CEPR-政策洞察129：如何降低風(fēng)險(xiǎn)：相互依存世界中的歐洲經(jīng)濟(jì)安全（英）

CEPRPOLICYINSIGHTNo.129

May2024

Howtode-risk:Europeaneconomicsecurityinaworldofinterdependence1

JeanPisani-Ferry,aceBeatriceWederdiMaurobdeand

JerominZettelmeyerae

aBruegel;bGenevaGraduateInstitute;cPIIE;dINSEAD;eCEPR

TheEUandotheradvancedeconomiesareattemptingto‘de-risk’theirexternal

economicrelationshipswithoutforegoingthebenefitsoftrade.Buthow?Drawing

onabodyofresearchpublishedinthejointCEPR/BruegelParisReport2,thisPolicy

Insightarguesthat(1)importde-riskingshouldfocusonafewproductcategorieswhere

thecostsofsupplyinterruptionsareunquestionablylarge;(2)riskexposuresthrough

exportsandfinancialchannelsrequiremoreattention;and(3)de-riskingmustbe

complementedbydeterrenceandbyraisingresiliencewithrespectalldisruptions,

wherevertheirorigin.ThelatterrequiresadeeperandbroaderEUSingleMarket.

1INTRODUCTION

Overaperiodofjust15years,Europehasbeenconfrontedwithafinancialshock

thatoriginatedintheUnitedStates,apandemicshockthatoriginatedinChinabut

couldhavecomefromanywhere,andanenergyshockprovokedbyRussia’sinvasion

ofUkraine.Theseeventshavepromptedare-examinationofefficiency/securitytrade-

offsthatariseasaresultofinternationalintegration,andparticularlyasaresultof

specialisationininternationaltradeandthevulnerabilitiesofglobalsupplychains.

Economistsandpolicymakershavelongworriedaboutsimilartrade-offs.Atthemost

fundamentallevel,suchtrade-offsarisefromthestandardtensionbetweengrowth

andeconomiccrises:highergrowthisoftenaccompaniedbygreaterinstability.

Forexample,regulationoffinancialandproductmarketsmaypreventormitigate

financialorenvironmentalhazardsatthecostofdampeningentryandgrowthof

firms.Similarly,inopeneconomies,tradeandfinancialintegrationmaybegoodfor

growth,butcanexposeeconomiestoimportedshocks.

Themostrecentsetofconcerns–asexemplified,forexample,byaseriesofEuropean

Commissionpapers(EuropeanCommission,2021,2022)andanassociatedlegislative

agenda(seeSection4andMcCaffreyandPoitiers,2024)–differsfromthesestandard

preoccupationsintworespects.

1ThisPolicyInsightisbasedonChapter1ofParisReport2(Pisani-Ferryetal.2024).WeareverygratefultoChad

Bown,JulianHinz,MorganKelly,ConorMcCaffrey,IsabelleMejean,KevinO’Rourke,NiclasPoitiers,PierreRousseaux

andMoritzSchularick,aswellastoShekarAiyar,AliciaGarciaHerrero,PetrosMavroidis,FrancescoPapadia,André

Sapir,FinoaScottMorton,NicolasVéron,LennardWelslauandGuntramWolffforhelpfuldiscussionsandcomments.

WeareparticularlygratefultoNiclasPoitiersforhiscontributionstothesurveyofEUeconomicsecurityinstruments

thatappearsinSection4ofthisPolicyInsight.

CEPRPOLICYINSIGHTNo.129May2024

2

First,economicrisksrelateincreasinglynotjusttocrisesorshocks,buttodeliberateeconomiccoercionbyforeigngovernmentsorevennon-governmentalentities(suchascriminalgroups).Thisisprobablythereasonwhytheterm‘security’–asopposedto‘stability’or‘resilience’–hasbecomepopulartodescribethemitigationofeconomic,ratherthanjustnational,securitythreats(wediscussthedifferenceinSection2).OnereasontobeconcernedwitheconomiccoercionisthatChina,anincreasinglypowerfulandauthoritariancountry,hasbeenapplyingcoercionregularlyinresponsetopoliticalactionsbytradepartners(forexample,Australia’scallforinvestigationsintotheoriginoftheCOVID-19pandemicandLithuania’sdecisiontoletTaiwanopenarepresentativeofficeinVilnius).2ButtheconcernisnotjustaboutChina:thepoliciesofPresidentTrumpbetween2017and2020showedthatevenone’sclosestallycanbetemptedtoleverageitsmarketpoweranditscontrolofthetechnicalandfinancialinfrastructuresofglobalisation.ThepossibilityofasecondTrumptermisnowpromptingareflectionontheneedforEuropetoprepareforsucharisk(GonzalesLayaetal.2024).

Second,recentconcernshavefocusedmostlyontrade-relatedratherthanfinancialvulnerabilities.Thisreflectsthefactthattrade-relatedvulnerabilitieshavebecomemoreprominentasaresultofspecialisationandthevulnerabilityofglobalsupplychainsthatmaximiseefficiency,butatthecostofcreatinghiddenfragilities.Buttheprominenceoftradeconcernsmayalsoreflectarathermyopicreasoning,asthelasttwoorthreeexternalshocksthatEurope(and,toalesserextent,theUS)hassufferedhavebeentrade-related:supplychaindisruptionsrelatedtoCOVID-19andenergypriceshocksfollowingtheRussianinvasionofUkraine.

Inlinewiththisconcern,wefocusmostlyontrade-relatedexternaleconomicsecurity.ThisshouldnotbetakentoimplythatEuropedoesnotneedtoworryaboutfinancialsecurity.Butunliketrade-relatedsecurity,financialriskscontinuetobemostlyofthefinancial-stabilityvariety,linkedtoshocksandfinancialvulnerabilitiesratherthancoercion.Totheextentthatfinancialcoercionisaseriousconcern,itislinkedtoonemainpotentialsource:theUnitedStatesifPresidentTrumpreturns(seeSection2).Incontrast,trade-relatedexternalsecurityrisksareubiquitous.

InthisPolicyInsight,weseektoanswertwocriticalquestions.First,howshouldtrade-relatedvulnerabilitiesbeidentified,andwhattraderelationshipsmakeEuropeparticularlyvulnerabletoshocksandcoercion?Second,howcanthesevulnerabilitiesbereducedwhileminimisingthecostsof‘de-risking’andreducingthechancesofunintendedconsequences?Foursuchpotentialcostscometomind:

?Foregoingsomeofthegainsfromtradespecialisationandtradeopenness.ThiscouldweighonEuropeangrowthandcompetitiveness,whichdependonexportspecialisationandonimportingrawmaterialsandintermediateinputsmorecheaplythantheycouldbeproducedathome(ifatall).Itcouldalsomakeithardertoattainemissionreductionobjectives,byraisingthecostofthetransitiontorenewableenergysources.Inturn,thiscouldexacerbatesocialandpoliticaldivisionsrelatedtoclimateaction.

?Becomingmorevulnerabletodomesticshocksincludingnaturaldisasters,epidemicsandhome-grownfinancialcrises–andmoregenerally,toanyshockwhoseconsequenceswouldbemitigatedbyinternationaltradeand/orcapitalflows.

?Damaginginternationalcooperation.ThiscouldincludeEUcooperationwithChinaonvitalmattersofcommoninterest,suchasclimate-changemitigation,aswellasrespectfortherulesofthemultilateraltradingsystem.NotwithstandingthedamagethattheWTOhassufferedoverthelastdecade,theserulescontinuetobelargelyrespected(MavroidisandSapir,

2See,forexample,“

ChinapunishesAustraliaforpromotinganinquiryintocovid-19

”,TheEconomist,21May2020;and

“Lithuaniacomplainsoftrade‘sanctions’byChinaafterTaiwandispute

”,FinancialTimes,3December2021.

CEPRPOLICYINSIGHTNo.129May2024

3

2024).Anaggressive‘de-risking’ofEuropeantraderelationshipsthroughtradepolicytoolsandsubsidiescouldtriggerprotectionistreactionsfromtradingpartners,particularlyifmeasuresviolateWTOrules.ItcouldalsobecomeanexcuseforprotectionistsintheEU,whomightuseeconomic-securityargumentstofurtherspecialinterests.

?DamagingcohesionwithintheEU.EUcountriesdifferintheirtradestructuresandtheirdependenceonspecificexportandimportmarkets.Asaresult,attemptstode-risktrademayhavenetbenefitsforsomeandnetcostsforothers.Ifde-riskingbecomesasourceofdivision,itmaybecounterproductive,asinternaldivisionsintheEUarepartlywhatanadversary–whetherChina,RussiaorPresidentTrump–mighttrytoexploit(andindeed,divisionsarewhatthesethreepowershavetriedtoexploitinthepast).

TheremainderofthisPolicyInsightsummarisesasbestwecantheanswerstothesequestions,drawingonasetofpaperscommissionedbyBruegelandCEPR(Pisani-Ferryetal.,2024).Section2defineswhatwemeanbyeconomicsecurity,andwhatrisksweshouldbeworryingabout.Section3discusseshowtheserisksshouldbeaddressedinprinciple.Whattraderelationshipsrequirede-risking?Section4discussestheinstruments.Howcanprotectionbebuiltthatpreservesthebenefitsoftrade?Aconcludingsectionsummarisesthemainlessons.

2DEFININGRISKSTOECONOMICSECURITY

AsnotedbyBown(2024),economicsecurityremainsanemergingconcept.Atitsmostabstractlevel,itcanbedefinedasbothpreventingbadeconomicoutcomesandmakingsurethatshouldrisksmaterialise,thedamagetheycauseisminimised.Societiescarebothaboutraisingexpectedwelfareandaboutreducingitsvolatility.Economicsecurityisconcernedwiththelatter.

Definedinthisbroadway,economicsecurityhasbeenastandardconcernofpolicymakersforcenturies–andnotjustofeconomicpolicymakers,sinceeconomicharmcanbeinflictedby‘non-economic’shocks,includingpoliticaldisruptionandwars.Theuseofstateinterventiontoaddresstheseconcerns,includingindustrialpolicyandtradepolicy,issimilarlynothingnew(KellyandO’Rourke,2024).Thequestion,then,ishowtheconceptof‘economicsecurity’differsfromthoseof‘economiccrisisprevention’or‘nationalsecurity’.Totheextentthattheperceivednatureoftheriskandriskpropagationhaschanged,itisimportanttounderstandhowithaschanged,toavoidduplication,andtopreventoverreactiontoperceivednewriskswhentheoldrisksandriskpropagationchannelsmightstillbethere.

Economistsconcernedwithcrisispreventionandmitigationtypicallyfocusonrisksandvulnerabilitiesrelatedtothefinancialsystemorthestructureofproduction.Forexample,creditcyclescanexposecountriestofinancialcrises,whicharepropagatedinternationally.Dependenceoncommodityexportsorimportsexposeseconomiestoswingsininternationalpricesandtodisruptiontodomesticproductionthatreliesoncommodityimports.

Militaryandsecurityexpertsworryaboutadifferenttypeofthreat:harmthatisinflictedpurposelybyoutsideactors,normallynationstates,butalsoterroristorcriminalorganisations.MurphyandTopel(2013)widenedthedefinitionofnationalsecuritytoincludeall“substantialthreats”tothesafetyandwelfareofanation’scitizens,forexampleincludingnationalcatastrophesandpublichealththreats.Definedthisbroadly,nationalsecuritywouldincludepreparednessandmitigationagainstanyharmfulactsconductedbyforeigngovernmentsornon-governmentalorganisationswithmilitaryornon-militarymeans,includingeconomicsanctions,andthreatsrelatedtophysicalandinformationinfrastructure.

CEPRPOLICYINSIGHTNo.129May2024

4

Therecentusageoftheterm‘economicsecurity’isattheintersectionofnon-financial

economiccrisesandnationalsecurityinthebroadsensedefinedbyMurphyandTopel.3Specifically,itfocusesonharminflictedthroughinternationaleconomicrelationships–andparticularlytraderelationships–whetherthesereflectexogenousshocks(suchasCOVID-relatedtradedisruption)ordeliberateactionsbyforeigngovernmentsornon-governmentalorganisations(Bown,2024;McCaffreyandPoitiers,2024;EuropeanCommission,2021,2022).TheserisksareparticularlyrelevanttodaybecauseofthecombinationofeconomicintegrationthroughtradeandFDI,specialisation,longsupplychainsandactorswillingtoengageincoercionthroughthesechannels.

Itisinthissensethattheterm‘economicsecurity’willbeusedintheremainderofthisPolicyInsight.Inthisdefinition,achievingeconomicsecurityinvolvesthepreventionandmitigationof:

?disruptiontocriticalimports,whetheraccidentalordeliberate;

?economiccoercionthroughrestrictionsorboycottsonspecificexports,alongthelinesofactionstakenbyChinaagainstAustralia,orthroughpressuresonforeigncompaniesevenwhentheyproducelocally(forexample,threatsofdeprivingthemfromaccesstothedomesticmarket,restrictionsonprofitrepatriation,orexpropriation);

?abroaddisruptionofglobaltradeatascalewithmacroeconomicimpact,forexample,asaresultofgeopoliticalconflictleadingtoeconomicsanctionsoraprotractedtariffwarwithamajortradingpartner.EventsthatcouldtriggersuchscenariosincludeaChineseattackonTaiwan,orthere-electionofPresidentTrumptriggeringasharpdeteriorationofthepoliticalrelationshipbetweentheUSandtheEU.

Itisimportanttoemphasisethatthisanarrow–perhapsinappropriatelynarrow–definitionofeconomicsecurity,fortworeasons.

First,itdisregardsthepossibilityofeconomicdisruptionsasaresultofdomesticshocks,whichhistoricallyhavebeenamajorsourceofeconomiccrises(Table1).Hence,abettertermforthetypeofeconomic-securityriskswediscusswouldbe‘externaleconomicsecurity’.Thisterminologyremindsusthattherecouldbetrade-offsnotjustbetweeneconomicsecurityandeconomicgrowth,butalsobetweenexternaleconomicsecurityandsecurityfromdomesticshocks.Internationalintegrationmayincreaseexposuretotheformer,butoffersprotectionagainstthelatter.

Table1Varietiesofwelfarethreatsandpropagationmechanisms

Origin

Domesticshock

Externalshock

Deliberateaction

Propagation

Tradeandinvestment

Financial

Disease

Military

Other

Economic

Externaleconomic

crises

securityrisks

National

Epidemics/pandemics

securityrisks

Source:Authors’illustration.

Note.ThecolumnsinTable1definetheoriginofabadevent–anexogenousshockoriginatingathomeorabroad(productiondisruption,naturalcatastrophe,transportationorinfrastructuredisruption,confidenceshock);oradeliberateactionbyaforeigngovernmentoranon-governmentalentity.Therowsdefinethe

propagationchannel:economicactivityrelatedtotradeorfinance,disease,militaryaction,orother(forexample,throughITinfrastructure).

3TheEuropeanCommission(2023)usesadefinitionwhichalsoincludes“risksrelatedtophysicalandcybersecurityofcriticalinfrastructure”and“risksrelatedtotechnologysecurityandtechnologyleakage”.Wewouldclassifythisaspartofnationalsecurity(withinthe‘other’categoryinTable1)ratherthaneconomicsecurity.

CEPRPOLICYINSIGHTNo.129May2024

5

Second,thenarrowdefinitionlargelyignoresexternaleconomicsecurityrisksthroughfinancialchannels.However,internationalfinance–includingtheinternationalpaymentssystemandtheconfiscationoffinancialassetslocatedinforeignjurisdictions–isanobviousinstrumentofeconomiccoercionandeconomicsanctions,asshownbyG7sanctionsagainstRussiasinceitsfull-scaleinvasionofUkraine.ThemainreasonwhyfinancialrisksdonotfeatureprominentlyintherecentliteratureonEuropeaneconomicsecurityisthatEuropeismuchlesslikelytobeonthereceivingendofsuchsanctions,giventhecontrolexertedbytheUSanditsalliesoverinternationalfinance.ButthiscouldrapidlychangeifPresidentTrumpisre-electedanddecidestousefinancialcoercionagainstEuropeforwhateverreason(forexample,toforceEuropetoalignitsforeignorcommercialpolicieswiththoseoftheUS,aswasthecasewhentheUSthreatenedEUfirmswith‘secondarysanctions’forviolatingUS-imposedsanctionsonIran).

AbroaderanalysisofEuropeaneconomicsecurityshouldtakeintoaccountsuchfinancialeconomicrisksandhowtomitigatethem.Fornow,theremainderofthisPolicyInsightfocusesontradeandinvestment-relatedrisks.TheseareparticularlyrelevantfortherelationshipwithChina,butcouldalsobecomerelevantintheeventofareturnofPresidentTrumpandarevivalofUStariffsagainstEurope,whetherimposedformercantilistorpoliticalreasons.

3WHATTODE-RISK

Firmshaveincentivestoavoidbecomingdependentononeorasmallnumberofsuppliersorcustomers,particularlywhenthosesuppliersorcustomersarevulnerabletohighrisksoutsidetheircontrol,includingpoliticallymotivatedinterference.Yet,asMejeanandRousseaux(2024)pointout,firms’privateinterestinsecuritymaynotbeenoughtotakecareofthecollectiveEUsecurityinterest.Firmsoftenfailtorealisetheextenttowhichsuppliersorcustomersarethemselvessubjecttorisks,simplybecausetheydonotknowtheentirevaluechain.Firmsalsodonotinternalisethepotentialcostsofsupplierorcustomerdependencyontheentirevaluechain,andultimatelythewelfareofcitizens.Ifasupplierrelationshiprepresentsacriticallinkinthatchain,thesocialcostsofthatlinkfailingmayfarexceedtheprivatecoststothefirm.Thisargument,whichisbroadlyconsistentwiththeevidencepresentedbyBown(2024),canjustifypolicy-ledde-risking.

Butwhatareasoftraderequirede-risking?Howcanpolicymakerstellwhentradedependenciesareexcessive,inthesensethattheeconomicsecurityrisksoftradeoutweighitsbenefits,bothforefficiencyandgrowthandasprotectionagainstdomesticdisruption?Theidealwaytoanswerthisquestionwouldbethroughafirm-levelmodeloftradeandsupplyrelationships,bothacrossbordersandwithintheEU.Themodelwould‘know’whotradeswithwhom,howspecificinputsentereachstageofproduction,andtowhomfirmssell.Itwouldalsohaveinformationabouttheeaseofswitchingsuppliersifasupplierfailsorsharplyraisesitsprices.SuchamodelcouldbeusedtostresstestEuropeaneconomiesinrelationtospecificsupplychainorcustomerrisks.Wherelargeeffectsarefound,itwouldbeusedtoidentifytradingrelationshipsworthde-risking.Unfortunately,suchamodeldoesnotexistandmayneverexistbecauseofdatalimitations.Wearethereforeconstrainedbytheavailableinformationandshouldmakethebestofit.

3.1Criticalgoodsandtheriskofimportdisruption

Supposeweweremainlyinterestedinrisksrelatedtoimportdisruption.Thiswouldbethecaseifexportsareeitherwelldiversifiedorgomainlytocountriesthatonewouldnotconsidertobemajorsourcesofshocks.Inthatcase,thefollowingapproachmightbeaclosesubstitutefortheperfectmodel.Usingthemostdisaggregateddatapossible,oneshouldidentifyproductsforwhich:

CEPRPOLICYINSIGHTNo.129May2024

6

1.alargeshareofEUconsumptionreliesonimportedinputs;

2.foreignsupplyofthesegoodsishighlyconcentrated;

3.findingalternativesuppliersintheeventofadisruptionisdifficult;and

4.disruptiontosupplywouldhavehigheconomiccosts.Unlikecriterion3,thiscriterionreflectsthesubstitutabilityofproductsineitherconsumptionor

production,asopposedtothesubstitutabilityofsupplierrelationships.

Productsthatmeetallfourcriteriawouldbeprimecandidatesforde-risking.

Thisapproach,whichbuildsonworkundertakenbytheEuropeanCommission(2021),approximatelydescribestheapproachtakeninMejeanandRousseaux(2024).TheirmaininnovationrelativetotheworkoftheCommissionandotherauthorsisstep3,whichtheyimplementbyeliminatingproductsforwhich‘relationshipstickiness’–thetypicaldurationoffirm-supplierrelationships–dropsbelowaspecificthreshold.Forexample,ifthestickinessthresholdissetatthesamplemedian,thenumberofproductsforwhichtheEUshouldconsideritselfimport-dependentdropsfrom378tojust105,andtojust49ifthe75%leastrelationship-stickyproductsareeliminated(Figure1).Focusingonlyonupstreamintermediateproducts–forwhichanexportbanwouldaffectmanysupplychainsandhencehavehigheconomiccosts–wouldreducethelistfurther,tojust21products.For12ofthese,themainsupplierisChina.

Figure1NumberofproductsforwhichtheEUisimportdependent

6000

5000

4000

3000

2000

1000

0

(100)

5381

(7.9)

306

(8.6)

378

(7.3)

228

(0.5)49

Imported

(CEPII-BACI

data)

BasedontheECmethodology

MatchedwithProdcomdata

BasedontheECBasedonEC

methodology+absorption criterion

methodology,

theabsorption

criterionand

thestickiness

criterion

Note:ThefigureshowsthenumbersofproductsforwhichtheEUisimport-dependentaccordingtovariousmethodologies,startingwiththatoftheEuropeanCommission(2021)(secondbluebar)andaddingthecriteriaproposedbyMejeanandRousseaux,basedontheratioofimportsoverdomesticabsorption(redbar)andthedegreeofproductstickiness(greenbar).NumbersinbracketsrefertopercentageofvalueofEUimports.

Source:MejeanandRousseaux(2024).

Tothese,MejeanandRousseaux(2024)suggestedaddingasmallnumberof“criticalgoods”that,ifinsufficientlysupplied,“canresultinhumanlossesandotherseverenon-economicconsequences”.Thesewouldincludebetweentwoand19pharmaceuticalproducts,dependingonwherethesubstitutabilitycut-offisset,aswellasinputstothegreentransition.Interestingly,mostoftheseinputs–includingmostcriticalrawmaterials,whichhavebeenamongthemainjustificationsforthedrivetode-riskimports,particularlyfromChina–currentlyfailoneormoreofMejean’sandRousseaux’sdependencytests.Whilehighlyrelationship-sticky,batteriesandtheircomponents,hydrogentechnologies,rareearthmetalsandsolarpanelsfailthe

CEPRPOLICYINSIGHTNo.129May2024

7

concentrationtest,andmostcomponentsofsolarpanelsfailboththeconcentrationtestandtherelationship-stickinesstest.Yet,MejeanandRousseauxurgedcautionwithrespecttotheseproducts,onthegroundsthatdemandforthemisdevelopingsofastthatthestructureofEUimportsduring2015-2019,onwhichconcentrationindicesandimportneedsarebased,maybeapoorproxyfortradedependenciesinthefuture.

MejeanandRousseaux’sworkrepresentsthemostexhaustiveanalysissofartoidentifydependenciesonthebasisofrankingcriticalimportswithrespecttoconcentrationandrelationshipsubstitutability,anddecidingonthresholdsaboveorbelowwhichconcentrationisdeemedtoohighorsubstitutabilitytoolow.Preciselybecauseitismorethoroughandcomprehensivethanpreviousattemptsinthisliterature,MejeanandRousseaux(2024)illustratestheintrinsiclimitationsofthisapproach.

?Wehavesofarnosystematicwayoftellingwhichimportsaregenuinely

critical.Focusingonupstreamproductsandpharmaceuticalsmaymiss

otherproducts(suchascomputerchips),theaccidentalscarcityofwhich

wouldcauselargeeconomicornon-economiclosses.Meanwhile,some

upstreamproductsandpharmaceuticalsmightnotbecriticaliftheycanbe

substitutedbyotherproducts.TheEuropeanCommission’s(2021)approach

ofdesignatingwhole"ecosystems"(sectors,suchashealth,energy,digital,

electronicsandaerospace)ascriticalseemsevenmoreproblematic,both

becausemanyproductsinthesesectorsarenotinfactcriticalandbecause

productsoutsidethesesectorsthatmaywellbecriticalcouldbemissed(for

example,mostofMejeanandRousseaux’supstreamproducts).

?AsbothMejeanandRousseaux(2024)andBown(2024)emphasise,datalimitationsimplythatimportdependencemeasuresdonotreflectindirectexposure.IftheEUhasanimportexposuretoacountrythatisitselfimportdependentonChinaforthisproduct(oranimportantintermediateinput),thendirectimportdependenceonChinamightsignificantlyunderstatetotalimportdependence.

?Thefinallistscanbeverysensitivetohowthecut-offsareset,whichissomewhatarbitrary.Forexample,whetherrelationshipsubstitutabilitythresholdsaresetatthe25th,50thor75thpercentileaddsorsubtractslargesharesofproductsfromthesample.

?Supplierrelationshipsinnormaltimestendtoberelativelylong(25and19months,respectively,forthe75thand50thpercentilesinMejeanandRousseaux’ssample).Thisimpliesthatunlessreplacementdurationissignificantlyshorterinacrisis,animportinterruptioncouldbeverydamagingevenforproductsthatarerelativelynon-relationship-stickyinnormaltimes.Buttheimpactofaforcedinterruptiononthereplacementperiodcouldgobothways.Firmsseekingtoreplacesuppliersunderduresswouldhaveincentivestodosomuchfasterthaninnormaltimes.However,findingnewsupplierswhenmanyotherfirmsaretryingtodosocouldtakelongerand/orresultinpricejumpsforscarcesupplies,whichcouldbeverydamaging.

3.2Riskfromexportdisruptionsandfromdecoupling

Anotherproblemisthatanapproachfocusedonreducingdependenceoncriticalimportsdoesnotconsiderdisruptionstoexports,whichcouldequallyhaveamacroeconomicimpactiftheywerehighlyconcentratedinanyonedestinationcountry.Forexample,20%ofEUexportsgototheUS,13%totheUKand9%toChina;while41%ofUKexportsgototheEU,21%totheUSand5%toChina.Furthermore,justasimportdependencynumbersignoreindirectexposures,sodoexportshares.Forexample,directUKexportdependencyonChinaisonly5%,buttheUK’sindirectexposureviatheEUalonecouldbelargerifUKproductsarepartofthevaluechainsofgoodsultimatelydestinedfortheChinesemarket.

CEPRPOLICYINSIGHTNo.129May2024

8

Whiledemandshocksviaexportsareastandardriskoftradeintegration,geopoliticalc