Knowledge Sources of Innovation in a Small Open Ec

KnowledgeSourcesofInnovationinaSmallOpenEconomy:TheCaseofSingaporeWongPoh-Kam([e-mail])HoYuen-Ping([e-mail])EntrepreneurshipCentreNationalUniversityofSingapore1KnowledgeSourcesofInnovationinaSmall,OpenEconomy:TheCaseofSingaporeAbstractBytracingtheflowsofpatentcitationofpriorpatentsandscientificjournalarticles,weinvestigatethesourcesofknowledgeforinnovationoutputinSingapore,asmall,highlyopeneconomythathastraditionallybeensignificantlydependentonforeignmultinationalcorporations(MNCs).WefoundthatthelocalproductionofnewknowledgebyindigenousSingaporeanfirmsdependsdisproportionatelyontechnologicalknowledgeproducedbyMNCswithoperationalpresenceinSingaporeandscientificknowledgegeneratedbyforeignuniversities.Locallyproducednewknowledgebyindigenousfirmsandlocaluniversitiesconstituteaninsignificant,albeitrapidlygrowing,sourceforinnovationinSingapore.JELCodes:033,031,053Keywords:innovationsystem,patentcitation,Singapore,knowledgesources21.IntroductionItiswellestablishedthatafirm’sinnovationispartiallydependentonR&Dandknowledgeflowsfromotherfirmsandindustries.AnimportantempiricalfindingbyJaffe(1986)andAcsetal.(1992,1994)isthatR&Dinvestmentbyprivatecorporationsanduniversitiesspillsovertootherthirdpartyfirms,whomaythenexploitthisR&Dspilloverintheirowninnovativeactivities.ThebodyofworkonR&Dspillovershasdevelopedsignificantly,asreviewedbyGriliches(1992).Economictheoristshaveatthesametimeincorporatedknowledgespilloversasadeterminantofaggregateeconomicgrowthinthenewclassofendogenousgrowthmodels,pioneeredbyRomer(1986,1990).Morerecently,literatureonknowledgerecombinationhasshownthatorganizationsinnovatethroughcombiningnewandexistingknowledge,includingknowledgefromexternalsources(KogutandZander,1992;Songetal.,2001).Theexistenceofspilloversledtoresearchersstudyingtheroleofgeographyandthespatialdimensioninthefirminnovationprocess(Jaffe,1989;Krugman,1991;Feldman,1994).StudiesofAmericanfirmsbyJaffeetal.(1993andAudretschandFeldman(1996),amongothers,empiricallyestablishedtheexistenceofgeographiclocalizationofR&D;ieknowledgespilloversaremostlikelyexploitedbyfirmsinthegeographicvicinityofthefirmfromwhichtheknowledgeoriginated.ThegeographiclocalizationofknowledgespilloversisbynowanacceptedfactintheUSA,andevidencetosupportthehypothesishasalsobeenfoundinlargecountriessuchasJapan(Branstetter,2001;GotoandNagata,1997),France(PiergiovanniandSantarelli,2001),Germany(Herrigel,1993)andItaly(Piergiovannietal.,1997).Therehavealsobeenanumberofstudiesongeographiclocalizationofknowledgeinspecificindustries,suchasthesemi-conductorsector(SorensenandStuart,2000;Almeida,1996).However,thereisnoequivalentempiricalliteratureonknowledgeflowsandsourcesofknowledgeforinnovationinsmallercountries.EarlyempiricalstudiesonknowledgespillovershavetypicallyuseddataonR&Dexpenditureandwerefocusedonidentifyingwherespilloversgo.Jaffeetal.(1993)pioneeredtheuseofpatentcitationsdatatotracetheflowofknowledge.Otherauthors,suchasFrost(2001)andAlmeida(1996)haveadoptedthisapproachofusingpatentscitationtostudyknowledgeflowsandspillovers.Anotherapproachistousepublicationscitationdata.ThishasbeennotablyaccomplishedintheUSAusingtheSmallBusinessAdministration’sInnovationDataBase(SBIDB).TheSBIDBhasbeenusedinstudiesonthegeographyofinnovationbyAcsetal.(1992,1994),Feldman(1994)andAudretschandFeldmand(1996).Manystudiesonknowledgespillovershavefocusedonthedestinationofoutgoingknowledge;establishingthattherearegeographicboundstothe“reach”ofspillovers.Theflipsideapproachistoexaminetherecipientfirmsandtheknowledgesourcesthatareincludedintheirsearchprocesswhenseekingtoexpandtheirknowledgestocks.Usingthisapproach,StuartandPodolny(1996)andRosenkopfandAlmeida(2003)havefoundthatfirms’searchfornewknowledgeisgeographicallyboundedtoitsimmediatevicinity.2.KnowledgesourcesforInnovationinSmall,OpenEconomies3Inthispaper,wepresentfindingsontheknowledgesourcesofinnovationbyindigenousfirmsinSingaporebytracingthecitationsmadebypatentsinventedinSingapore.SingaporerepresentsaveryinterestingcaseexampleofasmallandhighlyopeneconomythattraditionallyhasdependedheavilyondirectforeigninvestmentbyglobalMNCs.InthelightofincreasingglobalizationandthegrowingpresenceofforeignMNCsindevelopingcountries,thepotentialroleofforeignfirmsasasourceofknowledgeforinnovationinthehostcountryneedstobegivengreaterattention.BecauseofthesignificantpresenceofforeignMNCsinhernationalinnovationsystem,SingaporethusoffersarelativelyadvancedcaseforstudyingtheroleofforeignMNCsasaknowledgesourceforinnovationforlocalfirms.TherehasbeenlittleresearchonpatentingactivitiesinSingapore,dueinparttothelowlevelofpatentingactivityinSingaporeinthepast.WhileSingaporehasbeensteadilyincreasingherR&Dintensitysincetheearly-1980s,itisreallyonlyfromthemid-1990sthatthepaceoftechnologicalinnovationleadingtointellectualpropertystartedtoaccelerate.AscanbeseenfromTable1,Singapore’sR&DexpendituretoGDPratiorosesteadilyfromlessthan0.3%in1980andlessthan1%in1990to2.15%by2003,whichalreadyexceedsthelevelofmanyOECDcountries.Butthelevelofpatentingactivity,asmeasu