Exposure to Artificial Intelligence and Occupational Mobility: A Cross-Country Analysis
ExposuretoArtificialIntelligenceandOccupationalMobility: ACross-CountryAnalysis MauroCazzaniga,CarloPizzinelli,EmmaRockall,MarinaM.Tavares WP/24/116 IMFWorkingPapersdescriberesearchinprogressbytheauthor(s)andarepublishedtoelicitcommentsandtoencouragedebate. TheviewsexpressedinIMFWorkingPapersarethoseoftheauthor(s)anddonotnecessarilyrepresenttheviewsoftheIMF,itsExecutiveBoard,orIMFmanagement. 2024 JUN ©2024InternationalMonetaryFundWP/24/116 IMFWorkingPaper ResearchDepartment ExposuretoArtificialIntelligenceandOccupationalMobility:ACross-CountryAnalysisPreparedbyMauroCazzaniga,CarloPizzinelli,EmmaRockall,MarinaM.Tavares AuthorizedfordistributionbyFlorenceJaumotte June2024 IMFWorkingPapersdescriberesearchinprogressbytheauthor(s)andarepublishedtoelicitcommentsandtoencouragedebate.TheviewsexpressedinIMFWorkingPapersarethoseoftheauthor(s)anddonotnecessarilyrepresenttheviewsoftheIMF,itsExecutiveBoard,orIMFmanagement. ABSTRACT:Wedocumenthistoricalpatternsofworkers'transitionsacrossoccupationsandoverthelife-cyclefordifferentlevelsofexposureandcomplementaritytoArtificialIntelligence(AI)inBrazilandtheUK.Inbothcountries,college-educatedworkersfrequentlymovefromhigh-exposure,low-complementarityoccupations(thosemorelikelytobenegativelyaffectedbyAI)tohigh-exposure,high-complementarityones(thosemorelikelytobepositivelyaffectedbyAI).Thistransitionisespeciallycommonforyoungcollege-educatedworkersandisassociatedwithanincreaseinaveragesalaries.YounghighlyeducatedworkersthusrepresentthedemographicgroupforwhichAI-drivenstructuralchangecouldmostexpandopportunitiesforcareerprogressionbutalsohighlydisruptentryintothelabormarketbyremovingstepping-stonejobs.Thesepatternsof“upward”labormarkettransitionsforcollege-educatedworkerslookbroadlyalikeintheUKandBrazil,suggestingthattheimpactofAIadoptiononthehighlyeducatedlaborforcecouldbesimilaracrossadvancedeconomiesandemergingmarkets.Meanwhile,non-collegeworkersinBrazilfacemarkedlyhigherchancesofmovingfrombetter-paidhigh-exposureandlow-complementarityoccupationstolow-exposureones,suggestingahigherriskofincomelossifAIweretoreducelabordemandfortheformertypeofjobs. RECOMMENDEDCITATION:Cazzaniga,M.,C.Pizzinelli,E.Rockall,andM.M.Tavares(2024)“ExposuretoArtificialIntelligenceandOccupationalMobility:aCross-CountryAnalysis.”IMFWorkingPaper24/116 JELClassificationNumbers: J23,J23,O33 Keywords: Artificialintelligence;Employment;Occupations;EmergingMarkets Author’sE-MailAddress: mauro98cazzaniga@gmail.com,cpizzinelli@imf.org,erockall@stanford.edu,mmendestavares@imf.org INTERNATIONALMONETARYFUND2 ExposuretoArtificialIntelligenceandOccupationalMobility:ACross-CountryAnalysis MauroCazzaniga FGV-SP CarloPizzinelli IMF EmmaRockall StanfordUniversity MarinaM.Tavares IMF June3,2024 Abstract Wedocumenthistoricalpatternsofworkers’transitionsacrossoccupationsandoverthelife-cyclefordifferentlevelsofexposureandcomplementaritytoArtificialIntelligence(AI)inBrazilandtheUK.Inbothcountries,college-educatedworkersfrequentlymovefromhigh-exposure,low-complementarityoccupations(thosemorelikelytobenega-tivelyaffectedbyAI)tohigh-exposure,high-complementarityones(thosemorelikelytobepositivelyaffectedbyAI).Thistransitionisespeciallycommonforyoungcollege-educatedworkersandisassociatedwithanincreaseinaveragesalaries.YounghighlyeducatedworkersthusrepresentthedemographicgroupforwhichAI-drivenstructuralchangecouldmostexpandopportunitiesforcareerprogressionbutalsohighlydisruptentryintothelabormarketbyremovingstepping-stonejobs.Thesepatternsof“up-ward”labormarkettransitionsforcollege-educatedworkerslookbroadlyalikeintheUKandBrazil,suggestingthattheimpactofAIadoptiononthehighlyeducatedlaborforcecouldbesimilaracrossadvancedeconomiesandemergingmarkets.Meanwhile,non-collegeworkersinBrazilfacemarkedlyhigherchancesofmovingfrombetter-paidhigh-exposureandlow-complementarityoccupationstolow-exposureones,suggestingahigherriskofincomelossifAIweretoreducelabordemandfortheformertypeofjobs. Keywords:ArtificialIntelligence,Employment,Occupations,EmergingMarkets JELCodes:J23,J23,O33 1Correspondingauthors:MauroCazzanigaandMarinaM.Tavares.Email: mauro98cazzaniga@gmail.commmendestavares@imf.orgTheauthorswouldliketothankFlorenceJaumotte,GiovanniMelina,andAugustusPantonforhelpfulcomments.Disclaimer:Theviewsexpressed 1Introduction Theadoptionofartificialintelligence(AI)promisestobringaboutawaveofstruc-turaltransformationthatcouldreshapeboththeeconomyandsociety.Asindustriesin-creasinglyintegrateAItechnologies,therepercussionsofthisshiftareexpectedtobemostacutelyfeltinthelabormarket.Here,AIhasthepotentialtobothaugmenttheproduc-tivityofcertainworkersandcompetedirectlywithothers.Thisdualpotentialpointstoaperiodofsignificantdisruptionandadaptationintheworkforce,possiblychallengingtra-ditionalnotionsofemployment,skillrequirements,andjobsecurity.SeveralstudieshavesoughttocategorizethelikelyimpactofAIondifferentoccupations,offeringasnapshotoflabormarketopportunitiesandrisksbasedoncountries’currenteconomicstructures.Thispaperattemptstomovebeyondthese“sta
