INNOVATION AND MARKET STRUCTURE IN THE DYNAMICS OF

INNOVATIONANDMARKETSTRUCTUREINTHEDYNAMICSOFTHEPHARMACEUTICALINDUSTRYANDBIOTECHNOLOGY:TOWARDSAHISTORYFRIENDLYMODELFrancoMalerba*andLuigiOrsenigo***DepartmentofEconomicsandCESPRI,UniversitàL.Bocconi,Milan**UniversityofBresciaandCESPRI,UniversitàL.Bocconi,[e-mail]@uni-bocconi.itPresentedattheDRUIDNelsonandWinterConference,Aalborg,June12-15,2001WethankLucaBerga,ChristianGaravagliaandNicolaLaceterafortheirinvaluablecontributiontothedevelopmentofthemodel.WethanktheMerckFoundationforitssupportthroughtheEpris(EuropeanPharmaceuticalRegulationandInnovationSystems)Project.2I.IntroductionInthispaperwepresentafirstversionofa“history-friendly”modeloftheevolutionofthepharmaceuticalindustryandbiotechnology.Themotivationsunderlyingthismodellingstylehavebeendiscussedextensivelyinpreviouspapers(Malerbaetal1999,2001)andwewillnotcomebacktothisissuehere.Forthepurposesofthepresentpaper,sufficeittosaythatpharmaceuticalsconstitutesanidealsubjectforhistory-friendlyanalysis,forseveralreasonsandespeciallyincomparisontoourpreviouseffortsconcerningthecomputerindustry.Firstofall,itisworthremindingthatoneofthereasonsofinterestofhistory-friendlymodelsconsistsinthepossibilityofexaminingwhatkindoffactorsanddynamicprocessesaccountfortheevolutionofdifferentindustries.Inthisrespect,pharmaceuticalsareradicallydifferentfromthecomputerindustry.PharmaceuticalsaretraditionallyahighlyR&Dintensivesector,whichhasundergoneaseriesofradicaltechnologicalandinstitutional“shocks”.However,thecoreofleadinginnovativefirmsandcountrieshasremainedquitesmallandstableforaverylongperiodoftime,butthedegreeofconcentrationhasbeenconsistentlylow,whateverthelevelofaggregationisconsidered.Asweshallargue,thesepatternsofindustrialdynamicsareintimatelylinkedtotwomainfactors.First,thenatureoftheprocessesofdrugdiscovery,i.e.tothepropertiesofthespaceoftechnologicalopportunitiesandofthesearchproceduresthroughwhichfirmsexploreit.Second,thefragmentednatureoftherelevantmarkets.Specifically,innovationprocesseshavebeencharacterisedforaverylongtimebylowdegreeofcumulativenessandby“quasi-random”proceduresofsearch(randomscreening).Thus,innovationinonemarket(atherapeuticcategory)doesnotentailhigherprobabilitiesofsuccessinanotherone.Moreover,pharmaceuticalsrepresentsacasewherecompetitionislessdissimilartothemodelofpatentraces.Understandingiftheseintuitivefactorscanindeedexplaintheobservedpatternsofindustrialdynamicsandarticulatingthemechanismsthroughwhichtheyexerttheirimpactisinitselfaninterestingchallenge.Themoreso,ifthismodeliscomparedtotheanalysisofthecomputerindustry.Thecomparisonmightallowforsomegeneralisationsaboutthedeterminantsoftherelevantsimilaritiesanddifferencesinthepatternsofindustrialevolutionacrossindustries.Asecondreasonwhymodellingthepharmaceuticalindustryisparticularlyinterestingisthatthissectorisusuallyconsideredasstronglyscience-based.However,sciencehasinfluencedindustrialresearchinquitedifferentwaysovertime.Indeed,intherecentpasttheadventofanewscience–molecularbiology–hashadadramaticimpactonindustrystructure,theorganisationofinnovativeactivitiesandthecompetitivenessoffirmsandcountries.Thus,theanalysisofthepharmaceuticalindustrylendsitselftoastudyofaclassicalandextremelyimportantchapteroftheeconomicsofinnovation,i.e.therelationshipsbetweenscientificresearchandindustrialinnovation.Third,thepharmaceuticalindustry,eversinceitsinception,hasbeendeeplyaffectedbyalargevarietyofinstitutionalfactorsandpolicies,rangingfrompatents,differentformsofregulation(proceduresforproductapproval,pricecontrols,etc..),organisationofthepublicresearchsystems,etc..Fromthisperspective,pharmaceuticalsconstitutesanidealcaseforstudyingthedifferentialimpactandtheworkingofalternativepoliciesandnationalsystemsofinnovation.Inapreviouspaper(MalerbaandOrsenigo,2001)webegantoanalysethedynamicsofmarketstructureinpharmaceuticalsinthe“randomscreeningperiod”.Moreover,wealsoperformedsome3preliminaryexercisesconcerningtheeffectsofalternativepatentregimesoninnovationandmarketstructure.Inthisnewversionofthemodel,wefocusinsteadonthetransitionto“molecularbiology”.Thepaperisorganisedasfollows.SectionIIprovidesabriefhistoricalaccountoftheevolutionofthepharmaceuticals.Aparticularemphasisisassignedtothediscussionoftheevolutionoftheregimesofsearchofnewdrugsthathavecharacterisedtheindustryovertime.SectionIIIintroducesthemaintheoreticalissuesthatareraisedbytheprevioushistoricalaccountandpresentsthemodel.SectionIVdiscussestheresultsandSectionVconcludesthepaper.II.Innovationandtheevolutionofmarketstructureinthepharmaceuticalindustry:anoverviewThepatternsofdevelopmentofthepharmaceuticalindustryhavebeenextensivelyanalysedbyseveralscholars.Inwhatfollows,werelyespeciallyontheworkbyChandler1990and1998,GalambosandSewell1996,GalambosandSturchio1996and1998,Gambardella1995,Henderson,OrsenigoandPisano,1999,McKelvey,1996,Orsenigo1989,Schwartzman1976.Inverygeneralterms,thehistoryofthepharmaceuticalindustrycanbeanalysedasanevolutionaryprocessofadaptationtomajortechnologicalandinstitutional“shocks”.I