfoftheexperimentwhiletheyexplorenewstrategies.Second,theprospectofpayforperformancelateronencouragessubjectstolearnbetterwaysofperformingthetask.There-fore,eventhoughtheexplorationcontractcombineselementsofthe xed-wageandpay-for-performancecontracts,itisplausiblethatperformanceundertheexplorationcontractissuperiortoperformanceunderboththe xed-wageandthepay-for-performancecontracts.Ourresultsstronglysupportthemainhypothesisstatedabove.We ndthatsubjectsgiventheexplorationcontractendtheexperimentinthebestlocation80%ofthetime,whilesubjectsgiventhe xed-wageandthepay-for-performancecontractsendtheexperimentinthebestlocationonly60%and40%ofthetime,respectively.Toexplainthesedi erencesweexaminethereasonsbehindtherelativelypoorperformanceofsubjectsunderthe xedwageandpay-for-performancecontracts.Althoughsubjectsgiventhe xed-wagecontractexplorealot,theyarenotassystematicintheirexplorationassubjectswhoaregivenanexplorationcontract.Forexample,whenweanalyzethenotessubjectstakeinatableweprovidetothematthebeginningoftheexperiment,we ndthatonly55%ofthesubjectsunderthe xed-wagecontractcarefullykeeptrackoftheirchoicesandpro ts;whenfacingtheexplorationcontract,82%ofthesubjectskeeptrackoftheirchoicesandpro tsusingthetable.Subjectsunderthepay-for-performancecontract,ontheotherhand,tendtodirecttheire orttowards ne-tuningthepreviousmanager'sproductmixinsteadofsearchingforbetterlocations.Duringthe rst10periodsoftheexperiment,subjectsundertheexplorationcontractchoosealocationotherthantheinitialdefaultlocation80%ofthetime,whilesubjectsunderthepay-for-performancecontractdosoonly50%ofthetime.Di erentattitudestowardsrisk 6ThisexplorationcontractisbasedonManso(2011)whoshowsthatthecombinationoftoleranceforearlyfailureandrewardforlong-termsuccessisoptimaltomotivateinnovation.Theexplorationcontractcanbeimplementedinpracticethroughstockgrantsthatincreasewiththetenureofthemanagerinthe rmindependentlyofthemanager'sperformance.Alternatively,itcanbeimplementedviastockoptiongrantswithlongvestingperiods,andoptionrepricingincasethemanagerperformspoorlyearlyon.3 canalsoa ecttheoutcomeunderthedi erentcontracts.We ndthatriskaversionplaysanimportantroleinexplainingdi erencesintheexplorationbehaviorandperformanceofthesubjectsunderthepay-for-performancecontract.Underthepay-for-performancecontract,morerisk-aversesubjectsarelesslikelyto ndtheoptimalstrategyandtheyobtainloweraveragepro tsthanlessrisk-aversesubjects.Tostudythee ectsofterminationoninnovationandperformance,weintroducetwonewtreatmentgroupsinthelemonadestandexperiment:aterminationtreatmentgroupandaterminationwithgoldenparachutetreatmentgroup.Subjectsinbothgroupsreceivetheexplorationcontractandarealsotoldthattheexperimentwillendearlyiftheirpro tsinthe rst10periodsarelowerthanacertainthreshold.Subjectsintheterminationwithgoldenparachutetreatmentgrouparetoldthattheywillreceiveareparationpaymentiftheexperimentendsafter10periods.Ourhypothesisisthatsubjectsintheterminationtreatmentarelesslikelyto ndtheoptimallocationthansubjectsintheexplorationtreatment.Wefurtherhypothesizethatsubjectsintheterminationwithgoldenparachutetreatmentgrouparemorelikelyto ndtheoptimallocationthansubjectsinthepureterminationtreatmentgroup.Thishypothesisissupportedbythedata:only45%ofthesubjectsintheterminationtreatmentgroup ndtheoptimallocationwhereas65%ofthesubjectsintheterminationwithgoldenparachutetreatmentgroup ndtheoptimallocation.Finally,wedemonstratethatourresultsarerobusttomodi cationsintheexperimentaldesign.We rstaddresspotentialsignalinge ectsofincentivecontractsinthelemonadestandexperiment.Inaddition,weshowthatinasecondexperiment,thegoldprospectingexperiment,whichusesthechosene ortapproachandawithin-subjectsdesignandallowsustofullycontrolthebeliefsofsubjectsinaparametrizedsetting,incentivecontractsthattoler-ateearlyfailureencourageexplorationandaresuperiortobothstandardpay-for-performanceand xedwages.Severalrecentpapersinvestigatethee ectsofinstitutionalfeaturesoninnovationsuchasdebtor-friendlybankruptcylaws(AcharyaandSubramanian2009),stringentlaborlaws(Acharya,Baghai-Wadji,andSubramanian2009),takeoverpressure(Sapra,Subramanian,andSubramanian2008,Atanassov2008,ChemmanurandTian2011),leverage(LiuandWong2011)aswellasthefailure-tolerantattitudeofinstitutionalinvestors(Aghion,Reenen,andZingales2008)andofventurecapitalists(TianandWang2010,Chemmanur,Tian,andLoutskina2011).Anotherburgeoningliteratureinvestigatestheinterplaybetweennon-competeagreementsandinnovation,inparticularinrelationtoCaliforniaandSiliconValley(Gilson1999,AlmeidaandKogut1999,SorensonandStuart2003),butcausalevidenceremainsthin(Fallick,Fleischman,andRebitzer2006,Marx,Strumsky,andFleming2009).Ourworkismostcloselyrelatedtocontributionsthatinvestigatetherelationbetween4 explicitincentivestructuresandinnovation.7Long-termincentivesfortheheadsofresearchanddevelopmentdepartments(LernerandWulf2007),goldenparachutesforCEOs(Francis,Hasan,andSharma2010),andlongerstockoptionvestingperiodlengths(YanadoriandMarler2006)areassociatedwithmore(heavilycited)patents.Finally,Azoulay,Zivin,andManso(2011)addresswhetherfundingpolicieswithtoleranceforearlyfailureandlonghori-zonstoevaluateresultsmotivatecreativityinscienti cresearch.Allthesepapersprovidesomesupportforthethesisthattoleranceforearlyfailurecoupledwithrewardforlong-termsuccessmotivatesinnovation.Becausetheyusenaturallyoccurringdata,however,theyeitherrestrictthemselvestomerelyshowingcorrelationpatterns,ortheyaresubjecttothecriticismthatthevariationintheincentiveschemesmaynotbecompletelyexogenous.Inourpaper,weareabletostudythee ectsofincentivesoninnovationbyexogenouslyvaryingcompensationschemesinacontrolledlaboratoryenvironment.Acommonapproachtothestudyofincentivesusinglaboratoryexperimentsistogivesubjectsacostfunctionandrequirethemtochooseane ortlevel(Bull,Schotter,andWeigelt1987,Fehr,Gachter,andKirchsteiger1997,NalbantianandSchotter1997).MeyerandShi(1995)andBanks,Olson,andPorter(1997)investigatethetensionbetweenexploita-tionandexplorationinanexperimentalsettingusingmonetarye ortsandreportresultsinbroadaccordancewiththetheoreticalbanditmodel.Neitherofthesepapersconsidersthee ectofdi erentincentiveschemesonsubjectbehavior.8Experimentalresearchershavealsoconductedincentivestudiesinboththelabandthe eldinwhichsubjectshavetoexertreale orttocompleteroutinetaskssuchastypingletters(Dickinson1999),decodinganumberfromagridofletters(Sillamaa1999),crackingwalnuts(FahrandIrlenbusch2000),stu -inglettersintoenvelopes(FalkandIchino2006),orpickingfruit(Bandiera,Barankay,andRasul2005).Thesetasks,however,areinadequatetostudyincentivesforinnovation.Inthispaper,wethereforeintroduceanewtasktoinvestigatethecausalimpactofincentivesonexploration.2ExperimentalDesignWeestablishanenvironmentinwhichwecanmeasurethee ectsofdi erentincentiveschemesoninnovationandperformance.Forthispurposeweconductexperimentsinwhichpartici-pantshavetosolveareale orttaskinwhichthetrade-o betweenexplorationandexploita- 7RelatedtheoreticalcontributionsthatalsofocusontheinterplayofincentivesandinnovationincludeAghionandTirole(1994),Hellmann(2007)andHellmannandThiele(2011).8OtherrelatedexperimentalpapersusingsearchproblemsakintoourlemonadestandexperimentanddealingwithinnovationmodeledasacomputationallycomplexproblemincludeGabaix,Laibson,Moloche,andWeinberg(2006)andMeloso,Copic,andBossaerts(2009).5 tiontakescenterstage.2.1ProceduresandSubjectPoolTheexperimentswereprogrammedandconductedwiththesoftwarez-Tree(Fischbacher2007)attheHarvardBusinessSchoolComputerLaboratoryforEconomicResearch(HBSCLER).ParticipantswererecruitedfromtheHBSCLERsubjectpoolusinganonlinere-cruitmentsystem.Atotalof379subjectsparticipatedinourexperiments.Aftersubjectscompletedtheexperimentweelicitedtheirdegreeofriskaversionandam-biguityaversion.Wedescribetheexactproceduresforriskandambiguityaversionelicitationintheappendix.9Subjectswerethenprivatelypaid.Asessionlasted,onaverage,60minutes.Experimentalcurrencyunitscalledfrancswereused.Theexchangeratewassetat100francs=$1andtheshow-upfeewas$10.Subjectsonaverageearned$24.2.2TaskDescriptionSubjectstaketheroleofanindividualoperatingalemonadestand.Theexperimentlasts20periods.Ineachperiod,subjectsmakedecisionsonhowtorunthelemonadestand.Thesedecisionsinvolvethelocationofthestand,thesugarandthelemoncontent,thelemonadecolorandtheprice.Asisnaturalforthistypeoftask,someofthevariablesarediscretechoices(location,color)whileothersaremorecontinuous(sugar,lemoncontent,price)thusyielding6,181,806possiblecombinations.Theexactparametersofthegameareprovidedintheappendix.Attheendofeachperiod,subjectslearnthepro tstheyobtainedduringthatperiod.Theyalsolearncustomerreactionsthatcontaininformationabouttheirchoices.Customerfeedbackisimplementedbyhavingthecomputerrandomlyselectoneofthecontinuouschoicevariablestoprovideabinaryfeedbacktothesubject.Thisfeedbackisonlyinformativeforthelocationinwhichthesubjectchosetosellinthecurrentperiod.10Subjectsdonotknowthepro tsassociatedwitheachoftheavailablechoices.Attachedtotheinstructionsisaletterfromthepreviousmanagerwhichisreproducedintheappendix.Thelettergiveshintstothesubjectsaboutastrategythathasworkedwellforthismanagerando ersanaccuratedescriptionofagoodbusinessstrategyforoneparticularlemonade 9Forourstatisticalanalysiswesplitsubjectsintoalessandamorerisk/ambiguity-aversegroupbasedonthemedianobservationforeachmeasureinthesample.10Forexample,ifthesubjectchoseasugarcontentthatisabovetheoptimallevelfortheparticularsaleslocation,thefeedbacktakestheform:\Manyofyourcustomersatthislocationtoldyouthatthelemonadeistoosweet."Wechosetogiveonlylimitedfeedbackbecause,inpractice,itisexpensivefora rmtocollectinformationfromcustomersandfeedbackisonlylikelytobeforthcomingforcombinationsthathavebeentriedbythe rm.6 standlocation.Thestrategysuggestedbythepreviousmanagerinvolvessettingthestandinthebusinessdistrict,choosingahighlemoncontent,alowsugarcontent,ahighpriceandgreenlemonade.Themanager'sletteralsostatesthatthemanagerhastriedseveralcombinationsofvariablesinthebusinessdistrictlocation,butthathehasneverexperimentedwithsettingupthestandinadi erentlocation.Itfurthersuggeststhatdi erentlocationsmayrequireaverydi erentstrategy.Theparticipantsintheexperimentthusfacethechoicebetween ne-tuningtheproductchoicedecisionsgiventothembythepreviousmanager(exploitation)orchoosingadi erentlocationandradicallyalteringtheproductmixtodiscoveramorepro tablestrategy(ex-ploration).Thestrategyofthepreviousmanagerisnotthemostpro tablestrategy.Themostpro tablestrategyistosetthelemonadestandintheschooldistrict,andtochoosealowlemoncontent,ahighsugarcontent,alowpriceandpinklemonade.Thepayo sinthegamewerechoseninsuchawaythatwithoutchangingthedefaultlocationtheadditionalpro tsearnedfromimprovingthestrategyinthebusinessdistrictarerelativelysmall.Ontheotherhand,changingthelocationtotheschoolrequiredlargechangesinatleasttwoothervariablestoattainanequallyhighpro tassuggestedbythedefaultstrategy.Inadditiontothepreviousmanager'sletter,theinstructionscontainatableinwhichsubjectscaninputtheirchoices,pro ts,andfeedbackineachperiod.Subjectsaretoldthattheycanusethistabletokeeptrackoftheirchoicesandoutcomes.Weusetheinformationsubjectsrecordinthistableasonemeasureoftheire ortduringtheexperiment.2.3TreatmentGroupsandTestableHypothesesWeinitiallyimplementthreetreatmentconditionsinordertoexaminehowdi erentincentiveschemesa ectinnovationsuccess,explorationbehavior,timeallocationande ortchoices.Theonlydi erencebetweenthegroupsisthewaysubjectsarecompensated.Thethreeincentiveschemesareasfollows:IncentiveScheme1(FixedWage):\Youwillbepaida xedwageof50francsperperiod."IncentiveScheme2(Pay-for-Performance):\Youwillbepaid50%ofthepro tsyoumakeduringthe20periodsoftheexperiment."IncentiveScheme3(Exploration):\Youwillbepaid50%ofthepro tsyoumakeduringthelast10periodsoftheexperiment."Therewere51,46and47subjectsineachofthesethreetreatments.Ourmainhypothesisconcernstheextenttowhichthedi erentpaymentschemesconsideredinourtreatmentgroupsa ecttheexplorationactivityofsubjects.Inparticular,buildingonthetheoreticalpredictionsofManso(2011)wehypothesizethatsubjectsundertheexplorationcontract7 conditionshould ndtheoptimalbusinessstrategymoreoftenthansubjectsintheothertreatments.MainHypothesisSubjectsundertheexplorationcontractchooseabusinessstrategythatisclosertotheoptimalbusinessstrategythansubjectsunderthe xed-wageandpay-for-performancecontracts.Theabovehypothesisconstitutesthecentralpartofthepresentpaperandmayappearsurprisingat rstglance,sincetheexplorationtreatmentisahybridof xedwageandpay-for-performancetreatments.Thus,aplausiblenullhypothesiswouldbethatthechoicesandperformanceofsubjectsintheexplorationtreatmentlieinbetweenthoseofthe xedwageandthepay-for-performancetreatments.Similarly,anullhypothesisbasedonmodelsofrepeatede ort(Rogerson1985,HolmstromandMilgrom1987,Sannikov2008)wouldpredictthattheexplorationcontractinducesshirkingduringthe rsttenperiodsandhighe ortandgoodperformanceinthelattertenperiodsoftheexperiment.Thus,incentiveschemesthattolerateearlyfailureareinadequatebecausetheyleadtolowere ortandproductivitythanstandardpay-for-performanceincentiveschemes.However,asManso(2011)formallyshowsinatheoreticalmodelthatfeaturesthesametrade-o betweenexploitationandexplorationinadditiontoacostlye ortchoiceasinourexperiment,contractssuchastheexplorationcontractusedherearee ectiveinmotivatingexploration(andconsequentlyinnovation).11Therearetwoimportantreasonsforthedi erenceinperformancebetweensubjectsintheexplorationtreatmentandsubjectsintheothertwotreatments.First,sincethecompensationofsubjectsunderthepay-for-performancecontractdependsontheirperformancefromthevery rstperiod,theyarelesswillingtoexplorethansubjectsundertheexplorationcontract.Thisisbecauseasubjectwhoisgiventhepay-for-performancecontractanduseshis rstfewperiodstoexploredi erentstrategies,islikelytoobtainlowerpro tsandwouldonlylowerhiscompensationduringthoseperiodsascomparedwiththecompensationhewouldreceivefromthedefaultstrategy.Inanenvironmentwhereexplorationisdesirable,pay-for-performancecompensationthereforeleadstoinsucientexploration.Incontrast,theexplorationtreatmentwithitstoleranceforearlyfailureencouragessubjectstoexploreearlyon.Thus,relativetothepay-for-performancecontract,theexplorationcontracttiltsthescalestowardsexplorationforthesubjects'exploitation-explorationtrade-o .Second,subjectswhoaregivena xedwagecontractshouldalsobewillingtoexplorebecausetheydonotfaceanycostsfromfailingwhiletheyexploredi erentstrategies.Thus,onemightreasonablyexpectthatexplorationactivityisequallyhighamongsubjectsinthe 11ForafulldiscussionoftheparticularmodelandformalresultswerefertheinterestedreadertoManso(2011),SectionIV.B.Here,wefocusontheintuitionofthemain ndingsthatarerelevantforthepresentpaper.8 xedwagetreatment.However,undera xed-wagecontractsubjectsdonothaveanyexplicitincentivesforgoodperformanceandtheyshouldminimizethecostlycontemplatione ortnecessaryto ndthebestbusinessstrategy.Thus,thefailureofthe xedwagecontractisduetothelackofincentivestoexertcostlye ortthatitprovidestothesubjects.Inthetwofollowingsub-hypothesesweinvestigateindetailtheproblemsofthe xedwageandpay-for-performancecontractsinexplorationsettings.ExplorationSub-HypothesisSubjectsundertheexplorationcontractaremorelikelytoexplorethansubjectsunderthepay-for-performancecontract,whoaremorelikelytofocusonexploitationactivities.Sincethepay-for-performancecontractfailstoencourageexploration,wehypothesizethatsubjectswillexplorelessthansubjectsundertheexplorationcontract.Whiletheaboveexplorationhypothesisexplainsthedi erentiale ectsofexplorationandpay-for-performancecontracts,itdoesnotpredicthowsubjectsunderthe xed-wagecontractwillbehave.Subjectsunderthe xedwagecontractdonothaveexplicitincentivesforgoodperformance.Wethereforehypothesizethattheywillminimizetheire ort.ShirkingSub-HypothesisSubjectsunderthe xed-wagecontractexertlesse ortthansubjectsundertheexplorationcontract.Whilewepredictthatsubjectsundertheexplorationcontractaremorelikelytoexplorethansubjectsunderthepay-for-performancecontractandlesslikelytoshirkthansubjectsinthe xed-wagecontract,itneednotbethecasethattheyalsoproducebetteraverageperformancethansubjectsunderthesetwoothercontracts.InManso(2011),agentsareassumedtoberisk-neutral.Incontrast,inourexperimentsomeofthesubjectsarerisk-averseandthereisconsiderableheterogeneityinattitudesto-wardsrisk.Whilerisk(orloss)aversiondoesnotqualitativelychange(butactuallystrength-ens)themainresultsregardingtoleranceofearlyfailure,itintroducesinterestingheteroge-neoustreatmente ectsbetweenmoreandlessrisk-aversesubjects.Compensationispoten-tiallyquitevariableinthepay-for-performancetreatment,inparticularwhensubjectschoosetoexplorebydivergingfromthedefaultstrategy.Thus,inthepay-for-performancetreatmentweexpectmorerisk-aversesubjectstoexplorelessthanlessrisk-aversesubjects.Ontheotherhand,wepredictthatthereisnosigni cantdi erenceinexplorationbehaviorbetweenmoreandlessrisk-aversesubjectsinthe xedwagetreatmentwherecompensationisconstantbydesign.Wethenturntoinvestigatinghowthethreatofearlyterminationin\ruencesexplorationbehaviorandperformance.12Earlyterminationcanunderminetheexplorationbehaviorin- 12SeealsoManso(2011),SectionVI.9 ducedbytheexplorationcontractbyeliminatingthetoleranceforearlyfailure.Wefurthershowthatthise ectcanbemitigatedbytheuseof\goldenparachutes"orreparationpay-mentswhichsubjectsreceiveincaseofearlyterminationsincethesepaymentsreintroducethetoleranceforearlyfailurethatisrequiredtoencourageexploration.Weintroducetwoadditionaltreatmentgroupsthatenableustoinvestigatethee ectsofterminationandgoldenparachutes.IncentiveScheme4(Termination):\Youwillbepaid50%ofthepro tsyoumakeduringthelast10periodsoftheexperiment.However,ifthepro tsyoumakeduringthe rst10periodsoftheexperimentarebelow800francs,theexperimentwillendafter10periods."IncentiveScheme5(TerminationwithGoldenParachute):\Youwillbepaid50%ofthepro tsyoumakeduringthelast10periodsoftheexperiment.Ifthepro tsyoumakeduringthe rst10periodsoftheexperimentarebelow800francs,theexperimentwillendearlyandyouwillreceiveapaymentof250francs."Therewereatotalof71and78subjectswhoparticipatedintheterminationandthegoldenparachutetreatments.Pureterminationinhibitsexplorationactivitiesbecauseitdoesnoto ersucienttoleranceforearlyfailure.13Whilethethreatofterminationproducesstrongincentivesforgoodperformance,italsoforcesindividualstofocusonproducinggoodperformancefromtheverybeginningandthusreducestheincentivesforexploration.Incon-trast,inthegoldenparachutetreatmentweexpectsubjectstoexplorealittlemoreintensivelythanintheterminationtreatmentatthebeginningoftheexperimentdespitethependingthreatofterminationsincethegoldenparachutepaymentprovidesthemwithsomeinsuranceincaseoffailure.TerminationHypothesisSubjectsundertheterminationcontractarelesslikelyto ndtheoptimalbusinessstrategythansubjectsundertheexplorationtreatment.Furthermore,subjectsunderthegoldenparachutetreatmentaremorelikelyto ndtheoptimalbusinessstrategythansubjectsintheterminationtreatment.3ResultsWepresenttheresultsobtainedinourexperimentscomparingtheoutcomeacrossthe vedi erenttreatments( xed-wagecontract,pay-for-performancecontractandexplorationcon-tract,termination,terminationwithgoldenparachute).Subjectswererandomlyassignedto 13Thepredictionthatterminationhasanadversee ectonexplorationdependscruciallyonourchoiceoftheterminationthresholdwhichischosensuchthatitcanbeachievedwithoutexploring.10 thedi erenttreatmentsandtherearenosigni cantex-postdi erencesinage,gender,riskaversionandself-reportedincome.3.1Innovation,ExplorationBehaviorandE ortChoiceWe rstfocusontheexplorationbehaviorofsubjectsacrossthe rstthreedi erenttreat-ments.Subjectsdonotmakeanydecisionsthatin\ruenceanyothersubjectsandwehaveatotalof20perioddecisionsforeachofthe379individualsubjects.However,sincedecisionsarecorrelatedovertimeforeachsubject,eachindividualsubjectparticipatingintheexper-imentconstitutesonlyoneindependentobservation.Unlessotherwisenoted,nonparametrictestsarethereforebasedonaveragesoftherelevantvariablesoftheindividualsubjectandthroughoutouranalysisweclusterstandarderrorsattheleveloftheindividualsubject.Our rstresultshowsthatthepredictionthattheexplorationcontractleadstomoreinnovationthantheothertwocontractsiscon rmedbythedata.Result1(innovation):Subjectsunderthe xed-wageandpay-for-performancecontractsaresigni cantlylesslikelytochoosetosellattheschool(highestpro tlocation)inthe nalperiodoftheexperimentthansubjectsundertheexplorationcontract.Subjectsundertheexplorationcontractcomeclosest(intermsofmaximumandlast-periodpro t)to ndingtheoptimalbusinessstrategy.InitialsupportingevidenceforResult1comesfromFigure1whichshowstheproportionofsubjectsunderthe xed-wage,pay-for-performance,andexplorationcontractconditionschoosingtoselllemonadeinaparticularlocationinthe nalperiod.Consistentwithourexplorationhypothesis,subjectsundertheexplorationcontractsettingaremorelikelytosellattheschoolwhichisthelocationwiththehighestpro tsinthe nalperiodoftheexperimentthansubjectsunderthe xed-wageandpay-for-performanceconditions.Whereasintheexplorationcontractconditionmorethan80%ofsubjectschoosetoselllemonadeattheschool,only40%ofsubjectschoosetodosointhepay-for-performanceconditionand60%choosetodosounderthe xed-wagecontract.UsingMann-Whitney-Wilcoxonranksumtestsoftheindividualsubjectaverageswecanshowthatthesedi erencesarehighlysigni cantbetweentheexplorationcontractandthe xed-wagecontract(p-value0.0042)andtheexplorationandthepay-for-performancecontract(p-value0.0001).Thedi erenceislessmarkedbetweenthe xed-wageandthepay-for-performancecontract(p-value0.0865).14 14Inaddition,weestimatedasubject xed-e ectslogitmodelwherethedependentbinaryvariabletakesthevalue1ifthe nallocationchoiceistheschoolwhichistheoptimallocationchoiceintheexperiment,and0otherwise.Theindependentvariablesarebinaryvariablesforthethreedi erentcontracts.Asbefore,thecoecientestimatesshowthatsubjectsunderthepay-for-performance(p-value0.0001)and xed-wagecontract(p-value0.0054)aresigni cantlylesslikelytochoosetosellintheschoolinthe nalperiodofthe11 0 .2 .4 .6 .8Proportion of subjects by location in final period Fixed wageP-f-P contractExploration contract Business districtSchool Figure1:Proportionofsubjectsbylocationinthe nalperiodoftheexperimentforthe xed-wage,pay-for-performanceandexplorationcontracts.Finally,thedi erenceinperformancebetweenthe xedwageandthepay-for-performancetreatmentisinlinewiththenegativee ectsofperformancepayfoundinthepsychologyliterature.Wealsoexaminehowclosesubjectscometo ndingtheoptimalstrategyoverthecourseoftheexperimentintermsofthepro tstheyachieve.Thiscaneasilybemeasuredbyexaminingthemaximumperperiodpro tachievedbysubjectsthroughoutthecourseoftheexperiment.Perperiodpro tisamorecomprehensivemeasurethanlocationchoice.Itcapturesthemulti-dimensionalaspectofthetaskwhichinvolvesthechoiceofseveralvariables.Onaverage,subjectsundertheexplorationcontractachievethehighestmaximumperperiodpro ts(145francs)whilesubjectsunderthe xed-wage(128francs)andthepay-for-performance(117francs)contractsperformworseonthisdimension.Thesamepatternholdsfor nalperiodpro twheretherespectivevaluesare140(exploration),120( xedwage)and111francs(pay-for-performance).Asbeforethedi erencesinmaximumperperiodpro taswellas nalperiodpro tbetweentheexplorationcontractandtheothertwocontractsarehighlysigni cant(Mann-Whitney-Wilcoxontest:p-valuesof0.013and0.0001formaximumpro t,p-valuesof0.009and0.0001for nalperiodpro t)whilethedi erencebetweenthe xed-wageandthepay-for-performancecontractisnotstatisticallysigni cant(p-value0.1144formaximumpro t,p-value0.28for nalperiodpro t).Toexplainwhysubjectsundertheexplorationcontractaremorelikelyto ndtheopti-mallocationandbusinessstrategythansubjectsundertheothertwocontracts,weanalyze experimentthansubjectsintheexplorationcontract.Thenegativee ecton ndingtheoptimallocationinwhichtosellisparticularlypronouncedforthepay-for-performancecontractwhilethedi erencebetween xed-wageandpay-for-performancecontractsisnotassigni cant(p-value0.0865).12 di erentmeasuresofexplorationande ort.Thenextresultshowsthatsubjectsundertheexplorationcontractexploremorethansubjectsunderthe xed-wagecontractwhilesubjectsunderthepay-for-performancecontractexploretheleast.Result2(explorationbehavior):Subjectsunderthepay-for-performancecontractexplorelessthansubjectsunderthe xed-wagecontractandtheexplorationcontractwiththelatterexploringthemost.Usingthedi erentchoicevariablesavailabletotheagentswecanconstructseveralmea-suresofexplorationactivity.Subjectsinthepay-for-performanceconditionexplorelocationsotherthanthedefaultlocation(businessdistrict)lessoftenthansubjectsundertheothertwocontractswithsubjectsundertheexplorationcontractchoosingtoexplorethemostof-ten.Whilesubjectsundertheexplorationcontractchoosealocationotherthanthedefaultlocationin82%and85%ofcasesinthe rstandthelast10periods,subjectsunderthe xed-wagecontractchoosetodosoonlyin60%and63%ofcasesandtheproportionsareaslowas51%and48%forsubjectsinthepay-for-performancecontract.Thetoleranceforearlyfailureoftheexplorationcontractrelativetothe xed-wageandpay-for-performancecontractsencouragedindividualstoattemptnewuntriedapproachesinthe rst10periods.UsingMann-Whitney-Wilcoxonranksumtestsoftheindividualsubjectaveragesrevealsthatthisdi erenceinlocationchoicebehaviorbetweenthedi erentcontractsisstatisticallysig-ni cant.Inthe rst10periodssubjectsundertheexplorationcontractchoosetoexploreadi erentlocationmoreoftenthansubjectsunderthe xed-wagecontract(p-value0.0053)andthepay-for-performancecontract(p-value0.0001).Thedi erenceinexplorationbehaviorasmeasuredbylocationchoiceinthe rst10periodsisnotstatisticallysigni cantbetweensubjectsunderthe xed-wageandthepay-for-performancecontracts(p-value0.1482),butsubjectsunderthe xed-wagecontractchoosetoexploresigni cantlymoreoftenthansub-jectsunderthepay-for-performancecontractinthelast10periodsoftheexperiment(p-value0.0985).Thisparticularformofexplorationactivityisalsore\rectedinFigure2whichshowstheaveragesubject-speci cstandarddeviationinstrategychoicesforthethreecontinuouschoicevariables(sugarcontent,lemoncontentandprice)duringthe rstandlast10periodsoftheexperiment.Thisstandarddeviationmeasurecapturesvariationinallthevariablesofthismulti-dimensionalchoiceproblem.First,thevariabilityofactionchoicessigni cantlydeclinesoverthecourseoftheex-perimentinthepay-for-performance(p-value0.0005)andtheexplorationcontracts(p-value0.0001).Thisoccursbecauseinperiods11to20thebene ciallearninge ectsofexplorationrelativetoexploitationarenolongeraslargeasatthebeginningoftheexperimentsincethetimehorizonisshorter.Incontrast,thevariabilityofactionchoicesonlydecreasesslightlyin13 .4 .6 .8 1 1.2Standard deviation of strategy choices Fixed wageP-f-P contractExploration contract Periods 1-10Periods 11-20 Figure2:Averagesubject-speci cstandarddeviationofstrategychoicesforthethreecontinuousvariables(sugarcontent,lemoncontent,price)inperiods1-10and11-20oftheexperimentforthe xed-wage,pay-for-performanceandexplorationcontracts.the xed-wagecontractandthisdeclineisnotstatisticallysigni cant(p-value0.2194).Sinceagentsarenotpenalizedforlowpro ts,explorationbehaviorinthe xed-wagecontractisexclusivelydrivenbyintrinsicmotivesandsubjectsmaythereforecontinuetoexploreeventhoughtheadditionalbene tsofexplorationaresmall.Second,thevariabilityofactionchoicesinthe rst10periodsissigni cantlyhigherintheexplorationcontractthaninthepay-for-performance(p-value0.0012)andthe xed-wagecontracts(p-value0.0027).Thisshowsthatsubjectsundertheexplorationcontractexperimentandconsciouslymakeverydi erentactionchoicesinadirectedattemptto ndmorepromisingstrategies.Incontrast,inthepay-for-performancecontractthestandarddeviationofactionchoicesismuchlowerassubjectsoptto ne-tunethedefaultvalues.Thisisalsotrueforsubjectsunderthe xed-wagecontractwhoexplorelessthansubjectsundertheexplorationcontractduringthe rst10periods.However,becausesubjectsintheothertwotreatmentsexploremuchlessinthelaterperiodsoftheexperimentwhentheircompensationisdirectlylinkedtotheirperformance,thevariabilityofactionchoicesofsubjectsunderthe xed-wagecontractishigher(thoughnotalwayssigni cantlyso)thaninthepay-for-performance(p-value0.0246)andtheexplorationcontracts(p-value0.6567).Therelativelyhighexplorationbehaviorofsubjectsunderthe xed-wagecontractinthelast10periodsoftheexperimentalsoexplainswhytheyaremorelikelyto ndthehighest-pro tlocationthansubjectsunderthepay-for-performancecontractwhoexploretheleastovertheentirecourseoftheexperimentamongthethreecontracttreatmentgroups.Wealsoexpectthevariabilityofpro tstomirrorthevariabilityofactionchoices.Thisisindeedthecase.First,thevariabilityofpro tssigni cantlydeclinesovertimewiththe14 declineinvariabilitybeingparticularlymarkedfortheexplorationcontractandthepay-for-performancecontracts.Second,thevariabilityofpro tsinthe rst10periodsissigni cantlyhigherforsubjectsundertheexplorationcontractthansubjectsundertheothertwocontracts,whilethereisnosigni cantdi erenceinpro tvariabilityacrosssubjectsunderthethreecontractsinthelast10periods.WeuseCoxhazardratemodelstoanalyzethedynamicsthatgovernthestrategychoicesofindividualsintheexperiment.Thisallowsustotestwhetherthedi erenttreatmentcondi-tionsalsoin\ruencewhether,oncetheyhavedecidedtoexplore,subjectscontinuetoexploreandwhatotherfactorscontributetomakingthempersistintheirexplorationactivities.Weclassifysubjectsashavingenteredanexplorativephaseassoonastheychoosealocationotherthanthedefaultlocation(businessdistrict)suggestedbythepreviousmanager.Anexplorativephaseendswhensubjectsmakeonlysmallchangestostrategychoicesrelativetothepreviousperiodorswitchbacktothedefaultlocation.15Ascanbeseenfromcolumn1ofTable1,thehazardrateofendinganexplorativephaseissigni cantlyhigherunderthepay-for-performancecontractthanundertheexplorationcontract.Thehazardrateisalsohigherinthe xed-wagecontractalthoughthise ectisnotstatisticallysigni cant.Moreover,higherpro tssigni cantlydecreasethehazardrateassubjectsareencouragedtopersistintheirexploratione ort.Column2ofTable1showsthattheestimatesforthe rst10periodsarequalitativelysimilar.Finally,answersintheopen-endedpost-experimentalquestionnaireinwhichallsubjectswereaskedtodescribetheirstrategiesandthee ectthecompensationschemehadontheirchoicesalsore\rectedthedescribedexplorationpattern.Subjectsundertheexplorationcon-tractspontaneouslyarguedthatthetoleranceforearlyfailureofthecompensationschemeaswellasthestrongrewardsforsuccessinlaterperiodsin\ruencedtheirstrategicchoices,causingthemtoexperimentwithuntestedlocationsandactionchoicesearlyonandthentochooseand ne-tunethebestavailablestrategybeginninginperiod11.Sofar,ourresultshavelargelyfocusedonexplorationbehavior.However,wealsopredictedthatsubjectsunderthe xed-wagecontractshouldexertlesse ortthansubjectsundertheothertwocontractssincetheircompensationdoesnotdependontheirperformanceintheexperiment.Result3(timeallocationande ortchoice):Subjectsunderthe xed-wagecontractspendlesstimemakingandevaluatingdecisionsandexertlesse ortrecordingtheirpreviouschoicesandoutcomesintheexperimentthansubjectsunderthepay-for-performanceandexploration 15Inparticular,anexplorativephaseisde nedasendingwhenasubjectswitchesbacktothedefaultlocationorwhenasubjectdoesnotchangelocationandlemonadecolorandalsodoesnotchangelemoncontent,sugarcontentandpricebymorethan0.25units.Asarobustnesscheckwealsousedotherde nitionsthresholdsfortheendofanexplorationphase.Theresultingmagnitudesandsigni cancelevelsareverysimilar.15 CoxHazardRateModelsPeriod1-20Period1-10Period1-20Period1-10b/seb/seb/seb/se rhFixedWage0.2170.564***(0.136)(0.204)Pay-for-Performance0.334***0.632***(0.126)(0.201)Termination0.487***0.861***(0.110)(0.185)Parachute0.297**0.580***(0.123)(0.199)tPro ts-0.001***-0.001***-0.001***-0.001***(0.000)(0.000)(0.000)(0.000) Pseudo-R20.0110.0130.0140.015N2418106829951458 Table1:EstimatesfromaCoxhazardratemodelreportingthehazardratesforexitinganexplorationphasewiththeexplorationcontractasthebaseline.Separateestimationsareshownfortheentire20periodsoftheexperimentandthe rst10periods.Robuststandarderrorsarereportedinbrackets.Statisticalsigni canceusingclusteringattheindividualsubjectlevelattheten, veandonepercentlevelisindicatedby*,**and***.contracts.Aprincipaldecidingwhethertopayagentsa xedwagemightworrythatabsentanyintrinsicmotivationandimplicitincentivestheagentwillchoosetominimizecostlye ort.Similarly,inourexperiment|wheresubjectshavetomentallyfocusandrecordpastchoicestotrytomaximizetheirperformance|subjectswhosecompensationdoesnotdependontheirperformancemaychoosetominimizecostlyandtime-consumingcontemplationanddeliberatione ort.Indeed,manysubjectsunderthe xed-wagecontractclaimedinthepost-experimentalquestionnairethattheyattemptedtominimizethetimeande ortnecessarytocompletetheexperimentsincetheirperformancedidnota ecttheircompensation.Thispatternisalsoborneoutinourdata.Whilesubjectsunderthe xed-wagecontractspendonlyanaverageof24secondsonthedecisionscreen(wheresubjectsentertheirstrategychoices),subjectsundertheexplorationandthepay-for-performancecontractsspend31and30secondsrespectively.Thatis,overtheentiredurationoftheexperiment,subjectsundertheexplorationandthepay-for-performancecontractconditionspendalmost30%moretimeonthedecisionscreenthansubjectsunderthe xed-wagecondition.Thesedi erencesarestatisticallysigni cant(p-valuesof0.0014and0.0175)overthecourseoftheentireexperimentaswellasinsubperiods.Moreover,subjectsintheexplorationcontracttreatmentspendsigni cantlymoretimeonthedecisionscreenthansubjectsinthe xed-wagetreatment(p-value0.022)evenduringthe rst10periodsof16 theexperimentwhentheyreceivenocompensation.Thisdi erenceintimespentbetweentheexplorationandpay-for-performancecontractsisnotsigni cant(p-value0.8477).Thisevidencestandsincontrasttodynamicprincipal-agentmodelsofrepeatede ort(Rogerson1985,HolmstromandMilgrom1987,Sannikov2008),whichpredictthattheex-plorationcontractshouldinducemoreshirkingduringthe rsttenperiodsoftheexperimentthanthepay-for-performancecontract.Thesemodelsfailtoincorporatethelearningpro-ducedbytheexplorationofnewstrategies,whichpotentiallyenhancesperformanceinlaterperiods,andmaythusprovideincentivesfortheagenttoexerte ortinearlyperiods,evenwhenhiscompensationdoesnotdependonproductivityinthoseearlyperiods.Theresultsabovesuggestthatexperimentationandlearningcanindeedbeimportantcomponentsinincentiveproblems,andshouldbetakenintoaccountwhendesigningcompensationschemesforinnovativetasks.Furthermore,inadditiontospendinglesstimemakingdecisions,subjectsunderthe xed-wagecontractalsoexertlesse ortbyenteringlessinformationintothesheetgiventothemthansubjectsunderthepay-for-performanceandexplorationcontracts.Figure3showsthatacrossthethreecontractsthereisaconsiderablysmallerproportionofsubjectsunderthe xed-wagecontractwho llouthalformoreofthe eldsinthedecisiontablethanintheothertwocontracttreatments.Thisdi erenceine ortchoiceisstatisticallysigni cantbetweentheexplorationcontractandthe xed-wagecontract(p-value0.0053)aswellasbetweenthepay-for-performanceandthe xed-wagecontract(p-value0.0804).Inthe rst10periodsoftheexperimentsubjectsundertheexplorationcontractaresigni cantlymorelikelytorecordinformationthansubjectsinthe xed-wagecontract(p-value0.0111)therebyrefutingoncemoretheshirkingpredictionofthestandardrepeatedmoral-hazardmodel.Thedi erenceine ortexertedduringthe rst10periodsbetweensubjectsundertheexplorationcontractandthepay-for-performancecontractisnotsigni cant(p-value0.5782).Thedi erenceine ortchoicebetweentheexplorationcontractandthepay-for-performancecontractispositivebutnotstatisticallysigni cant(p-value0.29).Ontheonehand,subjectsunderthepay-for-performancecontractaregivenmorepowerfulincentivesoverallsincetheircompensationdependsonperformancebothinthe rstandthelast10periodsoftheex-periment.Ontheotherhand,sincesubjectsundertheexplorationcontractchoosetoex-perimentwithverydi erentstrategiesinthe rst10periodsasweshowedinResult3,theyneedtoexertmoree ortwhenevaluatingtheirdecisionsthansubjectsunderthepay-for-performancecontract.ThisisalsovisibleinFigure3whichshowsthate ortdeclinesinthepay-for-performancecontract.Thisoccurssincesubjectsinthepay-for-performancecontractessentiallystopexploringandexperimentingwithdi erentchoicesveryearlyintheexperi-mentandthereforetheybarelychangetheirchoicesinthelast10periods.Sincethereislittle17 .5 .6 .7 .8Proportion of subjects exerting high effort Fixed wageP-f-P contractExploration contract Periods 1-10Periods 11-20 Figure3:Proportionofsubjectswhocompletemorethanhalfofthe eldsinthedecisionrecordtableforthe xed-wage,pay-for-performanceandexplorationcontracts.change,theydonothavetorecordtheirchoicesascarefullyassubjectsintheexplorationcontracttreatment.Wealsonotethattimeallocationande ortchoiceinthe xed-wageisstrictlygreaterthanzerosincesomeofthesubjectsaresucientlymotivatedbyintrinsicrewardstoexerte ort.Aninspectionofe ortchoicesbysubjectsinthe xed-wagetreatmentrevealsabimodaldistribution.Subjectseitherfullyrecordordonotrecordanyoftheirpastchoices.Moreover,subjectsinthe xed-wagetreatmentwhoexertmoree ortaremorelikelytosuccessfullyinnovate:65%ofthemendupsellingattheschoolinthe nalperiodcomparedto47%ofthesubjectswhoexertlesse ort,butthisdi erenceisnotstatisticallysigni cant(p-value0.2047).However,maximumpro tsaresigni cantlyhigherforsubjectsinthe xed-wagetreatmentwhoexertmoree ort(p-value0.0298).163.1.1AveragePerformanceHavingcon rmedthattheinnovationsuccess,explorationbehavior,timeallocationande ortchoiceacrossthedi erentcontractsareinaccordancewithourtheoreticalpredictions,wenowshowthatsubjects'overallperformanceintheexperimentasmeasuredbyaveragepro tishighestintheexplorationcontract.Result4(performance):Subjectsundertheexplorationcontractproducehigheraverageprof-itsthansubjectsunderthepay-for-performanceand xed-wagecontracts. 16Forastudyofthee ectofintrinsicmotivationoninnovationproductivity,seeSauermannandCohen(2010).18 PreliminaryevidenceforResult4comesfrominspectingtheaveragepro tforthethreecontracts.Thisperformancemeasureishighestintheexplorationcontract(111francs)andthedi erenceinperformancebetweentheexplorationcontractandthepay-for-performance(96francs)andthe xed-wagecontract(102francs)isstatisticallysigni cant(Mann-Whitney-Wilcoxontest:p-valuesof0.0009and0.0253).Thisdi erenceinperformanceexistsdespitethefactthattheaveragewagereceivedbysubjectsundertheexplorationcontractislowerthanintheothertwocontracts.17Wenowinvestigatewhetherattitudestowardriskcanexplainthedi erencesinperfor-mancedocumentedinResult4.Result5(riskaversion):Underthepay-for-performancecontractmorerisk-aversesubjectsaresigni cantlylesslikelytoexploreandtochoosetosellintheoptimallocationinthe nalperiodoftheexperiment.Theyalsoproducesigni cantlylowerpro ts.Attitudestoriskhaveasimilar(thoughstatisticallyinsigni cant)e ectintheexplorationcontract,whilenosystematice ectsofriskarefoundforthe xed-wagecontract.Usingthedatafromtheseparateriskaversionexperimentweclassifysubjectsintomoreandlessrisk-aversegroups.TheleftpanelofFigure4providesa rstindicationforthesignandmagnitudeofthee ectofriskaversiononthelikelihoodof ndingthebeststrategy.Inthis gureweuseourriskaversionmeasurestofurtheranalyzethe nalperiodlocationchoiceaswedidinFigure1.Weseparatelypresent nallocationchoicesformoreandlessrisk-aversesubjectsforeachofthethreecontracts.Inthepay-for-performancecontract,morerisk-aversesubjectsarelesslikelyto ndtheoptimallocationastheyarelesslikelytoexplorethanthelessrisk-aversesubjects.Thisinnovation-reducinge ectofriskisstatisticallysigni cantinthepay-for-performancecontracttreatment(p-value0.0170),butitisnotstatisticallysigni cantintheothertwotreatments.18Thislowerrateofinnovationsuccesscausedbyriskaversionisdrivenbythelowerlevelsofexplorationunderthepay-for-performancecontractbecauseinthistreatmenttheproportionoflocationchoicesotherthanthedefaultlocation(p-value0.0075)aswellasthevariabilityofactionchoices(p-value0.0181)aresigni cantlylowerforsubjectswithhigherriskaversion.However,intheexplorationcontractwheresubjects'failureistoleratedinearlyperiodsoftheexperimentandcompensationhasamuchsmallerriskycomponentthee ectissmallerinmagnitudeandnotstatisticallysigni cant.Thesame 17Duringthe rst10periodsoftheexperimentthethreecontractsarequitesimilarintermsofaveragepro ts.Afterperiod10,theaveragepro tsunderthedi erentcontractsbegintodivergeassubjectsundertheexplorationcontractreverttoandsubsequently ne-tunethebeststrategytheyfoundduringthe rst10periodsoftheexperiment.18However,evenamonglessrisk-aversesubjectsthereexistsastatisticallysigni cantinnovation-reducinge ectofstandardpay-for-performancerelativetotheexplorationcontract(p-value0.0367).19 0 .2 .4 .6 Fixed wageP-f-P contractExploration contract less R-Amore R-Aless R-Amore R-Aless R-Amore R-A Business districtSchool 90 100 110 120 130 140 Fixed wageP-f-P contractExploration contract less R-Amore R-Aless R-Amore R-Aless R-Amore R-A Maximum profitLast period profit Figure4:Proportionofsubjectsbylocationinthe nalperiodoftheexperiment(left)andmaximumpro t,lastperiodpro tandaverageperperiodpro tofsubjects(right)forthe xed-wage,pay-for-performanceandexplorationcontractsadjustingfordi erencesinriskaversion.istrueinthe xed-wagecontractwherecompensationentailsnorisk.19Itisimportanttonoteatthispointthat,aswithallsubgroupanalysis,theheterogeneoustreatmente ectswithrespecttoriskaversionneedtobetreatedwithcaution.Itispossiblethatthesemaymerelybedrivenbycorrelatedomittedvariables(e.g.,cognitivere\rectionability(Frederick2005))aswecanonlyexogenouslyvarytheincentiveschemesadministeredtotheexperimentalsubjects.Sincemorerisk-aversesubjectsunderthepay-for-performancecontractarelesslikelytoexploreandthereforelesslikelytoselllemonadeintheoptimallocationinthe nalperiod,theyalsoproducelowerpro tsascanbeseenintherightpanelofFigure4.Thispro t-reducinge ectofriskaversioninthepay-for-performancecontractislargeinmagnitudeandstatisticallysigni cantformaximumpro t(p-value0.0563)and nalperiodpro t(p-value0.0382),butitisnotstatisticallysigni cantforaveragepro t(p-value0.1846).Furthermore,asinthecaseofthe nalperiodlocationchoice,riskaversionalsohasasmallnegativebutstatisticallyinsigni cante ectonpro tmeasuresintheexplorationandthe xed-wagecontracttreatment.20Therecouldbereasonsinadditiontoriskaversionforthedi erenceinaveragepro tsacrossthethreetreatmentgroups.Forexample,inourexperimentsubjectsarenotgiven 19Qualitativelysimilarresultsholdfortheambiguityaversionmeasure.Thee ectsareofthesamesignasthee ectsofriskaversion,buttheyaregenerallysmallerinmagnitudeandinsomecasesnotstaticallysigni cant.20Wealsospeculatethatlossaversion,similartoriskaversion,mayfurtherexacerbatetheexploration-reducinge ectsofpay-for-performancecompensation.Underthepay-for-performancecontractanyloss-averseagentmaysetthepayo ofthedefaultstrategyasareferencepointandwillnotdeviatefromthisstrategytoavoidanylossesrelativetothereferencepoint.20 preciseinformationaboutthepro tsassociatedwitheachoftheavailablechoices.21Thedi erencesinaveragepro tsacrossthethreetreatmentgroupscouldthusbeduetosubjectsbeingpessimisticaboutthereturnstoexploration.Theexplanationmirrorsthatgivenintheabovetwoparagraphswithpessimisminplaceofriskaversion.3.2TerminationWenowshowthatthethreatofterminationreducestheprobabilitythatsubjectssuccessfullyinnovatebecausethethreatofearlyterminationreducesexplorationactivities.Furthermore,theadversee ectsofterminationarelesspronouncedinthegoldenparachutetreatment.Result6(termination):Thethreatofterminationhasadversee ectsoninnovationsuccessandexplorationactivities,butgoldenparachutesalleviatethesenegativee ects.Riskaversionfurtherreducesinnovationsuccess,explorationactivitiesandperformanceintheterminationtreatment.Figure5shows nalperiodlocationchoicesintheexplorationcontract,terminationandgoldenparachutetreatmentswhereinthecaseofthelattertwotreatmentsweeliminatedsubjectsthatareterminatedafterthe rst10periods.Thethreatofterminationinthepureterminationandgoldenparachutetreatmentsigni cantlyreducestheprobabilitythatsubjectsendupchoosingtosellatthebestlocationinthe nalperiodoftheexperimentrelativetotheexplorationcontracttreatment(p-values0.0001and0.0200)whiletheuseofgoldenparachutesraisestheinnovationsuccessprobability(p-value0.0485)relativetotheterminationtreatment.22Focusingonsubjectsthatarenotsubjecttoterminationweagain ndthatterminationhasaninnovation-reducinge ectsinceaveragemaximumpro tintheexplorationcontracttreatment(145francs)issigni cantlyhigherthaninthetermination(126francs)andthegoldenparachutetreatments(134francs).Therespectivep-valuesforthedi erencesare0.0037and0.0772.Comparingthemaximumpro tsfortheterminationandgoldenparachutetreatmentsshowsthattheuseofgoldenparachutesslightlymitigatestheseadversee ects,thoughthee ectisnotsigni cant(p-value0.1784).Theadversee ectofterminationismorepronouncedifweconsiderthefullsampleofsub-jectsandonlyfocusonthe rst10periods.Theaveragemaximumpro tinthetermination 21Thisisalsothecaseinsomeoftheaforementionedpsychologyexperimentswhich ndthatsubjectsundera xed-wagecontractperformbetterthansubjectsunderapay-for-performancecontract.22Thesameresultsholdwhenfocusingexclusivelyonthe nallocationchoiceafterthe rst10periodsusingallthesubjectsintheterminationandgoldenparachutetreatments.Thethreatofterminationreducestheprobabilityof ndingthebestlocationrelativetotheexplorationtreatment(p-values0.0063and0.0562)andtheuseofreparationpaymentsincreases(thoughnotstatisticallysigni cantly)theinnovationsuccessprobabilityinthegoldenparachutetreatmentrelativetotheterminationtreatment(p-value0.3176).21 0 .2 .4 .6 .8Proportion of subjects by location in final period Exploration contractTerminationGolden parachute Business districtSchool Figure5:Proportionofsubjectsbylocationinthe nalperiodoftheexperimentfortheexplorationcontract,terminationandgoldenparachutetreatments.andthegoldenparachutetreatmentsisagainsigni cantlylowerthanintheexplorationcon-tracttreatment(p-values0.0032and0.0037).However,thedi erencebetweentheterminationandthegoldenparachutetreatmentsisnotstatisticallysigni cant(p-value0.7989).Asinouranalysisofthethreebaselinetreatments,wecantracethedi erencesinin-novationsuccessbacktodi erencesinexplorationbehavior.Tothisendweagaincomparethenumberoftimessubjectschoosetodeviatefromtheproposedstrategyandtoexplorealocationotherthanthebusinessdistrict.Toguardagainstpotentialselectione ectsarisingfromattritionwefocusexclusivelyonchoicesinthe rst10periods.Asexpected,explorationislowerintheterminationtreatmentwheresubjectsshyawayfromexploringotherlocationsinthe rst10periods.Whiletheaverageproportionoflocationchoicesotherthanthedefaultlocationis82%intheexplorationcontractitisonly47%intheterminationtreatmentand59%inthegoldenparachutetreatment.Thisexploration-reducinge ectofthethreatofter-minationisstatisticallysigni cant(p-values0.0001and0.0009).Moreover,goldenparachutesincreaseexplorationactivitiesrelativetothepureterminationtreatment(p-value0.0495).Inthepost-experimentquestionnairesubjectsarguedthatthethreatofterminationforcedthemtoconcentrateonsellinginthebusinessdistrictandleftnoleewayforexploration.Furtherevidencefortheexploration-reducinge ectofthethreatofterminationandtheexploration-increasinge ectofreparationpaymentscomesfromcomparingthevariabilityofactionchoicesinthe rst10periodsforthefullsampleofsubjects.Thesubject-speci cstandarddeviationofactionchoicesinthe rst10periodsishighestintheexplorationcontract(standarddeviation1.09).Thismeasureissigni cantlylowerintheterminationtreatment(standarddeviation0.74,p-value0.0014)andinthegoldenparachutetreatment(standarddeviation0.79,p-value0.0071).Theuseofgoldenparachutesslightlyincreasesexploration22 0 .2 .4 .6 Exploration contractTerminationGolden parachute less R-Amore R-Aless R-Amore R-Aless R-Amore R-A Business districtSchool 90 100 110 120 130 140 Exploration contractTerminationGolden parachute less R-Amore R-Aless R-Amore R-Aless R-Amore R-A Maximum profitLast period profit Figure6:Proportionofsubjectsbylocationinthe nalperiodoftheexperiment(left)andmaximumpro t,lastperiodpro tandaverageperperiodpro tofsubjects(right)fortheexplorationcontract,terminationandgoldenparachutetreatmentadjustingfordi erencesinriskaversion.activityrelativetotheterminationtreatment,butthise ectisnotstatisticallysigni cant(p-value0.2821).23Usingthesamehazardratemodelasinouranalysisofthebaselinetreatmentsthoughconcentratingexclusivelyonthe rst10periodswecaninvestigatehowlikelysubjectsaretopersistintheirexplorationactivitiesinthedi erenttreatments.Column3ofTable1showsthatbothintheterminationtreatmentandinthegoldenparachutetreatmentsubjectsaresigni cantlymorelikelytostopexploringthanintheexplorationcontract.Moreover,subjectsintheterminationtreatmentarealsosigni cantlymorelikelytostopexploringthansubjectsunderthegoldenparachutetreatment(p-value0.0663).Column4ofTable1reportsestimatesforthe rst10periodsshowingstatisticallysigni cantdi erencesinthehazardratebetweentheexplorationcontractandtheterminationaswellasthegoldenparachutetreatment.Thedi erencebetweenterminationandgoldenparachuteisalsostatisticallysigni cant(p-value0.0604).Riskaversionplaysanimportantroleintheterminationtreatmentascanbeseenintheleftpanel,whichshows nalperiodlocationchoice,andintherightpanelofFigure6,whichpresentsthedi erentpro tmeasures.Morerisk-aversesubjectsintheterminationtreatmentarelesslikelytosellintheschoolinthe nalperiodoftheexperimentandtheyachievelowermaximum, nalperiodandaveragepro ts.Throughout,thereisastatisticallysigni cantnegativee ectofriskaversionintheterminationtreatmentonthecorrect nalperiodlocation 23Thedi erentproportionsofsubjectswhoareterminatedintheterminationandthegoldenparachutetreatmentsarealsoinlinewithsubjectsexploringmoreinthelattercase.Whileintheterminationtreatment13outof71subjects(18%)donotmeetorexceedtheterminationthreshold,21outof78subjects(27%)areterminatedinthegoldenparachutetreatment,butthedi erenceisnotstatisticallysigni cant(p-value0.2124).23 choice(Mann-Whitney-Wilcoxontest:p-value0.0041)aswellasmaximumpro ts(p-value0.0023), nalperiodpro ts(p-value0.0041)andaveragepro ts(p-value0.0037).This ndingisinlinewithourpreviousanalysiswherewefoundsimilarlystronge ectsofriskaversionforthepay-for-performancecontractwhichalsoinducesindividualstoachievepro tsfromtheverybeginningoftheexperimentinsteadoflearningthroughexploration.Incontrast,likeforour ndingfortheexplorationcontracttreatment,thereisnostatisticallysigni cante ectofriskaversioninthegoldenparachutetreatment.Finally,intheterminationtreatmentahighdegreeofriskaversionsigni cantlydecreasessubjects'propensitytoexplore.Intheterminationtreatmentthenumberoftimessubjectschoosetodeviatefromtheproposedstrategyandtoexplorealocationotherthanthebusinessdistrictinthe rst10periodsissigni cantlylowerforsubjectswhoaremorerisk-averse(p-value0.0114).Similarly,intheterminationtreatmentthevariabilityofactionchoicesinthe rst10periodsisalsosigni cantlylowerformorerisk-aversesubjects(p-value0.0040).Therearealsosmallnegativee ectsofriskaversiononexplorationactivityinthegoldenparachutetreatment,butthesee ectsareneverstatisticallysigni cant.4RobustnessInthissubsection,weshowthatourresultsarerobusttomodi cationsintheexperimentaldesign.4.1SignalingE ectsofIncentiveContractsIntheanalysiswepreviouslyconducted,eachsubjectonlyeversawoneparticularincentivecontract.Thesubjectswerenotmadeawarethatavarietyofdi erentincentiveschemeswereadministeredtodi erentsubjects.Thismeansthatsubjectsmightmakedi erentinferencesfromthedi erentcontractstheyaregivenaboutwhatthebeststrategytoplayis.Forexample,whilesubjectsunderthepay-for-performancecontractmightinferthatthebeststrategyisnottoexplore,subjectsundertheexplorationcontractmightinferthatthebeststrategyistoexplore.Toaccountforthesepotentialsignalinge ectsweadministeredanothertreatmentinwhichsubjectswereabletoseethatbothpay-for-performanceandexplorationcontractswereavailable.Inthistreatment,afterhavingobservedthesetofpossiblecontracts(pay-for-performanceorexploration),theincentiveschemethatwouldbeappliedtoeachsubjectwasdeterminedbyhavingthesubjectrolladice.Afterhavingobservedtheoutcomeofthedicerolltheexperimentercircledtherelevantcompensationschemeandcrossedouttheirrelevantcompensationscheme.Atotalof70subjectsparticipatedinthistreatmentofwhich24 0 .2 .4 .6 .8Proportion of subjects by location in final period P-f-P contract (dice)Exploration contract (dice) Business districtSchool Figure7:Proportionofsubjectsbylocationinthe nalperiodoftheexperimentforthepay-for-performance(diceroll)andexploration(diceroll)contracts.32subjectsrolledthediceandreceivedapay-for-performancecontractand38subjectsrolledandreceivedanexplorationcontract.Figure7con rmsourresultsabouttheimportanceofcorrectlystructuredincentivesformotivatinginnovation.Asbefore,subjectswhoaregivenanexplorationcontractaresigni -cantlymorelikely(Mann-Whitney-Wilcoxontest:p-value0.0152)tochoosethebestlocationinthe nalperiodoftheexperimentthansubjectswhoreceiveapay-for-performancecon-tract.Subjectswithanexplorationcontractalsoagainachievesigni cantlyhighermaximumpro ts(138francs)andhigher nalperiodpro ts(134francs)thansubjectsunderapay-for-performancecontract(120francs,118francs).Therespectivep-valuesforthecomparisonsare0.0372and0.0654.Asbeforethisdi erenceininnovationsuccessisdrivenbythedi erencesinexplorationbehaviorthatincentiveschemesinduce.Theproportionoflocationchoicesotherthanthede-faultlocationissigni cantlyhigherforsubjectswhoobtainanexplorationcontractfollowingtheirdiceroll(p-value0.0045)andthevariabilityofstrategychoicesisalsohigher,althoughthisdi erenceisnotsigni cant(p-value0.1343)duetothesmallersamplesize.Mirroringourpreviousresults,subjectsunderthepay-for-performancecontractalsohavelowaveragepro tsalthoughthise ectisnotstatisticallysigni cant(p-value0.1591).Fur-thermore,riskaversionagainhasaninnovation-andpro t-reducinge ectinthepay-for-performancetreatment.Inthepay-for-performancetreatmentthereisastatisticallysigni -cantnegativee ectofriskaversiononthecorrect nalperiodlocationchoice(p-value0.0583),butthereisnosigni cante ectintheexplorationcontracttreatment.Thenegativee ectofriskaversionwhensubjectsobtainapay-for-performancecontractisalsoapparentinthelowerpro tsformorerisk-aversesubjects,butthise ectisnotstatisticallysigni cantdueto25 thesmallsamplesize.4.2GoldProspectingExperimentInthelemonadestandexperiment,subjectsexertedreale ortandhadtoformbeliefsabouttheoutcomesassociatedwitheachoftheavailableactions.Asanotherrobustnesscheckweconductedasecondexperimentusingthechosene ortapproachandawithin-subjectsdesign.Thisallowsustofullycontrolthebeliefsofsubjectsinaparametrizedsetting.4.2.1ProceduresandSubjectPoolTheexperimentwasprogrammedandconductedwiththesoftwarez-Tree(Fischbacher2007)attheCaliforniaSocialScienceExperimentalLaboratory(CASSEL)atUCLA.ParticipantswererecruitedfromtheCASSELsubjectpoolusinganonlinerecruitmentsystem.Atotalof30subjectsparticipatedintheseexperiments.Weemployedawithin-subjectsdesignsothatallsubjectsexperiencedallthreetreatmentconditions.Theexperimentalsessionlasted50minutes.Duringtheexperiment,experimentalcurrencyunitscalled\goldnuggets"wereusedtokeeptrackofmonetaryearnings.Theexchangeratewassetat3goldnuggets(gn)=$1,andtheshow-upfeewas$10.Onaverage,subjectsearned$33.4.2.2TheTaskSubjectsplayseveralseparateinstancesofagoldprospectinggamethatexactlymirrorsthetheoreticalmodelinManso(2011).Eachgoldprospectinggameconsistsof2periods.Ineachperiod,eachsubjectchoosestoprospectgoldinoneoftwomountainsortostayathome.Eachoftheseoptionshasanassociatedrevenueandcost.Theoption\Stayathome"alwaysyields0gnanditcoststhesubject0gntochoosethisoption.Incontrast,theoption\Mountain1"yields100gnwithaprobabilityof50%and0gnwithaprobabilityof50%anditcosts0.25gntochoosethisoption.Eachpayo realizationof\Mountain1"isanindependentdraw.Finally,theoption\Mountain2"allowsforlearning.Withprobabilityp,\Mountain2"isagold-richmountain,inwhichcaseityields100gninbothperiod1andperiod2wheneverthisoptionischosen.Withaprobabilityof1 p,\Mountain2"isagold-poormountain,inwhichcaseitalwaysyields0gnwheneverthisoptionischosen.Thecostofchoosing\Mountain2"is0.75gn.Eachinstanceofthegoldprospectinggameisanindependentdraw.Thus,eachsubjectcanonlylearnaboutthequalityof\Mountain2"forthatparticularinstanceofthegoldprospectinggame.26 Theparticipantsintheexperimentthusfacethechoicebetweenanoutsideoptionof0gn(\Stayathome"),anoptionwithaknownandconstantpayo distribution(\Mountain1")andanoptionwithanunknownpayo distribution(\Mountain2")thatallowsforlearning.Weelicitsubjectchoicesusingthestrategymethod:subjectsareaskedtoprovideafullcontingentplanofactionforallpossibleoutcomes.Thus,ontheinputscreenforeachgoldprospectinggamesubjectsenter(a)theirchoiceforperiod1,(b)theirchoiceforperiod2conditionalonapayo of0gninperiod1,and(c)theirchoiceforperiod2conditionalonapayo of100gninperiod1.Attheendofeachgoldprospectinggame,subjectslearnhowmuchrevenuetheyobtainedandareremindedofthecosttheyincurred.Inaddition,attheendofeachgoldprospectinggamesubjectsareinformedabouttheircumulativepayo balance.4.2.3TreatmentGroupsandPredictionsWeconductfoursetsoftreatmentswhichdi erintermsoftheprobabilitypthatMountain2isagold-richmountain.Theprobabilityptakesthevalues5%,35%,45%oritisunknowntothesubjects.Whenpisunknownwechooseitsuchthatpisequalto50%.Ineachofthefourtreatmentsets,eachsubjectplaysthegoldprospectinggameunderthreedi erentcompensationschemes.24Thus,thereareatotaloftwelveseparategoldprospectinggamesthateachsubjectplays.Inparticular,thethreecompensationschemesare:IncentiveScheme1(FixedWage):\Youwillbepaida xedwageof1gnperperiod."IncentiveScheme2(Pay-for-Performance):\Youwillbepaid10%ofthegoldrevenueduringthe1standthe2ndperiodoftheexperiment."IncentiveScheme3(Exploration):\Youwillbepaid10%ofthegoldrevenueinthe2ndperiodoftheexperiment."Theseincentivesystemsexactlymirrorthepaymentschemesinthelemonadestandex-perimentandgeneratethesamequalitativepredictions.Table2showstheoptimalchoicesforarisk-neutralagentineachtreatmentcondition.Alongthecolumnstherearedi erentvaluesofpwhilethedi erentincentiveschemesareshowninthedi erentrows.AsisapparentfromTable2,aspincreasesthechoiceof\Mountain2"becomesmoreattractive.Thismakesittheoptimalchoiceofactioninthe rstperiodforhighvaluesofpinthePay-for-PerformanceandExplorationcontracttreatments.Ourmainhypothesis, 24Toaccountforpotentialordere ects,foreachsubjectanyofthefourpconditionswasequallylikelyandindependentlychosenasthe rst,second,thirdorfourthtreatment.Furthermore,withineachptreatment,foreachsubjecteachcompensationschemewasequallylikelyandindependentlychosenasthe rst,secondorthirdcompensationscheme.27 OptimalStrategiesp=005p=035p=045punknown FixedWagePeriod1StayatHomeStayatHomeStayatHomen/aPeriod2after100gnStayatHomeStayatHomeStayatHomen/aPeriod2after0gnStayatHomeStayatHomeStayatHomen/a Pay-for-PerformancePeriod1Mountain1Mountain1Mountain2n/aPeriod2after100gnMountain1Mountain1Mountain2n/aPeriod2after0gnMountain1Mountain1Mountain1n/a ExplorationPeriod1StayatHomeMountain2Mountain2n/aPeriod2after100gnMountain1Mountain2Mountain2n/aPeriod2after0gnMountain1Mountain1Mountain1n/a Table2:Optimalstrategiesforarisk-neutralagentineachtreatmentcondition.however,concernstheextenttowhichthedi erentpaymentschemesencouragethechoiceof\Mountain2"inperiod1ofthegoldprospectinggameforagivenlevelofp.Ingeneral,duetothetoleranceforearlyfailureandrewardforlong-termsuccess,subjectsundertheexplorationcontractshouldbemorelikelytochoose\Mountain2"inperiod1thansubjectsunderthe xedwageandthepay-for-performancecontracts.Thecomparisonismostclearlyvisiblewhenp=035:inperiod1subjectsshouldchoose\StayatHome"underthe xedwagecontract,\Mountain1"underthepay-for-performancecontract,and\Mountain2"undertheexplorationcontract.4.2.4ResultsWe rstcomparestrategychoicesacrossthethreeincentivetreatmentsforthe4di erentlevelsofthesuccessprobabilityp.Eachofthe30subjectsmadechoicesunderallthe12di erenttreatmentswhichallowsustomakewithin-subjectcomparisons.Wefocusontheexplorationbehaviorofsubjectsacrossthedi erenttreatmentconditions.Table3showsthe rst-periodchoicesmadebysubjectsinthedi erentconditionsaswellastheassociatedtheoreticalprediction.Consistentwiththepredictionthattheexplorationcontractencouragesexploration,moresubjectschoose\Mountain2"underthisincentiveschemethanundertheothertwocontracts.Moreover,subjectsunderthe xedwagecontractoverwhelminglychoosetheshirkingoption\StayatHome",whilesubjectsunderthepay-for-performancecontractchoosetoexploit\Mountain1."25InTables2and3weshowthatforp=035,thepredictedchoiceofactionforarisk-neutralagentinthe rstperiodisdi erentbetweenthepay-for-performanceandtheexploration 25Asaresultofthisincreasedexplorationactivity,theexpectedprobabilityof nding100gninperiod2giventhesubjects'strategiesis60%undertheexplorationcontract,yetonly52.9%and8.2%underthepay-for-performanceand xed-wagecontracts.28 ChoicesinPeriod1p=005p=035p=045punknown FixedWageStayatHome23252426Mountain17233Mountain20331TheoreticalPredictionStayatHomeStayatHomeStayatHomen/a Pay-for-PerformanceStayatHome0111Mountain126201118Mountain2491811TheoreticalPredictionMountain1Mountain1Mountain2n/a ExplorationStayatHome17124Mountain111556Mountain22242320TheoreticalPredictionStayatHomeMountain2Mountain2n/a Table3:ChoicesinPeriod1foreachtreatmentcondition.contractwhileforp=045theoptimalchoicesforarisk-neutralagentunderthetwoincentivesystemscoincide.Finally,forp=005arisk-neutralagentundertheexplorationcontractisexpectedtochoose\StayatHome"sincethepayo fromtheexplorationoption\Mountain2"istoolowandsubjectsdonotearnenoughinthe rstperiodtojustifythecostsofchoosing\Mountain2".Whenp=035only9subjectschose\Mountain2"inperiod1underthepay-for-performancecontract,but24subjectschosetodosoundertheexplorationcontract.Duetoourwithin-subjectdesignweobservethechoicesofsubjectsunderallthedi erenttreat-mentconditions.AMcNemarmatchedsamplestestshowsthatthedi erenceinthechoiceof\Mountain2"forp=035underthepay-for-performanceandtheexplorationcontractishighlystatisticallysigni cant(p-value0.0007).Thestatisticalsigni canceofthisdi erenceisalsocon rmedbyasubject xed-e ectslogitregressionwherewecomparethe rst-periodchoicesforp=035underthepay-for-performanceandtheexplorationcontract.Thebinarydependentvariabletakesthevalue1ifasubjectchose\Mountain2"inperiod1andvalue0otherwiseandthedependentvariabletakesthevalue1ifthetreatmentwastheexplorationcontractand0ifitwaspay-for-performancecontract.Theassociatedp-valueofthecoecientofthedependentvariableis0.004indicatingagainthatthedi erenceishighlystatisticallysigni cant.AsisobviousfromtherawdatashowninTable3,thedi erenceinbehaviorbetweentheexplorationandthe xedwagecontractisalsostatisticallysigni cant.Underthe xedwagecontractsubjectsoverwhelminglychoosetheshirkingoption\StayatHome".Therelevantp-valuesfortheMcNemartestandthecoecientofthesubject xed-e ectslogitregressionare0.00001and0.001.Althoughwearenotabletoprovideprecisepredictionsforsubjectbehaviorinthetreat-29 mentconditionswherepisunknown,thissetoftreatmentconditionsisofparticularinterestasitsharesthefeatureoftheunknownsuccessprobabilitywiththemorerealisticlemon-adeexperiment.Table3showsthatsubjectbehaviorinthe rstperiodforunknownpissimilartowhenp=035andformalstatisticaltestscon rmthisimpression.Subjectsaresigni cantlymorelikelytochoose\Mountain2"inthe rstperiodundertheexplorationcontractthanunderthepay-for-performanceincentivescheme(p-value0.032, xed-e ectslogit).Thus,evenwhenthesuccessprobabilitypisunknowntheearlyfailuretoleranceoftheexplorationcontractrelativetothepay-for-performancecontractmotivatessubjectstochoosetheexplorationoption\Mountain2".Giventheobservedbehaviorofthesubjectstheexperimentalsoallowsustocalculatehowlargetheexpectedpro tofaprincipalwouldbewhoimplementsdi erentincentiveplans.Forp=005thepro tsoftheprincipalarehighestwhenheo ersapay-for-performancecontract(75.53)ratherthananexplorationcontract(59.47).Forhigherexpectedsuccessprobabilitiesoftheexploratoryoption,i.e.p=035,p=045orpunknowntothesubjects,theprincipalalwaysreapsahigherpro twheno eringanexploration(91.23,103.13,103.33)ratherthanapay-for-performancecontract(87.23,94.2,96).Finally,whenp=035andexplorationisdesirable,theprincipalwouldneedtoo ersubjects27%ofthetotalgoldrevenueinsteadofjust10%toinducerisk-neutralsubjectstoexploreunderastandardpay-for-performancecontract.5ConclusionInthispaper,wearguedthatappropriatelydesignedmanagerialcompensationise ectiveinenablingentrepreneurshipandmotivatinginnovation.Inareal-e orttaskthatinvolvesinnovationthroughexperimentation,we ndthatsubjectsunderapaymentschemethattoleratesearlyfailureandrewardslong-termsuccessexploremoreandarethusmorelikelytodiscoversuperiorstrategiesthansubjectsunder xed-wageorstandardpay-for-performanceincentiveschemes.Wealso ndthatthethreatofterminationcanundermineinnovation,andthatthise ectismitigatedbythepresenceofacontractthatmimicsagoldenparachute.Byusingacontrolledlaboratoryexperimentinwhichindividualsarerandomlyassignedtodi erentcontracts,weareabletoestablishacausalrelationbetweenparticularincentiveschemesandinnovationperformance.Ourresultscomplementothercontributionstotheliteratureonentrepreneurshipandinnovation,whichusenaturallyoccurringdatatoshowthattoleranceforfailureisassociatedwithinnovationactivity.Providedthatourresultsareexternallyvalid,themostdirectout-of-sampleimplicationofourresultsisthatinsituationsinwhichenablingentrepreneurshipandmotivatinginnovation30 areimportantconcerns,suchasthedesignofcompensationplansoftopexecutivesandmiddlemanagersinlargecorporationsaswellasofentrepreneursinstart-upcompanies,toleranceforearlyfailurecoupledwithrewardsforlong-termsuccessaree ectiveinmotivatinginnovation.Inthecontextofexecutivecompensation,forexample,implementingtoleranceforfailureinvolvestheuseofpracticesthatareoftencriticizedsuchasoptionrepricing,managerialentrenchment,andgoldenparachutes.Ourresultssuggestthatrestrictingthesepracticescanhaveadversee ectsoninnovation.Theframeworkandmethodsproposedinthispapercanalsobeusefulinstudyingotherimportantproblemsintheentrepreneurshipliterature.Forexample,howdoesthechoiceof nancinga ecttheentrepreneur'sattitudetowardsinnovation?Howcanwedesigncom-pensationpackagestoattractcreativeentrepreneurswhilekeepingshirkersandconventionalentrepreneursaway?Weleavethesequestionsforfutureresearch.31 AppendicesALemonadeStandExperimentA.1ExperimentalInstructionsInstructionsYouarenowtakingpartinaneconomicexperiment.Pleasereadthefollowinginstructionscarefully.Everythingthatyouneedtoknowinordertoparticipateinthisexperimentisexplainedbelow.Shouldyouhaveanydicultiesinunderstandingtheseinstructionspleasenotifyus.Wewillansweryourquestionsatyourcubicle.Duringthecourseoftheexperimentyoucanearnmoney.Theamountthatyouearnduringtheexperimentdependsonyourdecisions.Allthegainsthatyoumakeduringthecourseoftheexperimentwillbeexchangedintocashattheendoftheexperiment.Theexchangeratewillbe:100francs=$1Theexperimentisdividedinto20periods.Ineachperiodyouhavetomakedecisions,whichyouwillenteronacomputerscreen.Thedecisionsyoumakeandtheamountofmoneyyouearnwillnotbemadeknowntotheotherparticipants-onlyyouwillknowthem.Pleasenotethatcommunicationbetweenparticipantsisstrictlyprohibitedduringtheexperiment.Inadditionwewouldliketopointoutthatyoumayonlyusethecomputerfunctionswhicharerequiredfortheexperiment.Communicationbetweenparticipantsandunnecessaryinterferencewithcomputerswillleadtotheexclusionfromtheexperiment.Incaseyouhaveanyquestionsdon'thesitatetoaskus.ExperimentalProceduresInthisexperiment,youwilltakeontheroleofanindividualrunningalemonadestand.Therewillbe20periodsinwhichyouwillhavetomakedecisionsonhowtorunthebusiness.Thesedecisionswillinvolvethelocationofthestand,thesugarandlemoncontentandthelemonadecolorandprice.Thedecisionsyoumakeinoneperiod,willbethedefaultchoicesforthenextperiod.Attheendofeachperiod,youwilllearnwhatpro tsyoumadeduringthatperiod.Youwillalsohearsomecustomerreactionsthatmayhelpyouwithyourchoicesinthefollowingperiods.PreviousManagerGuidelinesDearX,Ihaveenclosedthefollowingguidelinesthatyoumay ndhelpfulinrunningyourlemonadestand.Theseguidelinesarebasedonmypreviousexperiencerunningthisstand.Whenrunningmybusiness,Ifollowedthesebasicguidelines:32 Location:BusinessDistrictSugarContent:3%LemonContent:7%LemonadeColor:GreenPrice:8.2francsWiththesechoices,Iwasabletomakeanaveragepro tofabout90francsperperiod.Ihaveexperimentedwithalternativechoicesofsugarandlemoncontent,aswellaslemonadecolorandprice.TheabovechoicesweretheonesIfoundtobethebest.Ihavenotexperimentedwithalternativechoicesoflocationthough.Theymayrequireverydi erentstrategies.Regards,PreviousManagerCompensation(Thefollowingparagraphisusedintheinstructionsforsubjectsinthetreatmentwiththe xedwagecontract.)Youwillgetpaida xedwageof50francsperperiodduringthe20periodsoftheexperiment.Your nalcompensationdoesnotdependonyourpro tsfromthelemonadestand.(Thefollowingparagraphisusedintheinstructionsforsubjectsinthetreatmentwiththepay-for-performancecontract.)Yourcompensationwillbebasedonthepro tsyoumakewithyourlemonadestand.Youwillgetpaid50%ofyourowntotallemonadestandpro tsduringthe20periodsoftheexperiment.(Thefollowingparagraphisusedintheinstructionsforsubjectsinthetreatmentwiththeexplorationcontract.)Yourcompensationwillbebasedonthepro tsyoumakewithyourlemonadestand.Youwillgetpaid50%ofyourownlemonadestandpro tsinthelast10periodsoftheexperiment.(Thefollowingparagraphisusedintheinstructionsforsubjectsinthetreatmentwiththeterminationcontract.)Yourcompensationwillbebasedonthepro tsyoumakewithyourlemonadestand.Youwillgetpaid50%ofthepro tsyoumakeduringthelast10periodsoftheexperiment.However,ifthepro tsyoumakeduringthe rst10periodsoftheexperimentarebelow800francs,theexperimentwillendearly.(Thefollowingparagraphisusedintheinstructionsforsubjectsinthetreatmentwiththegoldenparachutecontract.)Youwillgetpaid50%ofthepro tsyoumakeduringthelast10periodsoftheexperiment.Ifthepro tsyoumakeduringthe rst10periodsoftheexperimentarebelow800francs,theexperimentwillendearlyandyouwillreceiveapaymentof250francs.A.2ExperimentalDesignThesubjectswereabletomakethefollowingparameterchoices:33 Location=fBusinessDistrict,School,StadiumgSugarContent=f00102;:::;9910gLemonContent=f00102;:::;9910gLemonadeColor=fGreen,PinkgPrice=f00102;:::;9910gThetablebelowshowstheoptimalproductmixineachlocation. BusinessDistrictSchoolStadium Sugar1.5%9.5%5.5%Lemon7.5%1.5%5.5%LemonadeColorGreenPinkGreenPrice7.52.57.5 MaximumPro t10020060 Inordertocalculatethepro tsineachlocationwhenthechoicesaredi erentfromtheoptimalchoicesabove,weimplementedalinearpenaltyfunctionwitha\roorset0suchthatlossesinabsolutetermsforthesubjectwereimpossible.Ineachlocation,thepenaltyfactorsassociatedwithadeviationofoneunitforeachofthevariablesaregivenbythenexttable. BusinessDistrictSchoolStadium Sugar560.5Lemon560.5LemonadeColor20600.5Price560.5 BElicitingRiskandAmbiguityAversionWemeasuredthesubjects'riskandambiguityaversionbyobservingchoicesunderuncertaintyinanexperimentthattookplaceafterthebusinessgameexperiment.Aspartofthisstudy,thesubjectsparticipatedinaseriesoflotteriesofthefollowingform.RiskAversionLotteryA:Win$10withprobability1=2,orwin$2withprobability1=2.IfsubjectsrejectlotteryAtheyreceive$7.LotteryB:Win$10withprobability1=2,orwin$2withprobability1=2.IfsubjectsrejectlotteryBtheyreceive$6.34 LotteryC:Win$10withprobability1=2,orwin$2withprobability1=2.IfsubjectsrejectlotteryCtheyreceive$5.LotteryD:Win$10withprobability1=2,orwin$2withprobability1=2.IfsubjectsrejectlotteryDtheyreceive$4.LotteryE:Win$10withprobability1=2,orwin$2withprobability1=2.IfsubjectsrejectlotteryEtheyreceive$3.AmbiguityAversionIfaredballischosenyouwillwin$7,ifablueballischosenyouwillwin$2.CaseA:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing16redballsand4blueballs.CaseB:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing14redballsand6blueballs.CaseC:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing12redballsand8blueballs.CaseD:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing10redballsand10blueballs.CaseE:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing8redballsand12blueballs.CaseF:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing6redballsand14blueballs.CaseG:ChooseUrn1containing20ballsthatareeitherredorblueORchooseUrn2containing4redballsand16blueballs.Aftersubjectshadmadetheirchoicesforbothriskandambiguityaversiononelotterywaschosenatrandomandeachsubjectwascompensatedaccordingtohisorherchoice.Theabovelotteriesenableustoconstructindividualmeasuresofriskandambiguityaversion.Foreachmeasurewethenusedthemedianobservationtosplitthesampleintoamoreandalessrisk/ambiguity-aversegroup.CGoldProspectingExperimentInstructions35 Youarenowtakingpartinaseriesofeconomicexperiments.Pleasereadthefollowinginstructionscarefully.Everythingthatyouneedtoknowinordertoparticipateinthisseriesofexperimentsisexplainedbelow.Shouldyouhaveanydicultiesinunderstandingtheseinstructionspleasenotifyus.Wewillansweryourquestionsatyourcubicle.Atthebeginningoftheexperimentsyouwillreceiveaninitialendowmentof$5inadditiontoyourshow-upfee.Duringthecourseoftheexperimentsyoucanearnafurtheramountofmoneybygaininggoldnuggets.Theamountthatyougainduringtheexperimentsdependsonlyonyourowndecisions.Allthegainsthatyoumakeduringthecourseoftheexperimentswillbeexchangedintocashattheendoftheexperiments.Theexchangeratewillbe:3goldnuggets(gn)=$1Attheendoftheseriesofexperimentsyouwillreceivethemoneythatyouearnedduringtheexperiments.Duringeachexperimentyouhavetomakedecisions,whichyouwillenteronacomputerscreen.Thedecisionsyoumakeandtheamountofmoneyyougainwillnotbemadeknowntotheotherparticipants.Onlyyouwillknowthem.Pleasenotethatcommunicationbetweenparticipantsisstrictlyprohibitedduringtheexperiments.Inadditionwewouldliketopointoutthatyoumayonlyusethecomputerfunctionswhicharerequiredfortheexperiments.Communicationbetweenparticipantsandunnecessaryinterferencewithcomputerswillleadtotheimmediateexclusionfromtheexper-iments.Incaseyouhaveanyquestionspleasedon'thesitatetoaskus.OverviewoftheExperimentalProceduresInthisseriesofexperiments,youwillbeplayingseveralinstancesofagoldprospectinggame.Ineachexperimentthegamewillbethesame,butthepayo sofyourchoicesandthewayyouarecompensatedwilldi er.Inparticular,thechoicesoneachscreenconstituteanentirelynewgame.Eachgoldprospectinggameconsistsof2periods.Ineachperiodyoumaychoosetoprospectgoldinoneoftwomountainsortostayathome.Eachoftheseoptionshasanassociatedrevenueandcost:\Stayathome"alwaysyields0gnanditcostsyou0gntochoosethisoption.\Mountain1"isawell-exploredmountainthatisclosetoyourhome.Whenyouchoosetoprospectgoldatthismountain,ityields100gnwithaprobabilityof50%and0gnwithaprobabilityof50%.Thus,bychoosing\Mountain1"youcannotlearnwhetherityields0gnor100gn.Itcostsyou0.25gntotravelto\Mountain1"andtoprospectgoldthere.\Mountain2"isanunexploredmountainthatisfurtheraway.Withaprobabilityofp\Mountain2"isagold-richmountain,inwhichcaseitalwaysyields100gnwheneveryouchoosethisoption.Withaprobabilityof1 p,\Mountain2"isagold-poormountain,inwhichcaseitalwaysyields0gnwheneveryouchoosethisoption.Youonlylearnthe36 qualityof\Mountain2"ifyouchoosetoprospectgoldatthatmountain.Sinceitisfurtherawayitcostsyou0.75gntotravelto\Mountain2"andtoprospectgoldthere.Finally,notethateachgoldprospectinggameyouplayisindependent.Thus,youcanonlylearnaboutthequalityof\Mountain2"forthatparticularinstanceofthegoldprospectinggame.First,therewillbe3practicetrialswithdi erentvaluesofpwhichdonota ectyourcompensationandwillhelpyougetacquaintedwiththeinterface.Afterthat,therewillbeatotalof4experiments,whichdi erintermsoftheprobabilitypthat\Mountain2"isagold-richmountain.Atthebeginningofeachexperimentwewillannouncetheprobabilitypthatwillberelevantduringthatexperiment.Theprobabilitypcantakeonthreevalues:5%,35%,45%oritmaybeunknown.Ineachofthe4experiments,youwillbeplayingthegoldprospectinggameunder3di erentcompensationschemes.Thus,thereisatotalof43=12separategoldprospectinggameswhichareallindependent.Yourcompensationwilldependonyourgoldproductioninthetwoperiodsofeachgame.Inparticular,the3compensationschemesare:1.Youwillbepaida xedwageof0gnperperiodplus10%ofthegoldrevenueduringthe1standthe2ndperiodoftheexperiment.2.Youwillbepaida xedwageof0gnperperiodplus10%ofthegoldrevenueinthe2ndperiodoftheexperiment.3.Youwillbepaida xedwageof1gnperperiod.Thedi erentcompensationschemesandexperiments(12intotalafterthetrialphase)arecompletelyindependent.Thatis,yourchoicesineachgoldprospectinggamehavenoe ectonanyotherchoicesintheexperiment.Whatyoulearnaboutthepayo sisspeci ctoeachgame.Howisyourincomecalculated?Yourtotalincomeinthetwoperiodsiscalculatedinthefollowingway:totalincome= xedwagein1stperiod+ xedwagein2ndperiod+revenuein1stperiod*revenuepercentagein1stperiod+revenuein2ndperiod*revenuepercentagein2ndperiod-costsin1stperiod-costsin2ndperiodYourincomeisthereforehigherwhenyour xedwage,yourrevenueandyourrevenuepercentagearehigherandyourcostsarelower.Your xedwageineachperioddependsonthecompensationoftheexperimentyouarecurrentlyin.Itiseither0gnor1gn.Yourrevenueineachperioddependsonthechoiceyoumadeandthetypeofmountain.Therevenuepercentageis0%or10%anddependsonthecompensationoftheexperimentyouarecurrentlyin.Thatistosayyoumayreceivenoshareoftherevenueoryoureceive37 someoftherevenuedependingonthecompensationofthecurrentexperiment.Finally,asmentionedbeforethecostofchoosingoneofthetwomountainsis0.25gnor0.75gnwhereasstayingathomeiscostless.ExampleLet'sassumethefollowingscenario.Inthe1stperiodandthe2ndperiodyoureceivea xedwageof0gnandthattherevenuepercentageinthe1stperiodis0%and10%inthe2ndperiod.Let'sassumeyouchoose\Mountain2"inthe1stperiod.Ifthemountainrevenuewas0gninthe1stperiod,theninthe2ndperiodyouchoose\Mountain1".Ifthemountainrevenuewas100gninthe1stperiodyouchoose\Mountain2".CaseA:Revenueof\Mountain2"is0gnin1stperiodYourincomeinthe1stperiodthereforeis0+0*0-0.75=-0.75gnandinthe2ndperiodyouchoose\Mountain1"whichmayyield0gnor100gn(andyouknowforsurethat\Mountain2"yieldsarevenueof0gninthe2ndperiod).Let'sassumethat\Mountain1"yieldsarevenueof100gninthe2ndperiod.Henceyourincomeinthe2ndperiodis0+100*0.1-0.25=9.75gnCaseB:Revenueof\Mountain2"is100gnin1stperiodYourincomeinthe1stperiodthereforeis0+100*0-0.75=-0.75gnandinthe2ndperiodyouchoose\Mountain2".Since\Mountain2"hadarevenueof100gninthe1stperiodyouknowforsurethat\Mountain2"yieldsarevenueof100gninthe2ndperiod.Henceyourincomeinthe2ndperiodis0+100*0.1-0.75=9.25gn38 ReferencesAcharya,V.V.,R.Baghai-Wadji,andK.Subramanian,2009,\LaborLawsandInnovation,"SSRNeLibraryAcharya,V.V.,andK.Subramanian,2009,\BankruptcyCodesandInnovation,"ReviewofFinancialStudies,22,4949{4988.Aghion,P.,J.V.Reenen,andL.Zingales,2008,\InnovationandInstitutionalOwnership,"WorkingPaperAghion,P.,andJ.Tirole,1994,\TheManagementofInnovation,"QuarterlyJournalofEconomics,109,1185{1209.Almeida,P.,andB.Kogut,1999,\LocalizationofKnowledgeandtheMobilityofEngineersinRegionalNetworks,"ManagementScience,45,905{917.Amabile,T.M.,1996,CreativityinContext.WestviewPress,Boulder,CO.Argote,L.,B.McEvily,andR.Reagans,2003,\Introductiontothespecialissueonmanagingknowledgeinorganizations:Creating,retaining,andtransferringknowledge,"ManagementScience,49(4),5{8.Ariely,D.,U.Gneezy,G.Loewenstein,andN.Mazar,2009,\Largestakesandbigmistakes,"ReviewofEconomicStudies,76(2),451{469.Arrow,K.J.,1969,\Classi catoryNotesontheProductionandTransmissionofTechnologicalKnowledge,"TheAmericanEconomicReview,59(2),29{35.Atanassov,J.,2008,\QuietLifeorManagerialMyopia:IstheThreatofHostileTakeoversBene cialforTechnologicalInnovation?,"SSRNeLibraryAzoulay,P.,J.G.Zivin,andG.Manso,2011,\IncentivesandCreativity:EvidenceFromtheAcademicLifeSciences,"RandJournalofEconomics,42,527{554.Bandiera,O.,I.Barankay,andI.Rasul,2005,\Socialpreferencesandtheresponsetoincen-tives:Evidencefrompersonneldata,"QuarterlyJournalofEconomics,120(3),917.Banks,J.,M.Olson,andD.Porter,1997,\Anexperimentalanalysisofthebanditproblem,"EconomicTheory,10(1),55{77.Bebchuk,L.,andJ.Fried,2004,PayWithoutPerformance.HarvardUniversityPress,Cam-bridge,MA.Bull,C.,A.Schotter,andK.Weigelt,1987,\TournamentsandPieceRates:AnExperimentalStudy,"TheJournalofPoliticalEconomy,95(1),1{33.Chemmanur,T.,andX.Tian,2011,\DoAntitakeoverProvisionsSpurCorporateInnova-tion?,"WorkingPaperChemmanur,T.,X.Tian,andE.Loutskina,2011,\CorporateVentureCapital,ValueCre-ation,andInnovation,"WorkingPaperDickinson,D.L.,1999,\AnExperimentalExaminationofLaborSupplyandWorkIntensi-ties,"JournalofLaborEconomics,17(4),638{670.Fahr,R.,andB.Irlenbusch,2000,\Fairnessasaconstraintontrustinreciprocity:earnedpropertyrightsinareciprocalexchangeexperiment,"EconomicsLetters,66(3),275{282.Falk,A.,andA.Ichino,2006,\CleanEvidenceonPeerE ects,"JournalofLaborEconomics24(1),39{57.Fallick,B.,C.A.Fleischman,andJ.B.Rebitzer,2006,\Job-HoppinginSiliconValley:SomeEvidenceConcerningtheMicrofoundationsofaHigh-TechnologyCluster,"TheReviewof39 EconomicsandStatistics,88(3),472{481.Fehr,E.,S.Gachter,andG.Kirchsteiger,1997,\ReciprocityasaContractEnforcementDevice:ExperimentalEvidence,"Econometrica,65(4),833{860.Fischbacher,U.,2007,\z-Tree:Zurichtoolboxforready-madeeconomicexperiments,"Ex-perimentalEconomics,10(2),171{178.Francis,B.,I.Hasan,andZ.Sharma,2010,\DoIncentivesCreateInnovation?EvidencefromCEOCompensationContracts,"WorkingPaperFrederick,S.,2005,\Cognitivere\rectionanddecisionmaking,"JournalofEconomicPer-spectives,19(4),25{42.Frey,B.,andR.Jegen,2001,\Motivationcrowdingtheory,"JournalofEconomicSurveys15(5),589{611.Gabaix,X.,D.Laibson,G.Moloche,andS.Weinberg,2006,\CostlyInformationAcquisi-tion:ExperimentalAnalysisofaBoundedlyRationalModel,"AmericanEconomicReview96(4),1043{1068.Gilson,R.J.,1999,\TheLegalInfrastructureofHighTechnologyIndustrialDistricts:SiliconValley,Route128,andCovenantsNottoCompete,"NewYorkUniversityLawReview,74,575{629.Gneezy,U.,S.Meier,andP.Rey-Biel,2011,\WhenandWhyIncentives(Don't)WorktoModifyBehavior,"JournalofEconomicPerspectives,25(4),191{209.Gneezy,U.,andA.Rustichini,2000a,\AFineIsaPrice,"JournalofLegalStudies,29,1. ,2000b,\Payenoughordon'tpayatall,"TheQuarterlyJournalofEconomics115(3),791{810.Hellmann,T.,2007,\Whendoemployeesbecomeentrepreneurs?,"ManagementScience,53,919{933.Hellmann,T.,andV.Thiele,2011,\IncentivesandInnovation:AMulti-taskingApproach,"AmericanEconomicJournal:Microeconomics,3,78{128.Holmstrom,B.,andP.Milgrom,1987,\AggregationandLinearityintheProvisionofIn-tertemporalIncentives,"Econometrica,55(2),303{328.Kohn,A.,1993,Punishedbyrewards:thetroublewithgoldstars,incentiveplans,A's,praise,andotherbribes.HoughtonMiinCo.,Boston.Lazear,E.P.,2000,\PerformancePayandProductivity,"TheAmericanEconomicReview90(5),1346{1361.Lerner,J.,andJ.Wulf,2007,\InnnovationandIncentives:EvidencefromCorporateR&D,"ReviewofEconomicsandStatistics,89,634{644.Liu,Q.,andK.P.Wong,2011,\IntellectualCapitalandFinancingDecisions:EvidencefromtheU.S.PatentData,"ManagementScience,57(10),1861{1878.Manso,G.,2011,\MotivatingInnovation,"JournalofFinance,66,1823{1860.March,J.G.,1991,\ExplorationandExploitationinOrganizationalLearning,"OrganizationScience,2(1,SpecialIssue:OrganizationalLearning:PapersinHonorof(andby)JamesG.March),71{87.Marx,M.,D.Strumsky,andL.Fleming,2009,\Mobility,Skills,andtheMichiganNon-CompeteExperiment,"ManagementScience,55(6),875{889.McCullers,J.C.,1978,\IssuesinLearningandMotivation,"inTheHiddenCostsofRewarded.byM.R.Lepper,andD.Greene.LawrenceErlbaumAssociates,Hillsdale,NJ.40 McGraw,K.O.,1978,\TheDetrimentalE ectsofRewardonPerformance:ALiteratureReviewandaPredictionModel,"inTheHiddenCostsofReward,ed.byM.R.Lepper,andD.Greene.LawrenceErlbaumAssociates,Hillsdale,NJ.Meloso,D.,J.Copic,andP.Bossaerts,2009,\PromotingIntellectualDiscovery:PatentsVersusMarkets,"Science,323(5919),1335{1339.Meyer,R.J.,andY.Shi,1995,\SequentialChoiceunderAmbiguity:IntuitiveSolutionstotheArmed-BanditProblem,"ManagementScience,41(5),817{834.Nalbantian,H.R.,andA.Schotter,1997,\ProductivityUnderGroupIncentives:AnExper-imentalStudy,"TheAmericanEconomicReview,87(3),314{341.Rogerson,W.P.,1985,\RepeatedMoralHazard,"Econometrica,53(1),69{76.Sannikov,Y.,2008,\AContinuous-TimeVersionofthePrincipal-AgentProblem,"TheRe-viewofEconomicStudies,75,957{984.Sapra,H.,A.Subramanian,andK.Subramanian,2008,\CorporateGovernanceandInnova-tion:TheoryandEvidence,"SSRNeLibrarySauermann,H.,andW.M.Cohen,2010,\Whatmakesthemtick?EmployeeMotivesandIndustrialInnovation,"ManagementScience,56,2134{2153.Shearer,B.,2004,\PieceRates,FixedWagesandIncentives:EvidencefromaFieldExperi-ment,"TheReviewofEconomicStudies,71(2),513{534.Sillamaa,M.-A.,1999,\Howworke ortrespondstowagetaxation:Anexperimentaltestofazerotopmarginaltaxrate,"JournalofPublicEconomics,73(1),125{134.Sorenson,O.,andT.E.Stuart,2003,\Liquidityeventsandthegeographicdistributionofentrepreneurialactivity,"AdministrativeScienceQuarterly,48(175-201).Tian,X.,andT.Wang,2010,\ToleranceforFailureandCorporateInnovation,"WorkingPaperWeitzman,M.L.,1979,\OptimalSearchfortheBestAlternative,"Econometrica,47(3),641{654.Yanadori,Y.,andJ.H.Marler,2006,\Compensationstrategy:doesbusinessstrategyin-\ruencecompensationinhigh-technology rms?,"StrategicManagementJournal,27(6),559{570.41