To explore China’s role in Ukraine, the Global China project convened six Brookings experts with a range of backgrounds and areas of expertise.
Belt and Road Reboot
Beijing's Bid to De-Risk Its
Global Infrastructure Initiative
Full Report
November 2023
Bradley C. Parks, Ammar A. Malik,
Brooke Escobar, Sheng Zhang,
Rory Fedorochko, Kyra Solomon,
Fei Wang, Lydia Vlasto, Katherine
Walsh, and Seth Goodman
Acknowledgments
WethankSamuelBrazys,HarshDesai,CharlotteGardes-Landolfini,Sebastian
Horn,DavidLandry,MariaRepnikova,MarinaRudyak,BradSetser,L.Teal
Emory,andHaonanZhouforcommentsonanearlierversionofthisreport;John
Custer,SarinaPatterson,SariahHarmer,andAlexWooleyfromAidDataforthe
editing,formatting,andgraphicdesignofthefinalreport;DavideMonteleone
forgenerouslygrantinguspermissiontousehisphotographforthis
publication’scover;SashaTrubetskoyforcartographicsupport;andNicholas
Costa,TaylorLiegel,MiloYen-Goossens,RushikaDevineni,ShiyingCai,andEli
Svobodafordatamanagementsoftwaresupport.Wealsooweadebtof
gratitudetothelargeteamofresearchassistants(RAs)whohelpedconstructthe
datasetthatmadethisstudypossible—includingAbenaKonadu,AbigailTaylor,
AldrinYashko,AlecVeit,AlejandroSchnappMezerhane,AlexPark,AllySwindell,
AlysonReynolds,AndreaPhelps,AndyShufer,AnnaGarrett,AnnabelRichter,
AnnePaxton,AnnikaSuri,AriannaLagamma,AudreySun,BaileyGarber,Ben
Greenberg,BeyoncéLightfoot,BreeJiang,CaitlinConnor,CaitlinNoe,Camilla
Yeleussizova,CarinaRosenberg,CarolinHelmholz,CarolineElszy,Changlin
(Bob)Zhang,CheckKingTay,ChristianMoore,ClaireRudinsky,ClaireSchlick,
ClaireWyszynski,ClaraMcAllister,CollinAbsher,ConnorStanton,DavidZhu,
DorothyGao,DoryGilmer,EliseTsao,ElizabethOtt,EmmaBury,FalanKifle,
GelilaYonas,George(Teddy)Chunias,GujieShen,GwynethSmith,HarinOk,
HarrisonAbramsohn,HelenTieder,HenryTieder,HoliRaparaoelina,Jack
Mackey,JacksonHumphreys,JasonBao,JenniferLuo,JiajuDu,JiayiXu,Jiexi
Lin,JoanneLee,JoeyLindsay,JosephBohley,JuhwanCho,KateHentz,Katelyn
Oxer,KathrynSpurrier,KathrynZiccarelli,KaylaAviles,KellyShinners,Kerim
Gales,Kien-TamElston,LainaLomont,LauraLam,LexiOlsen,LiangGeng,
LillianButler,MaddieSharpe,MariannaBonilla,MaryTrotto,MaxBerckmueller,
MeverynChua,MiaMaccario,MicahKatahara,MichaelOlyniec,MujiaHe,
NancyLang,NateSpiers,NicoRullo,NoahTran,OliviaSchlamp,Peter
Awabdeh,PoojaMuthuraj,PoojaTanjore,PrestinTran,RedeitHailu,Riely
Capece,RileyCompanion,RinTsujikawa,RodrigoArias,RyanMcCormich,
SabrinaLevri,SailorMiao,SamLeBlanc,SamanthaRofman,SarahKennedy,
SarangaBansal,SeanNguyen,ShannonMcGuire,SophieAdams-Smith,Sophie
Alden,Sophie(Grace)McGehee,SophieStrauss,SophieTorres,Susannah
Matthews,TaylorPerry,TomMooney,VictorGedeck,WillWu,YeXiao,YouWu,
YufeiTang,Yunlang(Bella)Li,Yuxi(Felix)Tian,YuxinQin,Zheyi(Serena)Liu,
Zhiying(Cindy)Zhou,ZipuLiu,Ziqi(Samuel)Yan,andZupengZeng.
Overthelasttenyears,alargegroupofcoauthorsandcollaborators—including
ChristianBaehr,ArielBenYishay,RichardBluhm,MengfanCheng,Samantha
Custer,CarrieDolan,AxelDreher,Thai-BinhElston,SebastianHorn,Andreas
Fuchs,AlyshaGardner,AnnaGelpern,SiddharthaGhose,RolandHodler,Brook
Lautenslager,JoyceJiahuiLin,JacobHall,AlexandraJoosse,ScottMorris,
EdwinMuchapondwa,DanielNielson,JuliusN.Odhiambo,PaulRaschky,
CarmenReinhart,DanielRunfola,AustinStrange,MichaelTierney,Christoph
Trebesch,RachelTrichler,LukasWellner,andLincolnZaleski—hashelpedrefine
andexpandthedatasetuponwhichthisstudyisbasedbypilotingcoding
procedures,recommendingnewsourcesandmethods,andscrutinizing
preliminaryprojectrecords.
WegratefullyacknowledgefinancialsupportfromtheWilliamandFloraHewlett
Foundation,theFordFoundation,andtheSmithRichardsonFoundation.The
findings,interpretations,andconclusionsexpressedinthisstudyareentirely
thoseoftheauthors.TheydonotnecessarilyrepresenttheviewsoftheWilliam
andFloraHewlettFoundation,theFordFoundation,ortheSmithRichardson
Foundation.AidData’sresearchisguidedbytheprinciplesofindependence,
integrity,transparency,andrigor.AdiversegroupoffunderssupportsAidData’s
work,buttheydonotdetermineitsresearchfindingsorrecommendations.
Citation
Parks,B.C.,Malik,A.A.,Escobar,B.,Zhang,S.,Fedorochko,R.,Solomon,K.,
Wang,F.,Vlasto,L.,Walsh,K.&Goodman,S.2023.BeltandRoadReboot:
Beijing’sBidtoDe-RiskItsGlobalInfrastructureInitiative.Williamsburg,VA:
AidDataatWilliam&Mary
ExecutiveSummary
TheBeltandRoadRebootreportprovidesmyth-bustingevidenceaboutthechangingnature,scale,andscopeofChina’soverseasdevelopmentprogram.ItalsorevealsnewinsightsaboutBeijing’songoingbidtode-riskitsflagshipglobalinfrastructureinitiative—andoutflankitscompetitors.ThereportdrawsuponAidData’suniquelycomprehensiveandgranulardatasetofinternationaldevelopmentfinancefromChina,whichcaptures20,985projectsacross165low-andmiddle-incomecountriesfinancedwithgrantsandloansworth$1.34trillionovera22-yearperiod.1
IsChinastillthesinglelargestofficialsourceofaidandcredittothedevelopingworld?
Fourkeytakeaways
1.Contrarytotheconventionalwisdom,Beijing’sannualinternationaldevelopmentfinancecommitmentshavenotplummetedtonearlyzero.2Itremainstheworld’ssinglelargestofficialsourceofinternationaldevelopmentfinance.China’saidandcredit(ODAandOOF)commitmentstolow-andmiddle-incomecountriesarenowhoveringaround$80billionayear.3
2.WashingtonisbeginningtoclosethespendinggapwithBeijing.DueinlargeparttotheU.S.InternationalDevelopmentFinanceCorporation(DFC)’sfinancingofprivatesectorprojects,whichhasledtoafifteen-foldexpansioninU.S.OOF,Washingtonnowprovidesapproximately$60billionofdevelopmentfinanceeachyeartolow-andmiddle-incomecountries.
3BaseduponOECD-DACdefinitionsandmeasurementcriteria,AidDatacategorizeseachproject/activityinitsdatasetasOfficialDevelopmentAssistance(ODA)orOtherOfficialFlows(OOF).ODAmostlyconsistsofgrantsandhighlyconcessionalloansfordevelopmentprojectsandactivitiesthatarefinancedbyofficialsectorinstitutions.OOFmostlyconsistsofnon-concessionalloansthatareissuedbyofficialsectorinstitutions.Morethan90%ofChina’sannualinternationaldevelopmentfinancecommitmentsconsistofOOF.2Forexample,thelatestversionoftheChina’sOverseasDevelopmentFinance(CODF)databaseproducedbyBostonUniversity’sGlobalDevelopmentPolicyCentersuggeststhatoverseasdevelopmentfinancecommitmentsfromChinahaveplummetedby96%since2016,reachinganall-timelowof$3.7billionin2021.1Thelatest(3.0)versionofAidData’sGlobalChineseDevelopmentFinance(GCDF)datasetcapturesprojectsover22commitmentyears(2000-2021)andprovidesdetailsonthetimingofprojectimplementationovera24-yearperiod(2000-2023).Itcanbeaccessedviaaiddata.org/china.3.Intheshort-run,theG7isalsosteppingupitseffortstocompetewithBeijingthroughthePartnershipforGlobalInfrastructureandInvestment,theIndia-MiddleEast-EuropeEconomicCorridor,andotherinitiatives.AfterfailingtomatchChina’sannualODAandOOFcommitmentsduringtheearlyyearsoftheBeltandRoadInitiative(BRI),theG7outspentChinaby$84billionin2021.
4.However,inthelong-run,itisnotclearthattheU.S.anditsallieshavethefinancialfirepowertocompetedollar-for-dollarwithBeijing.TheG7hasahistoryofover-promisingandunder-deliveringnetincreasesininternationaldevelopmentspending.Beijing,bycontrast,hasarealsourceoffinancialstrengththatallowsittoavoidmakingpromisesthatitcannotkeep:foreignexchangereservesthatarevastlylargerthantheofficial,foreigncurrencyreserveholdingsofitscentralbank.4
HowhastheriskprofileofChina’sinternationaldevelopmentfinanceportfoliochanged?
Threekeytakeaways
1.Repaymentrisk:Beijingisnavigatinganunfamiliaranduncomfortablerole—astheworld’slargestofficialdebtcollector.55%ofitsloanstolow-andmiddle-incomecountrieshavealreadyenteredtheirprincipalrepaymentperiodsandthisfigurewillincreaseto75%by2030.Totaloutstandingdebt—includingprincipalbutexcludinginterest—fromborrowersinthedevelopingworldtoChinaisatleast$1.1trillionandpotentiallyevenashighas$1.5trillion(innominalUSD).5Beijingisfindingitsfootingasaninternationaldebtcollectoratatimewhenmanyofitsbiggestborrowersareilliquidorinsolvent.AidDataestimatesthat80%ofChina’s
5TotaloutstandingdebtfromborrowersindevelopedanddevelopingcountriestoChinaexceeds$2.6trillion(innominalUSD).4Asof2023,theofficial,foreigncurrencyreserveholdingsofChina’scentralbank(thePBOC)amountedto$3.1trillion.However,thisfigureexcludesforeigncurrencyreservesthatthePBOChasmovedoffofitsbalancesheetby,amongotherthings,entrustingthemtothecountry’sstate-ownedpolicybanks,state-ownedcommercialbanks,andstate-ownedfunds.BradSetseroftheCouncilonForeignRelationsarguesthatthese“hiddenreserves”maybeworthanadditional$3trillion.
1
overseaslendingportfoliointhedevelopingworldiscurrentlysupportingcountriesinfinancialdistress.OverduerepaymentstoChinaarealsosoaring—inabsolutetermsandasaproportionoftotaloverdueloanrepaymentstoofficial(i.e.,bilateralandmultilateral)creditors.
2.Projectperformancerisk:ThecumulativenumberofChinesegrant-andloan-financedinfrastructureprojectsinthedevelopingworldwithsignificantenvironmental,social,orgovernance(ESG)riskexposureskyrocketedfrom17projectsworth$420millionin2000to1,693projectsworth$470billionin2021.ThecumulativepercentageofChina’sgrant-andloan-financedinfrastructureprojectportfoliointhedevelopingworldwithsignificantESGriskexposureincreasedfrom12%to53%overthesame22-yearperiod.Infrastructureprojectsuspensionsandcancellationshavealsomounted—fromnearlyzeroattheturnofthecenturyto94projectsworth$56billionin49countries.However,BeijingissteppingupESGriskmitigationeffortstoshielditsoverseasinfrastructureportfoliofromthetypesofproblemsthathavepreviouslyplaguedtheBRI.
3.Reputationalrisk:Beijing’spublicapprovalratinginthedevelopingworldplungedfrom56%in2019to40%in2021.Washington,ontheotherhand,hasseenitspublicapprovalratingriseandopenedupa14percentagepointadvantageoverBeijing.Acrossthedevelopingworld,Chinahasalsostruggledtomaintainarazor-thinleadovertheU.S.inmediacoveragefavorability.Yetithasprovenverycapableofwinningandretainingtheforeignpolicysupportofgoverningelites.AcrossallU.N.GeneralAssemblyvotescastbetween2000and2021,thegovernmentsoflow-andmiddle-incomecountriesalignedtheirforeignpolicypositionswithChina75%ofthetime—ascomparedto23%withtheU.S.ThosewhovotewithChinaarerichlyrewarded:onaverage,ifaforeigngovernmentchoosestoincreasethealignmentofitsU.N.GeneralAssemblyvotingwithChinaby10%,itcanexpecttoseea276%increaseinaidandcreditfromBeijing.6
6Thisfindingisderivedfromas
...
Belt and Road Reboot
Beijing's Bid to De-Risk Its
Global Infrastructure Initiative
Full Report
November 2023
Bradley C. Parks, Ammar A. Malik,
Brooke Escobar, Sheng Zhang,
Rory Fedorochko, Kyra Solomon,
Fei Wang, Lydia Vlasto, Katherine
Walsh, and Seth Goodman
Acknowledgments
WethankSamuelBrazys,HarshDesai,CharlotteGardes-Landolfini,Sebastian
Horn,DavidLandry,MariaRepnikova,MarinaRudyak,BradSetser,L.Teal
Emory,andHaonanZhouforcommentsonanearlierversionofthisreport;John
Custer,SarinaPatterson,SariahHarmer,andAlexWooleyfromAidDataforthe
editing,formatting,andgraphicdesignofthefinalreport;DavideMonteleone
forgenerouslygrantinguspermissiontousehisphotographforthis
publication’scover;SashaTrubetskoyforcartographicsupport;andNicholas
Costa,TaylorLiegel,MiloYen-Goossens,RushikaDevineni,ShiyingCai,andEli
Svobodafordatamanagementsoftwaresupport.Wealsooweadebtof
gratitudetothelargeteamofresearchassistants(RAs)whohelpedconstructthe
datasetthatmadethisstudypossible—includingAbenaKonadu,AbigailTaylor,
AldrinYashko,AlecVeit,AlejandroSchnappMezerhane,AlexPark,AllySwindell,
AlysonReynolds,AndreaPhelps,AndyShufer,AnnaGarrett,AnnabelRichter,
AnnePaxton,AnnikaSuri,AriannaLagamma,AudreySun,BaileyGarber,Ben
Greenberg,BeyoncéLightfoot,BreeJiang,CaitlinConnor,CaitlinNoe,Camilla
Yeleussizova,CarinaRosenberg,CarolinHelmholz,CarolineElszy,Changlin
(Bob)Zhang,CheckKingTay,ChristianMoore,ClaireRudinsky,ClaireSchlick,
ClaireWyszynski,ClaraMcAllister,CollinAbsher,ConnorStanton,DavidZhu,
DorothyGao,DoryGilmer,EliseTsao,ElizabethOtt,EmmaBury,FalanKifle,
GelilaYonas,George(Teddy)Chunias,GujieShen,GwynethSmith,HarinOk,
HarrisonAbramsohn,HelenTieder,HenryTieder,HoliRaparaoelina,Jack
Mackey,JacksonHumphreys,JasonBao,JenniferLuo,JiajuDu,JiayiXu,Jiexi
Lin,JoanneLee,JoeyLindsay,JosephBohley,JuhwanCho,KateHentz,Katelyn
Oxer,KathrynSpurrier,KathrynZiccarelli,KaylaAviles,KellyShinners,Kerim
Gales,Kien-TamElston,LainaLomont,LauraLam,LexiOlsen,LiangGeng,
LillianButler,MaddieSharpe,MariannaBonilla,MaryTrotto,MaxBerckmueller,
MeverynChua,MiaMaccario,MicahKatahara,MichaelOlyniec,MujiaHe,
NancyLang,NateSpiers,NicoRullo,NoahTran,OliviaSchlamp,Peter
Awabdeh,PoojaMuthuraj,PoojaTanjore,PrestinTran,RedeitHailu,Riely
Capece,RileyCompanion,RinTsujikawa,RodrigoArias,RyanMcCormich,
SabrinaLevri,SailorMiao,SamLeBlanc,SamanthaRofman,SarahKennedy,
SarangaBansal,SeanNguyen,ShannonMcGuire,SophieAdams-Smith,Sophie
Alden,Sophie(Grace)McGehee,SophieStrauss,SophieTorres,Susannah
Matthews,TaylorPerry,TomMooney,VictorGedeck,WillWu,YeXiao,YouWu,
YufeiTang,Yunlang(Bella)Li,Yuxi(Felix)Tian,YuxinQin,Zheyi(Serena)Liu,
Zhiying(Cindy)Zhou,ZipuLiu,Ziqi(Samuel)Yan,andZupengZeng.
Overthelasttenyears,alargegroupofcoauthorsandcollaborators—including
ChristianBaehr,ArielBenYishay,RichardBluhm,MengfanCheng,Samantha
Custer,CarrieDolan,AxelDreher,Thai-BinhElston,SebastianHorn,Andreas
Fuchs,AlyshaGardner,AnnaGelpern,SiddharthaGhose,RolandHodler,Brook
Lautenslager,JoyceJiahuiLin,JacobHall,AlexandraJoosse,ScottMorris,
EdwinMuchapondwa,DanielNielson,JuliusN.Odhiambo,PaulRaschky,
CarmenReinhart,DanielRunfola,AustinStrange,MichaelTierney,Christoph
Trebesch,RachelTrichler,LukasWellner,andLincolnZaleski—hashelpedrefine
andexpandthedatasetuponwhichthisstudyisbasedbypilotingcoding
procedures,recommendingnewsourcesandmethods,andscrutinizing
preliminaryprojectrecords.
WegratefullyacknowledgefinancialsupportfromtheWilliamandFloraHewlett
Foundation,theFordFoundation,andtheSmithRichardsonFoundation.The
findings,interpretations,andconclusionsexpressedinthisstudyareentirely
thoseoftheauthors.TheydonotnecessarilyrepresenttheviewsoftheWilliam
andFloraHewlettFoundation,theFordFoundation,ortheSmithRichardson
Foundation.AidData’sresearchisguidedbytheprinciplesofindependence,
integrity,transparency,andrigor.AdiversegroupoffunderssupportsAidData’s
work,buttheydonotdetermineitsresearchfindingsorrecommendations.
Citation
Parks,B.C.,Malik,A.A.,Escobar,B.,Zhang,S.,Fedorochko,R.,Solomon,K.,
Wang,F.,Vlasto,L.,Walsh,K.&Goodman,S.2023.BeltandRoadReboot:
Beijing’sBidtoDe-RiskItsGlobalInfrastructureInitiative.Williamsburg,VA:
AidDataatWilliam&Mary
ExecutiveSummary
TheBeltandRoadRebootreportprovidesmyth-bustingevidenceaboutthechangingnature,scale,andscopeofChina’soverseasdevelopmentprogram.ItalsorevealsnewinsightsaboutBeijing’songoingbidtode-riskitsflagshipglobalinfrastructureinitiative—andoutflankitscompetitors.ThereportdrawsuponAidData’suniquelycomprehensiveandgranulardatasetofinternationaldevelopmentfinancefromChina,whichcaptures20,985projectsacross165low-andmiddle-incomecountriesfinancedwithgrantsandloansworth$1.34trillionovera22-yearperiod.1
IsChinastillthesinglelargestofficialsourceofaidandcredittothedevelopingworld?
Fourkeytakeaways
1.Contrarytotheconventionalwisdom,Beijing’sannualinternationaldevelopmentfinancecommitmentshavenotplummetedtonearlyzero.2Itremainstheworld’ssinglelargestofficialsourceofinternationaldevelopmentfinance.China’saidandcredit(ODAandOOF)commitmentstolow-andmiddle-incomecountriesarenowhoveringaround$80billionayear.3
2.WashingtonisbeginningtoclosethespendinggapwithBeijing.DueinlargeparttotheU.S.InternationalDevelopmentFinanceCorporation(DFC)’sfinancingofprivatesectorprojects,whichhasledtoafifteen-foldexpansioninU.S.OOF,Washingtonnowprovidesapproximately$60billionofdevelopmentfinanceeachyeartolow-andmiddle-incomecountries.
3BaseduponOECD-DACdefinitionsandmeasurementcriteria,AidDatacategorizeseachproject/activityinitsdatasetasOfficialDevelopmentAssistance(ODA)orOtherOfficialFlows(OOF).ODAmostlyconsistsofgrantsandhighlyconcessionalloansfordevelopmentprojectsandactivitiesthatarefinancedbyofficialsectorinstitutions.OOFmostlyconsistsofnon-concessionalloansthatareissuedbyofficialsectorinstitutions.Morethan90%ofChina’sannualinternationaldevelopmentfinancecommitmentsconsistofOOF.2Forexample,thelatestversionoftheChina’sOverseasDevelopmentFinance(CODF)databaseproducedbyBostonUniversity’sGlobalDevelopmentPolicyCentersuggeststhatoverseasdevelopmentfinancecommitmentsfromChinahaveplummetedby96%since2016,reachinganall-timelowof$3.7billionin2021.1Thelatest(3.0)versionofAidData’sGlobalChineseDevelopmentFinance(GCDF)datasetcapturesprojectsover22commitmentyears(2000-2021)andprovidesdetailsonthetimingofprojectimplementationovera24-yearperiod(2000-2023).Itcanbeaccessedviaaiddata.org/china.3.Intheshort-run,theG7isalsosteppingupitseffortstocompetewithBeijingthroughthePartnershipforGlobalInfrastructureandInvestment,theIndia-MiddleEast-EuropeEconomicCorridor,andotherinitiatives.AfterfailingtomatchChina’sannualODAandOOFcommitmentsduringtheearlyyearsoftheBeltandRoadInitiative(BRI),theG7outspentChinaby$84billionin2021.
4.However,inthelong-run,itisnotclearthattheU.S.anditsallieshavethefinancialfirepowertocompetedollar-for-dollarwithBeijing.TheG7hasahistoryofover-promisingandunder-deliveringnetincreasesininternationaldevelopmentspending.Beijing,bycontrast,hasarealsourceoffinancialstrengththatallowsittoavoidmakingpromisesthatitcannotkeep:foreignexchangereservesthatarevastlylargerthantheofficial,foreigncurrencyreserveholdingsofitscentralbank.4
HowhastheriskprofileofChina’sinternationaldevelopmentfinanceportfoliochanged?
Threekeytakeaways
1.Repaymentrisk:Beijingisnavigatinganunfamiliaranduncomfortablerole—astheworld’slargestofficialdebtcollector.55%ofitsloanstolow-andmiddle-incomecountrieshavealreadyenteredtheirprincipalrepaymentperiodsandthisfigurewillincreaseto75%by2030.Totaloutstandingdebt—includingprincipalbutexcludinginterest—fromborrowersinthedevelopingworldtoChinaisatleast$1.1trillionandpotentiallyevenashighas$1.5trillion(innominalUSD).5Beijingisfindingitsfootingasaninternationaldebtcollectoratatimewhenmanyofitsbiggestborrowersareilliquidorinsolvent.AidDataestimatesthat80%ofChina’s
5TotaloutstandingdebtfromborrowersindevelopedanddevelopingcountriestoChinaexceeds$2.6trillion(innominalUSD).4Asof2023,theofficial,foreigncurrencyreserveholdingsofChina’scentralbank(thePBOC)amountedto$3.1trillion.However,thisfigureexcludesforeigncurrencyreservesthatthePBOChasmovedoffofitsbalancesheetby,amongotherthings,entrustingthemtothecountry’sstate-ownedpolicybanks,state-ownedcommercialbanks,andstate-ownedfunds.BradSetseroftheCouncilonForeignRelationsarguesthatthese“hiddenreserves”maybeworthanadditional$3trillion.
1
overseaslendingportfoliointhedevelopingworldiscurrentlysupportingcountriesinfinancialdistress.OverduerepaymentstoChinaarealsosoaring—inabsolutetermsandasaproportionoftotaloverdueloanrepaymentstoofficial(i.e.,bilateralandmultilateral)creditors.
2.Projectperformancerisk:ThecumulativenumberofChinesegrant-andloan-financedinfrastructureprojectsinthedevelopingworldwithsignificantenvironmental,social,orgovernance(ESG)riskexposureskyrocketedfrom17projectsworth$420millionin2000to1,693projectsworth$470billionin2021.ThecumulativepercentageofChina’sgrant-andloan-financedinfrastructureprojectportfoliointhedevelopingworldwithsignificantESGriskexposureincreasedfrom12%to53%overthesame22-yearperiod.Infrastructureprojectsuspensionsandcancellationshavealsomounted—fromnearlyzeroattheturnofthecenturyto94projectsworth$56billionin49countries.However,BeijingissteppingupESGriskmitigationeffortstoshielditsoverseasinfrastructureportfoliofromthetypesofproblemsthathavepreviouslyplaguedtheBRI.
3.Reputationalrisk:Beijing’spublicapprovalratinginthedevelopingworldplungedfrom56%in2019to40%in2021.Washington,ontheotherhand,hasseenitspublicapprovalratingriseandopenedupa14percentagepointadvantageoverBeijing.Acrossthedevelopingworld,Chinahasalsostruggledtomaintainarazor-thinleadovertheU.S.inmediacoveragefavorability.Yetithasprovenverycapableofwinningandretainingtheforeignpolicysupportofgoverningelites.AcrossallU.N.GeneralAssemblyvotescastbetween2000and2021,thegovernmentsoflow-andmiddle-incomecountriesalignedtheirforeignpolicypositionswithChina75%ofthetime—ascomparedto23%withtheU.S.ThosewhovotewithChinaarerichlyrewarded:onaverage,ifaforeigngovernmentchoosestoincreasethealignmentofitsU.N.GeneralAssemblyvotingwithChinaby10%,itcanexpecttoseea276%increaseinaidandcreditfromBeijing.6
6ThisfindingisderivedfromastatisticalmodelinDreher,A.,Fuchs,A.,Parks,B.C.,Strange,A.,&Tierney,M.J.2022.BankingonBeijing:TheAimsandImpactsofChina’sOverseasDevelopmentProgram.Cambridge,UK:CambridgeUniversityPress.DoestheG7understandthedifferencebetweenBRI1.0andBRI2.0—orhowBeijing’srebootofits“projectofthecentury”hasalteredthecompetitivelandscape?
Threekeytakeaways
1.Beijinghaslaunchedafar-reachingefforttode-risktheBRIbyrefocusingitstime,money,andattentionondistressedborrowers,troubledprojects,andsourcesofpublicbacklashintheGlobalSouth.Itislearningfromitsmistakesandbecominganincreasinglyadeptinternationalcrisismanager.
2.NeithertheU.S.noritsG7alliesseemtohaveagoodunderstandingofhowChinaisrecalibratingitslendingandgrant-givingpracticesinresponsetochangingconditionsontheground.Consequently,thosewhomakeandshapepolicyinWashington,London,Paris,Berlin,Tokyo,Rome,andOttawaincreasinglyruntheriskofcompetingwithaversionoftheBRIthatnolongerexists—BRI1.0ratherthanBRI2.0.
3.TheG7shouldnotunderestimatetheambitionofChina’songoingefforttofuture-proofitsflagship,globalinfrastructureinitiative.Beijingisfocusedongivingleadersinthedevelopingworldexactlywhattheywant:rapiddeliveryoflarge-scaleinfrastructureprojectswithoutunreasonablyhighlevelsofESGrisk.IftheG7cannotcompeteonthisbasis,itsPartnershipforGlobalInfrastructureandInvestmentmayfaceacrisisofrelevance.
WhatmeasureshasBeijingtakentoreduceitsexposuretodistresseddebtinthedevelopingworld?
Sevenkeytakeaways
1.InrecognitionofthefactthatBRI1.0didnothavesufficientlyrobustriskmanagementguardrailsinplace,Beijingisfundamentallyalteringthecompositionofitsoverseaslendingportfolio.Itisrampingdowndollar-denominatedinfrastructureprojectlending,whilerampingupRMB-denominatedemergencyrescuelendingtofinanciallydistressed
2
borrowers.7Beijing’sstrategicobjectiveistoensurethatitslargestborrowershaveenoughcashonhandtoservicetheiroutstandinginfrastructureprojectdebts.
2.Beijing’spolicybanks(ChinaEximbankandChinaDevelopmentBank)haveparticularlyhighlevelsofexposuretonon-performingloansinlow-andmiddle-incomecountries.Insteadofreformingtheseinstitutionsfromwithin,Beijingisratchetingdownitsuseofthepolicybanks,whileratchetingupitsuseofstate-ownedcommercialbanks,suchasICBCandBankofChina.Inpreviousyears,approximatelythree-quartersofChina’slendingtolow-andmiddle-incomecountrieswaschanneledthroughthepolicybanks.However,thisfigurehasnowplummetedtolessthanone-quarter(22%).TheannuallendingcommitmentsofChina’sstate-ownedcommercialbankstolow-andmiddle-incomecountriesarenowonparwiththoseofitspolicybanks.8
3.Ratherthanrelyingonitsownbankstovetborrowinginstitutionsandproposedtransactions,Beijingisincreasinglyoutsourcingriskmanagementtolendinginstitutions—suchastheInternationalFinanceCorporation,theEuropeanBankforReconstructionandDevelopment,StandardCharteredBank,andBNPParibas—withstrongerduediligencestandardsandsafeguardpolicies.ItisdialingdownitsuseofbilaterallendinginstrumentsanddialinguptheprovisionofcreditthroughcollaborativelendingarrangementswithWesterncommercialbanksandmultilateralinstitutions.50%ofChina’snon-emergencylendingportfolioinlow-andmiddle-incomecountriesisnowprovidedviasyndicatedloanarrangements—andmorethan80%ofthesearrangementsinvolveWesterncommercialbanksandmultilateralinstitutions.
8Onaverage,duringthepre-BRIperiod(2000-2013),Beijingchanneled15%ofitsannuallendingcommitmentstolow-andmiddle-incomecountriesthroughitsstate-ownedcommercialbanks.Thisfigureincreasedto18%duringtheearlyBRI(2014-2017)periodand22%duringthelateBRIperiod(2018-2021).7InthefirstfullyearofBRIimplementation(2014),65%ofBeijing’slendingtolow-andmiddle-incomecountriessupportedinfrastructureprojects.By2021,thisfigureplummetedto31%.Emergencyrescuelendingrepresentedonly13%ofBeijing’sloanportfolioinlow-andmiddle-incomecountriesin2014.However,thisfiguresoaredto58%by2021.4.Beijingisputtinginplaceincreasinglystringentsafeguardstoshielditselffromtheriskofnotbeingrepaid.Attheturnofthecentury,only19%ofChina’soverseaslendingtolow-andmiddle-incomecountrieswascollateralized.Thisfigurenowstandsat72%.9TheabilitytoaccesscashcollateralwithoutborrowerconsenthasbecomeaparticularlyimportantsafeguardinChina’sbilaterallendingportfolio.Whenilliquidorinsolventborrowersfallbehindontheirrepayments,thepolicybanksare“payingthemselves”overdueprincipalandinterestbyunilaterallysweepingforeigncurrencyoutoftheescrowaccountsoftheirborrowers.Thesecashseizuresaremostlybeingexecutedinsecretandoutsidetheimmediatereachofdomesticoversightinstitutions—suchastheauditorgeneralandthepublicaccountscommitteewithinparliament—inlow-andmiddle-incomecountries.Aftermakingwithdrawalsthatsubstantiallydepletethebalanceofaborrower’sescrowaccount,anincreasinglycommonpracticeistorequirethattheborrowerreplenishtheaccountasaconditionforanyshort-termcashflowrelief.Escrowaccountreplenishmenthasbecomeamajorstickingpointindebtreschedulingnegotiationswiththepolicybanks,yetitisshroudedinsecrecybecauseofstrictconfidentialityrequirements.10
5.Asthenumberofborrowersfacingliquidityandsolvencycriseshassoared,Chinesestate-ownedcreditorshaveintroducedstrongerpenaltiesforlaterepayments.TheaveragepenaltyinterestratedoubledbetweentheearlyBRIperiod(2014-2017)andthelateBRIperiod(2018-2021).Themaximumpenaltyinterestratealsoincreasedfrom3%to8.7%betweenthesetwotimeperiods.Thesefindingscontradictthoseofapreviousstudy,whichclaimedthatthereisnoevidenceofpenaltyinterestratesinChina’soverseaslendingtodevelopingcountries.
10WhenasovereignborrowersignsanescrowaccountagreementordebtreschedulingagreementwithaChineselender,itisnotunusualforthepartiestoagreeuponanexpansivesetofconfidentialityobligationsthatgobeyondthoseinitsoriginalloanagreement.TheimplementationofAidData’sTrackingUnderreportedFinancialFlows(TUFF)methodologyhasfacilitatedtheretrievalandpublicationofasignificantnumberofunredactedescrowaccountanddebtreschedulingagreements.The3.0versionofAidData’sGCDFdatasetmakestheseagreementsavailableviastableURLs.9Beijingistakingspecialprecautionswithhigh-riskborrowers.Attheturnofthecentury,0%ofitscollateralizedlendingcommitmentstolow-andmiddle-incomecountriesweredirectedtodevelopingcountriesinfinancialdistress.By2021,thisfigureincreasedto74%.
3
6.TherepaymentriskmitigationmeasuresthatBeijingisputtinginplacepresentnewchallengesforborrowersinthedevelopingworld.ThosewhoseektorefinancetheirmaturingdebtstoChinabyacceptingemergencyrescueloanswithhighinterestratesandshortrepaymentperiodsmustbemindfulofthedangerofswappinglessexpensivedebtformoreexpensivedebt.ThosewhoseektorescheduletheirdebtstoChinamustbepreparedtoring-fenceforeigncurrencyforsomecreditorsbutnotothers.ThosewhocontractnewdebtfromBeijingmustbeawareofthedangerofcompoundingarrearsduetopenaltyinterest.
7.Beijing’sgo-it-aloneeffortstomitigaterepaymentriskmayunderminetheinternationalcommunity’seffortstoprovidecoordinateddebtrelieftosovereignborrowersinfinancialdistress.InNovember2020,ChinaagreedtoparticipateintheG-20CommonFrameworkforDebtTreatmentsandabidebytheso-called“comparabletreatment”principle(i.e.,reasonableburden-sharinginthewaythatfinanciallossesaredistributedacrosscreditors).However,Beijing’slatestactionssuggestthatitismusclingitswaytothefrontoftherepaymentlinebydemandingthatborrowersproviderecoursetocashcollateralthatotherslack.ParisClub,multilateral,andcommercialcreditorsfear—withsomejustification—thattheyarebecomingjuniorcreditorswhoseloanswillberepaidonalower-prioritybasis.IfBeijinginsistsuponbeingtreatedasaseniorcreditorwhosedebtsshouldbegivenfirstpriority,thencoordinateddebtreschedulingswithnon-Chinesecreditorswilllikelybecomemoredifficulttonegotiate.Thebiggestlosersinthisscenariowillbeordinarypeopleinthedevelopingworldwhoaredeniedbasicpublicservicesbecauseofacollectiveactionfailureamongforeigncreditors.
WhatmeasureshasBeijingtakentoreduceitsexposuretoESGrisk?HowareitsinfrastructureprojectswithstrongESGsafeguardsfaringduringimplementation?
Sevenkeytakeaways
1.Beijinghasearnedareputationforimplementingbrick-and-mortarprojectswithlightningspeed.IrrespectiveofESGsafeguardstringency,theaverageinfrastructureprojectfinancedwithChineseaidorcredittakesapproximatelythreeyearstocomplete.
2.Beijing’srivalsandcriticsclaimthatithasnottakenmeaningfulstepstosubjectitsoverseasinfrastructureprojectportfoliotomorestringentESGsafeguards.Thisclaimisfalse.By2021,57%ofChina’sgrant-andloan-financedinfrastructureprojectportfolioinlow-andmiddle-incomecountrieshadstrongdejureenvironmental,social,andgovernancesafeguardsinplace.Thisrepresentsamajordeparturefrompastpractice:attheturnofthecentury,China’sentiregrant-andloan-financedinfrastructureprojectportfolioinlow-andmiddle-incomecountrieshadweakdejureenvironmental,social,andgovernancesafeguardsinplace.
3.ThepaceofESGsafeguardreformacceleratedduringtheBRI2.0era—from2018to2021.11Overthesamefour-yearperiod,theannualESGriskprevalencerateinChina’sgrant-andloan-financedinfrastructureprojectportfoliosharplydeclinedfrom63%to33%.
4.Beijinghasde-riskedthecountry’soverseasinfrastructureprojectportfoliobyreiningintheactivitiesofdevelopmentfinanceinstitutionsthatlackstrongESGriskmanagementguardrails,increasingtheprovisionofinfrastructurefinancingviainstitutionsthathavestrongESGsafeguardsinplace,unwindingaidandcreditrelationshipswithcountriesthatpresenthighlevelsofESGrisk,andredirectingnewinfrastructurefinancingtocountriesthatpresentlowlevelsofESGrisk.
5.Chinesegrant-andloan-financedinfrastructureprojectswithstrongdejureESGsafeguardshavesubstantiallylowerlevelsofESGriskexposureinadefactosensethanthosewithoutsuchsafeguards.Theyarealsolessvulnerabletosuspensionandcancellation.
6.AparticularlyimportantfindingisthatChinesegrant-andloan-financedinfrastructureprojectssubjectedtostrongdejureESGsafeguardsdonotface
11In2018,26%ofChina’sgrant-andloan-financedinfrastructureprojectportfolioinlow-andmiddle-incomecountrieshadstrongdejureenvironmental,social,andgovernancesafeguardsinplace.By2021,thisfigurehadincreasedby31percentagepoints(to57%).
4
substantiallylongerimplementationdelaysthanthosesubjectedtoweakdejureESGsafeguards.12
SquaringthecirclebetweenspeedandsafetyisatthecenterofBeijing’sBRI2.0strategy.
7.Beijingenjoysastrongerpositionintheglobalinfrastructurefinancingmarketthanitsbilateralandmultilateralcompetitorsrealize.ThefactthatChinahasputinplaceincreasinglystringentESGsafeguards—withoutdamagingitsreputationforspeed—couldundermineG7effortstooutcompeteitonqualityorsafetygrounds.Developingcountriesprefertoworkwithlendersanddonorsthatbankrollbig-ticket,high-impactinfrastructureprojectswithreasonablyrobustESGsafeguardsbutwithoutexcessiveimplementationdelays.Beijingistakingmeasurestomeetthischallenge.WhethertheG7—andthemultilateraldevelopmentbanks—willdothesameisanopenquestion.
WhatmeasureshasBeijingtakentoreduceitsexposuretoreputationalrisk?
Sixkeytakeaways
1.Inatallyoftheannualnumberofsoftpower“gains”and“losses”thatChinahasexperiencedvis-à-vistheU.S.inlow-andmiddle-incomecountriessincethefirstfullyearofBRIimplementation(2014),Beijing’slossesoutnumbereditsgains—byasubstantialmargin.13Itexperiencedmorelossesthangainsvis-à-visWashingtononthreedifferentmeasuresofsoftpower:publicopinion,mediasentiment,andelitesupport.
2.Acrossallthreemeasuresofsoftpower,Beijingdevotednearlytwo-thirdsofitsentireinternational
13WemeasuretherelativegainsandlossesexperiencedbyChinaonacountry-by-countrybasisbetween2014and2021onthreedifferentmeasuresofsoftpower(publicapproval,mediasentiment,andelitesupport).Forexample,tomeasuretherelativegainsorlossesinpublicapproval,we(1)calculatethedifferencebetweenthepublicapprovalratingforChinainagivenyearandtheprioryear;(2)calculatethedifferencebetweenthepublicapprovalratingfortheU.S.inagivenyearandtheprioryear;and(3)calculatethe“doubledifference”between(1)and(2)todetermineifChinaexperiencedagreatergainorlossinpublicsupportthantheU.S.inthesamecountry-year.12Asapointofcomparison,WorldBankprojectssubjectedtotheorganization’smoststringentenvironmentalandsocialsafeguardstakemorethan7years,onaverage,tomovefromtheproposalstagetothecommencementstage.Onaverage,ittakesWorldBankprojectsanother6yearstomovefromthecommencementstagetothecompletionstage.developmentfinanceportfolioto“toss-up”countries—i.e.,competitivejurisdictionswhereneitherChinanortheU.S.openedupaninsurmountableleadvis-à-visitsprincipalrival.
3.Beijingseekstomaintainandbuilduponmomentum.InjurisdictionswhereitrecentlymadereputationalgainsattheexpenseoftheU.S.,itdoubleddownbyprovidingmoreaidandcredit.
4.Chinahasarelativelylowleveloftoleranceforriskinitspursuitofsoftpower.Itdevotedonly16%ofitsinternationaldevelopmentfinanceportfolioto“moonshot”countries—thosewhereitsprincipalrivalhadmomentumonitsside.14Aseparate,butrelated,findingisthatwhenreputationalassetsbecomereputationalliabilities,Beijingtendstodisengagefromdiscussionsaboutnewprojectsandfinancialcommitmentsandrefocusonmanagingriskswithinitsexistingportfolioofgrant-andloan-financedprojects.
5.Politicaltransitionsinhostcountriesarecriticaljunctureswhenthenature,levelandpaceofChina’sengagementcanchangesignificantly.IfanewleadercomestopowerandtakesalessadversarialposturetowardChina,Beijingtypicallyspringsintoactionandseekstocementbilateralrelationsbyhelpingincumbentstakecreditforhigh-profileinfrastructureprojects.
6.GiventhatBeijingtendstodisengageratherthandoubledownincountrieswheretherearestrongindicationsofBRIbacklash,Beijing’scompetitorsmaybeabletoluresuchcountriesbackintotheWest’sorbit.However,doingsowouldrequirethattheG7actquicklywhenthesewindowsofopportunitiesariseandadapttheirprogrammingtoaddresstheunmetneedsofpartnercountries.
14Similarly,Beijinghasassignedalowerlevelofpriorityto“toss-up”countrieswheremomentumrecentlyshiftedinfavoroftheU.S.
5
Acronyms
AEsAdvancedEconomies
AfDBAfricanDevelopmentBank
AGTFAfricaGrowingTogetherFund
AIArtificialIntelligence
AIIBAsianInfrastructureInvestmentBank
ALAwamiLeague
AsDBAsianDevelopmentBank
AVICAviationIndustryCorporationofChina
B3WBuildBackBetterWorld
BCRACentralBankofArgentina(Banco
CentraldelaRepúblicaArgentina)
BCSBilateralCurrencySwap
BISBankofInternationalSettlements
BNPBangladeshNationalistParty
BOPBalanceofpayments
BPCBotswanaPowerCorporation
BRIBeltandRoadInitiative
BRTBusRapidTransit
BUBostonUniversity
BUILDBetterUtilizationofInvestmentLeading
toDevelopmentAct
CBRCChinaBankingRegulatoryCommission
CCDICentralCommissionforDiscipline
Inspection
CCTVClosed-CircuitTelevision
CDBChinaDevelopmentBank
CFKCristinaFernándezdeKirchner
CGDCenterforGlobalDevelopment
CGSPChina-GlobalSouthProject
CGTCommonGroundTaxonomy
CHICOChinaHenanInternationalCooperation
GroupCompany
ChinaEximbankExport-ImportBankofChina
CICChinaInvestmentCorporationCIDCAChinaInternationalDevelopment
CooperationAgency
CLAChineseLoanstoAfrica
CMECChinaMachineryEngineering
Corporation
CNPCChinaNationalPetroleumCorporation
CODFChina’sOverseasDevelopmentFinance
CPCCommunistPartyofChina
CPECChina-PakistanEconomicCorridor
CRChinaStateRailwayGroupCo.Ltd.
CRSCreditorReportingSystem
CSISCenterforStrategicandInternational
Studies
DACDevelopmentAssistanceCommittee
DBZDevelopmentBankofZambia
DFCU.S.InternationalDevelopment
FinanceCorporation
DLPDebtLimitsPolicy
DRSDebtorReportingSystem
DSFDebtSustainabilityFramework
DSRADebtServiceReserveAccount
DSSIDebtServiceSuspensionInitiative
EBRDEuropeanBankforReconstructionand
Development
EIAEnvironmentalImpactAssessment
EMDEsEmergingMarketandDeveloping
Economies
EMPEnvironmentalManagementPlan
EPCEngineering,Procurement,and
Construction
EPCFEngineering,Procurement,
ConstructionandFinancing
ESGEnvironmental,Social,andGovernance
ESIAEnvironmentalandSocialImpact
Assessment
EUEuropeanUnion
EUREuros
EURIBOREuroInterbankOfferedRate
FASBFinancialAccountingStandardsBoard
FOCACForumonChina–AfricaCooperation
FPICFree,Prior,andInformedConsent
G-24TheGroupof24
G7TheGroupof7
GAAPGenerallyAcceptedAccounting
Principles
GBPPoundSterling
GCDFGlobalChineseDevelopmentFinance
Dataset
GCLGovernmentConcessionalLoan
GDELTGlobalDatabaseofEvents,Language,
andTone
GDPGrossDomesticProduct
GWPGallupWorldPoll
HIAHeritageImpactAssessment
HICsHigh-IncomeCountries
HSBCHongkongandShanghaiBanking
CorporationLimited
IADBInter-AmericanDevelopmentBank
ICBCIndustrialandCommercialBankof
China
ICTInformationandCommunications
Technology
IDSInternationalDebtStatistics
IFADInternationalFundforAgricultural
Development
IFCInternationalFinanceCorporation
IFIInternationalFinancialInstitution
IFRSInternationalFinancialReporting
StandardsIMECIndia-MiddleEast-EuropeEconomic
Corridor
IMFInternationalMonetaryFund
IPSFInternationalPlatformonSustainable
Finance
JPYJapaneseYen
JVJointVenture
KCHPNéstorKirchnerandJorgeCepernic
HydroelectricPowerPlantProject
KHPCKaribaHydroPowerCompany(Private)
Limited
KRCKenyaRailwaysCorporation
LACLatinAmericaandtheCaribbean
LBSLocationalBankingStatistics
LHCLahoreHighCourt
LIBORLondonInter-BankOfferedRate
LICsLow-IncomeCountries
LMICsLower-MiddleIncomecountries
LPRChinaLoanPrimeRate
LSFLiquiditySupportFacility
M&AMergersAndAcquisitions
MCDFMultilateralCooperationCenterfor
DevelopmentFinance
MCPPManagedCo-LendingPortfolio
Program
MDBMultilateralDevelopmentBank
MICsMiddle-IncomeCountries
MMDMovementforMulti-PartyDemocracy
MOFMinistryofFinance
MOFCOMMinistryofCommerce
MOUMemorandumofUnderstanding
MUFGMitsubishiUFJFinancialGroup
NATONorthAtlanticTreatyOrganization
NDRCNationalDevelopmentandReform
Commission
NGONon-GovernmentalOrganizations
NOCNoObjectionCertificate
NORINCOChinaNorthIndustriesGroup
CorporationLimited
NPVNetPresentValue
ODAOfficialDevelopmentAssistance
ODIOverseasDevelopmentInstitute
OECDOrganisationforEconomic
Co-operationandDevelopment
OFCOffshoreFinancialCenter
OOFOtherOfficialFlows
P4IPartnershipsforInfrastructure
PAPProject-AffectedPersons
PBCPreferentialBuyer’sCredit
PBOCPeople’sBankofChina
PFPatrioticFront
PGIIPartnershipforGlobalInfrastructure
andInvestment
PIIEPetersonInstituteforInternational
Economics
PIMCOPacificInvestmentManagement
Company
PLADPoliticalLeaders’AffiliationDatabase
PPGPublicandPubliclyGuaranteedDebt
PxFPre-ExportFinancing
QuadQuadrilateralSecurityDialogue
RAPResettlementActionPlan
RMBRenminbi
S&PStandard&PoorSAFEStateAdministrationofForeign
Exchange
SAIS-CARIChina-AfricaResearchInitiativeatthe
JohnsHopkinsSchoolofAdvanced
InternationalStudies
SCPSupremeCourtofPakistan
SDRSpecialDrawingRights
SHIBORShanghaiInterbankOfferedRate
SIAStructuralImpactAssessment
SinosureChinaExport&CreditInsurance
Corporation
SPVSpecialPurposeVehicle
TAZARATanzania-ZambiaRailwayAuthority
TPDCTanzaniaPetroleumDevelopment
Corporation
TUFFTrackingUnderreportedFinancialFlows
U.S.UnitedStates
UMICsUpperMiddle-IncomeCountries
UNUnitedNations
UNDRIPUNDeclarationontheRightsof
IndigenousPeoples
UNGAUnitedNationsGeneralAssembly
USDUnitedStatesDollars
WDPAWorldDatabaseonProtectedAreas
YoYYear-on-Year
TableofContents
Chapter1:BeltandRoadReconstruction—FromFire-FightingtoFuture-Proofing......................................................................................................1Section1:HasBeijing’sglobalinfrastructureinitiativebecomeanassetoraliability?.............................................................................................................1Section2:Myth-bustingevidenceaboutthescaleandcompositionofChina’soverseasdevelopmentprogram—andhowitsrivalshaveresponded............5Section3:RepaymentriskfromBeijing’sperspective....................................16Section4:ProjectperformanceriskfromBeijing’sperspective......................20Section5:ReputationalriskfromBeijing’sperspective..................................26Section6:Thebalancingactofportfolioriskmanagement...........................37Chapter2:TheRoadtoRepaymentfortheWorld’sLargestOfficialDebtCollector..............................................................................................................45Section1:DebunkingthemyththatBeijing’soverseaslendingprogramhascollapsed........................................................................................................45Section2:MajorchangesinBeijing’soverseaslendingportfolioduringtheBRI2.0era......................................................................................................64Section3:Beijing’scrisis-timeapproachtorepaymentriskmitigation—Foolmeonce,shameonyou.Foolmetwice,shameonme.................................81Section4:WhathavewelearnedaboutBeijing’sbidtorebootitsoverseaslendingprogram?.........................................................................................116Chapter3:RedesigningtheBeltandRoadforSafetyandSpeed....................119Section1:Beijing’sjourneyfromskeptictoadvocateofESGriskmanagement................................................................................................119Section2:MeasuringthescopeandseverityofESGriskexposureinChina'sinfrastructureprojectportfolio......................................................................126Section3:MeasuringthestringencyofESGsafeguardsinChina’sinfrastructureprojectportfoliowithnewsourcesofcontractualevidence...139Section4:DoESGsafeguardsproduceaprojectperformancepayofforpenalty?........................................................................................................157Section5:DecodingBeijing’sESGriskmitigationstrategy.........................164Section6:IsBeijingcourse-correctingorvirtue-signaling?..........................173Chapter4:ReputationalRehabilitationontheBeltandRoad..........................176Section1:China’squestforsoftpowerduringtheBRIera..........................176Section2:Vyingforsoftpower:China’splayforpublicopinion,mediasentiment,andelitesupport........................................................................182Section3:Fromobservationstoreactions...................................................192
Section4:Countrycasestudies...................................................................198Section5:Beijing’sapproachtoreputationalriskmanagement..................236References.........................................................................................................239Appendix:SupplementaryMaterial..................................................................270SectionA-1:Figuresandtablesreferencedinthereport............................270SectionA-2:HowAidDatameasuresconcessionalityandintent.................323SectionA-3:HowdoesAidDatameasurethecumulativestockofofficialfinancialflowsfromChinatoLICsandMICs?..............................................325SectionA-4:AdditionaldetailsontheBIS-basedestimatesofChina’sinternationallendingportfolio......................................................................326SectionA-5:HowdoesAidDatacategorizeChineselendingtodifferenttypesofborrowers?......................................................................................330SectionA-6:Howdidweidentifyprojectsthatrelyoncontractorssanctionedforfraudulentandcorruptbehavior?...........................................................331SectionA-7:HowdidweidentifyESGrisksandriskmitigationmeasuresinAidDataprojectdescriptions?......................................................................333SectionA-8:HowdidAidDataidentifyESGriskmitigationmeasuresininfrastructurefinancingagreements?...........................................................335SectionA-9:AssigningESGSafeguardStringencyRatingstoInfrastructureFinancingAgreementsandthe8InfrastructureFinancingInstrumentCategories....................................................................................................338SectionA-10:HowdoesthecompositionofthecodingsampleofinfrastructurefinancingagreementscomparetothecompositionofChina’sentiregrant-andloan-financedinfrastructureprojectportfolioinLICsandMICs?............................................................................................................354SectionA-11:ESGriskexposureanddejureESGsafeguardprotectioninChina’soverseasinfrastructureprojectportfoliotables................................355SectionA-12:SoftpowergainsandlossesduringearlyandlateBRIperiods.........................................................................................................................365SectionA-13:China’sofficialsectorlendingportfolioinLICsandMICs:acomparisonofAidData,IDS,andCODF......................................................379SectionA-14:ComparisonofAidDataandIDSestimatesofPPGandnon-PPGdebtexposuretoChina................................................................386
Boxes
Box1a:HowAidDataidentifieswhenChina’sborrowersareexperiencingfinancialdistress...................................................................................................19Box1b:HowAidDatameasuresgrassroots,media,andelitesupportforChinaandtheU.S.inthedevelopingworld..................................................................27Box2a:WhatisthetruescaleofChina’soverseaslendingprogram?................51Box2b:Theinvestmentagencybehindthecurtain—China’sStateAdministrationofForeignExchange(SAFE)................................................................................60Box2c:ShouldemergencyliquiditysupportfromPBOCswaplinesbetreatedasshort-termorlong-termdebt?.............................................................................97Box3a:China’sexperiencewiththeapplicationofWorldBanksocialsafeguardstotheWesternPovertyReductionProjectinQinghai.......................................150Box3b:DejureversusdefactoapplicationofESGsafeguardstotheLahoreOrangeLineMetroTrainProject........................................................................153Box4a:HowAidDatameasuresrelativegainsandlossesinChinesesoftpower..............................................................................................................................185
FiguresandTables
Figure1.1:TheglobaldistributionofChineseODA-andOOF-financedprojectsinLICsandMICs....................................................................................................6Figure1.2:OfficialfinancialflowsfromChinatothedevelopingworld,2000-2021..............................................................................................................8Figure1.3:CumulativeofficialfinancialflowsfromChinatothedevelopingworld,2000-2021.................................................................................................10Figure1.4:OfficialfinancialflowsfromChinaandtheG7tothedevelopingworldduringtheBRIera,2014-2021...................................................................11Figure1.5:OfficialfinancialflowsfromChina,theU.S.,andG7countriestothedevelopingworldduringtheBRIera,2014-2021................................................12Figure1.6:OfficialfinancialflowsfromtheU.S.tothedevelopingworld,2000-2021............................................................................................................13Figure1.7:ChangeinsectoralcompositionofofficialfinancialflowsfromChinatothedevelopingworld,2014-2021...................................................................15Figure1.8:PercentageofChina’sportfolioofloancommitmentssupportingcountriesinfinancialdistress,2000-2021............................................................17Figure1.9:China’sshareofoverduerepaymentsowedtoofficialcreditors,2000-2021............................................................................................................18Figure1.10:CanceledorsuspendedChineseinfrastructureprojects,2000-2021..21Figure1.11:EarlyversuslateBRI:Chinesegovernment-financedinfrastructureprojectsthatarebehindschedule.......................................................................22Figure1.12:EarlyversuslateBRI:Commencementandcompletiondelays.......23Figure1.13:Infrastructureprojectsfacingenvironmental,social,orgovernancerisks......................................................................................................................26Figure1.14:ChinaversustheU.S.:Publicapprovalrates...................................30Figure1.15:ChinaversusU.S.:Mediasentiment................................................31Figure1.16:ChinaversustheU.S.:UNvotingalignment...................................32Figure1.17:China’ssoftpowerlossesvis-à-vistheU.S.......................................34Figure1.18:PublicsupportforChina:BRIcountriesvs.non-BRIcountries........35Figure1.19:MediasentimenttowardChina:BRIcountriesvs.non-BRIcountries...............................................................................................................................36Figure1.20:UNvotingalignmentwithChina:BRIcountriesvs.non-BRIcountries.............................................................................................................................37Figure2.1:OfficialsectorlendingcommitmentsfromChinatoLICsandMICsbysource..................................................................................................................48Table2.1:ComparisonofAidData-andBIS-basedestimatesofChina’s
internationallendingportfolio.............................................................................53Figure2.2:CompositionofChina’sloanportfoliobycurrencyofdenomination59Figure2.3:CumulativeloancommitmentstoLICsandMICsbyfinancialinstitutionandcapitalinjectionsfromSAFE,2000-2021.....................................61Figure2.4:LoansfromChinawithintheirrepaymentperiods,2000-2050..........66Figure2.5:LoansfromChinareachingtheir(originallyscheduled)finalrepaymentdates,2000-2050...............................................................................67Figure2.6:Compositionofloanportfoliobyfinancialinstrument......................70Figure2.7:Compositionofloanportfoliobycreditorcategory..........................71Figure2.8:Chineseemergencyrescuelendingbycountries’levelofdebtexposuretoChina................................................................................................72Table2.2:AveragesovereignriskratingsandgrossreservesforrecipientsofdifferentChineseloantypes................................................................................73Figure2.9:Rescuelendinganddebtreschedulingeventsforthetop50borrowersinfinancialdistress,2000-2021...........................................................75Figure2.10:RMB-denominatedrescuelendingtocountriesindistressandcountriesnotindistress.......................................................................................78
Table2.3:IllustrativeescrowaccountbalanceslinkedtoloansfromChinaEximbank,CDB,andICBC...................................................................................82Figure2.11:Percentageofloansthathavereachedmaturityandnumberofloansrescheduled................................................................................................89Figure2.12:Loanreschedulingforcountriesinandnotinfinancialdistress......90Figure2.13:Percentageoflendingportfoliousingfixedorvariableinterestrates95Figure2.14:Compositionofvariableinterestratelendingportfolio..................96Figure2.15:Loanportfoliobackedbycollateral.................................................98Figure2.16:EarlyversuslateBRI:penaltyinterestrates.....................................99Figure2.17:Averagepenaltyinterestratesinselectcountries.........................101Figure2.18:Compositionofnon-emergencylendingportfolio........................104Figure2.19:Percentageofvariableinterestratelendingtocountriesinandnotinfinancialdistress.............................................................................................105Figure2.20:Compositionoflendingtocountriesinandnotinfinancialdistress..............................................................................................................................107Figure2.21:Repaymentriskinemergencylendingportfolioversusprojectlendingportfolio................................................................................................108Figure2.22:Repaymentriskinbilateral,multilateral,andsyndicatedlendingportfolio.............................................................................................................109Figure2.23:Compositionofnon-emergencylendingportfoliobychannelofdelivery..............................................................................................................110
Table2.4:IllustrativesetofsyndicatedloanagreementswithChineseandnon-Chinesebankparticipants..........................................................................112Figure3.1:Infrastructureprojectswithsignificantenvironmental,social,andgovernanceriskexposure..................................................................................131Figure3.2:ESGriskprevalenceinoverseasinfrastructureportfoliofromChinatoLICsandMICs....................................................................................................132Figure3.3:EnvironmentalriskprevalenceinoverseasinfrastructureportfoliofromChinatoLICsandMICs.............................................................................133Figure3.4:SocialriskprevalenceinoverseasinfrastructureportfoliofromChinatoLICsandMICs................................................................................................134Figure3.5:GovernanceriskprevalenceinoverseasinfrastructureportfoliofromChinatoLICsandMICs.....................................................................................135Figure3.6AglobalmapofChina’sinfrastructureprojectportfolioinLICsandMICswithsignificantenvironmental,social,governance(ESG)riskexposure..138Table3.1:DejureESGsafeguardstringencyinChina’soverseasinfrastructureportfoliobytypeoffinancinginstrument..........................................................144Figure3.7:DejureESGsafeguardstringencyinChina’soverseasinfrastructureportfolio.............................................................................................................147Figure3.8:DejuresafeguardstringencyinChina’soverseasinfrastructureportfoliobysafeguardtype...............................................................................148Figure3.9:Infrastructureprojectportfoliowithdejurevs.defactoESGriskmitigation...........................................................................................................153Figure3.10:MonetaryvalueofprojectsuspensionsandcancellationsbydejureESGsafeguardstrength.....................................................................................162Figure3.11:ProportionofinfrastructureprojectportfoliofacingsignificantESGriskexposurebylevelofsafeguardstringency..................................................164Figure3.12:Compositionofinfrastructureprojectportfolio:relianceuponthepolicybanksandweakdejureESGsafeguards................................................167Table3.2:China’sgrant-andloan-financedinfrastructureportfoliobytypeoffinancinginstrumentovertime..........................................................................167Figure3.13:Compositionofinfrastructureprojectportfolio:UseofsyndicatedloansandstrongdejureESGsafeguards..........................................................171Figure3.14:ProportionofinfrastructureportfolioallocatedtoLICs/MICswithhighESGriskprevalencerates..........................................................................172Figure3.15:ProportionofinfrastructureportfolioallocatedtoLICs/MICswithlowESGriskprevalencerates............................................................................173Figure4.1:Bangladeshexample:IsChinalosingorgaininggroundvis-à-vistheU.S.in2017?......................................................................................................185Table4.1:Bangladeshexample:Calculatinganaveragedouble-deltascoreforsoftpowermeasures..........................................................................................186
Figure4.2:China’spublicapprovalgainsandlossesvis-à-vistheU.S..............188Figure4.3:China’smediasentimentgainsandlossesvis-à-vistheU.S............190Figure4.4:China’selitealignmentgainsandlossesvis-à-vistheU.S...............191Table4.2:ExpectedversusobservedallocationofChinesedevelopmentfinancebysoftpowercohort..........................................................................................193Figure4.5:Beijing’saidandcreditprioritizationstrategy..................................196Figure4.6:OfficialfinancialflowsfromChinatoZambia,2014-2021...............203Figure4.7:ChinaversustheU.S.inZambia:Publicapprovalrates...................204Figure4.8:OfficialfinancialflowsfromChinatoBangladesh,2014-2021........215Figure4.9:ChinaversustheU.S.inBangladesh:Publicapprovalrates............216Figure4.10:ChinaversustheU.S.inBangladesh:Mediasentimentscores.....217Table4.3:DevelopmentfinancecommitmentsfromChinatoBangladeshbyexecutiveadministration....................................................................................218Table4.4:ChineseODAandOOFinfrastructureprojectmilestonesduringthesecondhalfof2018inBangladesh....................................................................221Figure4.11:OfficialfinancialflowsfromChinatoArgentina,2014-2021.........225Figure4.12:Argentina’sU.N.votingalignmentwithChinaandtheU.S...........227Table4.5:EstimatedpercentageofArgentina’sforeigncurrencyreservesderivedfromPBOCswapfacility....................................................................................231
AppendixFiguresandTables
TableA1:AidData’sGlobalChineseDevelopmentFinanceDataset,Version3.0ataglance.........................................................................................................270FigureA1:CompositionofofficialfinancialflowsfromChinatothedevelopingworld,2000-2021...............................................................................................271FigureA2:StockofofficialfinancialflowsfromChinaandtheG7tothedevelopingworldduringtheearlyandlateBRIperiods...................................272FigureA3:SectoralcompositionofofficialfinancialflowsfromChinatodevelopingworld,2014vs.2021.......................................................................273FigureA4:Numberoffinancially-distressedlow-andmiddle-incomecountrieswithoutstandingdebttoChina.........................................................................274FigureA5:SovereignoverduerepaymentsowedtoofficialsectorcreditorsinChina..................................................................................................................275FigureA6:Projectlifecycle...............................................................................276FigureA7:AveragelengthofcompletiondelaysacrossallChineseODA-andOOF-financedinfrastructureprojects................................................................277FigureA8:China’ssoftpowergainsandlossesvis-à-vistheU.S.......................278FigureA9:ChinaversustheU.S.:Publicdisapprovalrates...............................279FigureA10:CompositionofChina’soverseaslendingportfoliobylevelofpublicliability................................................................................................................280FigureA11:ProportionofChina’soverseaslendingportfolioprovidedviabuyer’scredits................................................................................................................281FigureA12:ProportionofChina’soverseaslendingportfolioprovidedviaGovernmentConcessionalLoans(GCLs)andPreferentialBuyer’sCredits(PBCs)..............................................................................................................................281FigureA13:ProportionofChina’soverseaslendingportfolioprovidedviainfrastructureprojectloans................................................................................282FigureA14:CompositionofChina’soverseaslendingportfoliobyemergencyandnon-emergencylendinginstrument............................................................283FigureA15:ProportionofRMB-denominatedrescuelendingtocountriesinandnotinfinancialdistress.......................................................................................284FigureA16:CompositionofChina’soverseaslendingportfoliobycurrencyofdenominationforcountriesinfinancialdistress.................................................285FigureA17:CompositionofChina’soverseasloanportfoliobycurrencyofdenominationforcountriesnotinfinancialdistress..........................................286FigureA18:EarlyversuslateBRI:weightedaverageinterestrates...................286FigureA19:EarlyversuslateBRI:weightedmaturitylengths...........................287FigureA20:EarlyversuslateBRI:weightedaveragegrantelement.................287
FigureA21:WeightedaveragegrantelementofoverseaslendingfromChina................................................................................................................................288FigureA22:Weightedaveragematuritylengthsforcountriesinandnotinfinancialdistress.................................................................................................289FigureA23:Weightedaverageinterestratesforcountriesinandnotinfinancialdistress...............................................................................................................290FigureA24:Weightedaveragegrantelementsforcountriesinandnotinfinancialdistress,2014-2021..............................................................................291FigureA25:GrantelementacrossChina’sportfoliotocountriesinandnotinfinancialdistress.................................................................................................291FigureA26:CompositionofChina’soverseaslendingportfoliobytypeoflendinginstrument.............................................................................................292FigureA27:CompositionofChina’soverseaslendingportfoliobytypebycreditorcategory...............................................................................................293FigureA28:ProportionofChina’soverseaslendingportfolioinvolvingamultilateralinstitution........................................................................................293FigureA29:CompositionofChina’soverseaslendingwithandwithoutcreditenhancementstocountriesinandnotinfinancialdistress...............................294FigureA30:CompositionofChina’snon-emergencyoverseasloanportfoliobyrepaymentrisk...................................................................................................294FigureA31:AnnualloancommitmentstoLICsandMICsbyfinancialinstitutionandcapitalinjectionsfromSAFE,2000-2021....................................................296FigureA32:ProportionofChina’ssyndicatedoverseaslendingthatinvolvesnon-Chinesecreditors........................................................................................297FigureA33:ProportionofChineseoverseassyndicatedlendinginvolvingamultilateralinstitution........................................................................................298FigureA34:AveragefinancialcommitmentsfromChinesestate-ownedcreditorstodifferenttypesofsyndicatedloans................................................................298FigureA35:Chinesestate-ownedbankcontributionstosyndicatedloansinLICsandMICsbyparticipantcohort.........................................................................299FigureA36:CompositionofChina’soverseasprojectlendingportfoliobychannelofdelivery.............................................................................................299FigureA37:Officiallendingcommitmentsbylendinginstitutiontype,2000-2021...........................................................................................................................300FigureA38:Cumulativepercentageofinfrastructureportfoliowithsignificantenvironmental,social,orgovernanceriskexposure..........................................301FigureA39:Infrastructureprojectslocatedwithinenvironmentallysensitiveareas302FigureA40:Infrastructureprojectslocatedwithinsociallysensitiveareas........303FigureA41:Infrastructureprojectslocatedingeographicalareasvulnerableto
politicalcaptureandmanipulation....................................................................304FigureA42:Infrastructureprojectsinvolvingcontractorssanctionedforfraudulentandcorruptbehavior........................................................................305FigureA43:Infrastructureprojectsthatencounteredsignificantenvironmental,social,orgovernancechallengesovertime.......................................................306FigureA44:ESGriskprevalenceinoverseasinfrastructureportfoliofromChinatoLICsandMICs................................................................................................307FigureA45:Environmentalriskprevalenceinoverseasinfrastructure...............308FigureA46:Socialriskprevalenceinoverseasinfrastructureportfolio.............309FigureA47:Governanceriskprevalenceinoverseasinfrastructureportfolio...310FigureA48:Infrastructureprojectssupportedbyoneormoredefactoenvironmental,social,orgovernanceriskmitigationefforts.............................311FigureA49:DejureESGsafeguardstringencyinChina’soverseasinfrastructureportfoliowithESGriskexposure........................................................................312FigureA50:Infrastructureprojectportfoliowithdejurevs.defactoESGriskmitigation...........................................................................................................313FigureA51:Infrastructureprojectportfoliowithdejurevs.defactoESGriskmitigation,byyearofprojectcompletion.........................................................314FigureA52:Infrastructureprojectportfoliowithdejurevs.defactoESGriskmitigation,byyearofprojectcommencement..................................................315FigureA53:ProportionofinfrastructureprojectfinancingfacingsignificantenvironmentalrisksbywhethertheprojectfinancinghadstrongdejureenvironmentalsafeguardsversusthosewithweakdejureESGsafeguards......316FigureA54:ProportionofinfrastructureprojectfinancingfacingsignificantsocialrisksbywhethertheprojectfinancinghadstrongdejureenvironmentalsafeguardsversusthosewithweakdejureESGsafeguards.............................317FigureA55:ProportionofinfrastructureprojectfinancingfacingsignificantgovernancerisksbywhethertheprojectfinancinghadstrongdejureenvironmentalsafeguardsversusthosewithweakdejureESGsafeguards......318FigureA56:Compositionofinfrastructureprojectportfolio:RelianceonChinaEximbankandweakdejureESGsafeguards.....................................................319FigureA57:Compositionofinfrastructureprojectportfolio:RelianceonChinaDevelopmentBankandweakdejureESGsafeguards......................................320FigureA58:ProportionofChina’sinfrastructureprojectportfoliobehindschedulebydejureESGsafeguardstrength.....................................................320FigureA59:AveragelengthofcommencementdelaysininfrastructureprojectsbydejureESGsafeguardstrength....................................................................321FigureA60:AveragelengthofcompletiondelaysininfrastructureprojectsbydejureESGsafeguardstrength..............................................................................321FigureA61:AveragetimetocompleteaninfrastructureprojectbydejureESG
safeguardstrength.............................................................................................322FigureA62:ESGmap-projectcount................................................................323FigureA63:OfficialsectorPRClendingtoLICsandMICsversusY1HICs.......329TableA2:ContractorssanctionedforfraudulentandcorruptbehaviorthatwereinvolvedinChineseODA-andOOF-financedinfrastructureprojectsduringtheirdebarmentperiods............................................................................................332TableA3:Criteriatoevaluateenvironmental,social,andgovernancesafeguardsinChinesegrantandloancontracts..................................................................336TableA4:SafeguardstringencyratingsforinfrastructurefinancingagreementsattheESGdomain-contractdyadlevel.................................................................339TableA5:SampleofinfrastructurefinancingagreementsusedtocodethedejurestringencyofESGsafeguards.....................................................................341TableA6:Summaryofenvironmental,social,andgovernancesafeguardclausesinthecontractsampleforthepre-BRIandearlyBRIperiod(2000-2017).........344TableA7:Summaryofenvironmental,social,andgovernancesafeguardclausesinthecontractsampleforthelateBRIperiod(2018-2021)...............................346TableA8:EnvironmentalSocialGovernanceSafeguardQuestions..................348TableA9:CompositionofcodingsampleofinfrastructurefinancingagreementsversusChina’sentiregrant-andloan-financedinfrastructureprojectportfolio.354TableA10:Regionaldistributionofenvironmental,social,andgovernanceriskexposureinChina’soverseasinfrastructureprojectportfolio............................355TableA11:Distributionofenvironmental,social,andgovernanceriskexposureinChina’soverseasinfrastructureprojectportfoliobyhostcountryincomelevel...356TableA12:Country-by-countrydistributionofESGriskexposureanddejureESGsafeguardprotectioninChina’soverseasinfrastructureprojectportfolio,2000-2021..........................................................................................................357TableA13:China’ssoftpowergainsandlossesvis-à-vistheU.S.duringearlyandlateBRIperiods...........................................................................................365TableA14:Country-by-countrybreakdownofexpectedversusobserveddevelopmentfinanceallocationsduringlateBRIperiod...................................372TableA15:OfficialsectorlendingcommitmentsfromChinatoLICsandMICs:Country-by-countrycomparisonofAidData,IDS,andCODFinconstant2021USDmillions.......................................................................................................380TableA16:Country-by-countrycomparisonofAidDataandIDSestimatesofPPGandnon-PPGdebtexposuretoChinainconstant2021USDmillions..............388
Chapter1:BeltandRoadReconstruction—From
Fire-FightingtoFuture-Proofing
Section1:HasBeijing’sglobalinfrastructureinitiativebecomeanassetoraliability?
TwocompetingnarrativesaboutChina’soverseasdevelopmentprogramare
gainingtraction.ThefirstisthatBeijingsitsatopamountainofforeignexchange
reservesandhasgainedtheupperhandinazero-sum,greatpowercompetition
byusingitsflagshipinfrastructureprogram—theBeltandRoadInitiative
(BRI)—towintheloyaltyofforeignleadersandgeneratefavorableinternational
publicsentiment.Accordingtothisnarrative,Chinaisaninternationallenderof
firstresortthatdevelopingcountriescannotaffordtoalienateorantagonize.It
bankrollsbig-ticketinfrastructureprojects—suchasroads,railways,bridges,
tunnels,powerplants,andtelecommunicationsystems—thatneithertheU.S.
noritsallieswillsupport.Itpliesforeignleaderswithlavishspendingonvanity
projectslikestatues,theaters,museums,conventioncenters,stadiums,
presidentialpalaces,andparliamentarybuildings.Itisalsoadeep-pocketed
lenderoflastresortthatiswillingandabletobailoutborrowerswhentheyare
underwaterorsailingintostrongmacroeconomicheadwinds.Strategistsand
decision-makersinWesterncapitalsofteninvokethisargument—aboutChinese
strength—tomakethecasethatBeijingis“outmaneuveringusinoverseas
theaters,”“eatingourlunch,”or“winningtheglobalbattleforheartsand
minds.”1
Butasecondnarrative—acounter-narrativeaboutChineseweakness—isgaining
currencyinWashington,London,andBrussels.TheBRI,inthisalternative
reading,ismoreofaliabilitythananasset,andBeijingisonaperilouspath,in
dangerofspendingitswayintointernationaldisrepute.Developingcountries
1The2018NationalDefenseStrategyoftheU.S.assertedthat“[t]hecentralchallengetoU.S.prosperityandsecurityisthereemergenceoflong-term,strategiccompetition…[with]revisionistpowers”(U.S.DepartmentofDefense2018).ItcalledupontheU.S.governmentto"out-think,out-maneuver,out-partner,andout-innovaterevisionistpowers”(U.S.DepartmentofDefense2018).InOctober2018,theU.S.CongresspassedtheBetterUtilizationofInvestmentLeadingtoDevelopment(BUILD)Act,establishinga“fullservice”developmentfinanceinstitutiontohelptheU.S.governmentcompetewithChinaaroundtheglobe.Then,inSeptember2019,itauthorizedthecreationofa“CounteringChineseInfluence”fund.
1
thatborrowedextensivelyfromChinaarenowsaddledwithdebtsforoversized
infrastructureprojectsthatgenerateinsufficientrevenue.Thegraceperiodson
theseloansareexpiring,forcingBeijingintoanunfamiliaranduncomfortable
roleasthedevelopingworld’slargestofficialdebtcollector.Ononehand,it
wantstopositionitselfatthefrontoftherepaymentlinebyrequiringthat
borrowersgrantitsourcesofleverage—suchascashcollateralinescrow
accounts—thatotherofficialcreditorsdonotpossess.Ontheotherhand,it
wantstocharacterizeitselfasaglobalchampionof“South-Southcooperation”
thatprivilegessolidaritywithlow-incomecountries(LICs)andmiddle-income
countries(MICs).AnothersourceofvulnerabilityisthatpoliticiansintheGlobal
Southincreasinglyfeelcompelledtocancelormothballhigh-profileBRIprojects
becauserisinglevelsofpublicantipathytowardChinaaremakingitdifficultto
maintaincloserelationswithBeijing.Internationalmediaoutletsarealsotraining
theirsightsonanarrayofproblemsintheBRIprojectportfolio,suchas
overpricing,corruption,habitatdestruction,andinvoluntarydisplacementof
vulnerableandmarginalizedpopulations.
ThefundamentaldifferencebetweenthesenarrativesisthatoneassumesChina
isplayingoffenseandtheotherassumesChinaisplayingdefense.Whichoneis
abettercharacterizationofreality?Orisitpossiblethatboth—orneither—are
true?BeijingisclearlyawarethatitfacesaBRI“buyer’sremorse”problem.2But
haveChineselendingandgrant-givinginstitutionslearnedfromtheirpast
mistakesandrecalibratedtheirpoliciesandpractices?Ifso,how?
OneofthefirstsignsthatBeijingwasconsideringamajorcoursecorrection
cameinOctober2016whenanofficialwithChina’sNationalDevelopmentand
ReformCommission(NDRC)toldaLondon-basednewspaperthat“thesedays
weneedviableprojectsandagoodreturn.Wedon’twanttobacklosers”
(FinancialTimes2016).Then,inMay2017,XiJinpingannouncedthat“[w]ewill
[...]strengtheninternationalcooperationonanti-corruptioninordertobuildthe
BeltandRoadInitiativewithintegrity.”3Hedeliveredasimilarmessagein
September2018:financingfromChinawas“nottobespentonanyvanity
projectsbutinplaceswheretheycountthemost.”4
4QuotedinShepherdandBlanchard2018.3BeltandRoadForumforInternationalCooperation2017.2Euractiv2023.
2
Callsforamajoroverhaulofthecountry’sflagshipglobalinfrastructure
initiative—dubbed“BRI2.0”—grewlouderoverthecourseofthenexttwo
years.InApril2019,attheSecondBeltandRoadForumforInternational
Cooperation,XiannouncedthatthenextphaseoftheBRI(“BRI2.0”)wouldbe
“open,greenandclean”(MinistryofForeignAffairsofthePeople’sRepublicof
China2019).HealsosaidthatChinawould“adoptwidelyacceptedrulesand
standardsandencourageparticipatingcompaniestofollowgeneral
internationalrulesandstandardsinprojectdevelopment,operation,
procurementandtenderingandbidding”(MinistryofForeignAffairsofthe
People'sRepublicofChina2019).Then,in2020,Beijingsignaledinterestin
“multilateralizing”theBRIandharmonizingsomeofitspoliciesandpractices
withprevailinginternationaldevelopmentfinancerulesandstandards.5
However,itisuncleariftheserhetoricalcommitmentshavetranslatedinto
significantchangesinthewaythatBeijingadministersitsoverseasdevelopment
program.Some“Chinawatchers”haveexpressedskepticismthatBeijingis
sincerelyinterestedinfinancing,designing,andimplementinginfrastructure
projectsincoordinationandcollaborationwithmultilateralinstitutionsand
Westernpowers.OnereasontoquestionthecredibilityofBeijing’scommitment
toBRI2.0isthatthecountry’scomparativeadvantages(scale,speed,and
near-termeconomicimpact)vis-à-visOECD-DACdonorsandmultilateral
developmentbanksmaynotbefullycompatiblewithBRImultilateralization.6
AnotherreasonisthatChinahasalongtrackrecordofengaginginstrictly
bilateral,“go-it-alone”effortstomanageoverseascrises.7Intensifying
competitionintheglobalinfrastructurefinancemarketpresentsanadditional
challenge:theU.S.,theU.K.,andtheothermembersoftheG7arenow
promotingaPartnershipforGlobalInfrastructureandInvestment
7Forexample,whenChina’sbiggestborrowershaveexperiencedfinancialdistress,ithasnotcalledforcoordinateddebtreschedulingswithallmajorcreditors(Gardneretal.2021;BonandCheng2021).Instead,ithasdiscreetlyprovidedemergencyrescueloanstoensurethatsuchborrowersaresufficientlyliquidtocontinueservicingtheirexistingBRIprojectdebts(Hornetal.2023a,2023b).BeijinghasalsospurnedmultipleinvitationstojointheParisClub(themainvenueforsovereigndebtrestructurings)andinsisteduponloancontractswithclausesthatexpresslyexcludeChinesedebtsfromtheParisClubandothercollectiverestructuringinitiatives,therebygrantingitsstate-ownedcreditorssolediscretiontodecideif,when,andhowtheywillgrantdebtrelief(Dreheretal.2022;Gelpernetal.2022).6WeaddressthisissueatgreaterlengthinChapter3.AlsoseeParks2019;Maliketal.2021;andDreheretal.2022.5Morrisetal.2020;Dreheretal.2022.In2020,BeijingteamedupwitheightmultilateralinstitutionstoestablishtheMultilateralCooperationCenterforDevelopmentFinance(AIIB2021).
3
(PGII)—previouslyknownastheBuildBackBetterWorld(B3W)initiative—that
theycharacterizeasanalternative,higherqualityoptionforcountriesthatwant
toundertakeinfrastructureprojectsbasedontheprinciplesofsustainableand
transparentfinancing,publicsectormobilizationofprivatecapital,consultation
andpartnershipwithlocalcommunities,andstrictadherencetoenvironmental,
social,andgovernance(ESG)safeguards.8Therefore,ifBeijingwantedto
protecttheBRIbrandbyspreadingreputationalriskacrossalargergroupof
donorsandlenders,itwouldhavetosecurethebuy-inofasetofactorswhodo
notseemtobeparticularlyinterestedincollaboration(Parks2019;Dreheretal.
2022).
RegardlessofwhetherChinacanconvincetraditionaldonorsandlendersto
jumpontheBRIbandwagon,thereareindicationsthatarebootofthe
infrastructureinitiativeisunderway.InNovember2017,thecountry'stop
bankingregulator—ChinaBankingRegulatoryCommission(CBRC)—introduced
newrulesthatrequireChinaDevelopmentBank(CDB)andChinaEximbankto
putinplacemorerobustenvironmentalandsocialriskmanagementprocedures
(CBRC2017a,2017b).Then,in2021,China’sStateAdministrationofForeign
Exchange(SAFE)—theultimatesourceoffundingformostofChina’s
state-ownedpolicybanks,commercialbanks,andinvestmentfunds(seeBox2b
inChapter2)—announcedthatitwouldprioritize“adoptingMDB’sESGcriteria”
and“incorporatingESGprinciplesintothewholeprojectinvestmentprocess
fromdecision-makingtopost-investmentmanagement”(SAFE2021:54).Atthe
sametime,SAFE’sparentorganization(thePBOC)calledforareductionin
collateralizedlendingtooverseasborrowersthatpresenthighlevelsof
repaymentrisk(Chen2023:1173).9
Inthischapter,wedocumenttheevolvingnature,scale,andcompositionof
China’soverseasdevelopmentprogram.WealsoexplainwhyChinahaspivoted
intoaninternationalcrisismanagerroleandhowitisseekingtomanagethree
9AtthethirdBeltandRoadsymposiuminNovember2021,XiJinpingalsocalledforbetterriskmanagementinoverseasinfrastructureprojectsanddiscouragedChinesecompaniesfromgoingto“chaoticanddangerous”places(StateCouncilofthePeople'sRepublicofChina2021).8TheEuropeanUnion,Japan,Australia,andIndiahavealsolaunchedasetofcomplementaryinitiatives—calledGlobalGateway,PartnershipsforInfrastructure(P4I),theBlueDotNetwork,andtheIndia-MiddleEast-EuropeEconomicCorridor(IMEC)—thataredesignedtoincreasechoiceandcompetitionintheglobalinfrastructurefinancemarket.
4
differenttypesofriskinitsoverseasprojectportfolio:(1)repaymentrisk,(2)
ESG-relatedprojectperformancerisk,and(3)reputationalrisk.
InChapters2and3,weidentifythespecificmeasuresthatBeijinghastakento
de-riskitsoverseasdevelopmentfinanceportfolio,andevaluatewhetherthese
changesareconsistentwithitsrhetoricalcommitmentstomultilateralizationand
harmonizationwiththeprevailinginternationaldevelopmentfinancerulesand
standards.Then,inChapter4,weexplainhowChinahasusedaidandcredit
instrumentstorespondtothesoftpowergainsandlossesithasexperiencedin
LICsandMICs.OurfindingssuggestthattheambitionofBeijing’seffortto
“future-proof”itsoverseasdevelopmentfinanceportfolio—anditsflagship,
globalinfrastructureinitiative—shouldnotbeunderestimated.
Section2:Myth-bustingevidenceaboutthescaleandcompositionofChina’soverseasdevelopmentprogram—andhowitsrivalshaveresponded
AidDatamaintainstheworld’smostcomprehensivedatasetofinternational
developmentfinancecommitmentsfromChinatothedevelopingworld.The
latest(3.0)versionofourGlobalChineseDevelopmentFinance(GCDF)dataset
captures20,985projectsandactivitiesin146countriessupportedbyfinancial
andin-kindtransfersworth$1.34trillionfromofficialsectorinstitutionsin
China.10Itcoverseverylow-income,lower-middleincome,andupper-middle
incomecountryandterritoryacrosseverymajorworldregion,includingAfrica,
Asia,Oceania,theMiddleEast,LatinAmericaandtheCaribbean,andCentral
andEasternEurope(including19additionalcountriesorterritorieswhere
systematicsearcheswereundertakenbutnoChinesegovernment-financed
projectsandactivitieswereidentified).Thedatasettracksprojectsandactivities
over22commitmentyears(2000-2021),anditincludesdetailsonthetimingof
project/activityimplementationovera24-yearperiod(2000-2023).17,957
recordsinthedatasetrepresentformallyapproved,active,andcompleted
projectsandactivities.Theremaining3,028recordsinthe3.0datasetrepresent
(1)projectsandactivitiesthatsecuredofficialfinancialorin-kindcommitments
10The$1.34trillionfigureexcludestheshort-term“rollover”facilitiesdescribedinBox2candSectionA-3).Itincreasesto$1.5trillionwhensuchfacilitiesareincludedinthetally.
5
fromChinabutweresubsequentlysuspendedorcanceled;(2)projectsand
activitiesthatsecuredpledgesoffinancialorin-kindsupportfromofficialsector
institutionsinChinabutneverreachedtheformalapproval(officialcommitment)
stage;and(3)so-called“umbrella”recordsthataredesignedtosupport
multiplesubsidiaryprojectsandactivities.Figure1.1presentstheglobal
distributionofapproved,active,andcompletedprojectsandactivitiesthatwere
financedwithOfficialDevelopmentAssistance(ODA)andOtherOfficialFlows
(OOF)fromChinabetween2000and2021.11
Figure
1.1:
The
global
distribution
of
Chinese
ODA-
and
OOF-financed
projects
in
LICs
and
MICsFigure1.1
The
global
distribution
of
Chinese
ODA-
and
OOF-financed
projects
in
LICs
and
MICs
Notes:ThismapshowsthegeographiclocationsofprojectssupportedbyChineseODAandOOF
commitmentsacrossallLICsandMICsbetween2000and2021.Onlyprojectsfromthe3.0versionof
11AllprojectsandactivitiesinAidData’sGCDFdatasetmustqualifyasofficialfinancialflows(ODAorOOF).Foreaseofexposition,intheremainderofthisreport,werefertoallsuchprojectsandactivitiesas“projects.”ThedefinitionsofandmeasurementstandardsforODAandOOFaredescribedinSectionA-2oftheAppendix.
6
AidData’sGCDFdatasetthathavephysicalfootprintsorinvolvespecificlocationsarerepresented.
Goodmanetal.(2023)describestheprocessbywhichthesepoint,polygon,andlinevectordatawere
generated.
WhenAidDatareleasedthe2.0versionoftheGCDFdatasetinSeptember
2021,itprovidedevidencethatChinawasoutspendingtheU.S.inthe
developingworldonatleasta2-to-1basis.12The3.0versionoftheGCDF
datasetdemonstratesthatChinaremainsthesinglelargestsourceof
internationaldevelopmentfinanceintheworld.In2021,officialfinancialflows
(ODAandOOFcommitments)fromChinatoLICsandMICsamountedto$79
billion(seeFigure1.2).13NoneofChina’speersorrivalshaveoverseaslending
andgrant-givingprogramsthatarecomparableinscale.In2021,nosingle
memberoftheG7providedofficialfinancialflowstoLICsandMICsinexcessof
$61billion.14Chinaalsooutspentallmultilateralsourcesofinternational
developmentfinance.Thesinglelargestmultilateralsourceofinternational
developmentfinancein2021wastheWorldBank,withinternational
developmentfinancecommitmentsworthapproximately$72billion.
Accordingtotheconventionalwisdomamongthinktanksandmediaoutlets,
BeijingmadeaconcertedefforttoreinintheBRIafter2017(Lu2023;Olander
2023;DoRosarioandSavage2023)anditsannualdevelopmentfinance
commitmentsplummetedtonearlyzeroby2021(GallagherandRay2020;
Hwangetal.2022;Ray2023;MyersandRay2023;Mosesetal.2023).15
However,the3.0versionofAidData’sGCDFdatasetdebunksthemyththat
Beijing’soverseasdevelopmentprogramhasexperiencedatotalcollapse.With
15Inarecentstock-takingexercise,NikkeiAsiaconcludedthat“initialoptimism[fortheBRI]hasbeenreplacedbydisappointmentovermismanagement,debtcrisesandcorruptionthathaveleftmanyprojectsunfinishedorincapableoffulfillingtheirpromisedpotential(Aamiretal.2022).14SincethefirstfullyearofBRIimplementationin2014,nomemberoftheG7hasoutspentChina.TheindividualmembersoftheG7spentbetween$2billionand$61billionayearonoverseasdevelopmentactivitiesbetween2014and2021.Chinaspentbetween$74billionand$142billionayearbetween2014and2021.13ToseetheannualpercentageofChina’sofficialfinancialflowstoLICsandMICsprovidedviaODAandOOF,seeFigureA1intheAppendix.12Atthetime,weestimatedthatChina’saverageannualdevelopmentfinancecommitmentsamountedto$85.4billionbetween2013and2017andaverageannualdevelopmentfinancecommitmentsfromtheU.S.amountedto$37billionoverthesamefive-yearperiod(Maliketal.2021).However,thelatest(3.0)versionoftheGCDFdatasetdemonstratesthatChinawasoutspendingitsrivalsbyawidermarginthanwepreviouslyunderstood:averageannualdevelopmentfinancecommitmentsfromChinaamountedto$117billionbetween2013and2017andaverageannualdevelopmentfinancecommitmentsfromtheU.S.amountedto$40.6billionoverthesamefive-yearperiod(inconstantUSD2021prices).ThesehistoricalrevisionsimplythatBeijingwasoutspendingWashingtononanearly3-to-1basisduringtheearlyyearsoftheBRI.
7
morecompletedataonthefullrangeofChina’slendingandgrant-giving
activitiesinLICsandMICs,itshowsafarlessdramaticdeclineinoverseas
spendingduringthelateBRIperiod:officialfinancialflows(ODAandOOF
commitments)fromChinatoLICsandMICsdeclinedfrom$115billionin2018
to$104billionin2019and$74billionin2020,beforeincreasingto$79billionin
2021(seeFigure1.2).16
Figure
1.2:
Official
financial
flows
from
China
to
the
developing
world,
2000-2021Figure1.2
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(see
SectionA-2oftheAppendixfordetails).TheVague(OfficialFinance)isaresidualcategoryforofficial
financialcommitmentsfromChinathatcouldnotbereliablycategorizedasODAorOOFbecauseof
insufficientlydetailedinformation.
InChapter2,wesettherecordstraightbydemonstratingthatasharpdeclinein
dollar-denominated,bilaterallendingforpublicinvestmentprojectsdidindeed
takeplacebetween2018and2021,butitwasoffsettoasignificantextentby
16InChapter2,weexplainwhyAidData’sestimatesofofficialsectorlendingcommitmentsfromChinatoLICsandMICsaresubstantiallydifferentfromthoserecordedinotherpubliclyavailabledatabases,suchastheChina'sOverseasDevelopmentFinanceDatabase,theChineseLoanstoLatinAmericaandtheCaribbeanDatabase,theChineseLoanstoAfricaDatabase,theChina'sGlobalEnergyFinanceDatabase,theChinaOverseasFinanceInventoryDatabase,andtheWorldBank'sInternationalDebtStatistics(IDS).
8
contemporaneousincreasesinRMB-denominated,bilateralemergencyrescue
lendingandincreasesinnon-emergencylendingviasyndicationand
multilateralization.Wealsoprovideevidencethat,althoughBeijingisreducing
itsrelianceonthepolicybanks,itisratchetingupitsuseofthecountry’s
state-ownedcommercialbanksandcentralbank.
ThescaleofChina’sambitionasaglobaldevelopmentfinancierbecomeseven
clearerinananalysisof“stocks”ratherthan“flows.”17Figure1.3provides
evidencethattheinflation-adjustedvalueofBeijing’s21st-centuryoverseas
developmentfinanceportfolioreachedthe$745billionmarkbythefirstfullyear
ofBRIimplementation(2014)andsurpassedthe$1trillionmarkbythefourthfull
yearofBRIimplementation(2017).Theportfoliocontinuedtoexpandin
subsequentyears—somuchsothatitexceeded$1.3trillionin2021.18Figure1.4
presentsthecumulativeamountofChineseODAandOOFfrom2014to2021
andcomparesittocumulativeamountsofODAandOOFfromeachmemberof
theG7overthesametimeperiod.China’soverseasdevelopmentprogramis
vastlylargerthanthatofanyG7member,includingtheU.S.Beijinghasoutspent
WashingtoninLICsandMICsonamorethantwo-to-onebasissincetheBRIwas
launched.19Between2014and2021,officialfinancialflowsfromChinatothe
developingworldamountedto$680billion,whilethosefromtheU.S.amounted
to$319billion.20Overthesametimeperiod,Chinaalsooutspentthesingle
largestsourceofmultilateraldevelopmentfinance—theWorldBank—bya
considerablemargin.21
21Totalinternationaldevelopmentfinance(ODAandOOF)commitmentsfromtheWorldBankamountedto$493billionfrom2014-2021.20The$680billionfigureexcludestheshort-term"rollover"facilitiesdescribedinBox2candSectionA-3.Itincreasesto$841billionwhensuchfacilitiesareincludedinthetally.19China’saverageannualinternationaldevelopmentfinance(ODA/OOF)commitmentsamountedto$85billionbetween2014and2021(excludingtheshort-term"rollover"facilitiesdescribedinBox2candSectionA-3).Thisfigurerisesto$105billionifoneincludesshort-term"rollover"facilities.Averageannualinternationaldevelopmentfinance(ODA/OOF)commitmentsfromtheU.S.amountedto$40billionbetween2014and2021.18SectionA-3intheAppendixprovidesguidanceforthosewhowishtousethe3.0versionofAidData'sGCDFdatasettoestimatecumulativestocksofofficialfinancialflowsfromChinatoLICsandMICs.17Therationaleforevaluating“stocks”(cumulativefinancialcommitments)ratherthan“flows”(annualfinancialcommitments)isthatgrantsandloansareissuedatspecificpointsintime,buttheydisburseovermanyyearsandsupportprojectsimplementedovermanyyears.
9
Figure
1.3:
Cumulative
official
financial
flows
from
China
to
the
developing
world,
2000-2021Figure1.3
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(as
describedinSectionA-2oftheAppendix).TheVague(OfficialFinance)isaresidualcategoryforofficial
financialcommitmentsfromChinathatcouldnotbereliablycategorizedasODAorOOFbecauseof
insufficientlydetailedinformation.Thisfigureexcludesshort-term"rollover"facilitiesfromthetallyof
officialfinancialcommitments(seeBox2candSectionA-3intheAppendix).
10
Figure
1.4:
Official
financial
flows
from
China
and
the
G7
to
the
developing
world
during
the
BRI
era,
2014-2021Figure1.4
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(as
describedinSectionA-2oftheAppendix).TheVague(OfficialFinance)isaresidualcategoryforofficial
financialcommitmentsfromChinathatcouldnotbereliablycategorizedasODAorOOFbecauseof
insufficientlydetailedinformation.G7ODAandOOFdatarepresentgrossdisbursementsfromthe
OECD-DAC.Thisfigureexcludesshort-term“rollover”facilitiesfromthetallyofofficialfinancial
commitments(seeBox2candSectionA-3intheAppendix).
Atthesametime,thereisevidencethattheU.S.isbeginningtoclosethe
overseasdevelopmentspendinggapwithChina(seeFigure1.5).Duringthe
earlyBRIperiod(2014-2017),ChinaoutspenttheU.S.onanearlythree-to-one
basis.However,duringthelateBRIperiod(2018-2021),Beijingspent$2.33for
everyoverseasdevelopmentdollarspentbyWashington.Thegapnarrowed
evenmoreduringthelastyearforwhichreliabledataareavailable:Beijingonly
spent$1.30foreveryoverseasdevelopmentdollarspentbyWashingtonin
2021.22
22Whileofficialfinancialflows(ODAandOOF)fromtheU.S.toLICsandMICsamountedto$61billionin2021,officialfinancialflowsfromChinatoLICsandMICsamountedto$79billionduringthesameyear.
11
Figure
1.5:
Official
financial
flows
from
China,
the
U.S.,
and
G7
countries
to
the
developing
world
during
the
BRI
era,
2014-2021Figure1.5
Notes:Thisfiguremeasuresofficialfinancialflows(ODAandOOFcommitments)fromChina,theU.S.and
G7countriestoLICsandMICsfrom2014to2021.AidDatareliesonOECD-DACmeasurementcriteriato
makeODAandOOFdeterminations(asdescribedinSectionA-2oftheAppendix).TheU.S.andG7ODA
andOOFdatarepresentgrossdisbursementsfromtheOECD-DAC.
Figure1.6demonstratesthattheU.S.gainedgroundonChinain2021duetoa
fifteen-fold(1,423%)increaseinOOFexpenditure.23In2020,OOFrepresented
just4%oftheU.S.internationaldevelopmentfinanceportfolio;theremaining
96%consistedofODA.However,oneyearlater,thepercentageofU.S.
internationaldevelopmentfinanceprovidedviaOOFsoaredto36%.Thismajor
compositionalchangeinU.S.developmentexpendituresuggeststhat
WashingtonisseekingtocompetewithBeijingviaemulationratherthan
differentiation.24
24Relatedly,Zeitz(2021)providesevidencethattheWorldBankisseekingtocompetewithChinaviaemulationratherthandifferentiation.23U.S.OOFamountedto$1.4billionin2020and$21.8billionin2021.
12
Figure
1.6:
Official
financial
flows
from
the
U.S.
to
the
developing
world,
2000-2021Figure1.6
Notes:U.S.ODAandOOFflowsreflectgrossdisbursements(asOOFcommitmentdataarenotpublished
bytheOECD-DACforindividualDACmembers).
The$20.3billionincreaseinOOFthattookplacein2021wastheresultofan
expansionintheoverseasactivitiesoftheU.S.InternationalDevelopment
FinanceCorporation(DFC).25InOctober2018,theU.S.Congresspassedthe
BetterUtilizationofInvestmentLeadingtoDevelopment(BUILD)Act,which
establishedtheDFCasa“fullservice”developmentfinanceinstitutiontohelp
theU.S.“outcompete”Chinaaroundtheglobe.26However,theDFCdidnot
becomefullyoperationaluntil2021.Oneofitsearliest(attempted)transactions
wasa$2.8billionloaninJanuary2021tohelptheGovernmentofEcuador
repaysomeofitsoutstandingdebtstoChinaaheadofschedule,inexchange
foracommitmenttoexcludeChinesecompaniesfromitstelecommunications
networks.27TheproposedborrowingtermsoftheDFCloanwere
27SevastopuloandLong(2021);Landersetal.(2021).26AkhtarandBrown(2021);Dreheretal.(2022).25InJuly2023correspondencewithAidData,theUSAIDofficeresponsibleforODAandOOFreportingtotheOECD-DACconfirmedthatthefulloperationoftheDFCin2021ledtothemajorincreaseinU.S.OOFin2021.SeealsoDFC2021aandDFC2022.TheDFC'stransaction-leveldata,whichisorganizedbyfiscalyearratherthancalendaryear,canbeaccessedviahttps://www.dfc.gov/our-impact/transaction-data.
13
non-concessionalinnature:an8-yearmaturity,a1-yeargraceperiod,andan
interestrateofLIBORplusa2.25%margin.28Atthetime,DFCCEOAdam
Boehlersaidthattheloanwould“refinancepredatoryChinesedebtandhelp
Ecuadorimprovethevalueofitsstrategicassets.”29
ButU.S.spendingpatternsdonotprovidethefullpicturebecauseWashington
isseekingtooutcompeteChinabypartneringwithitsalliesinLondon,Paris,
Berlin,Tokyo,Rome,andOttawa.G7effortstocompetewithChinaare
gatheringsteam.DuringtheearlyBRIperiod(2014-2017),ChinaandtheG7
wereeffectivelymatchingeachotherdollar-for-dollar:foreveryoverseas
developmentdollarthatChinaspent,theG7spent$1.09.30However,duringthe
lateBRIperiod(2018-2021),theG7steppedupitsefforts,spending$1.47for
everyoverseasdevelopmentdollarspentbyChina(seeFigureA2inthe
Appendix).31By2021,theG7wasoutspendingChinaonanearlytwo-to-one
basis(seeFigure1.5).32
AstheU.S.anditsalliesseektocompetewithChinabyrollingoutflagship
infrastructureprograms(likethePartnershipforGlobalInfrastructureand
InvestmentandtheIndia-MiddleEast-EuropeEconomicCorridorInitiative)and
rampingupnon-concessionallending(OOF),astrategicpivotisunderwayin
Beijing.Figure1.7providesevidenceofmajorchangesinthesectoral
compositionofChina’soverseasdevelopmentfinanceportfoliobetween2014
and2021.33Beijingwasforthemostpartfocusedonprovidingcreditfor
33AccordingtoFigureA3,thepercentageofChina’soverseasdevelopmentfinanceportfoliothatsupportedthree“hardware”sectors(Energy;Industry,Mining,andConstruction;andTransportandStorage)declinedfrom68.3%in2014(thefirstfullyearoftheearlyBRIperiod)to30.6%in2021(thelastyearofthelateBRIperiodforwhichwehavecompletedata).ThepercentageofChina’soverseasdevelopmentfinanceportfoliothatsupportedthe“BankingandFinancialServices”sectorand“GeneralBudgetSupport”sectorincreasedfrom13.0%to58.9%between2014and2021.ThesetwosectorscaptureemergencylendingfromChina’scentralbank(PBOC)toLICandMICcentralbanksviacurrencyswaparrangements,aswellasemergencylendingtofinanceLICandMICfinanceministriesviaChina’sStateAdministrationofForeignExchange(SAFE),state-ownedpolicybanks,state-ownedcommercialbanks,andstate-ownedenergycompanies32In2021,officialfinancialflows(ODAandOOF)fromallmembersoftheG7(combined)toLICsandMICsreached$163billion,whileofficialfinancialflowsfromChinatoLICsandMICsamountedto$79billion.31WhileChinaspent$93billionayearonaveragebetween2018and2021,theG7spent$137billionayearonaverage.30TheG7failedtomatchChina’sODAandOOFcommitmentsin2014and2016(seeFigure1.5).29DFC(2021b).28Bruni(2021).TheJanuary2021frameworkagreementbetweentheDFCandGovernmentofEcuadorcanbeaccessedinitsentiretyviahttps://www.dropbox.com/s/oy377uc6wz8u9oe/Ecuador%20DFC%20January%202021%20Framework%20Agreement.pdf?dl=0.
14
large-scaleinfrastructureprojectsduringtheearlyBRI(orBRI1.0)era.Yet,aswe
explainatgreaterlengthinChapter2,itrampeddowninfrastructureproject
lendingandrampedupemergencyrescuelendingduringthelateBRI(orBRI
2.0)period.Beijingmadethiscoursecorrectioninordertoadapttoanew
reality:thefactthatmanyofitslargestborrowerswerehavingseriousdifficulty
repayingtheirinfrastructureprojectdebts.34
Figure
1.7:
Change
in
sectoral
composition
of
official
financial
flows
from
China
to
the
developing
world,
2014-2021Figure1.7
Notes:Thisfigure,whichrelieson3-digitOECDsectorcodesfromthe3.0versionofAidData’sGCDF
dataset,presentsproportionalchangesinthesectoralcompositionofChineseODAandOOF
commitments(measuredinconstant2021USD)toLICsandMICsbetween2014and2021.FigureA3inthe
Appendixprovidessupplementaryevidenceonsectoralchangesoverthesametimeperiod.
Thereareglass-half-fullandglass-half-emptywaysofinterpretingG7effortsto
competewithChina.Theglass-half-fullviewisthatBeijingiscedingits
leadershipintheglobalinfrastructurespaceandtheG7isseekingtofillthe
vacuum,whichmayhelpaddressthelarge,unmetinfrastructurefinancingneeds
ofdevelopingcountries.Theglass-half-emptyviewisthattheG7ismisreading
thedemandsignalfromtheGlobalSouthandseekingtocompetewitha
versionoftheBRI(BRI1.0)thatnolongerexists.Afterrecentlygoingona
borrowingspreeforbig-ticketinfrastructureprojects,low-incomeand
middle-incomecountriesmayhavelessappetiteforexpensiveinfrastructureand
moreappetiteforbalanceofpaymentssupportthatwillkeepthemafloatduring
34BeijingeffectivelycreatedabackstopforhighlyexposedChinesecreditorsbychannelingemergencyrescueloanstothesubsetofBRIparticipantcountriesthatpresentmajorbalancesheetrisks(seeHornetal.2023a,2023baswellasthereplicationexercisethatweconductinChapter2).
15
atimewhenglobaleconomicconditionsarehighlyunfavorable(asinterestrates
rise,thedollarstrengthens,localcurrenciesweaken,andeconomicgrowth
slows).
Section3:RepaymentriskfromBeijing’sperspective
Beijing’spivotawayfrominfrastructureprojectlendingandtowardemergency
rescuelendingisasmuchaboutsupply-sidepressuresasitisabout
demand-sidepressures.AftertheBeltandRoadInitiativewaslaunched,Chinese
state-ownedcreditorswentonalendingspree,issuingthousandsofloansfor
big-ticketinfrastructureprojectsspreadacross129countriesinthedeveloping
world.However,theydidsowithoutstrongriskmanagementguardrailsinplace.
Theylenttoborrowerswithbadcreditratingsornocreditratings(likeLaos,
Tajikistan,Zambia,SouthSudan,Suriname,Zimbabwe,Pakistan,andArgentina);
bankedonborrowersbeingabletorepayloanswiththecashproceedsfrom
naturalresourceexports(likeAngola,EquatorialGuinea,Ecuador,Venezuela,
Congo-Brazzaville,andTurkmenistan);andissueddollar-andeuro-denominated
loanstocountries(likeRussia,Belarus,Myanmar,Sudan,Iran,andCuba)that
wouldlaterbeunabletotransactinthosecurrenciesduetointernational
sanctions.
Now,Chinesestate-ownedcreditorsaresaddledwithmanyunderperforming
loansandwanttoensurethattheiroverseasborrowersaresufficientlyliquidto
continueservicingtheirexistinginfrastructureprojectdebts.35Theyare
respondingtothischallengeintwoways,whichwedocumentingreaterdetail
inChapter2:debtreschedulingsthatprovideshort-termcashflowreliefto
borrowersandemergencyrescueloansthathelpborrowersshoreuptheir
foreignexchangereservesandrepayexisting(infrastructureproject)debts.
35Chinesestate-ownedcreditorshaveforthemostpartrespondedtothischallengeviacashflowrelief(emergencyrescueloans,graceperiodextensions,and/ormaturityextensions),whichsuggeststhattheythinktheirborrowersareilliquidbutnotinsolvent(Hornetal.2023a,2023b).InarecentinterviewwithMuyangChenofPekingUniversity,onepolicybankofficialcharacterizedtherisingtideofdebtdistressinthedevelopingworldas“essentiallyaproblemofliquidity”(Chen2023:1772).
16
Figure
1.8:
Percentage
of
China’s
portfolio
of
loan
commitments
supporting
countries
in
financial
distress,
2000-2021Figure1.8
Notes:MOFCOMinterest-freeloancommitments(whicharetypicallyissuedwithoutacredibleexpectation
ofrepayment)andemergencyrescueloancommitments(responsestofinancialdistress)areexcludedfrom
thecalculation.Todetermineifacountrywasexperiencingfinancialdistressinagivenyear,weusethe
binarymeasuredescribedinBox1a.
17
Figure
1.9:
China’s
share
of
overdue
repayments
owed
to
official
creditors,
2000-2021Figure1.9
Notes:Sovereignarrearscaptureprincipalandinterestarrears(i.e.,overduerepayments)onPPGdebtto
ChinaEximbank,ChinaDevelopmentBank,andChina’sMinistryofCommercecontractedbyasubsetof
LICsandMICsthatparticipateintheWorldBank’sDebtorReportingSystem(DRS).Eachcountry-year
observationisgivenequalweightinagivenyeartogenerateglobalaverages.Yearsinwhichacountry
maintaineddiplomaticrelationswithTaiwanareexcluded.ThedataaredrawnfromtheWorldBank’s
InternationalDebtStatistics.
Figures1.8,1.9,andA4intheAppendixhighlightthescopeandseverityofthe
challenge.Until2008,Beijingneverhadtodealwithmorethan10
financially-distressedcountrieswithunpaiddebtstoChinesestate-owned
creditors;however,by2021,atleast57countrieswithoutstandingdebttoChina
wereinfinancialdistress(seeFigureA4).36In2000,34%ofChina'soverseas
36AccordingtotheWorldBank’sInternationalDebtStatistics(IDS),sovereignarrearsfromLICsandMICstoofficialsectorcreditorsinChinahavealsosharplyincreasedinabsoluteterms(seeFigureA5).However,duetothecreditcoverageandunderreportingissuesthataffectIDSdata(seeFigure2.1,TableA15,Hornetal.2021,andMaliketal.2021),theabsoluteamountsinFigureA5shouldbeinterpretedwithcaution.
18
lendingportfoliosupportedborrowersinfinancialdistress.37By2021,thatfigure
skyrocketedto79%(seeFigure1.8).38
Box1a:HowAidDataidentifieswhenChina’sborrowersareexperiencing
financialdistress
FigureA4,Figure1.8andothergraphsandtablesinthisreportrequireameasureofwhencountrieshaveborrowedfromChinaundernormalcircumstancesandwhencountrieshaveborrowedfromChinaduringperiodsoffinancialdistress.Inordertodeterminewhetherandwhenaborrowercountryexperiencedafinancialdistressepisode,weidentifyifitmetanyoneofthefollowingconditionsinagivenyear:
●Itregisteredascoreof5orlessonthesovrateindexproducedbytheWorldBank;
●ItsoverduerepaymentsonpublicdebttoChinawereequaltoorgreaterthan10%ofitsoftotaloutstandingpublicdebttoChina(asmeasuredbytheWorldBank);or
●OneormoreofitsofficialsectorloansfromChina—thatarewithintheiroriginallyscheduledrepaymentperiods—showedsignsoffinancialdistress(asmeasuredbythe3.0versionofAidData’sGCDFdataset).
TheWorldBank’ssovrateindexisameasureofrepaymentrisk—basedonaveragesovereigncreditratingsfromMoody’s,Standard&Poor’s,andFitch—thatvariesfrom0to21,withhigherscoresindicatinglowerlevelsofsovereigncreditrisk(Koseetal.2022).Countrieswithscoresof0-5arein“CandDterritory”onthemeasuresproducedbytheBigThreecreditratingagencies—i.e.,theyareindefaultoratahighriskofdefault(Teixeiraetal.2018;Sérietal.2021).39ThedataonoverdueloanrepaymentstoChinaaredrawnfromtheWorldBank’sInternationalDebtStatistics(IDS),whichisbasedonvoluntaryreportingby119low-incomeandmiddle-incomecountries.Weusethesedatatomeasurewhetheracountry’sprincipalandinterestarrearsonpublicandpubliclyguaranteed(PPG)debttoofficialsectorcreditorsinChinaareequaltoorgreaterthan10%ofitstotaloutstandingPPGdebttoofficialsectorcreditorsinChina.40Wealsouseavariablefromthe3.0versionofAidData’sGCDFdatasetthatmeasureson
40WeexcludeallobservationsinyearswhenacountrymaintaineddiplomaticrelationswithTaiwan(sincethereportedarrearsmaybetoTaiwan).39ThesovereigncreditratingsproducedbyMoody’s,Standard&Poor’s,andFitchvarybetweenAAAandD.RatingsofBBorlowerareconsideredtobe“junkterritory.”38Duringthepre-BRIperiod(2000-2013),32%ofChina'soverseaslendingportfoliosupportedborrowersinfinancialdistressonaverageeachyear.Thisfigureincreasedto79.7%duringtheearlyBRIperiod(2014-2017)andremainedunchanged(79.7%)duringthelateBRIperiod(2018-2021).37Figure1.8drawsinspirationfromFigure1inHornetal.(2023b),whichreliesonanearlier(2.0)versionofAidData’sGCDFdataset.BothfiguresseektomeasurethepercentageofofficialsectorChineselendingtoLICandMICborrowercountriesindistress,butinsomewhatdifferentways.TheshapeofthelineinFigure1.8isdifferentfromtheoneinHornetal.(2023b),inthattheonsetofexceptionallyhighlevelsofdistress(inexcessof50%)beginsin2011ratherthan2022.Figure1.8alsosuggeststhatnearly80%ofChina’soverseaslendingtoLICsandMICsisnowsupportingcountriesindistress.Hornetal.(2023b)estimatethatthefigureiscloserto60%,althoughtheyuseaversionoftheGCDFdatasetwhichcaptures3,030officialsectorloansfromChinaratherthanthe4,890officialsectorloansfromChinacapturedinthe3.0version(including4,776forapproved,active,andcompletedprojects/activitiesand114forsuspendedandcanceledprojects/activities).
19
aloan-by-loanbasiswhethertheborrowerhaddifficultymakingrepaymentsorshowedsignsoffinancialdistressduringtherepaymentperiod.41
Ourmeasureofwhetheragivenborrowercountryexperiencedafinancialdistressepisodeisresetto0(“turnsoff”)inagivenyearifthecountry'ssovratescoreexceeds5.0,itsoverduerepaymentsonpublicdebttoChinaarenolongerequaltoorgreaterthan10%ofitsoftotaloutstandingpublicdebttoChina,andnoneofitsloansfromofficialsectorcreditorsinChina(thatarewithintheiroriginallyscheduledrepaymentperiods)showsignsoffinancialdistress.
AnotherusefulgaugeofthehealthofChina’soverseaslendingportfolioisthe
extenttowhichitsborrowersarefallingbehindontheirrepaymentstoChinese
creditors(incomparisontootherexternalcreditors).Figure1.9tracksthe
percentageofalloverduepayments(principalandinterestarrears)from
low-incomeandmiddle-incomegovernmentstoallofficialcreditorsthatare
owedtocreditorsinChina.Itshowsthatarapidlygrowingproportionof
overdueloanrepaymentsareowedtoChinesestate-ownedcreditors.This
figuremorethandoubled—from7%in2000to19%in2021.42
Section4:ProjectperformanceriskfromBeijing’sperspective
Beijinghaschanneledanextraordinaryamountofaidandcredittothe
developingworldforlarge-scaleinfrastructureprojects.Accordingtothe3.0
versionofAidData’sGCDFdataset,whichincludesanewmeasureofwhether
projectsinvolvetheconstruction,rehabilitation,expansion,ormaintenanceof
physicalinfrastructure,officialsectordonorsandlendersinChinaissued4,800
grantsandloans(worth$825billion)forinfrastructureprojectsin140developing
countriesbetween2000and2021.43Infrastructureprojectsarenotoriously
difficulttoimplement(forreasonsthatwediscussatgreaterlengthinChapter
43Formoreonthedefinitionandmeasurementofthe“infrastructure”variable,seeCusteretal.(2023).42SovereignarrearsfromLICsandMICstoofficialcreditorsinChinahavealsosharplyincreasedinabsoluteterms(seeFigureA5).41Oncethisdummyvariableissetto1(“turnson")inaparticularcountry-year,itretainsavalueof1(“stayson”)forthatcountryuntiltheendoftheloan’soriginallyscheduledrepaymentperiod.Therepaymentperiodiscalculatedforeachloancommitment(regardlessofwhetheritwassubsequentlysuspendedorcanceled)basedontheoriginallyscheduledfirstrepaymentdate(estimatedbyaddingthegraceperiodtothecommitmentdate)andtheoriginallyscheduledfinalrepayment(maturity)date(estimatedbyaddingthematuritytothecommitmentdate).Incaseswherethegraceperiodisunknown,theaveragegraceperiodacrossallofficialsectorloansfromChinatothesameborrowercountryisapplied.Thevariablenever“turnson”forMOFCOMinterest-freeloans(whicharetypicallyissuedwithoutacredibleexpectationofrepayment)oremergencyrescueloans(responsestofinancialdistress).
20
3).Therefore,animportantpartofBeijing’sportfoliomanagementchallengeis
dealingwithprojectperformancerisk.
The3.0versionofAidData’sGCDFdatasetisdifferentfromotherpublicly
availabledatasetsofChinesedevelopmentfinanceinthatitcapturesproject
suspensionsandcancellations(seeFigureA6).Wehavepreviouslyencountered
criticismforpublishingdataoninfrastructure—andnon-infrastructure—projects
backedbyofficialcommitmentsthataresubsequentlysuspendedorcanceled.44
However,wemaintainthatitisimportanttosystematicallytracktheseprojects
andsubjectthemtoanalysis.Shieldingsuspendedandcanceledprojectsfrom
publicscrutinyleavesanalystsanddecision-makerswithanincompletepicture
ofBeijing’soverseasdevelopmentprogram.Italsolimitsopportunitiestolearn
fromfailure.
Figure
1.10:
Canceled
or
suspended
Chinese
infrastructure
projects,
2000-2021Figure1.10
44Usersofthe3.0versionofAidData’sGCDFdatasetwhowishtoexcludesuspendedandcanceledprojectsfromtheiranalysiscaneasilydosobyusingthe“RecommendedforAggregates”filterorthe“Status”variable.
21
Notes:BasedonChineseODA-andOOF-financedprojects(includingthosecanceledorsuspendedsince
2000)markedas“infrastructure”inthe3.0versionofAidData’sGCDFdataset.Noprojectsthatwere
committedintheyear2000weresubsequentlycanceledorsuspended.
Figure1.10demonstratesthatsuspensionsandcancellationsofinfrastructure
projectshaverapidlyaccumulatedovertime.By2021,94infrastructureprojects
in49countriessecuredChinesegrantandloancommitmentsworth$56billion
thatweresubsequentlysuspendedorcanceled.Theseprojectsarelikelyviewed
byBeijingasevidenceofBRIbuyer’sremorse,as80%ofthesuspensionsand
cancellationsthattookplacebetween2014and2021wereinBRIparticipant
countries.45However,thismethodofmeasurementisaconservative,
lower-boundestimateofthetotalnumberofinfrastructureprojectsthatChina
unsuccessfullysoughttobankrollandbuild,asitexcludesthosethatwere
shelvedbeforesecuringfinancialcommitments(Luetal.2023b).46
Figure
1.11:
Early
versus
late
BRI:
Chinese
government-financed
infrastructure
projects
that
are
behind
scheduleFigure1.11
Notes:“Behindschedule”isdefinedasChineseODA-andOOF-financedprojectswheretheactual
implementationstartdatetookplace3monthsormoreafteritsoriginallyscheduledimplementationstart
date,aswellasprojectswheretheactualcompletiondatetookplace3months(ormore)afteritsoriginally
scheduledcompletiondate.
Anothersourceofimplementationriskandpotentialcauseforconcernisthe
increaseintheproportionofBeijing’sinfrastructureprojectportfoliorunning
46Takeforexampletheprojecttobuilda97-km“oilroad”connectingMasindi-Biso,Kabaale-KiziranfumbiandHohwa-Nyairongo-Kyarushesha-ButooleinUganda(capturedviaID#96073inthe3.0versionofAidData’sGCDFdataset).InDecember2021,Uganda’sMinistryofFinancewithdrewitsrequestforparliamentaryauthorizationtocontractaloanfromChinaConstructionBankfortheproject.Therefore,AidDatastatus-codestheprojectas“Pipeline:Pledge”ratherthanasuspensionorcancellationofafinancialcommitment.45Tocalculatethisfigure,wefirstdeterminewhichcountrieshadofficiallyjoinedtheBRIbytheendof2021,andthencalculatetheshareofsuspendedandcanceledprojectsinBRIparticipantcountriesbetween2014and2021.
22
behindschedulebetweentheearlyBRIperiodandthelateBRIperiod(see
Figure1.11).47However,thismeasureofwhetheraprojectisrunningbehind
schedulecapturestwodifferenttypesofdelays:commencementdelaysand
completiondelays.48Morespecifically,itmeasureswhetheraproject’s(a)actual
implementationstartdatetookplace3monthsafteritsscheduled
implementationstartdateorlonger,and/or(b)itsactualcompletiondatetook
place3monthsafteritsscheduledcompletiondateorlonger.Whenthis
summarymetricisunbundledintoitsconstituentparts,amorecomplexand
nuancedpictureofChina’soverseasinfrastructureprojectportfolioemerges.
Figure1.12demonstratesthat,whilecommencementdelayshaveincreased,
completiondelayshavenot.Theaveragecommencementdelayincreasedby32
daysbetweentheearlyBRIperiodandthelateBRIperiod,whiletheaverage
completiondelayshrankby59daysoverthesametwotimeperiods.49
Figure
1.12:
Early
versus
late
BRI:
Commencement
and
completion
delaysFigure1.12
Notes:Thisfigureisbasedonactiveandcompletedinfrastructureprojectsfinancedwith
ChineseODAandOOF.Delaysarecalculatedbytakingthedifference(incalendardays)
betweentheoriginallyscheduledprojectimplementationstartdate/completiondateandthe
actualprojectimplementationstartdate/completiondate(respectively).
Thefactthatashrinkingproportionofinfrastructureprojectslaunchontheir
originallyscheduledcommencementdatescouldberelatedtoseveraldifferent
49FigureA7providesevidenceofasteadydeclineinaveragecompletiondelays—from571daysin2000to220daysin2021.48Thismeasureisderivedfromthe“DeviationfromPlannedImplementationStartDate”and“DeviationfromPlannedCompletionDate”variablesinthe3.0versionofAidData’sGCDFdataset.47Figure1.11demonstratesthatnotonlytheproportionbutalsotheoverallsizeofBeijing’sinfrastructureprojectportfolio(inconstant2021USD)runningbehindscheduleincreasedbetweentheearlyBRIperiodandthelateBRIperiod.
23
factors.OnepossibilityisthatChinesecontractorsand/ortheirhostcountry
counterpartsareincreasinglyexpectedtocomplywithenvironmental,social,or
governance(ESG)standardspriortothestartofprojectimplementation—or
theyareunderestimatingthedifficultyofmeetingthesestandardsbefore
infrastructureprojectscangetunderway.50Aseparatebutrelatedpossibilityis
thatChineselendersareaskingtheirborrowerstomeetmoreESGconditions
(so-called“conditionsprecedent”)priortothedateofthefirstloan
disbursement,whichtypicallyprecedesthestartofprojectimplementation.We
addresstheseissuesatgreaterlengthinChapter3.Anotherpossibility,which
weaddressinChapter4,isthathostcountrypoliticiansareincreasinglyreluctant
to“claimcredit”forinfrastructureprojectsfinancedbyChinaviahigh-profile
groundbreakingceremonies.
Atthesametime,notallofthelightsonBeijing’sprojectperformance
dashboardare“flashingred.”Chineselendersandcontractorshaveevidently
learnedhowtoreducethelikelihoodthatoverseasinfrastructureprojectswill
notbefinishedontime—andthelengthofanydelaysthatdotakeplaceduring
projectimplementation.Theseimprovementscouldbetheresultofbetter
planning(morerealisticforecastingofhowlongittakestocompleteprojects)or
fewer/smallerscopeofworkdeviationsbythecontractorsresponsiblefor
projectimplementation.Theyalsocallattentiontoanimportantmeasurethat
BeijingcantaketosloworreversetherisingtideofBRI“buyer’sremorse”:
timelycompletionofprojectsthatimprovetheprovisionofpublicservicesin
hostcountries.InnewworkwithcollaboratorsfromHeidelbergUniversity,the
UniversityofGöttingen,andtheUniversityofHongKong,weprovidecausal
evidencethatthecompletionofChinesedevelopmentprojectsincreases
popularsupportfortheChinesegovernmentinhostcountries(Wellneretal.
forthcoming).51Wealsoshowinthesamestudythathostcountryresidentsare
morelikelytoreportsatisfactionwiththedeliveryofpublicservicesuponthe
completionofChinesedevelopmentprojects.
51Onepotentialreasonwhyprojectcompletiondatesareconsequentialforreputationsisthatthey"eraseanyuncertaintyaboutwhetheraprojectwillactuallyreachcompletion”(Wellneretal.forthcoming).Thesamestudydoesnotfindthatprojectcommencementconsistentlydeliversapublicopiniondividend.50Typically,theproceedsfromagrantorloanfromaChinesestate-ownedentityareusedbytherecipienttofinanceacommercialcontractbetweena“projectowner”inthehostcountryandacontractorfromthefinancier’scountryoforigin.Thecommercialcontracttypicallyspecifiesanexpectedimplementationstartdateandanexpectedcompletion(implementationend)date,butcontractorsand/orprojectownerscannegotiatescopeofworkandtimelinemodifications.
24
However,Beijingcannotaffordtorestonitslaurels.Anotherkeyfindingfrom
thesamestudyisthattheseeffectserodeovertime:onaverage,wefindthat
thecompletionofoneadditionalChinesedevelopmentprojectincreasespublic
supportfortheChinesegovernmentbyapproximately3percentagepointsin
theshortrunbutonly0.2percentagepointsinthelongerrun(Wellneretal.
forthcoming).
BeijingalsofacesadauntingsetofESGrisksinitsoverseasdevelopment
program.InChapter3,wedevelopandanalyzeanewcompositemeasureof
ESGriskthatcaptureswhetheragivenChinesegrant-orloan-financed
infrastructureproject(1)tookplaceinanareathatisenvironmentallysensitive,
sociallysensitive,orvulnerabletopoliticalcaptureandmanipulation;(2)relied
oncontractorssanctionedbyotherinternationalfinanciersforfraudulentand
corruptbehavior;or(3)encounteredasignificantenvironmental,social,or
governanceproblembefore,during,orafterimplementation.Figure1.13,which
drawsuponthismeasure,indicatesthatthecumulativenumberofChinese
grant-andloan-financedinfrastructureprojectsinLICsandMICswithsignificant
ESGriskexposureincreasedfrom17projectsin16countriesin2000to1,693
projectsin125countriesin2021.Asof2000,Beijinghadissuedgrantsand
loansworth$420millionforinfrastructureprojectsinLICsandMICsthatfaced
oneormoresignificantESGrisks.Thisfigureincreasedonacumulativebasisto
$470billionin2021.TheESGriskprevalencerate,whichwedefineasthe
annualpercentageofChina’sgrant-andloan-financedinfrastructureproject
portfolio(measuredinconstant2021USD)withsignificantenvironmental,social,
orgovernanceriskexposure,alsoincreasedfrom12%in2000to33%in2021
(seeFigure3.2).52
52TheaverageESGriskprevalenceratereached54%duringtheearlyBRI(2014-2017)periodand47%duringthelateBRI(2018-2021)period(seeChapter3andFigure44).
25
Figure
1.13:
Infrastructure
projects
facing
environmental,
social,
or
governance
risksFigure1.13
Notes:ThepresenceofsignificantESGriskexposureisbasedonaproject-levelcompositemeasurethatis
describedinSection2ofChapter3.
Section5:ReputationalriskfromBeijing’sperspective
ManagingtheriskprofileofChina'soverseasdevelopmentprogramalso
requiresafocusongrassroots,media,andelitesupportinhostcountries.53In
2014,XiJinpingmadethecaseforreputationalriskmanagementinChina’s
internationalactivities,arguingthat“[w]eshouldincreaseChina’ssoftpower,
giveagoodChinesenarrative,andbettercommunicateChina’smessagetothe
world”(People’sDaily2014).Reinforcingthispoint,aseniorofficialfromChina’s
MinistryofCommercenotedthat“theworkofforeignaidrelatestoChina’s
image.Wecannottolerateanynegligenceorprojectsofpoorquality”
(MOFCOM2014).Morerecently,ataSeptember2023conferenceonthe10th
anniversaryoftheBRI,China'sAmbassadortotheEuropeanUnion
acknowledgedthat"[w]henitcomestotheBRI,whetherit'sgoodorbad,we
53Foreignpowersundertakereputationalriskmanagementeffortsthatfocusonthegeneralpublicandthemediabecausetheybelievethatmorefavorablesentimentcan“filterupandinfluenceelitepolicytobemoreamenableto[theirown]interests”(BrazysandDukalskis2019:567).
26
needtolistenmoretothereceivingcountries.Weknowthereisalotofcriticism
intheWesternmediaandalsofromWesterngovernmentssometimesbutwe
caremoreaboutthereactionsfromthereceivingstates,especiallyintheGlobal
South”(Euractiv2023).
Figure1.14presentsdatafromtheGallupWorldPoll(GWP)onaveragelevelsof
publicsupportfortheChinesegovernmentandtheU.S.governmentacrossthe
developingworldduringtheearlyBRI(2014-2017)periodandthelateBRI
(2018-2021)period.TheGWPdatashowsimilarlevelsofpublicsupportfor
BeijingandWashingtonduringtheearlyBRIperiod.Bothcountriessawtheir
approvalratingsintheGlobalSouthfluctuatebetween50%and60%between
2014and2017.However,astheinitialmomentumbehindChina’sflagship
globalinfrastructureinitiativewanedandcountriesre-evaluatedtherisksand
rewardsoftheircontinuedparticipationduringthelateBRIperiod,globalpublic
opinionvis-à-visChinasoured.Beijingsuffereda16percentagepointloss
between2019and2021;itspublicapprovalratinginlow-incomeand
middle-incomecountriesplungedfrom56%in2019to40%in2021.54
Washington,bycontrast,sawitsapprovalratingintheGlobalSouthriseby7
percentagepointsin2021,therebyopeningupa14percentagepoint
advantageoveritsrival.
Box1b:HowAidDatameasuresgrassroots,media,andelitesupportforChina
andtheU.S.inthedevelopingworld
ThisreportreliesonthreedifferentmeasuresofChineseandAmericansoftpowerinlow-incomeandmiddle-incomecountries:(1)publicopinion,(2)mediasentiment,and(3)elitesupport.
WerelyontheGallupWorldPoll(GWP)fordataonpublicapprovalofChinaandtheU.S.TheGWPisthemostsystematicefforttoconsistentlycollectpublicopiniondataineverymajorworldregionovertime.Itprovidesannual(repeatedcross-section)datafrom2006to2021formorethan140countries.Thedataareprobability-basedandnationallyrepresentativeoftheresidentpopulationof15yearsandolder.Tofacilitateouranalysis,wefirsttransformtherespondent-leveldatafromWP156andWP151(“DoyouapproveordisapproveofthejobperformanceoftheleadershipofChina?”and“Doyouapproveordisapproveofthejob
54FigureA9intheAppendixprovidesevidencethatthisdeclineisnotduetoincreasedindifferencetowardChina(i.e.,theabsenceofactiveapproval).Itisbecauseofanincreaseinactivedisapproval—from44%in2019to60%in2021.Highlevelsofactivedisapprovallikelyreflectmultiplefactors,includingconcernsaboutthelocaleffectsofChinesedevelopmentprojectsandhowBeijinghandledtheCOVID-19pandemic(Silveretal.2020;Blairetal.2022a).
27
performanceoftheleadershipoftheUnitedStates?”)intotwobinaryindicatorsthatassumevaluesof1ifarespondentapprovesoftheleadershipofChinaortheleadershipoftheU.S.,respectively.55WethencalculatethepercentageofrespondentswhoapprovedoftheleadershipofChinaortheleadershipoftheU.S.,respectively,atthecountry-yearlevel.
WemeasuremediasentimenttowardChinaandtheU.S.bycalculatingtwosetsofaveragescores—oneforChinaandonefortheU.S.—atthecountry-yearfromtheGlobalDatabaseofEvents,Language,andTone(GDELT)1.0EventDatabase.WerelyontheAvgTonevariable,whichisalgorithmicallycalculatedbasedonthetoneofhundredsofmillionsofnewsarticlesacrossnearly200countries.Thismeasurevariesfrom-100to+100,withpositivescoresindicatingfavorablemediacoveragerelatedtogovernmentactorsinmainlandChina(ortheU.S.)andnegativescoresindicatingunfavorablecoveragerelatedtogovernmentactorsinmainlandChina(ortheU.S.).56Formostcountriesatmosttimes,AvgTonescoresvarybetween-10and+10,withvaluesof0indicatingneutralmediacoverage.
WemeasureelitesupportforChinaandtheU.S.byassessingtheextenttowhichothergovernmentsaligntheirvotesintheUnitedNationsGeneralAssembly(UNGA)witheachoftheseforeignpowers.TheUNGAisavenueinwhichgovernmentshaveanopportunitytostakeoutforeignpolicypositionsthataresimilarordissimilartothoseadoptedbyChina(ortheU.S.).OurmeasureofUNGAvotingalignmentwithChina(ortheU.S.)isbasedon“idealpointdistance”estimatesbetweeneachcountryinthe3.0versionofAidData’sGCDFdatasetandChina(ortheU.S.).TheseestimatesaredrawnfromVersion29.0oftheUnitedNationsGeneralAssemblyVotingData(Baileyetal.2017)andtheyareinverted,suchthathighervaluesindicatehigherlevelsofUNGAvotingalignmentwithChina(ortheU.S.).57
Theglobalcompetitionforfavorablemediacoveragedidnotplayoutinquite
thesameway.Figure1.15demonstratesthat,inabsoluteterms,China
outperformedtheU.S.onthisdimensionofsoftpowerinthedevelopingworld
duringboththeearlyBRIperiodandlateBRIperiod.58Thispatternisconsistent
withevidencethatBeijing’sgrassrootsimagemanagementstrategyinvolves
proactiveuseofpublicdiplomacytoolstogeneratemorefavorablemedia
58However,inrelativeterms,theU.S.gainedgroundonChinabetween2014and2021(seeFigure1.17.57Althoughthe“onecountry,onevote”ruleappliesinUNGA,wereportpopulation-weightedestimatesofUNGAvotingalignmentinchapters1and4becauseweusethismeasureasaproxyforelitesupportofChinaandtheU.S.(ratherthanasadirectmeasureofanempiricalphenomenonofinterest).Wealsoreportpopulation-weightedestimatesofgrassrootssupport(viaGallupWorldPoll)andmediasupport(viaGDELT)inchapters1and4.ThisapproachisbasedontheassumptionthatlargecountriesaremoreimportanttoChina(andtheU.S.)thansmallcountries,regardlessofthesoftpoweroutcomethatisbeingsought.56GDELTeventrecordsarerecordedinadyadicformat,withtwoactorsandanactionperformedbyActor1onActor2(e.g.,theprovisionofaidfromonecountrytoanothercountry,aleaderfromonecountryvisitinganothercountry).Forthepurposesofouranalysis,werestrictActor1totheLICsandMICsthatarecoveredbythe3.0versionoftheGCDFdatasetandActor2toChinaandtheU.S.Wealsorestrictouranalysistoeventrecordswheretheactor2type1codevariableissettoGOV(inordertoensurethatwearemeasuringmediasentimentabouttheChineseGovernmentandtheU.S.Government,respectively).55WealsogeneratedtwobinaryindicatorsthatassumevaluesofoneifarespondentdisapprovesoftheleadershipofChinaortheleadershipoftheU.S.,respectively.Wedroppedall"Don'tKnow"and"RefusedtoAnswer"observations.
28
reportingaboutChina(BrazysandDukalskis2019;Custeretal.2018,2019).But
Beijing’sadvantageoverWashingtonwashardlyinsurmountable;by2020and
2021,ChinahadlostgroundtotheU.S.andwasstrugglingtomaintaina
razor-thinleadinmediacoveragefavorability.
China’soutsizedinfluenceintheGlobalSouthismostclearlyevidentinthe
UnitedNationsGeneralAssembly(UNGA),wherevotingpatternsareoftenused
asaproxyfortheextenttowhichgoverningelitesindevelopingcountriesalign
theirforeignpolicypositionswiththoseoftheU.S.orChina.Figure1.16
demonstratesthatcountriesintheGlobalSouthconsistentlyvotewithChina
ratherthantheU.S.inUNGA.Althoughtherearesomenaturalforeignpolicy
affinitiesbetweenChinaandcountriesintheGlobalSouth,Beijinghasa
well-establishedtrackrecordofusingitslargessetobuyvotesininternational
organizations.59InjointworkwithourlongtimecollaboratorsfromHeidelberg
University,theUniversityofGöttingen,theUniversityofHongKong,andWilliam
&Mary,weshowinanewbookcalledBankingonBeijingthatwhencountries
votewithChinaintheUNGeneralAssembly,theyarerichlyrewarded.Our
statisticalmodelresultsimplythatifalow-incomeormiddle-income
governmentchoosestoincreasethealignmentofitsUNGAvotingwithChina
byjust10%,itcanexpecttoseea276%increaseinaidandcredit(ODAand
OOFcommitments)fromBeijing,onaverage(Dreheretal.2022).60
60WethankAxelDreher,AndreasFuchs,AustinStrange,andMikeTierneyforgeneratingandsharingsupplementaryevidencederivedfromastatisticalmodelinthefifthchapterofBankingonBeijing.59Tobesure,Chinaisnottheonlymajorpowerthathasusedforeignaidandcredittoinfluencetheforeignpolicypositionsofdevelopingcountries(AlesinaandDollar2000;KuziemkoandWerker2006;VreelandandDreher2014;Rose2018).
29
Figure
1.14:
China
versus
the
U.S.:
Public
approval
ratesFigure1.14
Notes:AveragepublicapprovalratingsforChinaandtheU.S.areweightedbycountrypopulation.The
constructionofthisvariableisdescribedingreaterdetailinBox1b.
30
Figure
1.15:
China
versus
U.S.:
Media
sentimentFigure1.15
Notes:AveragemediasentimentratingsforChinaandtheU.S.areweightedbycountrypopulation.The
constructionofthisvariableisdescribedingreaterdetailinBox1b.
31
Figure
1.16:
China
versus
the
U.S.:
UN
voting
alignmentFigure1.16
Notes:AverageUNGAvotingalignmentscoresforChinaandtheU.S.areweightedbycountrypopulation.
TheconstructionofthisvariableisdescribedingreaterdetailinBox1b.
Figure1.17talliestheannualnumberofsoftpower“gains”and“losses”that
Chinaexperiencedvis-à-vistheU.S.onacountry-by-countrybasisbetween
2014and2021.61Onallthreemeasuresofsoftpower(publicopinion,media
sentiment,andelitesupport),Chinahasexperiencedmorelossesthangains
vis-à-vistheU.S.since2014(thefirstfullyearofBRIimplementation).Public
opinioninthedevelopingworldhasmovedinaparticularlyunfavorable
directionforBeijing.DuringtheearlyBRIperiod,39%ofthecountry-levelpublic
61Figure1.17presentthepercentagesofLICsandMICsinwhichChinaexperiencedrelativegainsorlossesinpopularsupport,mediasentiment,andUNGAvotingalignmentvis-à-vistheU.S.Thepercentagesarereportedovertwotimeperiods:earlyBRI(2014-2017)andlateBRI(2018-2021).Tomeasuretherelativegainsorlossesinpopularsupport,wefollowathree-stepcalculationforeachcountry:(1)calculatethedifferencebetweenthepublicapprovalratingforChinainagivenyearandtheprioryear;(2)calculatethedifferencebetweenpublicapprovalratingfortheU.S.inagivenyearandtheprioryear;and(3)calculatethe“doubledifference”between(1)and(2)todetermineifChinaexperiencedagreatergainorlossinpublicsupportthantheU.S.inthesamecountry-year.ForrelativegainsandlossesinmediasentimentandUNGAvotingalignment,thesamethree-stepcalculationwasfollowedusingtheaveragemediasentimentscoreforeachcountry-yearfromtheGDELT1.0EventDatabase(relatedtogovernmentactorsfrommainlandChinaortheU.S.)andtheaverage“idealpointdistance”estimatebetweeneachcountryandChina(ortheU.S.)inagivenyear.
32
opinionchangesthatChinaexperiencedwererelativelossesratherthanrelative
gains(i.e.,publicopiniontowardtheU.S.improvedatafasterratethanpublic
opiniontowardChina,orpublicopiniontowardtheU.S.declinedataslower
ratethanpublicopiniontowardChina).However,duringthelateBRIperiod,this
figureshotupto66%.By2021,nearly85%ofthecountry-levelpublicopinion
changesthatChinaexperiencedwererelativelossesratherthanrelativegains
(seeFigureA8intheAppendix).Overtime,risinglevelsofpublicantipathy
towardChinaandexpandingpopularsupportfortheU.S.havewidened
Washington’ssoftpoweradvantageoverBeijing.Thebattleforheartsandminds
inthedevelopingworldwaseffectivelyatoss-upduringtheearlyyearsofthe
BRI: BeijingandWashingtonachievedasimilarnumberofpublicopiniongains
andlossesonacountry-by-countrybasis.62However,duringthelateBRIperiod
(2018-2021),Beijing’slossesoutnumbereditswins—byasignificantmargin.63
Figures1.18and1.19suggestthatBeijinghassufferedlessacutepublicopinion
andmediasentimentlossesinBRIparticipantcountries,whichisconsistentwith
newresearchontheinternationalimage-enhancingeffectsofChineseaidand
credit(Wellneretal.forthcoming,2023;BrazysandDukalskis2019).64However,
eveninBRIparticipantcountries,thetrendlineshavemovedinadirectionthat
shouldprovidecoldcomforttoBeijing.Figure1.20providesadditionalgrounds
forconcern,sinceitdemonstratesthatgoverningelitesinBRIparticipant
countriesaretakingforeignpolicypositionsthatareincreasinglyoutof
alignmentwiththoseofChina.Souringmediasentimentanddeclininglevelsof
publicsupportmaybemakingitmoredifficultforgoverningelitestomaintain
closerelationswithBeijing.
64BRIparticipantcountriesincludethosecountriesthathavesignedMOUswithChinatojoinitsBeltandRoadInitiative.AcountryisassignedtotheBRIparticipantcohortintheyearitsignedtheMOUandeveryyearthereafter.63DuringthelateBRIperiod(2018-2021),39%ofthepublicopinionchangesthatChinaexperiencedatthecountry-yearlevelvis-à-vistheU.S.wererelativegainsand61%wererelativelosses(seeFigureA8).62DuringthefirstthreeyearsoftheearlyBRIperiod(2014-2016),55%ofthepublicopinionchangesthatChinaexperiencedatthecountry-yearlevelvis-à-vistheU.S.wererelativegainsand45%wererelativelosses.OverthefullearlyBRIperiod(2014-2017),61%ofthepublicopinionchangesthatChinaexperiencedatthecountry-yearlevelvis-à-vistheU.S.wererelativegainsand39%wererelativelosses(seeFigureA8).
33
Figure
1.17:
China’s
soft
power
losses
vis-à-vis
the
U.SFigure1.17
Notes:China’srelativegainsandlossesinpopularsupport,mediasentiment,andUNGAvotingalignment
vis-à-vistheU.Sarecalculatedonacountry-by-countryandyear-by-yearbasis.Theconstructionofthese
measuresaredescribedingreaterdetailinBox1bandfootnote61.
34
Figure
1.18:
Public
support
for
China:
BRI
countries
vs.
non-BRI
countriesFigure1.18
Notes:AveragepublicapprovalratingsforChinaareweightedbycountrypopulation.The
constructionofthisvariableisdescribedingreaterdetailinBox1b.
35
Figure
1.19:
Media
sentiment
toward
China:
BRI
countries
vs.
non-BRI
countriesFigure1.19
Notes:AveragemediasentimentratingsforChinaareweightedbycountrypopulation.Theconstructionof
thisvariableisdescribedingreaterdetailinBox1b.
36
Figure
1.20:
UN
voting
alignment
with
China:
BRI
countries
vs.
non-BRI
countriesFigure1.20
Notes:AverageUNGAvotingalignmentscoresforChinaareweightedbycountrypopulation.The
constructionofthisvariableisdescribedingreaterdetailinBox1b.
Section6:Thebalancingactofportfolioriskmanagement
ManagingtheriskprofileofChina’soverseasdevelopmentprogramisa
multifacetedchallenge.First,Chinahasahighandrapidlyrisinglevelof
exposuretonon-performingloansinLICsandMICs.Itsexposuretodistressed
debtisheavilyconcentratedamongthemostimportantBeltandRoad
participants,whichraisesthequestionofwhetheritseffortstogetrepaidwill
conflictwithitsforeignpolicy(andsoftpower)objectives(HancockandHill
2022).65Thecruxofthematter,asdescribedbyaseniorofficialfromChina’s
centralbank,isthat“debtreductioncanshapeChina’simageasaresponsible,
majorglobalplayer,[but]itoftenleadstobig[financial]lossesandevencausesa
moralhazard—debtorsmayrefusetomakerepaymentsiftheyknow[thatwe
maywriteofftheirdebts]”(Chengjun2021).Second,alargeshareofChina’s
65Beijing’sexposuretodistresseddebtinLICsandMICsalsoraisesthequestionofwhethercertainborrowersare“toobigtofail”andwhetherChinesecreditorswillstrictlyadheretotheprinciplethateveryloanmustberepaidinfull(anissueweaddressatgreaterlengthinChapter2).
37
infrastructureprojectportfoliohassignificantESGriskexposure,andwhilethere
aremanygoodreasonstode-risktheportfoliobyadoptingstrongersafeguards,
doingsocouldresultinChinalosingacompetitiveedgeintheglobal
infrastructuremarket.Chinaisnolonger“theonlygameintown”forcountries
seekingexternalsourcesofinfrastructurefinancing,andtheothermajorplayers
inthemarketalreadyofferhigh-qualityinfrastructureprojectsthatbenefitfrom
strongESGsafeguards.ChinahashistoricallyoutcompetedG7countriesand
themultilateraldevelopmentbanksontwokeydimensions:speedand
convenience.Ifthesedifferentiatorsarenolongerapplicable,Chinamayfacea
differenttypeofproblem:insufficientdemandforChineseinfrastructure
financing.Third,risinglevelsofpublicantipathytowardBeijingandasouring
mediaenvironmenthaveleftChinaincreasinglydependentonthegoodgraces
ofpoliticalleadersintheGlobalSouth.Onewaytoaddressthischallengewould
betodustoffanoldplaybookandcatertotheparochialinterestsofhost
countrypoliticiansbyplyingthemwithlavishspendingonpetprojects—like
presidentialpalaces—andamenitiesinmajorurbancenters(suchasmuseums,
theaters,conventioncenters,andstadiums).Analternativeapproachwouldbe
todoubledownonpublicdiplomacyefforts—suchasscholarships,sistercity
initiatives,andcontent-sharingpartnershipswithlocalradiostations,television
channels,andnewspapers—togeneratemorefavorablemediacoverageand
influencepublicsentimentinBRIparticipantcountries.
Beijingisclearlyawareoftheneedtopivotandassumea“fire-fighting”role.It
israpidlyrefocusingitstime,money,andattentionondistressedborrowers,
troubledprojects,andsourcesofpublicbacklashintheGlobalSouth.However,
alonger-termreinventionoftheBRIisalsounderway.AsBeijinglearnsfrom
pastmistakes,itisrecalibratingitslendingandgrant-givingpracticesand
makingeffortstofuture-proofitsflagship,globalinfrastructureinitiative.The
ambitionofBRI2.0remainspoorlyunderstood—andunderestimated—bythose
whomakeandshapepolicyinG7countries.Washington,London,Paris,Berlin,
Tokyo,Rome,andOttawaare,forthemostpart,stillformulatingpolicyonthe
basisofevidencefromtheBRI1.0era.
TheprimarygoalofthisreportistoexplainhowChinaishandlingitsnewroleas
aninternationalcrisismanagerintheshort-runwhileengaginginalonger-run
38
efforttofuture-prooftheBRI.Inthenextthreechapters,weattempttoanswer
thefollowingquestions:
●HowisChinacopingwiththerisingtideofdebtdistress?Whatmeasures
isittakingtoreduceitsexposuretonon-performingloans?
●IsChinasteppingupitsESGriskmitigationefforts?Ifso,where,when,
andhow?AreitsinfrastructureprojectswithandwithoutstrongESG
safeguardsfaringdifferentlyduringimplementation?
●HowdoesChinamanagereputationalrisk?Whatmeasuresdoesittaketo
preservegrassroots,media,andelitesupportinhostcountries?Are
Chinesedevelopmentfinanceinstitutionslearningfromtheirpast
mistakesandrecalibratingtheirpoliciesandpracticesinBRI“buyer’s
remorse”countries?
●WhatareChina’stolerancelevelsforrepaymentrisk,projectperformance
risk,andreputationalrisk?
However,wealsowanttointroduceresearchers,policymakers,andjournaliststo
the“artofthepossible”withthe3.0versionofAidData’sGCDFdataset.There
aremanyadditionalquestionsthatcannowbeansweredbecauseofthe
dataset’suniquelycomprehensivescopeandunprecedentedgranularity.66
Severalnewandimprovedfeaturesofthedatasetmeritspecialattention:
1.Donorandlendercoverage:The3.0datasetcapturesprojectsand
activitiesinLICsandMICssupportedby791officialsectordonorsand
lendersinChina.Italsoidentifiestheparticipationof1,225co-financing
institutions—includingWesterncommercialbanks,multilateral
developmentbanks,andOECD-DACdevelopmentfinanceinstitutions
thathavechosentocollaborateorcoordinatewithBeijing—inChinese
grant-andloan-financedprojectsandactivities.Anewfeatureofthe3.0
datasetistheinclusionoftwo“flag”variablesthatallowforeasy
66TableA1presentsabroadviewofhowthe3.0versioncomparestothe2.0versionoftheGCDFdataset.
39
identificationofprojects/activitiesthatinvolve(a)non-Chinesefinanciers
or(b)multilateralinstitutions.
2.Borrowerandrecipientcoverage:The3.0datasetidentifies5,037
receiving(borrowing)institutionsandcategorizeseachonebytype
(governmentagency,state-ownedbank,state-ownedcompany,special
purposevehicle/jointventure,intergovernmentalorganization,private
sector,etc.),countryoforigin(recipientcountry,China,orathirdcountry),
and,whenapplicable,role(directborrowerorindirectborrowerthrough
anon-lendingarrangement).Inrecognitionofthefactthatspecial
purposevehiclesandjointventuresplayanimportantroleinChina’s
overseaslendingprogramandoftenblurthelinesbetweenpublicand
privatedebt(Maliketal.2021;MalikandParks2021),wehavealso
addedtwonewvariablestothelatestversionofthedataset:ameasureof
theextentofhostgovernmentownershipandameasureoftheextentof
Chinesegovernmentownershipwhenevertheborrowinginstitution
(receivingagency)isaspecialpurposevehicleorjointventure.67The3.0
datasetalsoidentifies422institutions(“accountableagencies”)thathave
supportedChineseloan-financedprojectsandactivitiesbyproviding
repaymentguarantees,insurancepolicies,andcollateralwhichcanbe
seizedintheeventofdefault.68
3.Financialinstrumentcoverage:The3.0datasetallowsuserstoeasily
differentiatebetweenthe10,291grant-financedprojects/activitiesand
4,776loan-financedprojects/activities.However,giventhatBeijingrelies
onanincreasinglydiversesetofdebtinstrumentstofinanceitsoverseas
developmentprograminLICsandMICs,AidDatahasintroducedanew
loancategorizationschemeinthelatestversionofthedatasetthatallows
userstoisolatespecifictypesofloaninstruments,includingbutnot
limitedtobilateralloans,syndicated/clubloans,interest-freeloans,
68Consistentwiththe2.0versionoftheGCDFdataset,eachaccountableagencyinthe3.0datasetisstillcategorizedbytypeandcountryoforigin.However,unlikethe2.0versionoftheGCDFdataset,eachaccountableagencyinthe3.0datasetisalsocategorizedbyrole(guarantor,insuranceprovider,orcollateralissuer).67Thesenewvariablesarecalled“JV/SPVHostGovernmentOwnership”and“JV/SPVChineseGovernmentOwnership.”The3.0datasetcaptures851loancommitmentsworth$315billion(inconstant2021USD)toborrowinginstitutionsthatarecategorizedasspecialpurposevehiclesorjointventures.
40
governmentconcessionalloans,preferentialbuyer’scredits,public
investmentloans,balanceofpayments(BOP)loans,M&Aloans,working
capitalloans,inter-bankloans,refinancingloans,deferredpayment
agreements,andpre-exportfinancing(PxF)agreements.
4.Borrowingtermsandconditions:Thereisnootherpubliclyavailable
datasetofChina’soverseasloancommitmentswithglobalcoveragefrom
2000-2021thatidentifiesborrowingtermsandconditionsatthe
transactionlevel.The3.0datasetidentifies2,699interestrates,3,315
maturitylengths,1,854graceperiods,498commitmentfees,480
managementfees,and2,537grantelementsacross4,776loansinAfrica,
Asia,Oceania,EasternandCentralEurope,theMiddleEast,andLatin
AmericaandtheCaribbean.Italsoidentifies668loansbackedby
third-partyrepaymentguarantees,529loanssupportedbycredit
insurancepolicies,and1,015loansunderpinnedbyoneormoresources
ofcollateral.Threeadditionaldatasetfeaturesareworthnoting.First,in
lightofrecentchangestotheOECD’sgrantelementmethodof
measurement(seeSectionA-2)andthegrowingimportanceoftheWorld
Bank/IMF-basedmethodofmeasurementtodetermineifandwhen
governmentshavecompliedwiththenon-concessionalborrowinglimits
specifiedintheirWorldBankandIMFprograms,wenowprovidethree
differentgrantelementmeasures:onebasedontheOECD’scash-flow
methodology,onebasedonthetheOECD’sgrant-equivalent
methodology,andanotherbasedonthelatest(post-2013)World
Bank-IMFmethodology.69Second,inordertofacilitatemoreaccurate
calculationofthe“all-in”priceofChinesedebt,wehaveintroduced
variablesthatmeasure(a)thedefault(penalty)interestratethatappliesto
aloanintheeventofdefault(i.e.,non-paymentofprincipal,interest,or
feesontheirscheduledpaymentdates),and(b)thecostofcredit
insurance.70Third,tosupportfutureresearchondebtservicetoChina,we
70Forexample,Argentina’sMinistryofEconomyandPublicFinancesigneda$4,714,350,000syndicatedtermfacility(loan)agreementonAugust1,2014withBankofChina,ChinaDevelopmentBank,andICBCforthe1,740MWNéstorKirchnerandJorgeCepernicHydroelectricPowerPlantConstructionProject.The69TheOECD’scash-flowmethodologyassumesafixed,10percentdiscountrate.Itsgrant-equivalentmethodologyusesfixeddiscountratesthatdependontheincomeleveloftheborrowingcountry(9%forLDCsandotherLICsand6%forUMICs).TheWorldBank-IMFmethodologyassumesafixed,5percentdiscountrate.FormoreontheIMF’sDebtLimitsPolicy(DLP),seehttps://www.imf.org/en/Topics/sovereign-debt/debt-limits-policy.
41
haveincludedtwonewmeasuresinthe3.0dataset:thecalendardayon
whichtheborrowerwasoriginallyscheduledtomakeitsfirstloan
repaymentandthecalendardayonwhichtheborrowerwasoriginally
scheduledtomakeitslastloanrepayment.Wheneverpossible,wehave
alsodocumenteddisbursements,repayments,andamountsoutstanding
inthedataset’s“description”field.
5.Spatialandtemporalgranularity:The3.0datasetprovidesan
unprecedentedlevelofdetailonprojectcommencement
(implementationstart)datesandprojectcompletion(implementation
end)dates.Itidentifiesprecise,calendarday-levelcommencementdates
for11,286projects(backedbyfinancialcommitmentsworth$767billion)
andcalendarday-levelcompletiondatesfor11,542projects(backedby
financialcommitmentsworth$606billion).Bywayofcomparison,the2.0
versionoftheGCDFdatasetidentifiedcalendarday-level
commencementdatesfor5,539projects(backedbyfinancial
commitmentsworth$504billioninconstant2021USD)andcalendar
day-levelcompletiondatesfor6,061projects(backedbyfinancial
commitmentsworth$383millioninconstant2021USD).The3.0dataset
alsoprovidesdataontheoriginallyscheduledprojectcommencement
datesandcompletiondates,whichhaspavedthewayforthe
introductionoftwonewmeasures(“DeviationfromPlanned
ImplementationStartDate”and“DeviationfromPlannedCompletion
Date”)oftowhatdegreeprojectsran(orarerunning)aheadofschedule
orbehindschedule.71Anotherimportantvalueadditiontothe3.0dataset
isthelevelofgeographicaldetailregardingwhereprojectstakeplace.As
wedescribeingreaterdetailinGoodmanetal.(2023),for9,497projects
71The“DeviationfromPlannedImplementationStartDate”variablecapturesthedifferencebetweenthe“PlannedImplementationStartDate”andthe“ActualImplementationStartDate”whenvaluesarerecordedforbothvariables.Itcapturesthedifferenceasthenumberofcalendardays,wherebypositivevaluesrepresentcaseswheretheprojectstartedimplementationaheadofscheduleandnegativevaluesrepresentcaseswheretheprojectstartedimplementationbehindschedule.The“DeviationfromPlannedCompletionDate”variablecapturesthedifferencebetweenthe“PlannedCompletionDate”andthe“ActualCompletionDate”whenvaluesarerecordedforbothvariables.Itcapturesthedifferenceasthenumberofcalendardays,wherebypositivevaluesrepresentcaseswheretheprojectwascompletedaheadofscheduleandnegativevaluesrepresentcaseswheretheprojectwascompletedbehindschedule.loanagreement,whichwaslateramendedonJanuary27,2015andagaininmid-2022,specifiesadefault(penalty)interestrateof1.5%.Asacreditenhancement,theborrowerpurchasedabuyer’screditinsurancepolicyfromSinosureworthapproximately7.1%ofthefacevalueoftheloan($502,976,000)(seeProjectID#59723,59724,37002inthe3.0versionoftheGCDFdataset).
42
thathavephysicalfootprintsorinvolvespecificlocations,the3.0dataset
extractspoint,polygon,andlinevectordataviaOpenStreetMapURLs
andprovidesacorrespondingsetofGeoJSONfilesandgeographic
precisioncodes.7272%(6,919)oftheseprojectsinclude“precise”or
“approximate”geocodes;theremaining28%(2,578projects)are
measuredatanadministrativeunitlevel.73Measuringthespatio-temporal
rolloutofprojectimplementationwithahighlevelofprecisionis
importantbecauseitcreatesnewopportunitiestoidentify
cause-and-effectrelationshipsinrigorousways.74
6.Sectoralcoverage:The3.0datasetsystematicallytracksprovisionof
officialfinancialflowsfromChinatoLICsandMICsacrossallsectors.
Everyproject/activityisassigneda3-digitsectorcodebasedonOECD
definitionsandmeasurementcriteria.This“methodologicalcrosswalk”is
importantbecauseitallowsforcross-financiercomparisons—atglobal,
regional,national,andsubnationalscales—sincemostofficialsourcesof
internationaldevelopmentfinance(includingOECD-DACmembersand
multilateralinstitutions)usethesamecriteria.Italsofacilitatesanalysisof
sectoralpatternsandtrendsoverspaceandtime.
7.Qualitativedetail:The3.0datasetprovidesdetailedprojectnarratives
that“tellthestory”ofeachprojectinthe“description”field.Theaverage
lengthofeachprojectnarrativeincreasedfrom144wordsin2.0dataset
to169wordsinthe3.0dataset.Whereastheprojectnarrativesinthe2.0
datasetconsistedof1.93millionwords(roughlythesamenumberof
wordsonewouldfindin19full-lengthbooks),theprojectnarrativesinthe
3.0datasetconsistof3.48millionwords(roughlythesamenumberof
wordsonewouldfindin34full-lengthbooks).75Aswewilldemonstratein
75Atypical,full-lengthbookincludes100,000words.74Tobetterunderstandhowhighlyprecisedataonthespatio-temporalrolloutofChinesegrant-andloan-financedprojectsmakeitpossibletoestimatethecausaleffectsofsuchprojectsonintendedandunintendedoutcomes,seeDreheretal.(2019,2022),Martyetal.(2019),Blairetal.(2022),Baehretal.(forthcoming),IsakssonandKotsadam(2018a,2018b),Isaksson(2020),Martoranoetal.(2020),Iacoellaetal.(2021),Maliketal.(2021),Bluhmetal.(2020),Anaxagorouetal.(2020),Wellneretal.(forthcoming),andAsmusetal.(forthcoming).73Aprojectwith“precise”geocodesisoneforwhichwehavehighlypreciseboundariesoftheproject’sgeofeature(s).Aprojectwith“approximate”geocodesisoneidentifiedwithina5-kmradiusofthepreciseboundariesoftheproject’sgeofeature(s).Formoredetails,seeGoodmanetal.(2023).72UserswhowishtoconductanalysisathigherlevelsofspatialaggregationcanfindtheADM1s(provinces)andADM2s(districts)thatcorrespondtotheseprojectlocationsinthe3.0versionoftheGCDFdataset.
43
Chapter3,thesenarrativesareusefulinthattheydocumenttherisksand
challengesthataroseduringprojectdesignandimplementation(e.g.,
bankruptcies,scandals,protests,laborstrikes,andcriminalinvestigation)
andhowfunding,receiving,implementing,andaccountableinstitutions
respondedtotheserisksandchallenges.
8.Scale,diversity,quality,andtransparencyofsourcing:The3.0dataset
wasassembledwith147,703sources(including99,393uniquesourcesin
morethanadozenlanguages,ofwhich51,597areofficialsources).By
wayofcomparison,the2.0datasetwasassembledwith91,356sources
(including63,464uniquesourcesinmorethanadozenlanguages,of
which34,075wereofficialsources).Whereastheaveragerecordinthe1.0
datasetwasbasedupon3.6sourcesandtheaveragerecordinthe2.0
datasetwasbasedon6.8sources,theaveragerecordinthe3.0datasetis
basedupon7.0sources.87%oftherecordsinthe3.0datasetare
underpinnedbyatleastoneofficialsource.Toexposeourcodingand
categorizationdeterminationstopublicscrutinyandpromotereplicable
researchfindings,wediscloseallofthesourcesthatwereusedto
constructthedatasetattheindividualrecordlevel,includinghundredsof
unredactedgrant,loan,debtforgiveness,anddebtrescheduling
agreements(thatAidDatahasneverpreviouslypublished).
44
Chapter2:TheRoadtoRepaymentforthe
World’sLargestOfficialDebtCollector
Section1:DebunkingthemyththatBeijing’soverseaslendingprogramhascollapsed
TheconventionalwisdomisthatBeijinghasrespondedtothedeteriorating
performanceofitsoverseaslendingprogramviaretreat.EricOlander,
co-founderoftheChina-GlobalSouthProject(CGSP),recentlysummarizedthe
stateofexpertopinion,notingthat“[t]herewasatimewhenChineselendingto
developingcountriesrivaledtheWorldBank”but“[t]hosedaysarenowlong
goneasChineseoverseasdevelopmentlendinghasbeenonasteady
downwardtrajectory”(Olander2023).ScottKennedyoftheCenterforStrategic
andInternationalStudies(CSIS)toldForeignPolicymagazineearlierthisyear
thattheBRIwasa“shadowofitsformerself”(Lu2023).76ElliotWilsonof
Euromoneymagazineclaimsthat“Chineseoverseaslendingtothedeveloping
worldhascollapsed”(Wilson2022).
However,theconventionalwisdomismostlywrong.Theempiricalevidencethat
wepresentinthischapterpaintsadifferentpicture—oneinwhichChinais
behavingmorelikeaninternationalcrisismanagerthanacountryadmiringits
problemsandstickingitsheadinthesand.Beijingisrebalancingitsoverseas
lendingportfoliobyadoptingawide-rangingsetofde-riskingmeasures.Itis
rampingdowntheprovisionoflong-term,dollar-denominatedbilateralloansto
sovereignborrowersforpublicinvestmentprojects,whileatthesametime
rampinguptheprovisionofloansthatareRMB-denominated,short-or
medium-terminnature,unrelatedtopublicinvestmentprojects,and/or
involvingmultipleChineseand/ornon-Chinesebanks.Itisratchetingdownits
useofthepolicybanks(ChinaDevelopmentBankandChinaEximbank),whileat
thesametimeratchetingupitsuseofthecentralbank(People’sBankofChina),
state-ownedcommercialbanks(suchasIndustrialandCommercialBankof
China,BankofChina,andChinaConstructionBank),andsyndicatedloan
76ThesameForeignPolicymagazinearticlenotesthat“expertssayChina'slendingforBRIprojectshasplummeted”(Lu2023).
45
arrangementswithnon-Chinesebanks(suchasStandardChartered,BNP
Paribas,theInternationalFinanceCorporation,andtheEuropeanBankof
ReconstructionandDevelopment).Itisalsoputtinginplacestrongersafeguards
toprotectitselffromborrowersthatpresenthighlevelsofrepaymentrisk.So,
anotherwayofreadingtheevidentiaryrecordisthatBeijingisbehavinglikea
yield-maximizinginvestmentportfoliomanager(seeBox2b).
Thede-riskingmeasuresthatBeijingisimplementingarepoorlyunderstood
amongthosewhomakeandshapepolicyinWesterncapitals,whichmeansthat
Washington,London,andBrusselsincreasinglyruntheriskofcompetingwitha
versionoftheBRIthatnolongerexists—BRI1.0ratherthanBRI2.0.Ouraimin
thischapteristosettherecordstraight.Wedosobyfirstdebunkingthe
popularmyththatChina’soverseaslendingprogramhasexperiencedatotal
collapse.Therestofthechapterisfocusedondemystifyingthepurpose,nature,
andscopeofBeijing’songoingBRIreboot.
The“Beijinginretreat”storylinehasgainedtractionbecauseitisconsistentwith
thetoplineChineselendingcommitmenttrendsthatarerecordedinseveral
publiclyavailabledatabases,suchastheChina’sOverseasDevelopmentFinance
Database,theChineseLoanstoLatinAmericaandtheCaribbeanDatabase,the
ChineseLoanstoAfricaDatabase,theChina’sGlobalEnergyFinanceDatabase,
theChinaOverseasFinanceInventoryDatabase,andtheWorldBank’s
InternationalDebtStatistics(IDS).Considerthefollowingclaimsthathavebeen
madebaseduponthesedatasources:
●InDecember2020,BostonUniversity’sGlobalDevelopmentPolicyCenter
releasedtheChina’sOverseasDevelopmentFinance(CODF)Database,
whichprovidesdataonChinaEximbankandCDBlendingcommitments
togovernments,inter-governmentalbodies,majoritystate-ownedentities
andminoritystate-ownedentitieswithsovereignguarantees.The
researchteamresponsibleforthedatabaseclaimedatthetimethatit
providedevidenceof“China’soverseasdevelopmentfinance
commitments”decliningfrom$75billionin2016to$3.9billionin2020
(GallagherandRay2020).Then,inJanuary2023,thesameresearchteam
usedanupdatedversionoftheCODFdatabasetoarguethatChina’s
46
overseasdevelopmentfinancecommitmentsremainedatan
exceptionallylowlevel($3.6billion)in2021(Ray2023).
●TheWorldBank’sInternationalDebtStatistics(IDS)showasharpdecline
inofficialsectorloancommitmentsfromChinatopublicsector
institutionsinlow-incomeandmiddle-incomecountries—from$31.5
billionin2016to$7billionin2021.
●AccordingtotheChineseLoanstoLatinAmericaandtheCaribbean
DatabasejointlyproducedbyInter-AmericanDialogueandBoston
University’sGlobalDevelopmentPolicyCenter,“theLACregionsawa
precipitousdeclineinloansfromCDBandEximbankbetween2015and
2020,whenlendingceasedaltogether”(MyersandRay2023:1).They
alsoconcludedthatChinesedevelopmentfinancecommitmentstothe
LACregionremainedstubbornlylowin2021,identifyingonlyone($204
million)CDBloancommitmentinthatyear.77
●AccordingtotheChineseLoanstoAfrica(CLA)database,whichwas
initiallydevelopedbytheChina-AfricaResearchInitiativeattheJohns
HopkinsSchoolofAdvancedInternationalStudies(SAIS-CARI)andisnow
maintainedbyBostonUniversity’sGlobalDevelopmentPolicyCenter,
ChineselendingcommitmentstoAfricangovernmentsandstate-owned
entitiesplungedfromahighof$28billionin2016toalowof$1.9billion
in2020.Inasummaryoftheirfindings,theresearchersresponsibleforthe
CLAdatabaseconcludethat“[c]omparedtopreviousyears,thenumber
ofloansandthetotalvalueofloancommitmentsdecreaseddramatically
in2020”(Hwangetal.(2022:2).78
Thelatest(3.0)versionofAidData’sGlobalChineseDevelopmentFinance
(GCDF)datasetpaintsadifferentpictureofChina’soverseasdevelopment
78AnupdateoftheCLAdatabasewaspublishedatthetimethatthisreportwasgoingtopress.ItshowsacontinueddeclineinChineselendingcommitment—to$1.2billionin2021and$994.4millionin2022(Mosesetal.2023).77AidDatahascategorizedthis$204million(EUR175million)CDBloanasaninformalpledgeoffinancialsupportratherthanaformalloancommitment,basedonevidencethatitgatheredviadirectcorrespondencewiththeDebtManagementDivisionofTrinidadandTobago'sMinistryofFinance.SeeID#95549inthe3.0versionofAidData’sGCDFdatasetformoredetails.
47
financeportfolio(seeFigure1.2inChapter1).ItdemonstratesthatBeijingisstill
thesinglelargestsourceofinternationaldevelopmentfinanceintheworld,with
annualODAandOOFcommitmentstoLICandMICsnowhoveringaround$80
billion.AlthoughitprovidesevidenceofChina’sannualinternational
developmentfinancecommitmentsfallingbetween2016and2020,italso
showsanincreasein2021(returningtoalevelthatisroughlycomparabletothe
firstfullyearofBRIimplementation).
Figure
2.1:
Official
sector
lending
commitments
from
China
to
LICs
and
MICs
by
sourceFigure2.1
Notes:Figure2.1comparesthetotalsizeofofficialsectorlendingcommitmentsfromChinatoLICsand
MICsacrossthreedatasets:the3.0versionofAidData’sGCDFdataset,BostonUniversity’sChina’s
OverseasDevelopmentFinanceDatabase(CODF),andtheWorldBank’sInternationalDebtStatistics
(representingcommitmentstoofficialcreditorsinChina).
Figure2.1providesacomparisonofofficiallendingcommitmentsfromChinato
borrowersinlow-incomeandmiddle-incomecountries,asmeasuredbythree
differentsources:the3.0versionofAidData’sGCDFdataset,theWorldBank’s
IDS,andtheCODFdatabaseproducedbyBostonUniversity'sGlobal
DevelopmentPolicyCenter.79AidDatacaptureslendingcommitmentsworth
79Toensurecomparability,weconverttheIDSdataseriesandtheCODFdataseriesintoconstant2021USD.
48
$1.28trillionbetween2000and2021,whileCODFandIDScapturelending
commitmentsworth$605billionand$378billion,respectively.80Whereas
AidDatarecordsa45%declineinlendingcommitmentsbetween2016and
2021,CODFandIDSrecordsubstantiallylargerdeclines—96%and78%,
respectively.CODFandIDSrecord$3.7billionand$7.1billion,respectively,in
newofficiallendingcommitmentsfromChinain2021.AidDatacaptures$75.5
billioninnewofficiallendingcommitmentsfromChinainthesameyear.
ThereareseveralreasonswhytheestimatesfromAidData’sGCDFdataset,BU’s
CODFdataset,andtheWorldBank’sIDSarewidelydivergent.First,althoughall
threesourcesprovidedataonpublicandpubliclyguaranteeddebt(PPG)debt
fromChinesestate-ownedcreditors,81theyprovidedifferentlevelsof
geographicalcoverage.82AidData’sGCDFdatasetcovers126countries,while
BU’sCODFdatasetcovers96countriesandtheWorldBank’sIDScovers89
countries.83Second,therearetemporalcoveragedifferences:whereasAidData’s
GCDFdatasetandtheWorldBank’sIDSprovidedatafor22commitmentyears
(2000-2021),BU’sCODFdatasetprovidesdatafor14commitmentyears
(2008-2021).Third,inthesubsetofLICsandMICsforwhichCODF,IDS,and
GCDFdataareavailable,therearedifferencesinhowmuchPPGdebtfrom
Chinesestate-ownedcreditorsiscaptured.TableA15demonstratesthat,inthe
subsetofLICsandMICsforwhichCODForIDSdataareavailable,AidData’s
GCDFdatasetcaptures$947billionoflendingcommitments84fromofficial
84The$947billionfigureexcludesshort-term“rollover”facilities.Whenshort-term“rollover”facilitiesareincludedinthetally,AidDatacaptureslendingcommitmentstoLICsandMICsthatqualifyasPPGdebtworth$1.09trillionbetween2000and2021.83TheIDSdatacaptureofficialsectorlendingcommitmentsfromChinato89low-incomeandmiddle-incomecountries(excludingthePeople’sRepublicofChina)from2000-2021.TheIDSdatadonotallowuserstodifferentiatebetweenthePeople’sRepublicofChina(“China”)andtheRepublicofChina(“Taiwan”).Assuch,weexcludeallloancommitmentsthat9additionalcountriesreportedduringtheyearswhentheymaintaineddiplomaticrelationswithTaiwan.82The3.0versionofAidData’sGCDFdatasetprovidescomprehensivecoverageofChineseODAandOOFcommitmentsacross165LICsandMICsbetween2000and2021.However,only134ofthesecountriessecuredloancommitmentsfromofficialsectorcreditorsinChinaoverthesametimeperiod.Anevensmallersubset(126countries)securedloancommitmentsbetween2000and2021fromofficialsectorcreditorsinChinathatqualifyasPPGdebt.81PPGdebtconsistsof(a)long-termexternalobligationsofpublicdebtors,includingthenationalgovernment,apoliticalsubdivision(oranagencyofeither),andautonomouspublicbodies;and(b)long-termexternalobligationsofprivatedebtorsthatareguaranteedforrepaymentbyapublicentity(WorldBank2000).80The$1.28trillionfigureexcludesshort-term“rollover”facilitiestorefinancematuringdebts.Whenshort-term“rollover”facilitiesareincludedinthetally,AidDatacaptureslendingcommitmentsfromChinaworth$1.44trillionbetween2000and2021.
49
sectorinstitutionsinChinatoLICsandMICsthatqualifyasPPGdebt.85Intotal,
BU’sCODFdatasetcaptures$605billionandtheWorldBank’sIDScaptures
$378billionoflendingfromofficialsectorinstitutionsinChinathatqualifiesas
PPGdebt.86Thesetoplinedifferencesreflectwidelyvaryinglevelsoflending
institution,borrowinginstitution,anddebtinstrumentcoverage(whichwe
discussatgreaterlengthbelow).TableA15callsattentiontothefourthandfinal
difference:neitherBU’sCODFdatasetnortheWorldBank’sIDSprovideany
coverageoflendingcommitmentsfromofficialsectorinstitutionsinChinato
LICsandMICsthatqualifyasnon-PPGdebt.AidData’sGCDFdataset,by
contrast,captures$333billionthatdoesnot(clearly)qualifyasPPGdebt:$67
billionofpotentialpublicsectordebt(i.e.,debtcontractedbyaminority
state-ownedinstitutionintheborrowingcountrywithoutapublicsector
repaymentguarantee),$216billionofprivatesectordebt,and$50billionthatis
notallocableduetoalackofsufficientinformationforcategorization
purposes.87
Oneadditionalsourceofinformationisworthconsidering:newlypublisheddata
fromtheBankofInternationalSettlements(BIS)ontotaloutstandingcreditfrom
ChinesebankstoLICsandMICsfrom2015to2021(seeBox2a).Thesedataare
particularlyvaluablebecausetheyarebasedondirectreportingfromChinese
banksabouttheircross-borderclaims.AlthoughtheBISdoesnotcurrentlyallow
foritsChineselendingdatatobedisaggregatedbyborrowercountries,onecan
derivelower-boundandupper-boundestimatesoftotaloutstandingcreditfrom
ChinesebankstoLICsandMICsbasedonnewresearchbyCeruttietal.(2023).
AswedescribeatgreaterlengthinBox2a,lower-boundestimatesbasedonthe
BISdataindicatethattotaloutstandingcreditfromChinesebankstoLICsand
87The“LevelofPublicLiability”variableinthe3.0versionofAidData’sGCDFdatasetcapturestheextenttowhichthehostgovernmentmayeventuallybeliablefordebtrepayment.Eachloanrecordisassignedtooneofsixcategories:(1)Centralgovernmentdebt,(2)Centralgovernment-guaranteeddebt,(3)Otherpublicsectordebt,(4)Potentialpublicsectordebt,(5)Privatedebt,or(6)Unallocable.ThesumofthefirstthreecategoriesisequivalenttoPPGdebt.Thefourthcategorycapturesloanstospecialpurposevehicles(SPV)orjointventures(JV)thatareminority-ownedbyoneormorepublicsectorinstitutionsinthehostcountryandthatdonotbenefitfromacentralgovernmentrepaymentguaranteeorarepaymentguaranteefromastate-ownedentityotherthanthecentralgovernmentinthehostcountry.Formoreinformationaboutthe“LevelofPublicLiability”variable,seeSectionA-5intheAppendix.86Foracountry-by-countrycomparisonofAidDataandIDSmeasuresofPPGdebtexposuretoChina(thatarenormalizedbyhostcountryGDP),seeTableA16.85Inthe96countriesthatarecoveredbytheCODFdataset,BUcaptures$605billionandAidDatacaptures$824millionoflendingcommitments(excludingshort-term“rollover”facilities)between2008and2021fromofficialsectorinstitutionsinChinathatqualifiesasPPGdebt.Whenshort-term“rollover”facilitiesareincludedinitstally,AidDatacaptures$987millioninthesamesetofcountries.
50
MICsincreasedinnominaltermsfrom$644billionin2015to$1.16trillionin
2021.Upper-boundestimatesbasedontheBISdataindicatethattotal
outstandingcreditfromChinesebankstoLICsandMICsincreasedinnominal
termsfrom$878billionin2015to$1.58trillionin2021.AidData’sestimatesof
thecumulativesizeofChina’soverseaslendingprograminLICsandMICsare
consistentwiththemoreconservative(lower-bound)BIS-basedestimates.88
Accordingtothe3.0versionoftheGCDFdataset,China’scumulativeoverseas
lendingcommitmentsincreasedfrom$620billionin2015to$1.03trillionin
2021(innominalUSD).89NeitherAidData’sGCDFdatasetnortheBIS-based
estimatessupporttheargumentthatChina’soverseaslendingprogramhas
experiencedatotalcollapse.
Box2a:WhatisthetruescaleofChina’soverseaslendingprogram?
TheoverallsizeofChina’soverseaslendingprogramisasubjectofongoingdebateandcontroversy.Hornetal.(2019)conductedpioneeringwork,producinga$393billionestimateoftotaloutstandingcross-bordercreditfromChinesebankstoLICsandMICborrowersin2017.Atthetime,theirestimatewascriticizedbytheIMFasbeingexaggerated(IMF2020).BräutigamandAcker(2020)alsopublishedacritique,concludingthatthey“agree[d]withtheIMF.”Hornetal.(2020a)thenissuedaresponsetotheircritics,arguingthat“ournumbersaresubstantiallybelowcomparisonfigures[fromthePBOCandotherofficialsources]andlikelyalowerboundestimateofthetrueextentofChineseoverseaslending.”90Theyexplainedthat“[d]espiteourbesteffortstomergedatafrommultiplesources,westillmisssubstantialamountsofChineseoverseaslending.”
TherewasalsoanimportantdevelopmentaroundthetimethattheHornetal.(2019)studywaspublished:ChinabeganreportingtotheBankofInternationalSettlements(BIS)onthecross-borderclaimsofitsbanks.91Atthetime,theLocationalBankingStatistics(LBS)publishedbytheBISindicatedthattotaloutstandingcreditfromChinesebankstooverseasborrowerswas
91AlthoughChinahasjoinedthelistofcountriesreportingtotheBIS,itsdata(unlikethedataofmanyofotherBISreportingcountries)arenotmadepubliclyavailableonabilateralbasis.90Atthetime,China’sStateAdministrationofForeignExchange(SAFE)hadpublished(2017)dataon“China’sInternationalInvestmentPosition”thatidentified$637billionoftotaloutstandingcross-bordercredit.By2021,thisfigureroseto$988billion.Seehttps://www.safe.gov.cn/en/2018/0928/1459.html.89ThesenominalUSDfiguresexcludeshort-term“rollover”facilitiestorefinancematuringdebts.Ifshort-term“rollover”facilitiesareincludedinthetally,China’scumulativeoverseaslendingcommitmentsincreasedfrom$630billionin2015to$1.17trillionin2021(innominalUSD).Ingeneral,weprefertoreportChina’soverseaslendingcommitmentsinconstant2021USD.However,sincetheBISdataarerecordedinnominalUSD,anapples-to-applescomparisonwiththe3.0versionofAidData’sGCDFdatasetrequiresuseofnominalUSD.InTable2.1,wereportChina’scumulativeoverseaslendingcommitmentstoLICsandMICsfrom2015-2021inbothnominalandconstant2021USD.88Akeycaveat,asweexplaininSectionA-4,isthattheBISdataarerepresentedasamountsoutstanding,whichiseffectivelyequivalenttocumulativedisbursementsminuscumulativerepayments(i.e.,creditstocksratherthancreditflows).Consequently,cumulativelendingcommitmentsusuallyexceedamountsoutstanding.
51
morethandouble($919million)thesizeoftheHornetal.(2019)estimate.92Sincethen,additionalBISreportingbyChinaandnewresearchbyCeruttietal.(2023)hasmadeitpossibletogenerateupdatedestimatesoftotaloutstandingcreditfromChinesebankstoborrowersinLICs,MICs,high-incomecountries(HICs),andotheroverseasjurisdictions.93InSectionA-4,weprovideastep-by-stepdescriptionofhowtheseBIS-basedmeasuresoftotaloutstandingcreditfromChinesebankstooverseasborrowersarederived.WealsodiscussanumberofimportantcaveatsandconsiderationsregardingtheBIScross-borderlendingdata.
Theestimates,whichwereportinTable2.1,demonstratethatChina’stotaloutstandingcredit—measuredinnominalterms—toborrowersinLICs,MICs,HICs,andotheroverseasjurisdictionssoaredfrom$1.45trillionin2015to$2.63trillionin2021.Conservative,lower-boundestimatesbasedonBISreportingalsoindicatethattotaloutstandingcreditfromChinesebankstoLICandMICborrowerseffectivelydoubledinnominaltermsbetween2015and2021—from$644billionto$1.16trillion.ThismeasureofthestockofLICandMICdebttoChinesebanksisremarkablysimilartoourownmeasureofthestockofLICandMICdebttoChinesebanks.94Accordingtothe3.0versionofAidData’sGCDFdataset,China’scumulativeoverseaslendingcommitmentstoLICandMICsincreasedinnominaltermsfrom$620billionin2015to$1.03trillionin2021.95
Table2.1providesseveraladditionalinsights.OneisthatChina’stotaloutstandingcredittoLICs,MICs,andHICssoaredfrom$1.45trillionin2015to$2.63trillionin2021.AnotheristhattotaloutstandingcreditfromChinesebankstoHICborrowerseffectivelydoubledoverthesamesix-yearperiod.Between2015and2021,thisfigurerosefromapproximately$330billionto$600billion.AthirdisthattotaloutstandingcreditfromChinesebanksto“other”borrowersincreasedfrom$314.3billionin2015to$568.3billionin2021.
WhiletheLBSdatafromtheBISareextremelyvaluableforcross-validationpurposes(sincetheyprovidecredibleestimatesoftotaloutstandingcreditfromChinesebankstoLIC,MICs,andHICs),theydonotmakeitpossibletotrackdisbursements,repayments,andamountsoutstandingonaloan-by-loanbasis.AidDatarecentlylaunchedanewdatacollectioninitiativeandresearchprogramtoaddressthismajorevidentiarygap.Forthetimebeing,inthe3.0versionoftheGCDFdataset,wehavedocumenteddisbursements,repayments,andamountsoutstandinginthe“description”fieldforasubsetofcountries.However,inthefuture,weintendtopublishloan-leveldataondisbursements,repayments,andamountsoutstandingforamorecompletesetofcountries.Wealsointendtomakethesedataavailableinamoreuser-friendlyformat.
95AccordingtoBU’sCODFdataset,China’scumulativeoverseaslendingcommitmentstoLICandMICsincreasedinnominaltermsfrom$308billionin2015to$498billionin2021.AccordingtotheWorldBank’sIDS,China’scumulativeoverseaslendingcommitmentstoLICandMICsincreasedinnominaltermsfrom$188billionin2015to$293billionin2021.94ThisBIS-basedmeasureofChina’slendingportfolioinLICsandMICsisbasedonamountsoutstanding.AidData’smeasureofChina’slendingportfolioinLICsandMICsisbasedoncumulativelendingcommitments,whichusuallyexceedamountsoutstanding.SeeSectionA-4.93BISclassifiesmostHICsas“AdvancedEconomies”(AEs)andmostLICsandMICsasEmergingMarketandDevelopingEconomies(EMDEs).Forthesakeofclarityandconsistency,weusetheLIC,MIC,andHICacronyms.FormoreontheBIScountryclassificationsystem,seeCeruttietal.(2023).92Onthispoint,seeZhouandCerutti2018.
52
Table
2.1:
Comparison
of
AidData-
and
BIS-based
estimates
of
China’s
international
lending
portfolioTable2.1
Comparison
of
AidData-
and
BIS-based
estimates
of
China’s
international
lending
portfolio
YearBISTotaltoLICsandMICs(Basedon60.4%Assumption)BISTotaltoLICsandMICs(Basedon44.31%Assumption)AidDataTotaltoLICsandMICsNominalUSD(Constant2021USD)BISTotaltoHICs(Basedon22.6%Assumption)BISTotaltoHICs(Basedon22.8%Assumption)BISTotalto"Other"Borrowers(Basedon21.6%Assumption)BISTotaltoOFCBorrowers(Basedon30.33%Assumption)BISTotaltoLICs,MICs,HICs,and“Other”Borrowers
2015878.7644.7620.7(823.4)328.8331.7314.3441.31,454.9
20161,046.1767.4721.4(942.2)391.4394.9374.1525.31,731.9
20171,197.7878.7805.7(1,039.3)448.2452.1428.3601.41,983.0
20181,348.3989.1889.4(1,130.6)504.5509.0482.2677.12,232.3
20191,367.01,002.8951.0(1,199.8)511.5516.0488.8686.42,263.2
20201,492.71,095.1986.8(1,239.8)558.5563.5533.8749.62,471.4
20211,589.21,165.91,027.1(1,280.1)594.6599.9568.3798.02,631.1
Notes:TheBISdataarereportedincurrent(nominal)USD.Forthesakeofcomparability,theamounts
recordedinthe“AidDataTotaltoLICsandMICs”columnarealsoreportedincurrent(nominal)USD,
thoughtheconstant2021USDamountsarereportedinparenthesis.ThetotalsfromAidDataexclude
short-term“rollover”facilities(seeBox2candSectionA-3intheAppendix).
TherearethreemainreasonswhyAidData’sGCDFdatasetchallengesthe
conventionalwisdomaboutthetotalcollapseofChina’soverseaslending
program.
1.Lendinginstitutioncoverage:Unlikeotherpubliclyavailabledatasetsthat
measureChinesedevelopmentfinance,the3.0versionofAidData’s
GCDFdatasetdoesnotexclusivelytracktheoverseaslendingactivitiesof
China’sgovernmentagencies(MOFCOMandCIDCA)anditsstate-owned
53
policybanks(ChinaEximbankandCDB).96BasedonOECD-DAC
reportingdirectives,ittrackstheoverseaslendingactivitiesofall
governmentandstate-ownedcreditorsinChina,includingstate-owned
commercialbanks(suchasBankofChina,ICBC,ChinaConstructionBank,
BankofCommunications,ChinaCITICBank,BankofShanghai,Postal
SavingsBankofChina,ChinaMerchantsBank,andHarbinBank),
state-ownedcompanies(suchasCNPC,CMEC,PolyTechnologies,
NORINCO,andAVIC),state-ownedfunds(suchastheSilkRoadFund
andChinaInvestmentCorporation),state-ownedpolicybanks(CDBand
ChinaEximbank),andgovernmentagencies(suchasMOFCOM,PBOC,
andSAFE).Intotal,180Chineselendinginstitutionsareincludedinthe
3.0versionofAidData’sGCDFdataset.Bywayofcomparison,two
ChineselendinginstitutionsareincludedinBostonUniversity’sCODF
database(ChinaEximbankandCDB)andonlyahandfulofChinese
lendinginstitutionsareincludedintheIDS.97ThebreadthofAidData’s
lendinginstitutioncoverageisparticularlyconsequentialbecause,in
recentyears,theLICandMIClendingoperationsofChina’scentralbank
(PBOC)andstate-ownedcommercialbankshaveexpandedwhilethoseof
CDBandChinaEximbankhavecontracted(seeFigure2.7).98
98Onaverage,duringthepre-BRIperiod(2000-2013),Beijingchanneled15%ofitsannuallendingcommitmentstolow-andmiddle-incomecountriesthroughitsstate-ownedcommercialbanks.Thisfigureincreasedto18%duringtheearlyBRI(2014-2017)periodand22%duringthelateBRIperiod(2018-2021).Onaverage,Beijingchanneled70%ofitsannuallendingcommitmentstolow-andmiddle-incomecountriesthroughitspolicybanksduringthepre-BRIperiod.Thisfiguredroppedto60%duringtheearlyBRIperiodand30%duringthelateBRIperiod.Onaverage,duringthepre-BRIperiod,Beijingchanneledonly3%ofitsannuallendingcommitmentstolow-andmiddle-incomecountriesthroughPBOC/SAFE.Thisfigureincreasedto14%duringtheearlyBRIperiodand43%duringthelateBRIperiod.SeeFigure2.7formoredetails.97IDSincludesloancommitmentsfromgovernmentagencies(suchasMOFCOMandCIDCA)andstate-ownedpolicybanks(suchasChinaEximbankandCDB),butexcludesloancommitmentsfromstate-ownedcompanies,state-ownedfunds,andstate-ownedcommercialbanks(Hornetal.2021:15).Forthemostpart,IDSalsoappearstoexcludeloancommitmentsfromthePBOCandSAFE,whichisaPBOCsubsidiary(seeBox2c).ThisistruedespitethefactthatIDSseekstocaptureallloansfrom“thegeneralgovernment,centralgovernment;stateandlocalgovernment;[and]centralbankandpublicenterprise”(WorldBank2020b:4).96TheChina’sOverseasDevelopmentFinanceDatabase,theChineseLoanstoLatinAmericaandtheCaribbeanDatabase,andChina’sGlobalEnergyFinanceDatabasetracktheoverseaslendingactivitiesoftwostate-ownedpolicybanks(CDBandChinaEximbank).TheydonottracktheoverseaslendingactivitiesofChina’sstate-ownedcommercialbanks.NordotheWorldBankIDSdatacaptureloansfromChinesestate-ownedcommercialbanks.AsexplainedbyHornetal.(2021:15),“theWorldBank'sdefinition[ofofficialsectorlending]doesnotcoverlendingbycommercialbankssuchastheIndustrialandCommercialBankofChina(ICBC)ortheBankofChina(BoC),despitethefactthattheyarestate-owned.Thesebanksareofficialcreditorsaccordingtoour(OECD)definition(theyareownedandcontrolledbytheChinesegovernment),buttheyarenotbilateralcreditorsaccordingtotheWorldBank’sdefinition,becausetheyarenota‘publicenterprise’inanarrowsense,incontrasttothepolicybankssuchasChinaEx-ImBankorCDB.”
54
2.Borrowinginstitutioncoverage:OECD-DACreportingguidelinesspecify
thatODAandOOFaredesignedtocaptureofficialsectorfinancialflows,
wherethe“officialsector”referstothesourcesratherthanthe
destinationsofsuchflows(Hornetal.2021:23).AidDataadherestothis
reportingstandardtoensurethatitsmeasuresofChinesedevelopment
financearecomparabletoOECD-DACsourcesofinternational
developmentfinance.Assuch,the3.0versionofAidData’sGCDFdataset
captureslendingfromallChinesegovernmentandstate-ownedcreditors
toallpublicsectorandprivatesectorborrowersinlow-incomeand
middle-incomecountries,regardlessofwhethertheysecuredsovereign
repaymentguarantees.Bycontrast,otherpubliclyavailable
datasets—suchasBostonUniversity’sCODFdatasetandtheWorldBank’s
IDS—exclusivelytrackChineseloanstogovernmentinstitutions,majority
state-ownedentities,andborrowinginstitutionsthatbenefitfrom
sovereignrepaymentguarantees(i.e.,publicandpubliclyguaranteed
debtowedtoChina).Thiscoveragedifferenceisconsequentialbecausea
significantpercentageofChina’soverseaslendingportfolioischanneled
tospecialpurposevehicles,jointventures,privatesectorinstitutions,and
minoritystate-ownedentities(seeFigures2.18andA10).Whilethese
loanstypicallydonotappearongovernmentbalancesheetsin
low-incomeandmiddle-incomecountries,manyofthembenefitfrom
implicithostgovernmentliabilityprotection,whichhasblurredthe
distinctionbetweenpublicdebtandprivatedebt.99Intotal,the3.0
versionoftheGCDFdatasetcaptures$525billionoflending
commitmentsto661borrowinginstitutionsthatqualifyascentral
governmentorcentralgovernment-guaranteeddebt,$421billionof
lendingcommitmentsto455borrowinginstitutionsthatqualifyas
anothertypeofpublicsectordebt,$67billionoflendingcommitmentsto
85borrowinginstitutionsthatqualifyaspotentialpublicsectordebt,$216
billionoflendingcommitmentsto724borrowinginstitutionsthatqualify
asprivatesectordebt,and$50billionthatisnotallocableduetoalack
99AsHornetal.(2021:4)explain,"[w]hilethedistinctionbetweenprivateandpublicsectorrecipientsisclearinprinciple,ittendstobeblurryinpractice,inparticularindevelopingcountriesandduringfinancialcrises.Privatedebtoftenturnsintopublicdebtonceacrisishitsandmanyoftheloans[toprivatesectorborrowers]mighthaveexplicitorimplicitgovernmentguarantees.”Onthispoint,alsoseeMaliketal.(2021)andMalikandParks(2021).
55
ofsufficientinformationforcategorizationpurposes.100Bycontrast,BU’s
CODFdatasetcaptures$605billionoflendingcommitmentsfromofficial
sectorinstitutionsinChinato150borrowinginstitutionsthatqualifyeither
ascentralgovernmentdebt,centralgovernment-guaranteeddebtor
anothertypeofpublicsectordebt.101
3.Debtinstrumentcoverage:Forthemostpart,existingChinese
developmentfinancedatasets—suchasBU’sCODFdatasetandthe
WorldBank’sIDS—tracklong-term,dollar-denominatedbilateralloansfor
publicinvestmentprojects.However,China’soverseaslendingprogramis
supportedbyamorediversesetoffinancinginstruments,someofwhich
areRMB-denominated,short-ormedium-terminnature,unrelatedto
publicinvestmentprojects,and/orinvolvingmultipleChineseand/or
non-Chinesebanks.TheyincludesyndicatedloanswithmultipleChinese
and/ornon-Chinesebankparticipants,loansentrustedtomultilateral
administrators,short-ormedium-termliquiditysupportfacilities(LSFs)to
providebalanceofpayments(BOP)support,currencyswapfacilities,
foreigncurrencydepositloans,pre-exportfinancing(PxF)facilities,
deferredpaymentfacilities,EPCFarrangements,interbankloan
agreements,andrevolvingcreditfacilities,amongotherthings.
ConsistentwithOECD-DACmeasurementstandards,the3.0versionof
AidData’sGCDFdatasettracksthefullsetofdebtinstrumentsusedby
Chinesestate-ownedentitiesinlow-incomeandmiddle-income
countries.102Hereagain,thebreadthofinstrumentcoverageisbecoming
increasinglyimportant,asChinesestate-ownedcreditorspivotawayfrom
thetraditionalapproachofissuinglong-term,dollar-denominated
102TheOECD-DACusesabroaddefinitionofwhatqualifiesasadebtinstrument.AccordingtothelatestversionofitsConvergedStatisticalReportingDirectivesfortheCreditorReportingSystem(CRS)andtheAnnualDACQuestionnaire,"[d]ebtinstrumentsrequirethepaymentofprincipaland/orinterestatsomepoint(s)inthefuture”(OECD2023b:12).101Foritspart,IDScaptures$378billionoflendingfromofficialsectorinstitutionsinChinatoborrowinginstitutionsthatqualifieseitherascentralgovernmentdebt,centralgovernment-guaranteeddebtoranothertypeofpublicsectordebt(i.e.,PPGdebt).However,thepubliclyavailableIDSdatacannotbedisaggregatedbyborrowinginstitution.Foracountry-by-countrycomparisonofAidDataandIDSmeasuresofPPGandnon-PPGdebtexposuretoChina,seeTableA16100Thesefiguresexcludetheshort-term“rollover”facilitiesthataredescribedinBox2candSectionA-3.Whensuchfacilitiesareincluded,the3.0versionoftheGCDFdatasetcapturesthefollowinglendingcommitmentamounts:centralgovernmentorcentralgovernmentguaranteeddebt($687billion),otherpublicsectordebt($422billion),potentialpublicsectordebt($67billion)andprivatesectordebt($216billion).Thereisanadditional$50billionindebtthatisnotallocableduetoalackofsufficientinformationforcategorizationpurposes.
56
bilateralbuyer’screditsandconcessionalloansforpublicinvestment
projects(seeFiguresA11,A12,andA13)andtowardshort-and
medium-termbilateralemergencyrescueloansaswellaslong-term
syndicatedloansandloansentrustedtomultilateraladministrators.103
The“scopeparameter”differencesbetweentheGCDFdatasetandother
publiclyavailabledatasetsmatterforasimplebutimportantreason:exclusively
trackingthelendingactivitiesofChinaEximbankandCDBandthesubsetof
loansfromthesetwopolicybanksthatqualifyasPPGdebtisnotausefulwayto
monitorthechangingscaleandcompositionofBeijing’soverseaslending
portfolioiftheportfoliohasshiftedtowardnewcreditors,newborrowers,and
newlendinginstruments.Inthischapter,wedemonstratethatsuchchanges
havealreadytakenplace,whichmeansthatacontinuedfocusonPPGdebtfrom
ChinaEximbankandChinaDevelopmentBankwouldbeanalogoustothe
proverbialdrunkardwhoinsistsuponsearchingforhiskeysbeneaththe
lamppost“becausethat’swherethelightis.”104
Thereareseveralsupply-sidefactorsthathavelikelypreventedChina’soverseas
lendingprogramfromcollapsing.First,Beijing’sstate-ownedbankshavehigh
levelsofinternationalexposuretodefaultrisk,whichmeanstheyhavean
interestinensuringthattheirbiggestborrowersaresufficientlyliquidtocontinue
servicingtheirexistingdebts.Onewayofprovidingliquidityrelieftoborrowers
isviashort-ormedium-termbridgeloans.105Second,eventhoughChinese
banksareincreasinglyreluctanttoissuelong-terminfrastructureprojectloansto
governmentborrowersduetotherisingtideofsovereigndebtdistress,Chinese
companieswithsignificantinternationaloperationshaveaninterestinsecuring
105Otheroptionsincludegraceperiodextensions,maturityextensions,andinterestratereductions.Formoreonthistopic,seeHornetal.(2023a,2023b).104Accordingtothe3.0versionofAidData’sGCDFdataset,thepercentageofofficialsectorlendingcommitmentsfromChinatoLICsandMICsthatwerechanneledthroughthepolicybanks(CDBandChinaEximbank)plummetedfrom87%in2009to22%in2021(seeFigureA27).AsMingeyandKratz(2021)putit“[t]heissueultimatelyisoneofscope.The[...]focusonpolicybankloansobscureschangesinChina’slendingpatterns—whetherashiftinthesourceofloanstoemergingmarketgovernmentsfrompolicycommercialbanks,orshiftsinthedestinationofloansfromgovernmentstoprivateinfrastructurevehiclesandcorporates.”103DuringtheearlyBRIperiod(2014-2017)andlateBRIperiod(2018-2021),anincreasingproportionofChina'sofficialsectorlendingtoLICsandMICsconsistedofemergencyrescueloans,includingthoseofthe“rollover”variety(seeFigures2.6andA14andBox2c).ThepercentageofChina’snon-emergencylendingprograminLICsandMICsthatwaschanneledviamultilateralinstitutionsorsyndicatedloanarrangementsincreasedfrom33.7%in2014to51.5%in2021(seeFigure2.23).
57
newbusinessandpreservingmarketshareinthecountrieswheretheyoperate.
Theythereforehaveincentivestoensurethatnewsourcesandtypesof
financing—suchassyndicatedloansthatpoolcreditriskacrossChineseand
non-Chineselenders,loanstospecialpurposevehiclesandjointventuresrather
thansovereigns,anddeferredpaymentorEPCFagreementsthatinvolvemore
risk-sharingbetweenChinesecompaniesandtheiroverseasclients—arebrought
tobearinsupportofnewinfrastructureprojectsinoverseasmarkets.Third,to
sustainhighlevelsofdomesticeconomicgrowth,Chinahasalong-runneedto
securenaturalresourcesthatitlacksinsufficientquantitiesathome,whichisan
importantreasonwhyBeijingallowsitsoverseasborrowerstocollateralizeand
repayloanswiththemoneytheyearnfromnaturalresourceexportstoChina.
Fourth,Chinaisseekingtopositionitselfasamajorworldpowerandproject
influencearoundtheglobe,whichservesasacounterweighttofinancial
pressuresforretrenchment.
However,thereisprobablynofactormoreimportantthantheoverallsizeof
China’sforeignexchangereserves.Chinawouldnothavebecometheworld’s
largestofficialcreditortothedevelopingworld—largerthantheWorldBank,the
IMF,andallParisClubcreditorscombined—ifnotforitsmassivestockpileof
foreignexchangereserves(Dreheretal.2021,2022).WhenBeijingadoptedthe
“GoingOut”strategyattheturnofthecentury,itprioritized
dollar-denominatedlendingtooverseasborrowerstodealwithaforeign
currencyoversupplychallenge:annualtradesurplusesledtoarapid
accumulationofdollarreserves,whichcreatedariskofcurrency(RMB)
appreciationandpromptedChina’sStateAdministrationofForeignExchange
(SAFE)tosearchforinternationalassetswhereitcouldinvestitssurplusdollar
reservesandgetagoodreturn.106SAFEusedthesefundsintheearly2000sto
recapitalizeseveralstate-ownedbanks,whichinturnrampedup
dollar-denominatedlendingtooverseasborrowers(seeFigure2.2andBox2b).
However,untilthe2008GlobalFinancialCrisis,SAFEparkedthelion’sshareof
itssurplusdollarreservesinU.S.governmentsecurities.Then,in2008and2009,
internationalassetpricesplummetedandtheU.S.FederalReserveweakened
thedollarviaquantitativeeasing.Beijing’straditionalinvestmentstrategyof
106Fordecades,Chinahassoughttoavoidarapidappreciationofitscurrencyinordertosustainhighlevelsofexportgrowth,whichitregardsasessentialtoachievehigh-incomecountrystatus.
58
parkingsurplusdollarreservesinU.S.governmentsecuritiesbecameless
attractiveanditlaunchedasearchforhigher-yield(undervalued)overseas
assets.SAFEentrustedalargerproportionofitssurplusdollarreservestothe
country’sstate-ownedpolicybanks,state-ownedcommercialbanks,and
state-ownedinvestmentfundsandtaskedthemwiththepursuitofhigher
investmentreturnsviadollar-denominatedinternationallending(seeBox2b).107
Figure
2.2:
Composition
of
China’s
loan
portfolio
by
currency
of
denominationFigure2.2
Notes:The“Other”categoryincludesallothercurrenciesofdenomination,includingEUR,GBP,
andlocalcurrenciesinlow-incomeandmiddle-incomecountries.
Thecentralrolethatforeignexchangereserveshaveplayedinthedramatic
expansionofChina’s21stcenturyoverseaslendingprogramraisesseveral
importantquestions.AreChina’sforeignexchangereservesrisingorfalling?If
theyarerising,howaretheybeinginvested?AreChinesestate-ownedlenders
beingrecapitalizedwiththesereservesandtaskedwithusingthesereservesto
extendforeigncurrency-denominatedloanstooverseasborrowers?
107However,SAFEisnottheonlysourceofforeignexchangethatChina'sstate-ownedcommercialbankshavedrawnupontosupportforeignexchange-denominatedoverseaslendingactivities.Theirbalancesheetdatademonstratesthattheyalsohaveaccesstodomesticforeignexchangedeposits(Setser2023a).
59
China’sforeigncurrencyreservesremainvast—approximately$3.1trillionasof
2023.However,thisfigureonlyreferstotheofficial,foreigncurrencyreserve
holdingsofChina’scentralbank(thePBOC).Itdoesnotaccountforthe
country’sso-called“hiddenreserves,”whichincludeforeigncurrencythatthe
PBOChasmovedoutofitsofficialreserveholdings(andoffofitsbalancesheet)
byentrustingthemtoChinesestate-ownedpolicybanks(likeCDBandChina
Eximbank),state-ownedcommercialbanks(likeBOC,ICBC,andChina
ConstructionBank),andstate-ownedfunds(liketheSilkRoadFundandCIC).
BradSetseroftheCouncilonForeignRelationsarguesthatthecountry’s
“hiddenreserves”maybeworthanadditional$3trillion.108Healsoprovides
evidencethattheseadditionalreserveshaverapidlyincreasedoverthelast
decade,whichmayexplainwhythe3.0versionoftheGCDFdatasetdoesnot
showamajorcontractionintheoverallsizeofChina’soverseaslendingprogram
inLICsandMICs.109Additionally,itmayexplainwhyChina’soverseaslending
programinhigh-incomecountries(HICs)andoffshorefinancialcenters(OFCs)
nearlydoubledbetween2015and2021(seeBox2aandTable2.1).
Box2b:Theinvestmentagencybehindthecurtain—China’sState
AdministrationofForeignExchange(SAFE)
China’sStateAdministrationofForeignExchange(SAFE)isaviceministry-levelinstitutionandasubsidiaryofthePBOC;itsoriginalmandatein1955wastoactasthecountry’sforeignexchangeregulatoryauthority.However,priortotheopeningupofChina’seconomyinthelate1970s,thecountryhadlimitedforeignexchangereservesandthePBOChadlimitedcentralbankresponsibilities.110
Despiteitsinauspiciousbeginnings,SAFEwaseventuallymaderesponsibleforthemanagementoftheworld’slargeststockpileofforeignexchangereserves.Italsobecameoneofthemostimportantinvestorsintheworld.111Duetorecurringcurrentaccountsurplusesandcapitalaccountsurpluses,China’sstockofforeignexchangereservesskyrocketedfromfrom$200billionin2000to$1trillionin2006,$2trillionin2009,and$3trillionin2011.Keepingallofthesereserves(mostlyUSD)onshoreposedacurrencyappreciationriskandthreatenedtoundermine
111Wei2013;Liu2023.110Atthattime,thePBOCwasastate-ownedcommercialbankoperatingunderthesupervisionoftheMinistryofFinance.ThePBOCdidnotofficiallybecomeChina’scentralbankuntil1983.109SeeSetser2023a,2023c.108Setser(2023a)arguesthatChina’s“hiddenreserves”consistof(a)non-reserveforeignexchangeassetsthatSAFEhasprovidedtothepolicybanks,state-ownedcommercialbanks,andstate-ownedinvestmentfundstolendandinvestabroad;(b)foreignassetsthatstate-ownedcommercialbankshavepurchasedtomatchtheirdomesticforeigncurrencydepositbase;and(c)foreignexchangethatCICpurchasedoffthebalancesheetofSAFE.
60
thecountry’sexport-ledeconomicgrowthstrategy,soSAFEwastaskedwithinvestingsurplusdollarsinoverseasassetsthatwouldyieldattractivereturnswithinacceptablerisklevels.112
Figure2.3belowpresentscumulativelendingcommitmentsofChina’sstate-ownedpolicybanks,state-ownedcommercialbanks,andtheSilkRoadFundtoLICsandMICs,inrelationtothetimingofSAFE’sinvestmentsintheseorganizations.ItdoesthesameforSinosure-backedlendingcommitmentstoLICsandMICs.OnecanseethatlargecashinjectionsfromSAFEhavegenerallyprecededincreasesintheoverseaslendingactivitiesofChina’sstate-ownedpolicybanks,state-ownedcommercialbanks,andtheSilkRoadFund.AnincreaseinSinosure-backedlendingtooverseasborrowersalsofollowedSAFE’srecapitalizationofthestate-ownedcreditinsuranceagencyin2010.
Figure
2.3:
Cumulative
loan
commitments
to
LICs
and
MICs
by
financial
institution
and
capital
injections
from
SAFE,
2000-2021
Figure2.3
CumulativeloancommitmentstoLICsandMICsbyfinancialinstitutionandcapitalinjections
fromSAFE,2000-2021
112China’sforeignexchangemanagementlawrequiresthatthecountry’sofficialreservesbeinvestedinhighlyliquidandlow-riskassetsthatcanbeusedtoaddressbalanceofpaymentneeds.However,foreignexchangereservesthatareentrustedtoastate-ownedentityforinvestmentpurposesfalloutsidetheofficial(IMF)definitionofforeignexchangereserves.Theycanthereforebeinvestedinhigher-risk,illiquidassets(Liu2023).
61
Notes:ThistablepresentscumulativelendingcommitmentstoLICsandMICs(inconstant2021USD)from
selectedChinesestate-ownedpolicybanks,state-ownedcommercialbanks,andstate-ownedfunds.Italso
presentscumulativelendingcommitmentstoLICsandMICs(inconstant2021USD)thatarebackedby
creditinsurancefromSinosure.TheverticaldashedlinesrepresentyearsinwhichaSAFEcapitalinjectionis
knowntohavetakenplace.Thesefiguresexcludeshort-term"rollover"facilitiestorefinancematuring
debts(seeBox2candSectionA-3intheAppendix).
SAFE’sfirstmajorinvestmentcameinDecember2003,whenitcapitalizedtheCentralHuijinInvestmentCorporation(CentralHuijin)with$45billion,whichinturnboughtequitystakesintwostate-ownedcommercialbanks:BankofChina($22.5billion)andChinaConstructionBank($22.5billion).InApril2005,CentralHuijinalsoboughtanequitystakeinICBCfor$15billion.SAFEinjectedanadditional$150billionintoChina’sstate-ownedcommercialbanks—byswappingUSDforRMBheldbythebanks—inlate2005and2006.113Thefirstknownrecapitalizationofastate-ownedpolicybankcameinJuly2005,whenSAFEinjected$5billionintoChinaEximbank.Then,inDecember2007,CentralHuijin—awholly-ownedsubsidiaryofSAFE—injected$20billionintoCDB.Sixmonthslater,SAFEagreedtoprovideadditionalfunding(worthanestimated$166.5billion)toCDBviaentrustedloanagreementstosupporttheoverseasactivitiesofChinesecompanies.114Undertheseagreements,CDBactedasacustodianoffundsforSAFE,disbursingloanstoborrowers,supervisingtheuseofthefunds,andmanagingrepayments.115SAFEeventuallyexpandeditsuseofentrustedloanagreementstootherstate-ownedbanks.116By2010,ithad“takeninitialstepstowardgivingpolicyandcommercialbanksauthoritytohandleloansforintergovernmentalcooperationprojects”andmoved“beyonditstraditionalroleofmanagingforeignexchangereserves,effectively[becoming]aforeign-currencylender”(Yuzhe2010).117
Then,inApril2010,CentralHuijininjected$11.7billion(RMB80billion)intoChinaEximbankandSinosuretohelpthemcleanupbadloansontheirbalancesheets.Fourteenmonthslater,inJune2011,ChinaInvestmentCorporation—anotherstate-ownedentityresponsibleforinvestingthecountry'sforeignexchangereserves—injectedanadditional$3.15billion(RMB20billion)intoSinosure.Then,inDecember2014,SAFEinjected$40billionintotheSilkRoadFundthroughasubsidiaryknownasButtonwoodInvestmentHoldingCompanyLtd.118SAFEinjected$48billionintoCDBand$45billionintoChinaEximbankinJuly2015throughwholly-ownedsubsidiaries(knownasWutongshuInvestmentPlatformCo.Ltd,SycamoreTreeInvestmentPlatform,andButtonwoodInvestmentHoldingCompanyLtd.).SAFEalsopurchasedequity
118InMay2017,theChineseGovernmentannouncedthatitwouldinjectanotherRMB100billionintotheSilkRoadFund.Theentityresponsiblefortheinjectionisunknown.117Atthattime,SAFEalsoagreedto"actastheorganizerandprimaryarrangerofsyndicatedloansunderentrustagreements”(Yuzhe2010).116AccordingtoLiu(2023:174),“SAFEdoeshaveaminimumreturntargetfortheforeignexchangeentrustedloansofabout2.5percent,calculatedasaspreadofseveralbasispointsabovetheinternationalbenchmarkbank-lendingrate,LIBOR.In2012and2013,aninterestrateof2.5percentwasnotparticularlylowinanenvironmentofquantitativeeasingintheEUandtheUnitedStates;attheendof2012theUSten-yearTreasurynote,aproxyfortherisk-freerate,wasonly1.5percent.”115However,asafiduciaryactingonbehalfofSAFE(inexchangeforacommission),itdoesnotassumeanyoftherisksorrewardsoftheentrustedloans.Assuch,CDBrecordsentrustedloansasoff-balancesheetitems.114NineforeignreserveentrustedloanagreementsweresignedbySAFEandCDBinMay2010.113CentralHuijininjectedRMB130billion($19billion)intotheAgriculturalBankofChinainNovember2008.However,itwasnolongerawholly-ownedsubsidiaryofSAFEatthetime.
62
stakesinICBCandBankofChina—viaWutongshuInvestmentPlatformCo.Ltd—duringthelastquarterof2015.119
Since2015,SAFEhasnotpubliclyannouncedanyadditionalcashinfusionsintoChina’sstate-ownedpolicybanks,state-ownedcommercialbanks,oritsstate-ownedcreditinsuranceagency.AnongoingsourceofspeculationanddebateiswhetherSAFEisworriedabout“throwinggoodmoneyafterbad.”120Therearesomeindicationsthatthecountry’slargeststate-ownedbanksandSinosure,whicharestewardingforeignexchangereservesfromSAFE,mayhavehighlevelsofcross-borderexposuretonon-performingloansonthebalancesheetsofChinesebanks.121InNovember2017,ChinaBankingRegulatoryCommission(CBRC)—thecountry'stopbankingregulator—publiclycalleduponCDBandChinaEximbanktoputinplacemorerobustriskmanagementprocedures(Xueqing2017).122Then,in2018,Sinosure’sChiefEconomisttooktheextraordinarystepofpubliclyadmonishingChina’spolicybanksfortheir“downrightinadequate”duediligenceprocedures(PillingandFeng2018).AnothersignofpotentialperilisthefactthattheannualoverseaslendingcommitmentsofCDB,ChinaEximbank,BankofChina,andICBCsharplydeclinedaftertheyreceivedlargecashinfusionsfromSAFEin2015(seeFigureA31).OnepotentialexplanationforthisunusualpatternisthatSAFE’smoneywasusedtocleanupbaddebtsonthebalancesheetsofthesebanksinsteadofsupportingnewoverseaslendingcommitments.
SAFEisdiscreetaboutthereturnsthatithasearnedviaoverseasinvestments,duetodomesticpoliticalconcernsaboutwhetherChina’s“xuèhànqián”(thehard-earnedmoneyofChineseworkers)isbeingeffectivelystewarded.It“triestolimititsinvestmentsoutside[U.S.governmentsecurities]toamountssmallenoughtohidefromthepublicincasethebetsgobad”(Wei2013).However,in2021,SAFEdisclosedthatithadearnedanaverageannualreturnonitsforeignexchangereserveportfolioof3.35%between2008and2017and4.11%between1998and
122ZhouMinyuan,theheadofCBRC'spolicybankssupervisiondepartment,announcedatthetimethatit“requiredbothbankstofullyidentifyoverseasbusinessrisks,stepupcompliancemanagement,completelyunderstandtheoperationalandfinancialstatusoftheirclientsaswellasthelawsandregulationsofhostcountries,strictlyobservethelocalenvironmentalandindustrialregulations,andstrengthencommunicationwithlocalregulators”(Xueqing2017).CBRCalso“demandedthebanksenhancecapitalsupervisionviaon-siteinspectionsandinvestigations,effectivelypreventandcontroloverseasbusinessrisksbytakingrisk-sharingmeasures,prudentiallyevaluatethefeasibilityandcomplianceofrelevantguaranteemeasures,andimprovetheiremergencyresponsemechanism”(Xueqing2017).121Duringtheprimarydatacollectioneffortforthe3.0versionoftheGCDFdataset,AidDatauncoveredevidenceofSinosureinrecentyearsprovidingcreditinsuranceforloansissuedbynon-ChinesebanksforprojectsbeingimplementedbyChinesecontractors.Theseloancommitmentsarenotincludedinthe3.0versionofAidData’sGCDFdataset.120SAFEhasaparticularlyhighlevelofexposuretonon-performingloansonChinaEximbank’sbalancesheet.AccordingtoabondprospectusthatChinaEximbankissuedinMarch2017,“theownershipoftheMinistryofFinanceinChinaEximbankisapproximately10.7%while[...]SAFEownsapproximately89.3%ofChinaEximbankthroughitsinvestmentplatform”(Export-ImportBankofChina2017:20).SAFEisalsoexposedtonon-performingCDBloans,butitsexposureisrelatedtoitsuseofCDBasafiduciaryforentrustedloansandCDB’sownbalancesheet.SAFEholdsa27.1%ownershipstakeinCDBthroughButtonwoodInvestmentHoldingCompanyLtd.ItpurchasedthisstakeonJuly15,2015,whenButtonwoodInvestmentHoldingCompanyLtd.injected$48billionofsharecapitalintoCDB(CDB2020).119Thebank(re)capitalizationinformationinBox2bwasdrawnfromSAFE2004,2017a,2017b;Ma2006;PBOC2007,2012;Ying2008;ChinaDaily2009;Reuters2009,2010a,2010b,2016;Parson2010;Shan2011;Yuan2014;Tangjunetal.2014;Jia2015;Xinhua2015;Xiao2016;XieandLamar2016;Chen2014;KongandGallagher2017;OfficeoftheLeadingGroupforPromotingtheBeltandRoadInitiative2019;EmbassyofthePeople'sRepublicofChinaintheRepublicofthePhilippines2019;Liu2023;andSetser2023a.
63
2017(SAFE2021:53).Atthetime,itcharacterizedtherateofreturnasbeingata“relativelygoodlevel”(SAFE2021:53).SAFEhasnotreleasedanydataonaverageannualreturnsfor2018oranysubsequentyears.
AnotherimportantquestionthatremainsunansweredistheextenttowhichSAFEownsthebondeddebtoftheLICandMICgovernmentsofEmergingMarketandDevelopingEconomies(EMDEs)—eitherthroughdirectpurchasesorsecondarymarketpurchases(i.e.,viainvestmentfundslikePIMCO,Blackrock,AllianceBernstein,FidelityInvestments,andAmundiAssetManagement).ThereareseveralreasonstobelievethatSAFE’sholdingsofbondedEMDEsovereigndebtmaybesignificant.First,whenSAFEagreedtobuy$300millionofbondeddebtfromtheGovernmentofCostaRicain2008,itattemptedtohidethepurchase(Anderlini2008).TheDeputyAdministratorofSAFE,FangShangpu,soughtandsecuredawrittenassurancefromtheGovernmentofCostaRicathatitwould“takenecessarymeasurestopreventthedisclosureofthefinancialtermsofthisoperationandofSAFEasapurchaserofthebonds”(SAFE2008).123
Thepurchasewasonlymadepublicbecauseofacourtorder.WhenSAFEwasaskedtocommentonthematter,itsaidthattherewasnothingunusualaboutthepurchasebecauseit“ownsbondsissuedbymanyothergovernments”(Batson2008).Second,SAFE’sChiefInvestmentOfficerfrom2010to2014wasabondtraderatPIMCO(andright-handmanofPIMCOco-founderBillGross)from2006to2009(Wei2013).Third,SAFEopenedanofficeonFifthAvenueinNewYorkCityin2013,andshortlythereafteritbecameanopensecretamongtheworld’slargestassetmanagers—likePIMCOandBlackRock—thatSAFEwasoneoftheirmostimportantconfidentialclients.
Section2:MajorchangesinBeijing’soverseaslendingportfolioduringtheBRI2.0era
OurgoalintheremainderofthischapteristoexplainhowChinaisresponding
totherisingtideofdebtdistressinthedevelopingworldandidentifythe
measuresitistakingtode-riskitsoverseaslendingportfolio.However,before
wedoso,itisimportanttounderstandsomeofthemajorchangesinBeijing’s
overseaslendingportfoliothathaverecentlytranspired.Threechanges,which
havebecomehallmarksoftheBRI2.0era,meritspecialattention:
1.China’snewandunfamiliarroleastheworld’slargestofficialdebt
collector
123AlsoseeGovernmentofCostaRica(2007,2008).WehavemadetheofficialcorrespondencebetweenSAFEandCostaRica’sMinistryofFinanceinJanuary2008accessibleviahttps://www.dropbox.com/scl/fi/a20fdxb2lecowt8tp5lym/2-January-2008-SAFE-Letter-to-Minister-of-Finance-of-Costa-Rica.pdf?rlkey=y9ywdrrhap4rdkt0cyqj4mj4y&dl=0andhttps://www.dropbox.com/scl/fi/48qdeslzwim9t6dqhyk0m/7-January-2008-Letter-from-Minister-of-Finance-of-Costa-Rica-to-SAFE.pdf?rlkey=pr3p4advi1wfzzwjr5t8gva5m&dl=0.
64
2.Theriseofemergencyrescuelendingandthefallofinfrastructureproject
lending
3.AstrategicpivotfromUSD-denominatedbilaterallendingto
RMB-denominatedbilaterallending
China’snewandunfamiliarroleastheworld’slargestofficialdebt
collector
WhentheBRIwasinitiallylaunched,Beijingdifferentiateditselffromitspeers
andcompetitorsbyfinancingbig-ticketinfrastructureprojects—likehigh-speed
railwaysandnext-generationtelecommunicationnetworks—thatvirtuallynoone
elsewaswillingtobankroll.Manyoftheloansthatwereissuedfortheseprojects
includedlengthygraceperiods,somostborrowersdidn’tneedtoworryabout
makinglargedebtservicepaymentsforfive,six,orsevenyears.Withalotof
“easymoney”sloshingaround,manyborrowercountriessawtheireconomic
growthratessoar(Dreheretal.2021,2022).China’spopularityaroundtheglobe
alsosoared:thehigh-watermarkofpublicsupportforChinainthedeveloping
world—aftertheintroductionoftheBRI—wasin2019.124
However,asFigure2.4demonstrates,theageofeasymoneyisintherear-view
mirrorforanexpandingsetofborrowers.By2020,40%ofofficialsectorloans
fromChinatolow-incomeandmiddle-incomecountrieshadenteredtheir
principalrepaymentperiods(followingtheexpirationoftheirgraceperiods).
Thisfigureincreasedto55%in2023,andweexpectitwillreachapproximately
75%by2030and100%by2049.
Anotherwayofthinkingabouttheloomingrepaymentchallengeistotrackthe
percentageofloansinChina’sLICandMICportfoliothathavealreadyreached
their(originallyscheduled)finalrepaymentdates.In2014(thefirstfullyearof
BRIimplementation),thisfigurestoodatonly17%(seeFigure2.5).By2023(the
tenthfullyearofBRIimplementation),ithadincreasedto44%.Weexpectthis
124Atthattime,thepopulation-weightedaveragelevelofpublicsupportforChinainthedevelopingworldwas55.66%.OnthecausalrelationshipbetweenthereceiptofChineseaidandcreditandpublicsupportforChina,seeWellneretal.(forthcoming,2023).
65
figuretoriseto52%by2025andapproachnearly100%by2040(seeFigure
2.5).
Figure
2.4:
Loans
from
China
within
their
repayment
periods,
2000-2050Figure2.4
Notes:ThisgraphshowsthecumulativepercentageofofficialsectorloansfromChinatoLICsandMICs(as
measuredbythenumberofloans)thatarewithintheirrepaymentperiods).Todeterminewheneachloan
willenterrepayment,eachloan’sgraceperiodisaddedtoitscommitmentdate.Thisfigurerepresents
whenloanswillreachtheirrepaymentperiodaccordingtotheiroriginalborrowingterms,althoughmany
loanshavebeenrescheduled(ofteninvolvinganextensionoftheloan’sgraceperiodand/ormaturity).
MOFCOMinterest-freeloancommitments(whicharetypicallyissuedwithoutacredibleexpectationof
repayment)areexcludedfromthecalculation.
66
Figure
2.5:
Loans
from
China
reaching
their
(originally
scheduled)
final
repayment
dates,
2000-2050Figure2.5
Notes:ThisgraphshowsthecumulativepercentageofofficialsectorloansfromChinatoLICsandMICs(as
measuredbythenumberofloans)thathavereachedmaturity.Todeterminewheneachloanwillreach
maturity,eachloan’smaturityperiodisaddedtoitscommitmentdate.Thisfigurerepresentswhenloans
willreachtheirmaturityaccordingtotheoriginalborrowingterms,althoughmanyloanshavebeen
rescheduled(ofteninvolvinganextensionoftheloan’sgraceperiodand/ormaturity).MOFCOM
interest-freeloancommitments(whicharetypicallyissuedwithoutacredibleexpectationofrepayment)are
excludedfromthecalculation.
Whatdoesallofthismeaninpracticalterms?Itmeansthatarapidlygrowing
percentageofborrowersintheGlobalSoutharemakinglargedebtservice
payments(thatareforthemostpartdenominatedinU.S.dollars)toBeijingata
timewheninterestratesarerising,theU.S.dollarisstrengthening,local
currenciesareweakening,andglobalgrowthisslowing.Manyborrowersdonot
haveenoughhardcurrencyonhandtomeettheirrepayment
obligations—especiallyonloanswithLIBOR-basedinterestratesthatincreased
byfive-and-a-halfpercentagepointsbetweenJanuary2022andSeptember
2023.125
125Average6-monthLIBORskyrocketedfrom0.426%inJanuary2022to5.892%inSeptember2023.
67
ItalsomeansthatBeijingfindsitselfinanunfamiliaranduncomfortablerole—as
theworld’slargestofficialdebtcollector.SomeofChina’sstate-sponsored
tabloidsandresearchinstitutionsareattemptingtodeflectcriticismbyblaming
theU.S.FederalReserveforthesharpincreaseinLIBOR-basedinterestrates
(e.g.,Xueqing2022;Qing2023).126However,thispositionwillbedifficultto
defend.WhenBeijingdecidedtomakeLIBORcentraltoitsdollar-denominated
overseaslendingprogram(seeFigure2.14),itdidnotdosoatthebehestofthe
U.S.oranyotherforeignpower.127Itdidsoonitsown—andforitsown
profit-makingpurposes(seeBox2b).Now,thegraceperiodsonmostofChina’s
LIBOR-basedloansareexpiring,andBeijingislearningadifficultlesson:it’s
easierforgreatpowerstobehavelikecommercialcreditorsintimesofplenty
thanintimesofwant.IfChinawasacommercialbank,itwouldbeeasierto
demandthatfinancially-distressedLICandMICborrowersdrawupontheir
limitedU.S.dollarreservestomakeincreasinglylargedebtservicepayments.
However,Chinaisaglobalsuperpowerthathastobalanceawiderangeof
competingequities,includingreputationalriskandrepaymentrisk(seeChapter
4formoreonthiskeytensionandtradeoff).
Theriseofemergencyrescuelendingandthefallofinfrastructure
projectlending
China’soverseaslendingprogramhasbecomesynonymouswiththeBRI—an
overland“Belt”ofroad,rail,port,andpipelineprojectsthatseekstocreatean
infrastructurecorridorfromChinatoCentralAsiaandEurope,anda“Maritime
SilkRoad”thatseekstolinkChinatoSouthandSoutheastAsia,theMiddleEast,
andAfricathroughdeep-waterseaportconstructionprojectsalongthelittoral
areasoftheIndianOcean.Howeverthe3.0versionofAidData’sGCDFdataset
127Accordingtothe3.0versionofAidData’sGCDFdataset,45%ofChina’sdollar-denominatedofficialsectorlendingtoLICsandMICsisbasedon6-monthLIBORoranotherLIBOR-basedinterestrateand67%ofChina'svariableinterestrateofficialsectorlendingtoLICsandMICsisbasedon6-monthLIBORoranotherLIBOR-basedinterestrate(seeFigure2.14).126Forexample,inApril2023,China’sstate-ownedtabloid,GlobalTimes,publishedanop-edidentifying“theUS’irresponsiblemonetarypolicy[as]therootofAfricandebtproblems.”Itarguedthat“[r]elyingondollarhegemony,theUShasimplementedthreeroundsofquantitativeeasing,cutinterestratestonearzero,andfloodedAfricaandemergingmarketswithlow-interestdollars.Itthenarbitrarilyandaggressivelyraisedinterestrates,boostedtheU.S.dollarexchangerate,attractedthereturnofdollars,asaresult,Africancountrieshavetofaceliquidityshortages,brokenfundingchains,currencydepreciation,skyrocketingdebtrepaymentcostsdenominatedindollars,asurgeinsovereigndebt,andexacerbateddebtproblems.TheunfairfinancialsystemledbytheUSistherootofAfrica’sdebtproblems.”(Qing2023).
68
highlightstheimportanceofnotconflatingChina’sflagship,globalinfrastructure
initiativewithitsoverseaslendingprogram.128
Duringthepre-BRIperiod(2000-2013),Beijingprovidedanextraordinary
amountofcreditforinfrastructureprojectsinLICsandMICs.129Intotal,the3.0
versionofAidData’sGCDFdatasetidentifies2,217loan-financedprojectsworth
$628billionduringthisfourteen-yearperiod.13066%oftheseprojects(worth
$412billion)soughttoconstruct,rehabilitate,expand,ormaintainphysical
infrastructure.131However,Figures2.6andA13provideevidencethat,sincethe
BRIwaslaunched,arapidlyshrinkingproportionofChina’soverseaslendingto
LICsandMICshassupportedinfrastructureprojects.132Thiswastrueduringthe
earlyBRIperiod(2014-2017)andduringthelateBRI(2018-2021)period.
Infrastructureprojectlendingcommitmentsasashareoftotallending
commitmentstoLICsandMICsfellfrom65%in2014,to50%in2017,49%in
2018,and31%in2021.
132FigureA26intheAppendixprovidesanotherversionofthisgraphthatpresentsprojectlendingratherthaninfrastructureprojectlendingovertime.Itshowsaverysimilarpattern:projectlendingcommitmentsasashareoftotallendingcommitmentsfellfrom78%in2014to56%in2017,andthenfrom50%in2018to32%in2021.131Togeneratethisestimate,weidentifyallloansinthe3.0versionoftheGCDFdatasetthatarecategorizedas“investmentprojectloans”andthatinvolvetheconstruction,rehabilitation,expansion,ormaintenanceofphysicalinfrastructure.Formoreonthedefinitionandmeasurementoftheinfrastructurevariableandtheinvestmentprojectloanvariable,seeCusteretal.(2023).130Togeneratethisestimate,weidentifyallloansinthe3.0versionoftheGCDFdatasetthatarecategorizedas“investmentprojectloans.”Formoreonthedefinitionandmeasurementoftheinvestmentprojectloanvariable,seeCusteretal.(2023).12984%ofChina’sinfrastructureprojectlendingfrom2000-2014wasprovidedbyCDBandChinaEximbank,and32%wasprovidedviabuyer’scredits.However,FiguresA11and2.7demonstratethatarapidlyshrinkingpercentageofChina’soverseaslendingprograminthedevelopingworldisprovidedviathecountry’spolicybanksandbuyer’scredits.128Table2.1alsocallsattentiontothispointbyspotlightingthevastscaleofChina’soverseaslendingtohigh-incomecountries(HICs)andoffshorefinancialcenters(OFCs).
69
Figure
2.6:
Composition
of
loan
portfolio
by
financial
instrumentFigure2.6
Notes:Fordetailsonhowinfrastructureprojectlendingismeasured,seefootnote131.
GiventhatBeijingdidnotdramaticallyscalebacktheoverallsizeofitsoverseas
lendingprograminLICsandMICs,howandwhydiditcontinuetolendrecord
amountstodevelopingcountriesatatimewhenitwasratchetingdownBeltand
Roadprojectlending?Figure2.6providesaclearanswer:Beijingwasramping
upitsemergencyrescuelendingactivitieswhileitwasrampingdownits
infrastructureprojectlendingactivities.133In2013,oneyearbeforethefirstfull
yearofBRIimplementation,emergencyrescuelendingrepresentedonly5%of
China’soverseaslendingtoLICsandMICs.By2021,58%ofChina’soverseas
lendingtoLICsandMICsconsistedofemergencyrescuelending.ThePeople’s
BankofChina(PBOC)—China’scentralbank—isbyfarthemostimportant
financierofinternationalemergencyrescuelendingoperations,whichexplains
133ConsistentwithHornetal.(2023a,2023b),wedefineemergencyrescueloansasallloansfromChinesestate-ownedentitiestogovernmentborrowinginstitutionsinlow-incomeandmiddle-incomecountriesthatcanbeusedforatleastoneofthreepurposes:repayingexistingdebts,financinggeneralpublicexpenditures,orshoringupforeignexchangereserves.Suchloansincludeborrowingsviacurrencyswapagreements,liquiditysupportfacilities,foreigncurrencytermfinancingfacilityagreements,depositloans,commodityprepaymentfacilities,andso-called“sovereignloans”
(
主
权
贷
).
70
whyithadassumedadominantroleinBeijing’sLICandMIClendingportfolio
by2020(seeFigure2.7).134
Figure
2.7:
Composition
of
loan
portfolio
by
creditor
categoryFigure2.7
InMarch2023,ateamofresearchersfromAidData,theWorldBank,theHarvard
KennedySchool,andtheKielInstitutefortheWorldEconomypublishedastudy
thatexplainswhyBeijinghasundertakenrescuelendingoperationsworthnearly
$250billionin22countries(Hornetal.2023a).135Theyfindthatmostofthese
operationshavetakenplaceinBRIparticipantcountrieswithhighlevelsof
outstanding(infrastructureproject)debttoChinesebanksandcompanies.They
alsofindthatbailoutsfromBeijingaredirectedtodistressedgovernment
borrowersattimeswhentheirforeignexchangereservelevelsarelowandtheir
creditratingsareweak.
135TheauthorsofthestudyincludeSebastianHornoftheWorldBank;BradParks,ExecutiveDirectorofAidDataandResearchProfessoratWilliam&Mary’sGlobalResearchInstitute;CarmenReinhart,formerWorldBankGroupChiefEconomistandcurrentProfessorattheHarvardKennedySchool;andChristophTrebesch,DirectorattheKielInstitutefortheWorldEconomy.134In2013,thePBOCandSAFE(itssubsidiary)wereresponsibleforonly6%ofChina’sofficialsectorlendingcommitmentstoLICsandMICs.By2021,thatfigurereached54%(seeFigure2.7andFigureA37).
71
Figure
2.8:
Chinese
emergency
rescue
lending
by
countries’
level
of
debt
exposure
to
ChinaFigure2.8
Notes:Thethreecategories(high,medium,orlowdebtexposuretoChina)areconstructedbyranking
countriesaccordingtototalnon-emergencylendingcommitmentsfromofficialsectorinstitutionsinChina
between2000and2021.Countriesthatdidnotreceiveanynon-emergencylendingcommitmentsare
excludedfromthecalculation.Thecalculationofaverageemergencyrescueloancommitmentsexcludes
short-term"rollover"facilitiestorefinancematuringdebts(seeBox2candSectionA-3intheAppendix).
The3.0versionofAidData’sGCDFdatasetprovidesanopportunitytoconducta
basicreplicationexercisewithfouryearsofadditionaldata(2018-2021),
substantiallyreviseddatafor2000-2017,andmoreprecisemeasurementsof
China’sproject,infrastructureproject,andnon-emergencylendingactivities.In
Figure2.8,wereproduceFigure5fromtheHornetal.(2023a)studywith
updateddataonemergencyandnon-emergencylendingcommitmentsand
confirmthatChina’sinternationalemergencylendingoperationsare
concentratedincountriesthataccumulatedlargeamountsofdebttoChinafor
non-emergencypurposesbetween2000and2021.136Wecanalsoconfirmthat
everycountrythatreceivedanemergencyrescueloanfromChinaisa
participantintheBRI.137InTable2.2,wereplicateTableA4fromtheHornetal.
(2023a)studyandconfirmthatChineseemergencyrescueloansareissuedto
sovereignsattimeswhenreservesarelowandattimeswhenborrowershave
137Asof2021,thiswastrueofallemergencyrescueloanrecipientsotherthanArgentinaandMalawi.Then,in2022,ArgentinaandMalawijoinedtheBRI.136Figure2.8usesanewdummyvariable(“rescue”)inthe3.0versionoftheGCDFdatasetforemergencyrescueloans,whichcapturesanyloanthatallowsasovereigndebtorto(1)serviceexistingdebts,(2)financegeneralbudgetaryexpendituresand/or(3)shoreupforeignreserves.
72
veryweakcreditratings.138Bycontrast,wefindthatChineseprojectloansand
Chineseinfrastructureprojectloansareissuedtoborrowersattimeswhenthey
haverelativelystrongcreditratingsandreserveadequacyratios.139Allofthese
findingsareconsistentwiththoseofHornetal.(2023a)andsupporttheir
characterizationofBeijing’scrisismanagementresponseasoneof“Bailoutson
theBeltandRoad.”140
Table
2.2:
Average
sovereign
risk
ratings
and
gross
reserves
for
recipients
of
different
Chinese
loan
typesTable2.2
Average
sovereign
risk
ratings
and
gross
reserves
for
recipients
of
different
Chinese
loan
types
TypeofLoanAverageSovereignRiskRatingMoody’sRatingFitchRatingS&PRatingGrossReserves(inMonthsofImports)
RescueLoans5.7Caa1CCCCCC+4.7
ProjectLoans9.4Ba3BB-BB-6.2
InfrastructureProjectLoans9.2Ba3BB-BB-5.9
Notes:Thistablepresentstheaveragesovereignriskratingandlevelofgrossreserves(inmonthsofimport
cover)forcountriesthatreceivedemergencyrescueloans,projectloansandinfrastructureprojectloans
fromChinabetween2000and2021.Theaveragesareweightedbythecommitmentamountsofthe
emergencyrescue,project,andinfrastructureprojectloansfromthe3.0versionofAidData’sGCDFdataset.
ThedataongrossreservesinmonthsofimportsaredrawnfromtheWorldBank’sWorldDevelopment
Indicators.ThesovereignriskratingsdataarederivedfromtheWorldBank’ssovrateindexusingthe
conversionscaleinSéri(2021).Sovrateisameasureofrepaymentriskthatvariesfrom0to21,withhigher
scoresindicatinglowerlevelsofsovereigncreditrisk(Koseetal.2022).
140Emergencyrescueloansanddebtreschedulingsaresimilarinthattheybothprovidecashflowrelieftoinsufficientlyliquidborrowers.Inthisway,theycanbothbeusedto“bailout”aborrower(Hornetal.2023b).139Hornetal.(2023a)usesloancommitmentsfornon-emergencypurposesasaproxyfor“projectloans”andreliesonthe2.0versionofAidData’sGCDFdataset.Table2.2usesthe3.0versionofAidData’sGCDFdataset,whichallowsformoreprecisemeasurementofprojectloansandinfrastructureprojectloans.Formoreonthesemeasurements,seefootnote130andfootnote131.138ThesovereignriskratingsproducedbyMoody’stakeoneof21categoricalvalues,whereAaarepresentsthelowestlevelofriskandCrepresentsthehighestlevelofrisk.ThesovereignriskratingsproducedbyFitchtakeoneof21categoricalvalues,whereAAArepresentsthelowestlevelofriskandDrepresentsthehighestlevelofrisk.ThesovereignriskratingsproducedbyS&Ptakeoneof21categoricalvalues,whereAAArepresentsthelowestlevelofriskandC/Drepresentsthehighestlevelofrisk.Formoredetails,seeSéri(2021).
73
InFigure2.9,wetaketheanalysisonestepfurther.Wefirstusethecountry-year
levelmeasureoffinancialdistress(thatweintroducedinChapter1)toidentify
Beijing’s50largestLICandMICborrowersthatexperiencedfinancialdistressat
somepointbetween2000and2021.141Wethenidentifythetimingofbailouts
(emergencyrescueloans)anddebtreschedulings(cashflowrelief)fromBeijing
inrelationtotheonsetanddurationoffinancialdistressepisodes.
141Inordertodifferentiatebetweenrepaymentrisksandrepaymentriskmitigationefforts,wemodifythecriteriaforthefinancialdistressmeasure.Insteadofusingallloanrecordswherethedescriptionfieldinthe3.0GCDFdatasetindicatesthattheborrowerhaddifficultymakingrepaymentsorexperiencedfinancialdistress,weexcludeallobservationsforwhichtheonlysourceofevidenceoftheborrowerhavingdifficultymakingrepaymentsorexperiencingfinancialdistressisanattemptedoractualdebtrescheduling.
74
Figure
2.9:
Rescue
lending
and
debt
rescheduling
events
for
the
top
50
borrowers
in
financial
distress,
2000-2021Figure2.9
Rescue
lending
and
debt
rescheduling
events
for
the
top
50
borrowers
in
financial
distress,
2000-2021
Notes:Thisfigureprovidesanoverviewofthetimelineofwhencountriesexperiencedfinancialdistress
(blueshading),whenChinaprovidedrescuelending(circles),andwhenChinarescheduledexistingloan
repayments(X’s).AcircleindicatesthatatleastonerescueloanwasprovidedbyChinatotherespective
countrythatyear,andanXindicatesthatatleastoneloanwasrescheduledbyChinafortherespective
countrythatyear.Countriesincludedinthislistrepresentthetop50borrowersindistress,orderedbythe
75
sizeoftheircumulativelendingportfolioasof2021.Seefootnote141fordetailsonhowthefinancial
distressindexwasmodifiedtodifferentiatebetweenrepaymentrisksandrepaymentriskmitigationefforts.
Figure2.9demonstratesthat83%ofChina’semergencyrescueloans(including
short-term“rollover”facilities)wereissuedinyearswhentherecipientsofthese
loanswereinfinancialdistress.142Similarly,80%ofChina’sdebtreschedulings
tookplaceinyearswhenborrowersintheparticipatingcountriesexperienced
financialdistress.143Figure2.9alsoprovidesevidenceofBeijingrepeatedly
targetingdebtreschedulingsandemergencyrescueloanstothesameBRI
participantcountrieswithhighlevelsofdebtexposuretoChina.144Theserial
natureofthesecashflowreliefeffortssuggeststhatBeijing’sbiggestborrowers
maynotonlyhaveshort-termliquidityproblems,butalsolong-termsolvency
problems.
AnotherimportantimplicationoftheseresultsisthatChinaisincreasingly
behavinglikeaninternationalcrisismanager.Ithaseffectivelycreatedasafety
netforfinanciallydistressedsovereignsthatareparticipatingintheBRI—and,by
extension,theirhighlyexposedChinesecreditors.145Ithasalsotakena
differentiatedapproachacrosscountriesthatpresentvaryinglevelsofrisktothe
Chinesebankingsector,wherebycountriesthatpresentahighlevelofbalance
sheetexposuregetnewmoney(viabalanceofpaymentssupport)andcountries
thatpresentalowlevelofbalancesheetexposuregetcashflowrelief(viadebt
reschedulings)butnonewmoney(Hornetal.2023a,2023b).146Theseactions
146Formoreonthispoint,seeHornetal.(2023a)andHornetal.(2023b).145Itisimportanttokeepinmindthat,fromahistoricalperspective,countercyclicalofficiallendingisthenormratherthantheexception.AsHornetal.(2020b)explain,“[d]uringthecourseofthe1930s,theUnitedStatesjoinedEuropeanstatesinextendingofficialloanstostateswithbalance-of-paymentsproblems,inparticularthroughtheUSExport-ImportBankandtheUSTreasury’sExchangeStabilizationFund,whichwasestablishedin1934”(Hornetal.2020b:7).144AccordingtotheunderlyingdatathatwasusedtoconstructFigure2.9,100%ofthecountriesthatreceivedserialdebtreschedulingsand100%ofthecountriesthatreceivedserialemergencyrescueloansareformalparticipantsintheBRI.AllofthesecountriesrankamongBeijing’s50largestLICandMICborrowersandeitherbenefitedfromdebtreschedulingsintwoormoreconsecutiveyearsoremergencyrescueloansintwoormoreconsecutiveyears.143ThisfindingisalsoconsistentwithHornetal.(2023a)andHornetal.(2023b).142Bycomparison,49%ofChina’snon-emergencyloanswereissuedinyearswhentherecipientsoftheseloanswereinfinancialdistress.ThisdifferenceisalsoobservableduringtheBRIera(2014-2021).Overthiseight-yearperiod,83%ofChina’semergencyrescueloans(representing86%ofChina’semergencyrescuelendingportfolioinmonetaryterms)wereissuedinyearswhentherecipientsoftheseloanswereinfinancialdistress.DuringtheBRIera(2014-2021),63%ofChina’snon-emergencyloans(representing67%ofChina’snon-emergencyrescuelendingportfolioinmonetaryterms)wereissuedinyearswhentherecipientsoftheseloanswereinfinancialdistress.Thesepercentagesreflectthedistressmarkerthatwasmodifiedtodifferentiatebetweenrepaymentrisksandrepaymentriskmitigationefforts(seefootnote141).
76
aredifficulttoreconcilewiththeincreasinglypopular“Beijinginretreat”
narrativethatwepreviouslydescribed.
AstrategicpivotfromUSD-denominatedbilaterallendingto
RMB-denominatedbilaterallending
AswenotedinSection1ofChapter2,Chinadramaticallyincreaseditsforeign,
dollar-denominatedlendingactivitiesintheimmediateaftermathofthe2008
GlobalFinancialCrisis.Figure1.2inChapter1demonstratesthatthesingle
largestyear-on-yearincreaseinofficialsectorlendingcommitmentsfromChina
toLICsandMICswasbetween2008and2009,andFigure2.2provides
evidencethatnearlyallofthesecommitmentsweredenominatedindollars.
Butthe3.0versionofAidData’sGCDFdatasetalsocallsattentiontothree
additionalchangesthattookplaceafterthe2008GlobalFinancialCrisis.First,
Chinesestate-ownedcreditorswerealreadyholdingasubstantialamountof
distressed,dollar-denominateddebtbythetimetheBRIwasannouncedin2013
(seeFigures1.8andA4).Second,from2013onward,thedollar’simportancein
China’soverseaslendingportfoliosteadilydeclined:theshareofnewlending
commitmentsdenominatedindollarsfellsharply,from93%in2013to44%in
2021(seeFigure2.2).Dollar-denominatedloanswerereplacedby
RMB-denominatedloans:theshareofnewlendingcommitmentsdenominated
inRMBsoaredfrom6%in2013to50%in2021(seeFigure2.2).Third,the
RMB-denominatedloansthatBeijingissuedwerepredominantlyemergency
rescueloanstocountriesinfinancialdistress(seeFigures2.9,2.10,andA15).147
147IntheAppendix,wereplicateFigure2.2fortwodifferentcohorts:oneforcountriesinfinancialdistressandanotherforcountriesnotinfinancialdistress(seeFiguresA16andA17).OnecanseeanincreaseinRMB-denominatedlendingacrossbothcohorts,buttheincreaseismoresubstantialforcountriesinfinancialdistress.Figure2.10,whichincludesshort-term,roll-overloanamounts,showsasimilarpatterninChina’sRMB-denominatedrescuelendingportfolio.
77
Figure
2.10:
RMB-denominated
rescue
lending
to
countries
in
distress
and
countries
not
in
distressFigure2.10
Notes:Todetermineifacountryexperiencedfinancialdistressinagivenyear,weusethebinarymeasure
thatisdescribedinBox1ainChapter1.
Ononehand,thisstrategymakessense.Duringthe1930sandafterWorldWar
II,theU.S.becameamajorinternationallenderoflastresort,providing
dollar-denominatedemergencyrescueloanstoborrowerswithlarge
outstandingdollar-denominateddebtstoU.S.companiesandbanksthroughthe
U.S.FederalReserve,theU.S.Treasury’sExchangeStabilizationFund,andthe
U.S.Ex-ImBank(Hornetal.2020b).Theseactivitieshelpedthedollareventually
becomeadominantcurrencyforreserveholdingsandinternationalfinancial
transactions.Now,Chinesestate-ownedpolicybanks,state-ownedcommercial
banks,andstate-ownedenterpriseshavehighlevelsofexposuretooverseas
borrowersthatareindefaultorteeteringontheedgeofdefault,andthe
institutionwithamandatetoprotectthehealthofChina’sfinancialsectorand
internationalizetheRMB(thePBOC)isrampinguptheprovisionofemergency
rescueloanstoensurethatitsoverseasborrowersaresufficientlyliquidto
continueservicingtheiroutstandingdebtstoChinesecreditors.
78
Ontheotherhand,thePBOC’sdecisiontoprovideRMB-denominatedrescue
loansispuzzlingbecausetheborrowersbeingbailedoutneedUSDmorethan
RMBtorepaytheiroutstandingdebtstoChinesecreditors.Onepotential
explanationisthatthePBOCisheedingWinston’sChurchill'sadviceto“never
letagoodcrisisgotowaste.”Formanyyears,itsoughttointernationalizethe
RMB—withoutachievingmuchprogressduetotightcapitalcontrolsandan
insufficientlydeepandliquidRMBbondmarketoutsideofmainlandChina.148
However,therisingtideofdebtdistressintheGlobalSouthhasmadetwo
groupsofcountriesmorewillingtoincreasetheirRMBreserveholdings:(1)
countriesfacingsevereliquidityand/orsolvencyproblemsthatwouldliketo
avoidborrowingfromthetraditionallenderofthelastresort(theIMF)because
ofconcernsaboutonerouspolicyconditionalities(likeVenezuela,Belarus,and
Laos);and(2)countriesfacingsevereliquidityand/orsolvencyproblemsthat
havedecidedtoseekIMFsupportbutneedadditionalsupportinordertostay
afloat(likeArgentina,Mongolia,andSriLanka).149
WhyhasthePBOCchanneledRMB-denominatedemergencyrescueloansto
bothgroupsofcountriessince2013?Thelogicofdoingsointhefirstcountry
cohortseemsrelativelystraightforward:theproceedsfroman
RMB-denominatedemergencyloancanbeusedtoservicepreviouslycontracted
loansfromChinesecreditorsthatweredenominatedinRMB.Forthesecond
cohortofcountries,Beijing’sintentionsareopaqueandpoorly-understood.But
recenteventsinArgentinamayprovideaclue.Inmid-2023,thedollarreserve
holdingsofthecountry’scentralbank(BCRA)wereperilouslylow,anditwas
urgentlyseekingbridgefundingtoavoiddefaultingonitsrepayment
obligationsundera$44billionloanagreementwiththeIMF.ThePBOCstepped
intothebreach,helpingBCRAmakethreelargedebtservicepaymentstothe
IMFinJune2023,July2023,andOctober2023(Nugent2023a,2023b;Do
Rosario2023b;DoRosarioandStrohecker2023).ItdidsobyallowingBCRAto
useRMBdrawingsunderaswaplinebetweenthetwocentralbanksworth
approximately$9.3billion.BCRAwasabletouseshort-termRMBswapdebt
149SeeHornetal.(2023a,2023b).148Asof2022,theRMBaccountedforlessthan3%ofglobalcurrencyreservesandlessthan2.5%ofglobalpayments.KeyimpedimentstoRMBinternationalizationincludetightcapitalcontrolsthatinhibitthefreemovementofthecurrencyandtheabsenceofalargeoffshoremarketforinvestorstopurchasesafe,RMB-denominatedassetsthatarecomparabletoU.S.Treasurybonds(Bertautetal.2023).
79
fromthePBOCtorepaytheIMF“withouttouching[its]dollarreserves”(do
RosarioandOtaola2023)fortworeasons:(1)moneyisfungible,and(2)IMF
loanscanberepaidwithmultiplecurrencies(includingUSD,EUR,RMB,JPY,
GBP,andSDR).Inotherwords,BCRArepaiditsdebttotheIMFinRMB,which
allowedittopreserveitsdollarreserveholdings.150ThecaseofArgentinaalso
callsattentiontoaseparate,butcloselyrelated,point:Beijinghasan
encompassinginterestinensuringthatitslargestborrowerswith
dollar-denominateddebtstoChinesecreditorsdonotexhausttheirdollar
reserveholdings.151
Inthisway,thePBOC’swillingnesstoserveasaninternationallenderoflast
resortsuggeststhatitmaybeseekingtokillseveralbirdswithonestone:
preservingthedollarreserveholdingsofitsbiggestborrowers,encouraging
greateruseoftheRMBincross-bordertransactions,andlayingthegroundwork
fortheRMBtoeventuallyachieveglobalreservecurrencystatus.152Thefactthat
thePBOCplayedaninstrumentalroleinhelpingtheIMF’ssinglelargest
borrowermeetitsrepaymentobligationswithRMBalsosuggeststhatformer
U.S.TreasurySecretaryLarrySummersmayhavebeenoverlyoptimisticwhenhe
wasaskedabouttheriskofde-dollarizationandrespondedthat:“youcannot
replacesomethingwithnothing.[…]Europe’samuseum,Japan’sanursing
home,andChina’sajail.Wedon’tneedtoworryaboutthosecurrenciesbeing
somekindofmajorthreattous.”153
153Indeed,thereisevidencethatsigningacurrencyswapagreementwiththePBOCresultsina14%increaseintheprobabilityofacountryusingtheRMBforinternationalpayments(BahajandReis2022).Large-scaleborrowingviaPBOCswaplinescanalsohavethedirecteffectofchangingthecurrencycompositionofacountry’sreserveholdings.AfternearlyadecadeofRMBdrawdownsthroughitsswaplinewithPBOC,approximately50%ofArgentina’sreserveholdingsconsistedofRMB(Douglas2022).152ItisalsonoteworthythatBeijinghaschangeditspublicmessagingabouttheadvisabilityofdollar-denominatedoverseaslendingandborrowing(e.g.,Qing2023).151Thereisalsoapotentialdemand-sideexplanationforwhysovereignsinfinancialdistress—likeArgentinaandSriLanka—arewillingtocontractRMB-denominatedswapdebtfromthePBOC.DespitesignificantrestrictionsonthefreeandflexibleuseofPBOCswapdrawings,centralbankscanusetheseRMBdrawingsasa“windowdressing”devicetotemporarilyinflatetheirgrossinternationalreserves(Hornetal.2023a).Thisapproachmighthelpavertcreditratingdowngradesandborrowingcostincreases.However,itcanalsofreeupotherwiseencumberedforeignexchangereservestofacilitatedollar-denominateddebtservicetoChinesecreditors(otherthanPBOC)andnon-Chinesecreditors.Forexample,BradSetserhasargued(seehttps://twitter.com/Brad_Setser/status/1602151579150438401)thatthereceiptofPBOCswapdebtinSriLankafreedupforeignexchangereservestofacilitatea$400millionpaymentonamaturingEurobondanddollar-denominateddebtservicetoChinaEximbank.150InAugust2023,Argentina’scentralbank(BCRA)decidedtorepaysomeofitsoutstandingswapdebttothePBOCwiththeproceedsfromanIMFloandisbursement(DoRosario2023a).
80
Section3:Beijing’scrisis-timeapproachtorepaymentriskmitigation—Foolmeonce,shameonyou.Foolmetwice,shameonme.
Inthenextsectionofthischapter,weanalyzethe3.0versionofAidData’sGCDF
datasettobetterunderstandhowBeijingisseekingtode-riskitsoverseasloan
portfolio.WeseeevidenceofChinesestate-ownedlenderstakingthefollowing
riskmitigationefforts:
1.Sweepingcashoutofescrowaccounts
2.Providingshort-termcashflowreliefinexchangeforescrowaccount
replenishment
3.Lendingwithhigherinterestrates,shorterrepaymentperiods,more
safeguards,andmoreseverepenaltiesfordefault
4.Takingadifferentiatedapproachwithborrowersthatpresenthighandlow
levelsofrepaymentrisk
5.Scalingdownbilaterallendingoperationsandscalinguplending
operationsviasyndicationandmultilateralization
Riskmitigationstrategy#1:Sweepingcashoutofescrowaccounts
InMarch2021,ateamofresearchersfromAidData,theCenterforGlobal
Development(CGD),theKielInstitutefortheWorldEconomy,andthePeterson
InstituteforInternationalEconomics(PIIE)publishedastudyentitled“How
ChinaLends,”whichdemonstratedthatChinesestate-ownedlendershavea
preferenceforsourcesofcollateralthatdonotrequireliquidationthrougha
costly,time-consuming,anduncertainjudicialprocess(Gelpernetal.2021,
2022).Morespecifically,thestudyfoundthatChinesestate-ownedlenders
prefertocollateralizeonforeigncurrencydepositsinescrowaccountsthatthey
controlandcanunilaterallydebit(withouthavingtoinitiatejudicialproceedings
totrytorecoveranoverduedebtbyseizingorliquidatingaphysicalasset).It
alsofoundthatborrowersaretypicallyrequiredtomaintainaminimumcash
81
balanceinaspecialtypeofescrowaccount—knownasa“RepaymentReserve
Account”or“DebtServiceReserveAccount”(DSRA)—equivalenttooneyear’s
worthofprincipalandinterestrepayments.
Atthetimethatthestudywaspublished,therewasnohardevidenceof
borrowerscomplyingwiththeseescrowaccountconditions.Norwasthereany
hardevidenceofChineselenderssweepingcashoutoftheseescrowaccounts
inordertodealwithnonpaymentorlatepaymentbyoverseasborrowers.The
3.0versionoftheGCDFdatasetprovidessuchevidence.154
Table
2.3:
Illustrative
escrow
account
balances
linked
to
loans
from
China
Eximbank,
CDB,
and
ICBCTable2.3
Illustrative
escrow
account
balances
linked
to
loans
from
China
Eximbank,
CDB,
and
ICBC
CountryLenderBorrowerAggregateCashBalanceofEscrowAccounts(Maximum)CorrespondingLoan(s)
TanzaniaChinaEximbankMinistryofFinance,TPDC$60.3million$920million(2012),$275million(2012)
GuineaICBCCentralGovernment€76.35million€559.4million(2018)
RepublicofCongoChinaEximbankCentralGovernment$338million~20loansunder$1.6billion(2006)and$1billion(2012)frameworkagreementsSurinameChinaEximbankCentralGovernment,Telesur$9.3million$98.4million(2016)
GhanaCDBCentralGovernment$71.2million$850million(2012),$150million(2013),$210.6million(2019),$185.5million(2019)MalawiICBCReserveBankofMalawi$32million$66million(2021)
MyanmarCDBMyanmarOilandGasEnterprise$77.1million(in€equivalent)€452.7million(2010)
ZimbabweChinaEximbankMinistryofFinance,KHPC$17.2million$319million(2013)
AngolaCDBMinistryofFinance$1.5billion$15billion(2015)
KenyaChinaEximbankNationalTreasury,KRC$250million$1.9billion(2014),$1.6billion(2014)
154Escrowaccountcashbalancesarerecordedinthe3.0versionoftheGCDFdataset’s“collateral”and“description”fields.CashsweepsoutofescrowaccountsrecordedintheGCDFdataset’s“description”field.
82
CountryLenderBorrowerAggregateCashBalanceofEscrowAccounts(Maximum)CorrespondingLoan(s)
EcuadorCDBMinistryofFinance$113million$1billion(2010)
GhanaChinaEximbankCentralGovernment$27.2million$293.5million(2007)
ZimbabweCDBEconetWirelessZimbabwe$12.4million$93million(2014)
BotswanaICBCBPC$33million$825million(2009)
ZambiaCDBDBZ$6million$30million(2015)
Notes:ThistableprovidesexamplesofescrowaccountcashbalanceslinkedtoloansfromEximbank,CDB,
andICBC(projectID#59752,59733,65116,65115,60219,59273,55437,73140,30578,58586,60039,
98520,34468,62674,66847,31777,37103,35865,183,62601,40,52190inthe3.0versionofAidData’s
GCDFdataset).Theescrowaccountbalanceinformationisdrawnfromthecollateralfieldanddescription
field.Escrowaccountbalancesvaryovertime.Thistablerecordsthemaximumobservedaccountbalances.
Table2.3providesanillustrativesetofescrowaccountcashbalanceslinkedto
ChinaEximbank,CDB,andICBCloansin15countries.Thereareseveral
importantpointstokeepinmindaboutthesebalances.First,althoughthereis
someevidenceofborrowernoncompliancewiththeescrowaccountconditions
intheirChineseloanagreements,complianceseemstobethenormratherthan
theexception.Borrowerssubjecttosuchconditionstypicallymaintainescrow
accountcashbalancesthataresufficienttocoverthecostof1to3semi-annual
principalandinterestpayments.Theseamountsareusuallyequivalentto5-10%
ofthefacevalueoftheloansupportedbytheescrowaccount.Second,itisnot
uncommonforminimumcashbalances—andminimumcashbalance
requirements—tochangebasedonaloan’sactualorexpectedamount
outstandingatdifferentpointsintimeoverthelifetimeoftheloan.Thecash
balance(requirement)isusuallyatitslowestpointduringthegraceperiodwhen
theloanhasnotfullydisbursedandatitshighestpointwhentheloanhasfully
disbursedbutnorepaymentshavebeenmade.Some,butnotall,Chinese
lendersallowborrowerstoincrementallyreducetheamountofcashintheir
escrowaccountsasrepaymentsaremadeandthetotalamountsoutstanding
shrink.155
155Inthe3.0versionofAidData’sGCDFdataset,thereismoreevidencethatCDBfollowsthispracticethanChinaEximbank.See,forexample,projectID#37103,55437,58839,and58842.
83
Third,notwithstandingthe“1to3semi-annualdebtservicepayments”ruleof
thumb,minimumcashbalancerequirementscanbeadjustedbasedonthe
borrower’sriskprofileand/orthelender’slevelofexposure.156Chinese
state-ownedcreditorsmay,forexample,usea“portfolio-wideapproach”to
compensateforahighlevelofexposuretoariskyborrower.ChinaEximbank’s
collateralizedlendingarrangementwiththeGovernmentofCongo-Brazzavilleis
acaseinpoint.Theborrowerisrequiredtokeepaminimumcashbalance
equivalentto20%ofitstotaloutstandingdebtundermultipleChinaEximbank
loanagreementsinanoffshore,lender-controlledescrowaccount(République
duCongo2018).The3.0versionofAidData’sGCDFdatasetdemonstratesthat
theCongoleseauthoritieshaveforthemostpartcompliedwiththis
requirement:thecashbalancesintheirChinaEximbank-controlledescrow
accountwere$338millionin2017and$266.6millionin2020.157Fourth,while
theamountsofforeigncurrencythatChinesestate-ownedcreditorsask
borrowerstoring-fenceinescrowaccountsarenotnecessarilylargeenoughto
beconsequentialduringnormaltimes,thesignificanceofthesefundscan
increaseduringperiodsoffinancialdistress,whenborrowersarestrappedfor
hardcurrencyandseekingacoordinateddebtrestructuringwithmultiple
creditors.Non-Chinesecreditorsoftenlackaccesstoforeigncurrencythatis
ring-fencedfortheirexclusiveuseandtheyfear—withsomejustification—that
Chinesecreditorshavepositionedthemselvesatthefrontoftherepaymentline
bydemandingthatborrowersgrantthemaccesstocashcollateralthatother
creditorslack(andthatcanbeunilaterallydebitedinamoment’snotice).
Consequently,theymaynotbewillingtoparticipateinacoordinateddebt
reschedulingunlessallcreditorsagreetoabidebytheso-called“comparable
treatment”principle—i.e.,ensurethatthereisreasonableburdensharinginthe
157SeeID#60219and#59273inthe3.0versionoftheGCDFdataset.156Gelpernetal.(2021,2022)analyzethetermsandconditionsin100loancontractsissuedbyChinesestate-ownedcreditorsand142foreignloancontractsissuedby28non-Chinese(commercial,bilateral,andmultilateral)creditorstogovernmentborrowersinLICsandMICs.Theyfindthat30%oftheChineseloancontractsincludeescroworrevenueaccountprovisions,butonly2%ofthenon-Chineseloancontracts(onefromAfDB,onefromCommerzbank,andonefromAgenceFrançaisedeDéveloppement)includesuchprovisions.Intherarecaseswhennon-Chinesecreditorsrequiregovernmentborrowerstomaintainminimumcashbalancesinescrowaccounts,theamountsthattheyrequirearesimilartotheamountsrequiredbyChinesecreditors.Thereisonlyonecontractinthesampleof142non-ChineseloancontractsanalyzedbyGelpernetal.(2021,2022)thatclearlyspecifiesaminimumcashbalancerequirement:a$56.6millionloanagreementbetweenCommerzbankAGParisBranchandtheGovernmentofCameroonthatrequirestheborrowerto(initially)maintainaescrowaccountcashbalanceequivalenttooneyear’sworthofprincipalandinterestpayments(i.e.,twosemi-annualdebtservicepayments).Theunredactedloanagreementcanbeaccessedinitsentiretyviahttps://docs.aiddata.org/ad4/pdfs/how_china_lends/CMR_2015_121.pdf.
84
waythatfinanciallossesaredistributedacrosscreditors(BuchheitandGulati
2023).158
Duringourreviewoftheprimarysourcesthatunderpinthedataset(including
escrowaccountagreements,repaymentmechanismagreements,andthe
auditedfinancialstatementsofborrowinginstitutions),wealsodiscoveredthat
Chineselendershaveputinplaceseveraldifferentsafeguards(or“linesof
defense”)tominimizetheriskthattheseescrowaccountswillnotfulfilltheir
intendedriskmitigationpurposes.Thefirstsafeguardisarequirementthatthe
borrowersinitiallymeettheirminimumcashbalancerequirementsbydepositing
fundsintotherepaymentreserve(orDSRA)accountswithrevenuesthatare
alreadyattheirdisposal(ratherthanprojectrevenues,whicharetypically
minimalornonexistentwhenprojectloansarefirstissued).Thesecond
safeguardisautomaticityinthewaythattherepaymentreserve(orDSRA)
accountsarereplenishedafterwithdrawalshavetakenplace.Inatypicalescrow
accountagreementbetweenaChineselender,borrower,andescrowaccount
bank,iftheborrowermissesaprincipaland/orinterestpaymentandthelender
sweepscashoutoftherepaymentreserveaccount(inordertosatisfyitsdesire
toberepaidinatimelymanner),theescrowaccountbankisresponsiblefor
immediatelyreplenishingtherepaymentreserveaccountwithcashfromanother
escrowaccountthatisoftenreferredtoasthe“revenueaccount”or“sales
collectionaccount.”Chineselenderstypicallyrequirethattheirborrowers
depositallprojectrevenues—oralloftherevenuesgeneratedbytheunderlying
infrastructureasset(e.g.,atollroad,anairport,atelecommunicationsnetwork)
supportedbytheproject—intothisadditionalescrowaccount.159Insomecases,
Chineselenderswillgoonestepfurtherandrequirethatafixedpercentageof
allrevenuesoftheborrowinginstitutionbedepositedintherevenueaccount.160
ThethirdsafeguardisthatChineselendersusuallypossessexclusiveauthorityto
freezetherevenueaccount(withouttheconsentoftheborrower)andprevent
160See,forexample,ID#55437inthe3.0versionofAidData’sGCDFdataset.159See,forexample,ID#59753inthe3.0versionofAidData’sGCDFdataset.Alternatively,theborrowermayberequiredtodepositintothisaccountcashproceedsfromsalesunderalong-termcommoditypurchaseagreement.See,forexample,ID#35865inthe3.0versionofAidData’sGCDFdataset.158IfthereisaperceptionthatChinawantstobetreatedasaseniorcreditorwhosedebtsneedtobegivenfirstpriorityandothercreditorsarebeingpushedtothebackoftherepaymentline,acollectiveactionfailureamongcreditors(i.e.,nocoordinateddebtrescheduling)becomesmorelikely(WigglesworthandYu2023;FerryandZeitz2023).
85
theborrowerfrommakingwithdrawalsfromtheaccount.161Thefourthsafeguard
isthat,intheeventofamissedinterestorprincipalpayment,Chineselenders
areentitledunderthetermsofmostescrowaccountagreementsto“pay
themselves”bywithdrawinganequivalentamountofcashoutoftherepayment
reserveaccountand/ortherevenueaccount(withoutborrowerconsent).162
ThelatestversionoftheGCDFdatasetalsoprovidesevidencethat,when
borrowersdefaultontheirrepaymentobligations,Chineselendersdoinfact
“paythemselves”byunilaterallymakingcashwithdrawalsfromtheescrow
accountsthattheyestablishedforriskmitigationpurposes.Consider,for
example,the$98.4millionloanthatChinaEximbankissuedtotheGovernment
ofSurinameandTelesur—thestate-ownedtelecommunicationscompanyof
Suriname—in2016foraNationalBroadbandNetworkProject.163Asasourceof
cashcollateral,thelenderaskeditsborrowertomaintainaminimumbalancein
aUSDrepaymentreserveaccountof$2.9million(equivalenttototalpayable
interestforoneyearortwosemi-annualinterestpayments)duringtheloan’s
graceperiodand$9.3million(equivalenttototalpayableprincipalforoneyear
ortwosemi-annualprincipalpayments)duringtheloan’srepaymentperiod.It
alsorequiredthatTelesurdepositatleast50%ofitsorganizationalfunding
(includingbroadbanduserrevenuesfromtheNationalBroadbandNetwork
Project)inalocalcurrencyrevenueaccount.Then,duringtheCOVID-19
pandemic,theGovernmentofSurinamedefaultedonitssovereigndebt
obligationsmultipletimesin2020and2021.Bytheendof2021,the
GovernmentofSurinamehadaccumulatedprincipalandinterestarrearsto
ChinaEximbankworthapproximately$61million(IMF2021).Then,inFebruary
2022,theGovernmentofSurinameandTelesurmissedascheduledinterest
paymentontheChinaEximbankloanfortheNationalBroadbandNetwork
163OnDecember30,2016,ChinaEximbankandtheGovernmentoftheRepublicofSurinamesigneda$98.4millionpreferentialbuyer’scredit(PBC)agreementfortheSurinameNationalBroadbandNetworkProject.TheloanofficiallywentintoeffectinNovember2017afterseveralpreconditions(so-called“conditionsprecedent”)weremetbytheborrower,includingbutnotlimitedtothesigningofarepaymentmechanism(escrowaccount)agreement.Theloancarriesthefollowingborrowingterms:a15-yearmaturity,a5-yeargraceperiod,a3%interestrate,a0.4%managementfee,anda0.4%commitmentfee.TheGovernmentofSurinameon-lenttheproceedsoftheloantoTelesuronMay12,2017.Formoredetails,seeID#55437inthe3.0versionofAidData’sGCDFdataset.162Bywayofexample,seeID#37103,#31777,#59753,and#35865inthe3.0versionofAidData’sGCDFdataset.161Moreprecisely,Chineselenderstypicallypossesstheexclusiveauthoritytoinstructtheescrowaccountbanktofreezetherevenueaccountandpreventborrowerwithdrawals.See,forexample,ID#59753inthe3.0versionofAidData’sGCDFdataset
86
Project.Thelenderrespondedbyimmediatelywithdrawing$1.47million(the
monetaryvalueofonesemi-annualinterestpayment)fromtherepayment
reserveaccountinordertocoverthecostofthemissedinterestpayment.Italso
instructedalocalSurinameseescrowaccountbank(FinaBankN.V)to
immediately(a)blockTelesur’saccesstothelocalcurrencyrevenueaccount,and
(b)replenishtherepaymentreserveaccountwithfundsintherevenue
account.164Theseactionsprovedconsequentialfordomesticandinternational
reasons.Thedecisiontodenythestate-ownedtelecommunicationscompany
accesstoatleast50%ofitsorganizationalfundinginstigatedquestionsamong
localstakeholdersaboutwhyaforeignlenderpossessedsuchextraordinary
authority.TherevelationthatChinaEximbankpaiditselfbyexecutingacash
sweepoutofanoffshore,lender-controlledescrowaccountalsoproved
controversialbecausetheGovernmentofSurinamewaspursuingacoordinated
debtreschedulingwithallofitsmajorexternalcreditorsatthetime,manyof
whomwereunawarethatChinaEximbankhadrecoursetoring-fencedforeign
currencyreserves(andundertheimpressionthattheGovernmentofSuriname
wascash-strappedandunabletomakeloanrepaymentstoanyexternal
creditors).165
Mostofthesecashsweepsaredoneinsecret.Theseareexceptionallydifficult
tomonitorbecausethelenderisdebitingcashfromanescrowaccountthatis
typicallydomiciledoutsideoftheborrowercountryorinsidetheborrower
countrybutbeyondtheimmediatereachofdomesticoversight
institutions—suchastheauditorgeneralandthepublicaccountscommittee
withinparliament.166Also,thelegalagreementsthatgrantChineselendersthe
authoritytoconductthesecashsweepseffectivelyrepresent“sideagreements.”
166Inmostsovereigndebttransactions,financeministriesaretheborrowercountrycounterpartstoescrowaccountagreementswithChinesecreditors.Theseagreementstypicallyimposeexpansiveconfidentialityobligationsonborrowers(financeministries).Bywayofillustration,theescrowaccountagreementfortheChinaEximbankloancapturedinProjectID#59753ofAidData’sGCDFdataset(Version3.0)saysthat“[t]heobligationofconfidentialityshallendureinperpetuity.[…]TheParties[totheescrowaccountagreement]shallnotatanytimeduringthetermsofthisDeedreleaseanystatementtothepressormakeanyotherpublicstatementofanynaturewhichcouldreasonablybeexpectedtobepublishedinanymediaregardingtherelationshiporthesubjectmatterofthisDeed[…].”165InMarch2022,theIMFreportedthattheFebruary2022“[p]aymentfromtherepaymentreserveaccountfortheTelesurloanwillbereflectedintheeventualdebtrestructuringwith[ChinaEximbank]toensurethereiscomparabilityoftreatmentwithotherofficialcreditors”(IMF2022).164Undertheoriginalescrowaccountarrangementthatwasfinalizedin2017,Telesurwasrequiredtodepositatleast50%ofitsfunding(includingbroadbanduserrevenuesfromtheNationalBroadbandNetworkProject)inthelocalcurrency-denominatedrevenueaccount.SeeID#55437inthe3.0versionofAidData’sGCDFdataset.
87
Theauthoritytoconductacashsweepisalmostneveridentifiedinloan
agreementsbetweenChineselendersandtheiroverseasborrowers,whichare
morereadilyaccessibletodomesticoversightinstitutions.Instead,loan
agreementswithChinesestate-ownedcreditorstypicallycross-referenceanother
agreement—oftenknownasan“accountmanagementagreement,”an“escrow
accountagreement,”ora“repaymentmechanismagreement”—thatgrantsthe
lendersuchauthority.Financeministriesrarelydisclosethesesideagreements
toauditorsintheexecutivebranch,overseersinthelegislativebranch,or
internationalorganizations(liketheIMF)withsurveillanceresponsibilitiesunless
theyarepressedtodoso.Theseagreementsareshroudedinsecrecyfora
purpose:collateralizingoncashdepositsinlender-controlledescrowaccountsis
rareamongofficialcreditors,so(sovereign)borrowersarereluctanttodisclose
thattheyhavegrantedthesesourcesofleverage(debtseniority)toChinese
state-ownedcreditorsbutnototherofficialcreditors.Inthevanishinglyrare
instancesinwhichcashsweepsarediscovered,itcanbecomemoredifficultfora
distressedsovereigntogetallmajorcreditorstoparticipateinacoordinated
debtreschedulinggovernedbythe“comparabletreatment”principle.Ifsome
creditorshaverecoursetoring-fencedsourcesofforeigncurrencyandothersdo
not,anypromisebythesovereignborrowertoabidebythe“comparable
treatment”principleisrenderedlesscredible.167
Riskmitigationstrategy#2:Providingshort-termcashflowreliefin
exchangeforescrowaccountreplenishment
Figure2.11demonstratesthatwehaveenteredaneweraof“reschedulings
galore.”The3.0versionoftheGCDFdatasetrecords98Chinesedebt
reschedulingsin2020and2021,whichismorethandoublethenumberof
ChinesedebtrestructuringsrecordedintheRhodiumGroupdatabaseoverthe
sametimeperiod(MingeyandWright2023).Chinesedebtreschedulingsfor
LICsandMICshaveincreasedintandemwiththepercentageofloansinChina’s
LICandMICportfoliothathavealreadyreachedtheir(originallyscheduled)final
repaymentdates(seeFigure2.11).Theyhavealsoforthemostparttakenplace
infinanciallydistressedcountries(seeFigures2.9and2.12).
167Surinameisacaseinpoint.
88
Figure
2.11:
Percentage
of
loans
that
have
reached
maturity
and
number
of
loans
rescheduledFigure2.11
Notes:Todeterminewheneachloanwillreachmaturity,eachloan’smaturityperiodisaddedtoits
commitmentdate.Thisfigurerepresentswhenloansreachedtheirfinalmaturitydatesaccordingtothe
originalborrowingterms,althoughmanyloanshavebeenrescheduled(ofteninvolvinganextensionofthe
loan’sgraceperiodand/ormaturity).MOFCOMinterest-freeloancommitments(whicharetypicallyissued
withoutacredibleexpectationofrepayment)areexcludedfromthecalculation.
89
Figure
2.12:
Loan
rescheduling
for
countries
in
and
not
in
financial
distressFigure2.12
Notes:Todetermineifacountryexperiencedfinancialdistressinagivenyear,weusethebinarymeasure
thatisdescribedinBox1ainChapter1.Seefootnote141fordetailsonhowthefinancialdistressindex
wasmodifiedtodifferentiatebetweenrepaymentrisksandrepaymentriskmitigationefforts.
TheseempiricalpatternsunderlinethefactthatmanyofChina’soverseas
borrowersareinsufficientlyliquidtomeettheirrepaymentobligationsandin
needofdebtrelief.HowisBeijingrespondingtothischallenge?The3.0version
ofAidData’sGCDFdatasetdemonstratesthatChineselendersaregenerally
willingtodeferprincipaland/orinterestpaymentsforseveralyears,thereby
providingshort-termcashflowrelief(i.e.,“breathingroom”)totheir
borrowers.168However,asyield-maximizingsurrogatesofthestate,Chinese
lendersareultimatelyfocusedonprotectingthebottomline(Chen2020a,
168Thereisanimportantdistinctionbetweencashflowreliefanddebtrelief.Chinesestate-ownedcreditorsaresubstantiallymorewillingtoprovidecashflowreliefthandebtrelief,whichistraditionallydefinedintheacademicliteratureas“thereductioninthenetpresentvalueofthedebtor’soutstandingobligationsduetotherestructuringagreement”(Hornetal.2022a:14).TheimportanceofthisdistinctioncameintosharperresolutionafterChinaagreedtoparticipateintheG20’sDebtServiceSuspensionInitiative(DSSI).InApril2021,thePresidentofChinaEximbankpubliclyclarifiedthat“debtsuspension[...]isneitherdebtreductionnordebtforgiveness.Oneshouldnottaketheopportunity[oftheCovid-19pandemic]toharmChina’sinterestsandtakeadvantageofChina”(TheExport-ImportBankofChina2021).
90
2020b,2023;Dreheretal.2021,2022).169Assuch,theyaregenerallyunwilling
toacceptsignificantfinanciallossesinnetpresentvalue(NPV)terms(Ministryof
FinanceofthePeople'sRepublicofChina2020;BonandCheng2021;Gardner
etal.2021;Hornetal.2022b).170
The3.0versionofAidData’sGCDFdatasetalsoshedslightonthekeyrolethat
escrowaccountreplenishmentplaysindebtreschedulingnegotiationswith
Chinesecreditors.Onceaborrowerhasdefaultedonitsrepaymentobligations
anditsChineselenderhasexhaustedthefundsinarepaymentreserveaccount
(DSRA),theChineselenderwilltypicallyinstructtheescrowbankto(a)
immediatelyreplenishtheaccountwithfundsfromtherevenue(salescollection)
account,and(b)blocktheborrowerfrommakinganywithdrawalsfromthe
revenue(salescollection)account.However,thisapproachisnotfoolproof,since
afinanciallydistressedborrowercanstopmakingdepositsintotherevenue
(salescollection)account,therebyeliminatingtheChineselender’ssecondline
ofdefense.Consequently,asmoreborrowersseekdebtrelief,Chineselenders
aredemandingacredibleprotectionagainst(another)defaultinexchangefor
short-termcashflowrelief.
TheGovernmentofAngola’s2020debtreschedulingwithChinaDevelopment
Bankisacaseinpoint.171Inlate2015,CDBissueda$15billionloantothe
GovernmentofAngolaandrequiredthattheborrowermaintainaminimum
balanceof$1.5billioninanescrowaccountasasourceofcashcollateral.Then,
theAngolanGovernmentfacedacashcrunch—duetoasharpdeclineinoil
prices—andhaddifficultyservicingitsdollar-denominateddebts.CDBagreedto
deferprincipalpaymentsaspartofadebtreprofilingagreementin2020and
171SeeID#95415andID#53063inthe3.0versionofAidData’sGCDFdataset.170ReinhartandTrebesch(2016)studydebtreliefoperationsduringtwoperiods—1920-1939and1978-2010—andfindthateconomicgrowthgenerallyincreasedfollowingdebtstockreductions(facevaluereductionstoloanprincipal)butnotdebtflowreductions(cashflowreliefviamaturity/graceperiodextensionsandinterestratereductions).169Duringdebtreschedulingnegotiations,Chineselendershavetraditionallyprovidedcashflowrelieftoborrowerssolongastherewasnosignificantnetpresentvalue(NPV)reductionintotalrepaymentstothelenderoverthelifetimeoftheloan(i.e.,theNPVofdebtservicepaymentsafterreschedulingwasnotlowerthantheNPVofdebtservicepaymentsbeforerescheduling).However,changemaybeafoot.InJanuary2023,IMFManagingDirectorKristalinaGeorgievaannouncedthat"[w]ehavereachedanunderstandinginprinciplethatChinawilldefactoacceptNPVreductiononthebasisofsignificantstretchingofthematuritiesandreductionofinterest.[...]InChinathereisnotyetaconsensustotakeupfronthaircuts”(Mfula2023).Reductionsinaloan’sNPVcanbeachievedviasubstantialinterestratereductions,lengthymaturity/graceperiodextensions,and/orfacevaluereductionstoloanprincipal(so-called“facevaluehaircuts”or“principalhaircuts”).
91
usethecashintheescrowaccounttosatisfytheborrower’sinterestpayment
obligationsfrom2020to2022.However,inanticipationoftheescrowaccount
balancebeingdepletedtonearlyzerobymid-2022,CDBdemandedthatthe
borrowerreplenishtheaccountto$1.5billionby2023.
Riskmitigationstrategy#3:Lendingwithhigherinterestrates,
shorterrepaymentperiods,moresafeguards,andmoresevere
penaltiesfordefault
Onewaytogoaboutde-riskinganoverseaslendingportfolio—onagoing
forwardbasis(“future-proofing”)—istoreducetheprovisionofconcessional
credit.Lendingonbelow-markettermsisriskyfromacreditorperspectivefor
multiplereasons.172First,anysubsidyfromacreditortoaborrowerimpliesthat
thecreditorwillachieveasmallerinvestmentreturnthanitwouldintheabsence
ofthesubsidy.Second,mostconcessionalloanscarrylowinterestratesthatare
fixedratherthanfloating,soacreditorthatprovidessuchloansmustbearallof
theriskassociatedwithits“costoffunds”increasingovertime.173Asthe“cost
offunds”increases(forexample,duetoanincreaseininterbanklendingratesor
anincreaseinthecostofborrowingviabondsissuedoncapitalmarkets),the
creditor’sincentivetoofferconcessionalloanswithlow,fixedinterestrates
weakens.Third,concessionalloansfromofficialcreditorsaresubjectto
substantiallylargerhaircuts(investorlosses)thanloansfromcommercial
creditors.174Therefore,ifSAFE—acentralgovernmentinstitutionwith
discretionaryauthoritytoentrustsurplusforeignexchangeto“officialcreditors”
likeChinaEximbankor“commercialcreditors”likeBankofChina(seeBox2b
formoreonSAFE’srole)—wantedtode-riskthecountry’soverseasloan
portfolio,itmightseektominimizefuturelosses(i.e.,future-prooftheportfolio)
174Dependingonthediscountingapproachthatisused,thereisa19.4percentagepointminimumdifferenceand38.6percentagepointmaximumdifferenceinaveragehaircuts(investorlosses)forofficialcreditorsversuscommercialcreditors(Schlegletal.2019).173Relatedly,thebankbearsallofthedownsideriskassociatedwithanincreaseintheopportunitycostoffunds.172Ifacreditorissuesaloantoaborrowerataninterestratethatislowerthanitsownborrowingterms(i.e.,“costoffunds”)atthetimetheloanisissued,thecreditorislendingtotheborrowerata“below-market”rate.Thereissignificant“costoffunds”variationacrossChinesebanksandloansdenominatedindifferentcurrencies(Chen2020a).WithrespecttoRMB-denominatedloans,thestate-ownedcommercialbanksmobilizefundsbyacceptingdeposits(withthedepositrateat1-2%),whilethestate-ownedpolicybanksmobilizefundsbyissuingbondswithyieldsof3-5%.Withrespecttoforeigncurrency-denominatedloans,thestate-ownedpolicybankscanborrowfromPBOCata2-3%interestrate.
92
byreiningintheamountofconcessionalcreditprovidedviaofficialcreditorsand
prioritizingtheprovisionofnon-concessionalcreditviacommercialcreditors.175
The3.0versionofAidData’sGCDFdatasetprovidesevidencethatBeijingis
indeedmovinginthisdirection.FigureA18demonstratesthattheweighted
averageinterestrateofofficialsectorlendingfromChinatoLICsandMICs
increasedfrom4.5%duringtheearlyBRIperiodto5%duringthelateBRI
period.Consistentwiththisshifttowardharderlendingterms,theweighted
averagerepaymentperiod(maturity)declinedfrom11yearsduringtheearlyBRI
periodto6yearsduringthelateBRIperiod(seeFigureA19).Theweighted
averagegrantelement—asummarymeasureoffinancialconcessionality
(discussedatgreaterlengthinSectionA-2intheAppendix)—fellby9
percentagepointsbetweentheearlyBRIperiodandthelateBRIperiod:from
25%to16%(seeFigureA20).176
InordertogaugewhetherBeijingismakingaconcertedefforttoreininthe
provisionofconcessionalcross-bordercredit,itisalsoworthconsideringifithas
increasedorreduceduseofthetwoprimaryinstrumentsthatithasestablished
forconcessionallendingtooverseasborrowers:governmentconcessionalloans
(GCLs)andpreferentialbuyer’scredits(PBCs)fromChinaEximbank.177Figure
A12providesevidencethatGCLsandPBCshaveindeedfallenoutoffavor,as
theyrepresentarapidlydwindlingpercentageofnewoverseaslending
commitmentsfromChinatoLICsandMICs(18%in2014and3%in2021).178
178Thefactthatthesetwoconcessionallendingwindowsareincreasinglyinaccessiblehassignificantimplicationsforsovereignborrowers.ThetermsandconditionsinPBCandGCLcontractsarefairlystandardized(Gelpernetal.2022)andreasonablywell-understoodamongdebtmanagementofficersandtransactionlawyersinLICandMICfinanceministries(e.g.,BancoCentraldeBolivia2016;NEDA2017;EconomicRelationsDivisionoftheGovernmentofthePeople'sRepublicofBangladesh2023).Theyincludelow(fixed)interestrates,longmaturities,generousgraceperiods,nopenaltyinterestprovisions,andstrictrequirementstouseChinesecontractorsandkeepalltermsandconditionsconfidential,among177GCLsareRMB-denominatedloansthatareissuedtogovernmentinstitutionsandprovidedonbelow-marketterms(usually20-yearmaturities,5-yeargraceperiods,and2%interestrates).PBCsareUSD-denominatedloansthatareissuedtogovernmentinstitutionsontermsthataremoregenerousthanprevailingmarketterms,butslightlymoreexpensive(higherinterestrates,shortermaturities,andshortergraceperiods)thanGCLs(Morrisetal.2020;Hornetal.2021;Dreheretal.2022).176TheweightedaveragegrantelementofChineselendingtooverseasborrowersdeclinedby19percentagepoints—from30%in2014to11%in2021(seeFigureA21).175SomeChinesestate-ownedlendersparticipateindebtrestructuringsthatare(loosely)coordinatedwithofficialcreditorsfromothercountries,whileothersdonotandwishtobetreatedascommercialcreditors(Gardneretal.2021;Hornetal.2022b).RecenteventsinZambiasuggestthatBeijingconsiderstheclaimsofitsofficialcreditorstoincludethoseofChinaEximbankandthoseofotherChinesestate-ownedcreditorsthatareinsured(“guaranteed”)bySinosure.BeijingevidentlyconsidersCDBdebtsthatarenotbackedbyaSinosurecreditinsurancepolicytobetheclaimsofacommercialcreditor(Setser2023b).
93
Duringthepre-BRIandearlyBRIera,GCLsandPBCswereBeijing’s“workhorse”
lendinginstruments.ChinaEximbankissued669GCLsandPBCsworth$121
billionfrom2000to2017.179Alloftheseloanscarrylow,fixedinterestratesand
nearly100%ofthemqualifyasconcessionalloansunderOECD-DAC
measurementstandards.180However,Figure2.13providesevidenceofashift
awayfromfixedinterestratelendingandtowardvariableinterestratelending:
whereas60%ofnewlendingcommitmentsfromChinatoLICandMIC
borrowerscarriedvariableinterestratesin2014,thisfigureincreasedto90%by
2021.Theseportfolio-levelsummarystatisticssuggestthataneffortisunderway
totransfermoreriskfromlenderstoborrowers.Risk-basedloanpricingmodels
usuallyrelyonavariableinterestrate,whichinturnisbasedonafloatingmarket
referencerate—suchastheLondonInterbankOfferedRate(LIBOR)orEuro
InterbankOfferedRate(EURIBOR)—andapremiumthataccountsfortherisk
profileoftheborrower.ItisalsoimportanttokeepinmindthatSAFEhastasked
Chinesestate-ownedlenderswithmaximizinginvestmentreturnswithin
acceptablerisklevels,whichismoredifficulttodoviafixedinterestlending
whenvariable(floatingmarket)interestratesareincreasingandpushingupthe
opportunitycostoffundsforlenders(seeBox2b).
18099.6%(260of261)oftheGCLsthatwereissuedbetween2000and2017withgrantelementobservationsmettheOECD’s25%grantelementthresholdofconcessionality.99.6%(247of248)ofthePBCsthatwereissuedbetween2000and2017withgrantelementobservationsmettheOECD's25%grantelementthresholdofconcessionality.179Between2018and2021,ChinaEximbankissued49GCLsworth$9billionand40PBCsworth$22billion.otherthings.However,thetermsandconditionsthatareincludedinmore“exotic”Chineselendinginstruments(likePBOCswapcontracts,CDBliquiditysupportfacilities,andaccountsreceivablefinancingarrangementswithChinesestate-ownedcommercialbanks)arenotwell-understood.
94
Figure
2.13:
Percentage
of
lending
portfolio
using
fixed
or
variable
interest
ratesFigure2.13
Notes:VariableinterestratesthatChinesestate-ownedcreditorsuseasbenchmarksincludeLIBOR,
EURIBOR,SHIBOR,BADLAR,CIRR,JIBOR,LPRandBADCOR.Weexcludeallloansforwhichwecannot
determineifafixedorvariableinterestratewasapplied.
Together,theshiftfromdollar-denominatedlendingtoRMB-denominated
lendingandtheshiftfromfixedinterestratelendingtovariableinterestrate
lendinghaveresultedinadramaticturnawayfromLIBOR-basedlendingand
towardSHIBOR-basedlending.Figure2.14demonstratesthatthepercentageof
China’svariableinterestratelendingtoLICsandMICsbasedonLIBORsharply
declinedfrom100%in2008to29%in2021.181ThefallofLIBORwas
accompaniedbytheriseofSHIBOR.ThepercentageofChina’svariableinterest
ratelendingtoLICsandMICsbasedonSHIBORsoaredfrom0%in2008to70%
in2021.182
182SHIBORandLPRarebothvariableinterestratessetbythePBOC.Figure2.14combinesallloanswithSHIBOR-andLPR-basedinterestratesintoasinglecategory.However,LPRisnotyetwidelyusedinChina’soverseaslendingportfolio.Thereareonly2LPR-basedloanrecordsinthe3.0versionoftheGCDFdataset.TheshiftthathastakenplaceislargelyashiftfromLIBORtoSHIBOR.181China'sLICandMIClendingprogramwasalmostexclusivelydollar-denominatedin2009(seeFigure2.2).
95
Figure
2.14:
Composition
of
variable
interest
rate
lending
portfolioFigure2.14
Notes:LIBORreferstotheLondonInterbankOfferedRate.SHIBORreferstotheShanghaiInterbank
OfferedRate.LPRreferstotheChinaLoanPrimeRate.The“other”categoryincludesloanswithvariable
interestrates,suchasEURIBOR,BADLAR,CIRR,JIBOR,andBADCOR.
Anotherimportantreasonwhyconcessionalityattheportfoliolevelhasfallen
overtimeisthatsomeofBeijing’sbiggestsovereignborrowershaveswapped
lessexpensivedebtformoreexpensivedebt.Hornetal.(2023a)provide
evidencethattheinterestratesonChina’semergencyrescueloansexceedthe
interestratesontheexistingdebtstocksoftheborrowers.Theyalsoprovide
evidencethatChina’semergencyloanshaveveryshortmaturities(inmany
cases,1yearorless),althoughitisnotunusualforsuchloanstobe“rolled
over”whentheyreachtheirofficialmaturitydates(seeBox2cformoreonthis
issue).183Assuch,refinancingwithbailoutsfromBeijingtypicallydoesnot
reducethenetpresentvalueofaborrower’sdebtstock,whichhighlightsan
emergingtensionbetweenthoseprovidingandthosereceivingnewloans:
183Rolloverdebtcomesintwovarieties:(1)loansthatreachtheiroriginalcontractualmaturitydatesandsecurefinalmaturitydateextensions;and(2)loansthatarerepaidontheiroriginalcontractualmaturitydatesandreissued(withsimilarordifferentfacevaluesandborrowingterms)andassignednewmaturitydates.
96
financiallydistressedborrowerswantcheapcreditthatwillhelpthem“rideout
thestorm,”butChinesestate-ownedcreditorsareunsureiftheirbiggestBelt
andRoadborrowersareilliquidorinsolvent,sotheyarepricinginahigherrisk
ofdefaultwhentheyprovideadditionalcredit.
Box2c:ShouldemergencyliquiditysupportfromPBOCswaplinesbetreated
asshort-termorlong-termdebt?
Between2016and2021,thePBOCuseditsswaplinenetworktoprovidenearly$150billioninemergencyliquiditysupporttocentralbanksinLICsandMICs(Hornetal.2023a).Theseborrowingshavecreatedanewmeasurementchallengefororganizationsthat(a)tracktheinternationallendingactivitiesofexternalcreditors,and(b)monitorlevelsofpublicdebtexposureinthedevelopingworld.
ThereportingdirectivesoftheOECD’sCreditorReportingSystem(CRS)specifythat“[l]oanswithamaturityofoneyearorlessarenotreportableinDACstatistics”(OECD2021:51).184Similarly,governmentsthatparticipateintheWorldBank’sDebtorReportingSystem(DRS)areaskedtoreporttheirlong-termdebtrepaymentobligationstoexternalcreditorsonanannualbasis.185
Long-termdebtisdefinedintheDRSreportingmanualasdebt“withanoriginalcontractualorextendedmaturityofmorethanoneyear[…]”(WorldBank2000:4).
Basedonanarrowinterpretationoftheprevailinginternationalreportingrules,PBOCswaplineborrowingsshouldnotbereportedtotheDRSortheCRS.Nearlyalloftheseborrowingscarrydejurematuritiesofoneyearorless(i.e.,theyareinitiallyscheduledforrepaymentin12monthsorless).However,centralbanksthatborrowfromthePBOCfrequentlyseetheirfinalmaturitydatesextended—ortheyrepeatedlyreceiveshort-termloanstorefinancematuringdebts.Hornetal.(2023a)provideevidencethatthedefactomaturityoftheaveragePBOCswaplineborrowingis3.5years.
ThecustodiansoftheDRSareawareofthegapbetweenthedejureanddefactomaturitiesofPBOCswaplineborrowingsandtheunderreportingofPBOCswapdebt.InOctober2020,theysoundedthealarm,questioningwhether“currencyswaparrangementsthatrepresentloansfromothercentralbanksarereflectedinexternaldebtstocksoflow-andmiddle-incomecountries”(WorldBank2020a:13).Atthattime,theyclarifiedthat“[t]heDRS[…]considersone-year[centralbank]depositsthatareconsistentlyrolledover(defacto)tobelong-termdebt”(WorldBank2020a:13).Theyalsoemphasizedthat“[t]hetransparencyofdataondebtmustevolvetokeeppacewithanever-changingcreditorlandscapeandwithnewandincreasinglycomplexdebt-likeinstrumentsanddatarequirements,”andthatoneoftheirtopprioritiesis
185TheWorldBank’sDebtorReportingSystem(DRS)hasservedastheprimaryinternationalreportingsystemforpublicdebtsince1951.Itsupportsthepublicationofawidelyuseddatasource:theWorldBank’sInternationalDebtStatistics(IDS).184Since1973,theOECD’sCreditorReportingSystem(CRS)hascollectedandpublisheddataonofficialsectorfinancialflows(ODAandOOF)fromDACandnon-DACcountries.
97
“incorporatingCentralBankdepositsandcurrencyswapslinesintotheDRSdataset”(WorldBank2021:29).186
The3.0versionofAidData’sGCDFdatasetcapturesthefullrangeofChina’sinternationalrescuelendingoperations.FigureA14demonstratesthatanincreasingproportionofChina’sofficialsectorlendingtoLICsandMICsconsistedof“rollover”emergencyrescueloansduringtheearlyBRIperiod(8%)andthelateBRIperiod(34%).
Anotherwayofmitigatingelevatedlevelsofdefaultriskistoattachstronger
repaymentsafeguardstonewloans.Beijingappearstobefollowingthis
strategy.Figure2.15demonstratesthatthepercentageofChina’soverseas
lendingportfoliosupportedbycollateralincreasedfrom42%duringthepre-BRI
andearlyBRIperiod(2000-2017)to56%inthelateBRIperiod(2018-2021).By
2021,nearlythree-quarters(72%)ofChina’soverseaslendingtoLICsandMICs
wascollateralized.
Figure
2.15:
Loan
portfolio
backed
by
collateralFigure2.15
Notes:Collateralizedlendingcommitmentsareidentifiedwiththe“collateralized”fieldinthe
3.0versionoftheAidData’sGCDFdataset.
186ThecustodiansoftheCRS,bycontrast,havenotaddressedthisissue—mostlikelybecauseChinadoesnotparticipateinitsreportingsystem.
98
Thereisalsosomeevidencethat,inordertoreducethelikelihoodofdefault
and/orminimizepost-defaultlosses,Chinesestate-ownedcreditorsare
includingstifferpenaltiesfordefaultintheircontractswithborrowers.The3.0
versionofAidData’sGCDFdatasetincludesanewvariablethatmeasuresthe
default(penalty)interestratesthatapplytoindividualloansfromChinese
state-ownedentities.Figure2.16,whichdrawsuponthedata,demonstratesthat
theaveragedefault(penalty)interestratemorethandoubledbetweentheearly
BRIperiodandthelateBRIperiod(0.584%from2014-2017and1.175%from
2018-2021).187Themaximum(observed)default(penalty)interestratealso
increased—from3%duringtheearlyBRIperiodto8.7%duringthelateBRI
period.188ThesefindingsaredifficulttoreconcilewiththoseofSAIS-CARI,which
hasconcludedthatthey“seenoevidenceofpenaltyinterestrates”inChina’s
overseaslendingtodevelopingcountries(Ackeretal.2020:31).189
Figure
2.16:
Early
versus
late
BRI:
penalty
interest
ratesFigure2.16
Notes:Default(penalty)interestratesareidentifiedwiththe“DefaultInterestRate”fieldinthe3.0version
oftheAidData’sGCDFdataset.
189LesttherebeanyconfusionaboutwhethersovereignborrowersareresponsibleformakingpenaltyinterestpaymentstotheirChinesecreditors,theGovernmentofSudandisclosedthatitowed$127millionofpenaltyinteresttoChinesecreditorsasofMarch31,2022(CBOS2023).TheGovernmentofZimbabweandtheGovernmentofSerbiahavealsoacknowledgedincurringpenaltyinterestundertheirloanagreementswithChinaEximbank(NBS2007;MOFED2022).188ItisalsoimportanttokeepinmindthatthecreditorcompositionofChina’soverseaslendingportfolioisshiftingawayfromfinanciersthatusenoorlowpenaltyinterestrates.Interest-freeloansfromMOFCOM,governmentconcessionalloansfromChinaEximbank,andpreferentialbuyer’screditsfromChinaEximbankdonotcarrypenaltyinterestrates—andtheyareallonthedecline(seeFigures2.7andA12).However,lendingfromChinesestate-ownedcommercialbankscarriespenaltyinterestratesinthe0.5%to3%range—anditisontherise(seeFigure2.7).187InmostChineseloanagreementsthatincludesuchprovisions,thedefault(penalty)interestrateappliestooverdueprincipaland/oroverdueinterestamounts.See,forexample,Section6.9oftheChinaEximbankbuyer’screditloanagreementforPhase1oftheStandardGaugeRailwayProject(accessibleviahttps://www.dropbox.com/s/5j3alwun2tv8wk2/SGR%20BCL%202014.pdf).
99
Figure2.17zoomsinonfoursovereignborrowers—Ghana,Uganda,Guyana,
andSerbia—forwhomwehavereasonablycompletedataovertimeonthe
default(penalty)interestratesattachedtoloansfromChinesestate-owned
creditors.Onecanseethat,inallofthesecountries,default(penalty)interest
ratesvariedbetween0-0.33%until2015.Theseratesremainedmostlystable
from2016to2020,althoughtherewasaslightuptickintwocountries.However,
by2021or2022,default(penalty)interestratesreached2%inSerbia,2%in
Uganda,3.5%inGuyana,and4.8%inGhana.Thesenewriskmitigation
measuresbyBeijingposeachallengetoborrowersinthedevelopingworldthat
israrelytakenintoconsideration:thosewhocontinuetocontractnewdebtfrom
Beijingmustbeawareofthedangerofcompoundingarrearsduetopenalty
interest.
100
Figure
2.17:
Average
penalty
interest
rates
in
select
countriesFigure2.17
Notes:Thischartshowsunweightedaveragedefault(penalty)interestratesonloansfromofficialsector
institutionsinChinatogovernmentandstate-ownedborrowinginstitutionsinGhana,Serbia,Guyanaand
Uganda.Theabsenceofavalueinagivencountry-yearindicatesmissingpenaltyinterestratedata.
Observationsaredrawnfrom2000-2021GCDF(Version3.0)dataandpreliminary2022GCDFdata.In
additiontoloancommitments,pledgedloansandsuspended/canceledloancommitmentsareincluded
sinceborrowersmaybemorelikelytorejectloanofferswithhighpenaltyinterestratesandsuspendor
cancelloancommitmentswithtermsthattheyperceivetobeonerous.
101
The3.0versionofAidData’sGCDFdatasetalsoprovidesevidencethatwhen
borrowersfailtohonortheirrepaymentobligations,Chineselenderswillseekto
recoverthepenaltyinterestthattheyareowedbysweepingcashoutofescrow
accounts(whentheyhaverecoursetosuchcollateral).Bywayofillustration,
considerhowChinaEximbankrespondedtoanoverduedebtrepaymentfrom
TanzaniaPetroleumDevelopmentCorporation(TPDC)—aparastatalthatis
wholly-ownedbytheGovernmentofTanzania—fortheSongoSongotoDarEs
SalaamGasPipelineandNaturalGasProcessingPlantsatMnaziBayandSongo
SongoProject.Asasourceofcollateralfora$275millionbuyer'screditloan
(BCL)thatChinaEximbankissuedin2012,TPDCdepositedapproximately$60
millioninescrowaccountsaccessibletothelender.190However,bythefirsthalf
of2017,TPDChadaccumulatedarrearstoChinaEximbank.Thelender
respondedbysweepingcashoutofoneoftheescrowaccountsbetweenJuly
2017andJune2018.AccordingtoTPDC’sauditedfinancialstatements,it
withdrewthefunds“asapenaltyfor[theborrower’s]laterepaymentofdue
installment”(TheControllerandAuditorGeneraloftheNationalAuditOfficeof
Tanzania2018:76).191
Riskmitigationstrategy#4:Takingadifferentiatedapproachwith
borrowersthatpresenthighandlowlevelsofrepaymentrisk
Previousresearchindicatesthatwhencreditorsareconfrontedwithasudden
waveofdefaultsorasharpincreaseinnon-performingloans,theymayseekto
rebalancetheriskprofilesoftheirassetportfoliosbyissuingfewerloansto
potentiallyriskyborrowers—orbypullingbackfromanentireassetclasswitha
newriskprofile(KaminskyandReinhart2000;Longstaffetal.2011;Brooksetal.
2015;Gilchristetal.2022).Consistentwiththisexpectation,the3.0versionof
AidData’sGCDFdatasetprovidesevidenceofacompositionalshiftin
191InJune20,2012,ChinaEximbankandTanzania’sMinistryofFinancesigneda$1,225,327,000financingagreementfortheSongoSongotoDarEsSalaamGasPipelineandNaturalGasProcessingPlantsatMnaziBayandSongoSongoProject.TwoloanswereissuedtoTanzania’sMinistryofFinance:(1)a$275millionbuyer’screditloan(BCL)withaninterestrateof6-monthLIBORplus430basispoints,a1.83-year(22-month)graceperiod,anda12.83-year(154-month)maturity,and(2)a$920millionpreferentialbuyer’scredit(PBC)witha20-yearmaturity,a7-yeargraceperiod,anda2%interestrate.TheBCLandPBCproceedswerethenon-lentfromTanzania’sMinistryofFinancetotheTanzaniaPetroleumDevelopmentCorporation(TPDC).SeeID#59733and59752inthe3.0versionofAidData’sGCDFdataset.190AccordingtoTPDC’sauditedfinancialstatements,theescrowaccounts“wereopenedtosecurerepaymentofprincipalandpaymentofinterestandfeesundertheloanagreements”andtheminimumcashbalancesintheseaccountsfunctionedassourcesof“collateral.”
102
non-emergencylendingfrompublicsectortoprivatesectorborrowers.This
moveawayfromthesovereigndebt“assetclass”isparticularlynoticeable
between2020and2021—whentheeffectsoftheCOVID-19pandemicwere
mostacute.TheshareofChina’snon-emergencylendingcommitmentsto
privatesectorborrowersinLICsandMICssoaredfrom21%in2020to54%in
2021,whilethesharedevotedtopublicsectorborrowersshrankfrom67%in
2020to43%in2021(seeFigure2.18).192
192Herewedefinepublicsectorloansasthesumofcentralgovernmentdebt,centralgovernment-guaranteeddebt,andotherpublicsectordebt(asdescribedinSectionA-5intheAppendix),whichisconsistentwiththeIDSdefinitionofPPGdebt.IntheAppendix,wereplicatethisgraphforChina’stotal(emergencyandnon-emergency)lendingcommitmentstoborrowersinLICandMICs(seeFigureA10).TheincorporationofChina’semergencylendingcommitmentsdisguisesthepivotawayfrompublicsectorlendingandtowardprivatesectorlending(asemergencylendingcommitmentsarelargeandexclusivelychanneledtosovereignborrowers).
103
Figure
2.18:
Composition
of
non-emergency
lending
portfolioFigure2.18
Notes:ThisgraphshowsthecompositionofChina’snon-emergencylendingportfolio(asmeasuredin2021
constantUSD)inLICsandMICsaccordingtotheextenttowhichhostgovernmentsmayeventuallybe
liablefordebtrepayment.Centralgovernmentdebtandotherpublicsectordebtrepresentloanswhere
theborrowerisagovernmentagencyorawholly-ormajority-ownedstateentity.Centralgovernmentdebt
representsloansthathaveasovereignguaranteefromthehostgovernment.Potentialpublicdebt
representsloanstoentities(includingspecialpurposevehiclesorjointventures)wherethehost
governmenthasaminoritystake.Privatedebtcapturesloanstoprivateentities.
Anotherpotentialde-riskingstrategyistotakeadifferentiatedapproachacross
borrowercountriesthatpresentvaryinglevelsofrepaymentrisk.FiguresA22
andA23provideevidencethatsuggestsBeijingisadoptingthisapproach:
Chinesestate-ownedcreditorshavelenttofinanciallydistressedcountriesat
higherinterestratesandwithshorterrepaymentperiodsthancountriesthatare
104
notinfinancialdistress.Consequently,theconcessionalitylevel(weighted
averagegrantelement)ofofficialsectorlendingfromChinaisconsistentlylower
incountriesexperiencingfinancialdistressthanincountriesnotexperiencing
financialdistress(seeFigureA24).193
Figure
2.19:
Percentage
of
variable
interest
rate
lending
to
countries
in
and
not
in
financial
distressFigure2.19
Notes:VariableinterestratesincludeLIBOR,EURIBOR,SHIBOR,BADLAR,CIRR,JIBOR,LPRandBADCOR.
Todetermineifacountryexperiencedfinancialdistressinagivenyear,weusethebinarymeasurethatis
describedinBox1a.
Chinesestate-ownedcreditorshavealsochangedthewaythattheylendto
financiallydistressedcountriesovertimebyshiftingtowardvariableinterestrate
lending(seeFigure2.19).Thisapproachtolendingfollowsarisk-basedpricing
modelbyaddingaborrower-specificmargin—thataccountsforthecreditprofile
characteristicsoftheborrower—toamarket-basedreferenceinterestrate.194In
194Risk-basedloanpricingmodels(a)chargeriskyborrowershigherprices(i.e.,attachlargerriskpremiatoborrowersthatpresentahighprobabilityofdefault),and(b)accountfortheexpectedmagnitudeofpost-defaultlosses(i.e.,howmuchmoneythecreditorexpectstoloseiftheborrowerdefaults).193Between2000and2021,theweightedaveragegrantelementofofficialsectorlendingfromChinatocountriesexperiencingfinancialdistresswastenpercentagepointslowerthanincountriesnotexperiencingfinancialdistress(seeFigureA25intheAppendix).
105
theearly2000s,therewasnoevidencewhatsoeverofanyvariableinterestrate
lendingbyChinesestate-ownedcreditorstoLICsorMICsinfinancialdistress.
However,by2021,morethan80%ofChina’svariableinterestratelendingwas
directedtocountriesinfinancialdistress.
The3.0versionofAidData’sGCDFdatasetalsorevealsthatChinaisincreasingly
collateralizingloanstocountriesindirefinancialstraits:theshareofChina’s
collateralizedlendingportfoliodirectedtocountriesinfinancialdistress
increasedfromzeroattheturnofthecenturyto74%by2021(seeFigure
2.20).195ThischangeispartofabroaderpatterninChina’suseofcredit
enhancements,whichAidDatadefinesaslendingbackedbyacreditinsurance
policy,athird-partyrepaymentguarantee,and/orcollateral.FigureA29divides
China’sLICandMIClendingportfoliointothreecategories:(1)credit-enhanced
lendingtocountriesinfinancialdistress;(2)credit-enhancedlendingtocountries
notinfinancialdistress;and(3)lendingthatisnotcredit-enhanced.Itshows
almostnoproportionalincreaseincredit-enhancedlendingtocountriesthatare
notinfinancialdistress,butalargeproportionalincreaseincredit-enhanced
lendingtocountriesthatareinfinancialdistress.
195Figure2.20showstheproportionofofficialsectorlendingcommitmentsfromChina(inconstant2021USD)toLICsandMICsthatfallsintothreecategories:(1)loansthatarecollateralizedandaregoingtoaborrowingcountrythatisindistressatthetimeoftheloancommitment,(2)loansthatarecollateralizedbutaregoingtoaborrowingcountrythatisnotexperiencingfinancialdistressatthetimeoftheloancommitment,and(3)loansthatarenotcollateralizedatthetimeofcommitment.TheunderlyingdatafromFigure2.20demonstratethat,while51%ofChina'scollateralizedlendingcommitmentstoLICsandMICsweredirectedtodevelopingcountriesinfinancialdistressduringthepre-BRIperiod(2000-2013),thisfigureincreasedto82%intheearlyandlateBRIperiods(2014-2021).
106
Figure
2.20:
Composition
of
lending
to
countries
in
and
not
in
financial
distressFigure2.20
Notes:Todetermineifacountryexperiencedfinancialdistressinagivenyear,weusethebinarymeasure
thatisdescribedinBox1a.
Giventhatthelarge-scaleprovisionofemergencyrescueloanstofinancially
distressedborrowershasincreasedChinesebankexposuretorepaymentrisk,
anotherpotentialportfoliorebalancingstrategyistodirectproject(or
nonemergency)loanstolessriskyborrowers.Asapreliminarytestofwhether
Chinesestate-ownedcreditorsmayberespondinginthisway,wefirstmeasure
thepercentageofChina’semergencyrescueloanportfoliosupportingborrower
countrieswithcreditratingsin“CandDterritory”—i.e.,countriesthatMoody’s,
Standard&Poor’s,andFitchidentifyasindefaultorpresentingahighriskof
default.WethenmeasurethepercentageofChina’sprojectloanportfolio
supportingborrowercountrieswithcreditratingsinCandDterritory.Figure
2.21providesthesesummarystatistics,whichdemonstratethattheoverallcredit
qualityofChina’sprojectloanportfolioissubstantiallybetterthanthatofits
emergencyrescueloanportfolio.196
196Projectloansaredefinedasthoseinthe3.0versionofAidData’sGCDFdatasetthatarecategorizedasinvestmentprojectloans(IPLs).
107
Figure
2.21:
Repayment
risk
in
emergency
lending
portfolio
versus
project
lending
portfolioFigure2.21
Notes:CountriesindefaultoratahighriskofdefaultrepresentLICsandMICswithscoresof5orlesson
thesovrateindex(seeBox1a).CountriesnotatahighriskofdefaultrepresentLICsandMICswithscores
above5onthesovrateindex.TheWorldBank’ssovrateindexisameasureofrepaymentriskthatvaries
from0to21,withhigherscoresindicatinglowerlevelsofsovereigncreditrisk(Koseetal.2022).
Country-yearobservationswithoutofficialsectorChineselendingcommitmentsorsovratescoresare
excludedfromthefigure.MOFCOMinterest-freeloancommitments(whicharetypicallyissuedwithouta
credibleexpectationofrepayment)areexcludedfromthecalculation.
Wealsofindevidenceofportfoliorebalancingovertime.AccordingtoFigure
A30,Chinareducedthepercentageofitsnon-emergencyloanportfolio
supportingborrowercountrieswithcreditratingsinCandDterritory—from
7.9%duringtheearlyBRIperiodto0.8%duringthelateBRIperiod.One
potentialwayofshiftingthenon-emergency(andproject)lendingportfolioaway
fromcountriesthatpresenthighlevelsofrepaymentriskisviasyndicationand
multilateralization.Figure2.22demonstratesthatwhenBeijingoutsourcesrisk
management(includingtheloanoriginationprocess),lesscreditischanneledto
countriesthatareindefaultorpresentahighprobabilityofdefault.However,is
thereanyevidencethatBeijingisactuallypivotingawayfromitsbilateral
lendinginstitutionsandtowardmultilaterallendinginstitutionsandsyndicated
loanarrangements?Thisisthenextquestionthatweattempttoanswer.
108
Figure
2.22:
Repayment
risk
in
bilateral,
multilateral,
and
syndicated
lending
portfolioFigure2.22
Notes:CountriesindefaultoratahighriskofdefaultrepresentLICsandMICswithscoresof5orlesson
thesovrateindex(seeBox1a).CountriesnotatahighriskofdefaultrepresentLICsandMICswithscores
above5onthesovrateindex.TheWorldBank’ssovrateindexisameasureofrepaymentriskthatvaries
from0to21,withhigherscoresindicatinglowerlevelsofsovereigncreditrisk(Koseetal.2022).
Country-yearobservationswithoutofficialsectorChineselendingcommitmentsorsovratescoresare
excludedfromthefigure.MOFCOMinterest-freeloancommitments(whicharetypicallyissuedwithouta
credibleexpectationofrepayment)areexcludedfromthecalculation.
Riskmitigationstrategy#5:Scalingdownbilaterallending
operationsandscalinguplendingoperationsviasyndicationand
multilateralization
IfChina’sbilateraldevelopmentfinanceinstitutionsdonothavesufficiently
strongriskmanagementguardrailsinplace,oneoptionistoreformthese
institutionsfromwithin.However,theseinstitutionsarepoliticallypowerfuland
slowtochange(Chen2020a,2020b,2023).Anotheroptiontoimprovetherisk
profileofthecountry’soverseaslendingprogramistodialdowntheprovisionof
creditthroughbilateralchannelsanddialuptheprovisionofcreditthrough
syndicatedandmultilaterallendingchannels.
Syndicatedloansallowagroupoflenders(a“syndicate”)topooltheirfundsand
sharecreditrisk.Whenatransactionisfinancedthroughasyndicated
arrangement,allmembersofthesyndicatemustagreetoacommonsetof
contractualtermsandconditions,includingthestandardsandsafeguardsthat
willgovernthetransaction.Giventhateachlenderhasitsownstandardsand
safeguards,themembersofasyndicatecanseektoreconcile(hybridize)their
109
respectivestandardsandsafeguards.However,itismorecommonforthe
membersofthesyndicatetodefertothestandardsandsafeguardsofone
memberofthesyndicate.Inmostsyndicatedloanarrangements,a“lead
arranger”—sometimesreferredtoasthe“arranginglender”“leadlender,”
“leadmanager,”or“underwriter”—establishesthetransaction’skeytermsand
conditions,whichcannotbeamendedwithouttheconsentofallmembersofthe
syndicate.197Itisalsocustomaryforallmembersofthesyndicatetousea
commonsetofduediligencestandardstovettheborrowinginstitutionandthe
proposedtransaction.198Hereagain,theleadarrangerisusuallyresponsiblefor
identifyingandapplyingtheduediligencestandards(DennisandMullineaux
2000;Ivashina2009).
Figure
2.23:
Composition
of
non-emergency
lending
portfolio
by
channel
of
deliveryFigure2.23
Notes:Allemergencyrescueloansareexcluded.The“loansthatinvolvemultilateralinstitutions”category
includeloanswhereamultilateralagencyhassomeformalrole,suchasanentrustedloanagreementora
co-financingarrangement.
198Sufi(2007)demonstratesthatleadarrangersreducethecostsofduediligenceforallothersyndicateparticipants.197However,othertermsandconditionscanusuallybeamendedwiththeconsentofthe“majoritylenders.”
110
Figure2.23providesevidencethat,attheturnofthecentury,China’s
non-emergencylendingprograminLICsandMICsexclusivelyconsistedof
bilateralloans—i.e.,loansissuedbyasinglelendertoasingleborrower.
However,overtime,Beijinghasmovedawayfromthisapproach,rampingupits
useofsyndicatedloanarrangements.Itbeganexperimentingwiththismore
collaborativewayofissuingcreditduringthepre-BRIperiodandearlyBRI
period,butmadesyndicationcentraltothecountry’soverseaslendingstrategy
duringthelateBRIperiod.By2021,50%ofChina’snon-emergencylending
programinLICsandMICsconsistedofsyndicatedloancommitments.199
Thisempiricalpatternfliesinthefaceoftheconventionalwisdomthat
“[e]mergingeconomies’debttoChinais[...]non-marketable”(Kondoetal.
2022).200Thatwascertainlytrue25yearsago,butitisnolongerthecase:halfof
China’snon-emergencylendingportfoliointhedevelopingworldnowconsists
ofsyndicatedloans(Figure2.23).
China'sstate-ownedcommercialbanksareparticularlyfocusedonthisapproach
tocross-borderlending.The3.0versionofAidData’sGCDFdataset
demonstratesthattheyaremoreheavilyengagedthanthecountry’spolicy
banksinsyndicatedlendingtoLICsandMICs.In2021,84%ofChina’s
state-ownedcommercialbanklendingtoLICsandMICsreliedonsyndicated
loaninstrumentsandtheremaining16%reliedonbilateralloaninstruments.By
comparison,only36%ofChina’spolicybanklendingtoLICsandMICsreliedon
syndicatedloaninstrumentsandtheremaining64%reliedonbilateralloan
instruments.
Anotherimportantaspectofthecountry’spivotfrombilaterallendingto
syndicatedloansisthegrowinginvolvementofnon-Chinesebanksand
multilateralinstitutionsinthesyndicatesthatarebeingestablished.FigureA32
demonstratesthatroughly80%ofthesyndicatedloansfromofficialsector
200Asyndicatedloanconstitutesmarketabledebtinthatthearranginglenderresponsiblefororiginatingandstructuringthetransactioncandistribute/sellpartoftheloantootherbanksandnonbankinstitutionsthroughamarketingandsyndicationprocessandtheloancanbetradedonsecondarymarkets.19949%ofChina’snon-emergencylendingprograminLICsandMICsconsistedofbilateralloancommitmentsin2021.Beijing’spivottowardlendingviasyndicatedloansisespeciallynoticeableintheprojectlendingportfolio(seeFigureA36forsupplementaryevidence).
111
creditorsinChinatoLICsandMICsinvolvenon-Chinesebankparticipants.201A
non-trivialpercentageoftheseloansalsoinvolvemultilateralinstitutions(see
FigureA33).202Multilateralandnon-Chinesebankparticipationcouldhave
far-reachingconsequencesiftheirstandardsandsafeguardsprevailastheones
thatallothersyndicatemembersmustfollow.203InTable2.4,weprovide
metadataforanillustrativesetofsyndicatedloanagreementsinvolvingChinese
state-ownedcreditorsandnon-Chinesecreditors.Onecanseethatsyndicated
loanswithChineseandnon-ChineseparticipantsfrequentlyrelyonWestern
commercialbanksandmultilateralinstitutionstoserveasleadarrangers,which
isconsistentwiththenotionthatBeijingisde-riskingitsoverseasloanportfolio
byoutsourcingriskmanagement.204InChapter4,whereweaddressthisissueat
greaterlength,wefindthatChineseparticipationinsyndicatedloanagreements
withnon-Chinesebanksandmultilateralinstitutionsconsistentlyresultsin
strongerratherthanweakerriskmanagementstandardsandsafeguards.
Table
2.4:
Illustrative
set
of
syndicated
loan
agreements
with
Chinese
and
non-Chinese
bank
participantsTable2.4
Illustrative
set
of
syndicated
loan
agreements
with
Chinese
and
non-Chinese
bank
participants
CountryProjectYearValueofLoanLeadArrangerChineseParticipantsTotalNumberofParticipantsSierraLeonePortElizabethIIUpgradingandExpansionProject2017$659millionICBCICBC,ChinaEximbank2
204WithanearliervintageoftheGCDFdataset,Joosseetal.(2023)usesocialnetworkanalysistomaptheinternationalnetworkofnon-ChinesefinanciersthatfacilitatetheparticipationofChinesestate-ownedcreditorsinsyndicatedloanagreements.Theyconcludethatthetenmostimportantnetwork“brokers”includetwomultilateralinstitutions(Inter-AmericanDevelopmentBankandAfricanDevelopmentBank)andeightWesterncommercialbanks(BNPParibas,StandardChartered,SumitomoMitsui,MUFG,Citibank,ING,DeutscheBank,andHSBC).203WeprovideevidencethatsuggeststhisisindeedthecaseinChapter3.202Whenoneaccountsformultilateralparticipationinsyndicatedloansandentrustedloanagreementswithmultilateralinstitutions,themultilateralizationofChina’sLICandMIClendingportfoliocomesintosharperresolution(seeFigureA28).201AlthoughtherearemoresyndicatedloanswithChineseandnon-Chineseparticipants(833)thantherearesyndicatedloanswithexclusivelyChineseparticipants(312),syndicatedloanswithChineseandnon-Chineseparticipantstendtobesmaller($138.8milliononaverage)thansyndicatedloanswithexclusivelyChineseparticipants($545.5milliononaverage).ThesesummarystatisticsaredrawnfromFiguresA34andA35.
112
CountryProjectYearValueofLoanLeadArrangerChineseParticipantsTotalNumberofParticipantsIraqBasrahNaturalGasLiquidsFacilityConstructionProject2021$260millionInternationalFinanceCorporation(IFC)ICBC,BankofChina9
KazakhstanAlmatyRingRoadProject2020$585millionEuropeanBankforReconstructionandDevelopment(EBRD)BankofChina5
ArgentinaLaCastellanaWindPowerProject2017$64.05millionInternationalFinanceCorporation(IFC)SAFEthroughtheManagedCo-lendingPortfolioProgram2
BangladeshUnit3of220MWSirajganjCombinedCyclePowerPlantProject2017$196.7millionStandardCharteredBankBankofChina4
Notes:ThistableprovidesexamplesofsyndicatedloanswithChinesestate-ownedparticipants(ID#62223,
62224,95921,92613,98022,and69033inthe3.0versionofAidData’sGCDFdataset).The“Year”column
capturesthefinancialcommitmentyear.The“ValueofLoan”columncapturestheaggregatemonetary
valueofallsyndicatedloantranches/contributions.The“ChineseParticipants”columncapturesallofficial
sectorChineseparticipantsinthesyndicate.The“TotalNumberofParticipants”columncapturesthetotal
numberofChineseandnon-Chinesecreditorsthatparticipatedinthesyndicate.
Figure2.23abovehighlightsanotherinterestingtrend:amodestincreasein
ChineselendingthatischanneledviamultilateralinstitutionsduringtheBRIera
(2014-2021).DuringthefiveyearperiodthatprecededtheBRI(2009-2013),2%
ofChina’snon-emergencylendingtoLICsandMICswaschanneledvia
multilateralinstitutions.Thisfiguredoubled(to4%)duringthefirsteightyearsof
BRIimplementation(2014-2021).205
OneofChina’sinitialforaysintoentrustedlendingviamultilateralinstitutions
beganin2013and2014,withthecreationoftheAfricaGrowingTogetherFund
(AGTF)attheAfricanDevelopmentBank,theChinaCo-financingFundforLatin
AmericaandtheCaribbeanattheInter-AmericanDevelopmentBank,andthe
ManagedCo-LendingPortfolioProgram(MCPP)attheInternationalFinance
Corporation(IFC).TheMCPP,whichwaslaunchedbytheIFCinpartnershipwith
China’sStateAdministrationforForeignExchange(SAFE),helpsillustratethe
logicofanentrustedloanagreementwithamultilateralinstitution.Ratherthan
directlylendingtoborrowersindevelopingcountries,SAFEentrusted$3billion
totheIFCand“leverage[d]IFC’sprojectpipelineandduediligenceskillsto[...]
205Figure2.23excludesemergencyloanssincetheyareexclusivelyprovidedviabilateralchannels.
113
co-lendtoprojectsorgroupsofprojectsalongsideIFConcommercialterms”
(WorldBankGroup2020:35).Beijing’sdecisiontooutsourceloanmanagement
toamultilateralinstitutionwasevidentlymotivatedbyadesireforstronger
safeguardsandattractiveinvestmentreturns(seeBox2bformoreonSAFE’s
mandatetomaximizeinvestmentreturnsonthecountry’ssurplusforeign
exchangereserves).Anex-postevaluationoftheMCPPconcludedthat“[m]ost
oftheseborrowers[had]thecapacitytomeettheirfinancialcommitmentsand
[were]lessvulnerabletononpaymentthanotherspeculativeprojects"(World
BankGroup2020:36).
Thinktanks,researchinstitutions,andmediaoutletsinG7countrieshavein
recentyearstrainedtheirsightsonChina’sbilaterallendingactivities—in
particular,infrastructureprojectsfinancedbyChinaEximbankandChina
DevelopmentBank(e.g.,Dyeretal.2011;SandersonandForsythe2013;
GallagherandRay2020;MyersandRay2023;Ray2023;DoRosarioandSavage
2023).YettherehasbeenvirtuallynodiscussionaboutBeijing’sstrategicpivot
awayfrombilaterallendingandtowardcollaborativelendingarrangements
involvingChineseandnon-Chinesebanks.Thereasonwhythischangein
China’soverseaslendingportfoliohasnotreceivedseriousattentionissimple:
Chinesebankparticipationinsyndicatedloans(withnon-Chineseparticipants)
andloansentrustedtomultilateralinstitutionshavegoneundetectedbecause
theyaremostlyexcludedfromotherpubliclyavailableChinesedevelopment
financedatasets.206Beijinghasnotbeenespeciallysecretiveaboutthischange
inthecompositionofitsoverseaslendingstrategy.Itwashidinginplainsight;
thoseseekingtotrackChina’soverseaslendingactivitieseitherdidnotdetectit
ordidnotthinkitwasimportantenoughtotrack.
206Hereagain,oneisremindedoftheproverbialdrunkardwhoinsistsuponsearchingforhiskeysbeneaththelamppost“becausethat’swherethelightis.”ChinesecreditorcontributionstosyndicatedloansarenotsystematicallytrackedintheChina’sOverseasDevelopmentFinanceDatabase,theChineseLoanstoLatinAmericaandtheCaribbeanDatabase,theChineseLoanstoAfricaDatabase,theChina’sGlobalEnergyFinanceDatabase,theChinaOverseasFinanceInventoryDatabase,ortheWorldBank’sInternationalDebtStatistics(IDS).However,FigureA32demonstratesthatofficialsectorlendingfromChinaisincreasinglychanneledviabanksyndicatesthatincludeChineseandnon-Chinesebanks.
114
ThereareseveralreasonswhyBeijing’spivotawayfrombilaterallending
instrumentsiscrucialtounderstanditsde-riskingstrategy.207First,sincemostof
thesyndicatedloansinquestioninvolvenon-Chinesebankparticipants(see
FigureA32),Beijingisincreasinglyabletooutsourceriskmanagementto
lendinginstitutionswithstrongerrulesandstandards.208Second,multilateral
institutionshaveparticularlystrongriskmanagementguardrailsinplace(see
Chapter3),sotheuseofentrustedloanagreementswithmultilateralinstitutions
necessarilyinvolvestheapplicationofamorestringentsetofsafeguards.These
areeffectivelyde-risking“shortcuts.”Bilateraldevelopmentfinanceinstitutions
haverulesandstandardsthathaveevolvedoverdecadesviaaccretion.Itis
unlikelythattheseinstitutionsaregoingtodispensewiththesesafeguardsor
dramaticallychangethemintheshort-run.So,delegatingborrowerselection
andloanpreparationtoacrediblethirdpartyisanattractiveshortcuttode-riska
loanportfolio(onagoingforwardbasis).209Second,participationinasyndicated
loanagreementoranentrustedloanagreementwithamultilateraldevelopment
bank(MDB)isarapidandreliablede-riskingstrategybecauseMDBsenjoyde
factopreferredcreditorstatus(Schlegletal.2019)andconferthisbenefittoall
otherlendersthatparticipateintheirsyndicatedloanagreementsandentrusted
loanagreements(Guraraetal.2020).210Anaddedbenefitofthisapproachis
thatloansinvolvingMDBsaregenerallyexemptfromreschedulingandshielded
fromlargehaircuts(investorlosses).211Fromariskmitigationperspective,
syndicatedloanagreementsalsohavetwoperverselyattractivefeaturesto
officialcreditors:(1)unanimousconsentrequirementscanmakethemmore
difficultandtime-consumingtoreschedule(Buchheit1985,1991;Gelpern
211Onthispoint,seeCordellaandPowell(2021).210InatypicalA/BsyndicatedloanarrangementinvolvinganMDB,thelender-of-recordistheMDB;itkeepsapartoftheloanforitsownaccount(the“A-loan”)anditsellsparticipationintheremainderoftheloan(the“B-loan”).Asthelender-of-record,theMDBconfersitsstatusasadefactopreferredcreditortoallB-loanparticipants.209Onthispoint,seeDennisandMullineaux(2000).208InNovember2017,thecountry'stopbankingregulator—theChinaBankingRegulatoryCommission(CBRC)—calleduponChinesestate-ownedbanksto“preventandcontroloverseasbusinessrisksbytakingrisk-sharingmeasures”(Xueqing2017,emphasisadded).TheshifttowardsyndicationandmultilateralizationduringthelateBRIperiod(highlightedaboveinFigure2.23)mayconstituteevidenceofbankresponsivenesstoCBRC’spolicyguidance.207AsweexplainatgreaterlengthinChapter3,BeijinghaspositioneditsbilaterallendinginstitutionsasalternativesourcesoffinancingforLICsandMICsthatwouldprefernottodealwithOECD-DACdonorsormultilateraldevelopmentbanks(Parks2019;Maliketal.2021;Dreheretal.2022).Therefore,thepivottowardsyndicationandmultilateralizationrepresentsamajordeparture.
115
2016);and(2)theyaregenerallysubjecttosmallerhaircuts(financiallosses)than
bilateralloans(CrucesandTrebesch2013;Schlegletal.2019).212
Section4:WhathavewelearnedaboutBeijing’sbidtorebootitsoverseaslendingprogram?
Inthischapter,wehavepresentedevidencethatchallengestheconventional
wisdomaboutthetotalcollapseofChina’soverseaslendingprogram.With
morecompletedataonthefullrangeofChina’slendingactivitiesinLICsand
MICs,wehaveshownthatBeijingremainstheworld’slargestofficialcreditor:in
nominalterms,theworldowesChinaabout$2.6trillionandthedeveloping
worldowesChinaatleast$1.1trillionandperhapsasmuchas$1.5trillion.We
havealsoshedlightonthewaysinwhichBeijingisrebalancingitscross-border
creditportfolio—muchlikeayield-maximizinginvestmentportfoliomanagerthat
isnavigatinganenvironmentinwhichagrowingnumberofborrowersare
illiquidorinsolvent.Beijinghasrampeddowntheprovisionoflong-term,
dollar-denominatedbilateralloanstosovereignborrowersforpublicinvestment
projects,whileatthesametimerampinguptheprovisionofRMB-denominated
emergencyrescueloansthatareshort-ormedium-terminnature.Ithasreduced
itsrelianceuponthepolicybanks,whilemakinggreateruseofstate-owned
commercialbanks,thecentralbank,syndicatedloanarrangementswith
non-Chinesebanks,andmultilateraladministrators.
Whatcanwesay,bywayofconclusion,aboutBeijing’seffortstode-riskits
overseaslendingportfolioanditsdeterminationtoensurethatLICandMIC
borrowersrepaytheirdebts?Theevidenceathanddoesnotsuggestthatitis
readytotakefinanciallossesinordertominimizediplomaticblowbackand
reputationaldamage.213Quitetheopposite:Beijingappearstobestiffeningits
resolveandpreparingforalonganddifficultslog.Itissweepingcashoutofthe
escrowaccountsofitsoverseasborrowers,requiringthatborrowersreplenish
213GongChen,thefounderofAnbound(aBeijing-basedthinktank)andaBRIadvisertothecentralgovernment,recentlytoldNikkeiAsiathat“widespreaddebtevasionandavoidance[byBRIparticipants]wouldhaveasignificantimpactonChina'sfinancialstability”and“weareconcernedthatsomecountriesmaytrytoavoidpayingbacktheirdebtbyutilizinggeopoliticsandtheideologicalcompetitionbetweenEastandWest”(Aamiretal.2022).212BilateralloansfromofficialcreditorsaretypicallyrescheduledthroughtheParisClub,whilesyndicatedloansaretypicallyrescheduledthroughLondonClubreschedulings(regardlessofwhetherthesyndicatemembersincludeofficialcreditorsorcommercialcreditors).
116
escrowaccountsinexchangeforshort-termcashflowrelief,introducingstronger
penaltiesforlaterepayments,andchannelingemergencyrescueloanswithhigh
interestratesandshortrepaymentperiodstofinanciallydistressedborrowers(to
makesuretheyhaveenoughcashonhandtoservicetheirexistinginfrastructure
projectdebts).OnlytimewilltellifBeijinghasenough“steelinitsspine”tostay
thecourse,butitsactionstodatesuggestthatitintendstodowhateverittakes
toprotectthebottomline.
Afinalpointbearsemphasis.Wefreelyconcedethatthischapterhasonly
scratchedthesurfaceofwhatcanbedonewiththe3.0versionofAidData’s
GCDFdatasettouncovernewinsightsaboutBeijing’scontemporarylending
activitiesandpracticesintheGlobalSouth.Thedatasetcan—andshould—be
usedtoansweranarrayofadditionalquestions,suchas:
●HowhavethetermsandconditionsinChineseloancontractschanged(or
not)sinceBeijingendorsedtheCommonFrameworkinNovember2020?
●Whataretheconsequencesofborrowernon-compliancewithescrow
accountconditions?
●HowdoChinesecreditorsselecttheirpreferredpartnersinsyndicated
loanagreements?
●WhydoesBeijingprefertochannelloansthroughsomemultilateral
institutionsratherthanothers?
●Whatarethesupply-sideanddemand-sidedriversofChina’semergency
lendingprogram?Hasthischangedovertime,andifso,howandwhy?
●Whatarethesupply-sideanddemand-sidedriversofChina’s
non-emergencylendingprogram?Hasthischangedovertime,andifso,
howandwhy?
AlthoughtherearecommercialincentivestoputtheGCDFdatasetbehinda
paywallandprofessionalincentivestowithholdreleaseuntilitisintroducedin
leading,peer-reviewedacademicjournals,weremainfullycommittedtothe
117
principlethatallpast,present,andfutureversionsoftheGCDFdatasetshould
betreatedaspublicgoodsratherthanprivategoods.214Ourhopeisthatthe
datasetwillcatalyzeaknowledgemultipliereffectandfacilitateevidence-based
decision-making.215
215Ifpastisprologue,weexpectthatthisapproachwilldeliverasignificantpayoff.Todate,theGCDFdatasethasbeenusedinmorethan500researchpublications(Wooley2023).214Inthisrespect,wearecarryingforwardthe“openresearch”traditionoftheoriginal,interdisciplinarygroupofresearcherswhodevelopedtheTrackingUnderreportedFinancialFlows(TUFF)methodologythatunderpinsAidData’sGCDFdataset.SeeChapter3ofDreheretal.(2022).
118
Chapter3:RedesigningtheBeltandRoadfor
SafetyandSpeed
Section1:Beijing’sjourneyfromskeptictoadvocateofESGriskmanagement
Forthebetterpartofthelasttwodecades,Beijingsoughttopositionitselfasan
alternativesourceofinfrastructurefinancingforgovernmentsthatwouldprefer
nottodealwithOECD-DACdonorsormultilateraldevelopmentbanks.It
characterizeditselfasademand-drivenfinancierofSouth-Southcooperation
andtacitlyencourageditsforeigncounterpartstobypasstherulesand
standardsofOECD-DACdonorsandmultilateraldevelopmentbanksby
fast-trackinglarge-scaleinfrastructureprojectswithaslittle“redtape”and
“hasslefactor”aspossible.Itofferedfastandflexibleprojectpreparation
procedures;astreamlinedsetofloanandgrantapprovalprocesses;theability
toissuecontractswithoutcompetitivebiddingrequirements;andlessstringent
environmental,social,andgovernance(ESG)safeguards.216
China’svaluepropositionwascompelling.Between2000and2021,140LICs
andMICsaccepted$825billionofaidandcreditfromBeijingfor4,800
infrastructureprojects.217Chinabecamethedevelopingworld’sgo-tobankerfor
big-ticketinfrastructureprojectsbecauseitdemonstratedthreecomparative
advantagesvis-à-visOECD-DACdonorsandmultilateraldevelopmentbanks
(MDBs):scale,speed,andimpact.Beijingbankrolledlarge-scaleinfrastructure
projectsthatitspeersandcompetitorswereunwillingorunabletosupport.218It
financed1,385infrastructureprojectswithgrantsandloansworth$100million
218Chinawasfasterthanitscompetitorsatfinalizingloanagreementsforlarge-scaleinfrastructureprojects.In2008,Senegal’sthen-President,AbdoulayeWade,wroteintheFinancialTimesthat“withdirectaid,creditlinesandreasonablecontracts,ChinahashelpedAfricannationsbuildinfrastructureprojectsinrecordtime.…Ihavefoundthatacontractthatwouldtakefiveyearstodiscuss,negotiateandsignwiththeWorldBanktakesthreemonthswhenwehavedealtwithChineseauthorities”(Wade2008).217Thesefiguresonlyrefertoactiveandcompletedprojects.Between2000and2021,Beijingalsoissuedgrantsandloansworth$56billionfor94infrastructureprojectsin49LICsandMICsthatweresubsequentlysuspendedorcanceled.216Atthesametime,Beijingspurnednearlyallinvitationstofollowtheprevailingsetofinternationaldevelopmentfinancerulesandnorms,anditadmonishedWesterndonorsandmultilaterallendersfortheir“one-size-fits-all”policies(Maliketal.2021;Dreheretal.2022).
119
ormorebetween2000and2021.219Chinaalsoearnedareputationfor
implementingbrick-and-mortarprojectswithlightningspeed:theaverage
Chinesegovernment-financedinfrastructureprojectbetween2000-2021took
only2.7yearstocomplete.220SimilarprojectsfinancedbytheWorldBankand
regionaldevelopmentbanksusuallytook5-10yearstocomplete(Bulmanetal.
2017;Lagardaetal.2018;Dugganetal.2020;WorldBank2023c).China’s
overseasdevelopmentprojects—inparticular,thoseinvolvingtheconstruction
ofinfrastructuresuchasroads,bridges,tunnels,railways,andports—also
generatedsignificanteconomicbenefitsinapoliticallyrelevanttimeframe.221In
anevaluationof4,304projectsin138LICsandMICsovera15-yearperiod,
Dreheretal.(2021)findthattheaverageprojectincreasedeconomicgrowthby
0.95percentagepointstwoyearsaftersecuringfundingapprovalfrom
Beijing.222Bluhmetal.(2020)provideevidencethatChina’sconnective
infrastructureprojectportfoliowasespeciallyeffectiveatpromoting
spatially-inclusiveeconomicdevelopment—bydecentralizingeconomicactivity
withintheprovincesanddistrictswheretheywereimplemented.223
223Bluhmetal.(2020)findthat,onaverage,ChineseODAandOOF-financedconnectiveinfrastructureprojectsreduceeconomicconcentration(asmeasuredbytheGinicoefficientofnighttimelightoutputinall9.3kmsquaregridcellswithinaparticularsubnationallocality)byabout2.2percentagepoints.Theyalsoprovideevidencethattheseprojectshaveeffectivelyrelocatedeconomicoutputfromdenseareaslikecitycenterstotheirimmediateperipheries(i.e.,peri-urbanandsuburbanareas).Theinstallationofconnectiveinfrastructurecanacceleratespatially-inclusiveeconomicdevelopmentbymakingiteasierforfirmstoreachmoredistantmarketsandindividualstocommuteorrelocatetoplacesofwork.Italsocanalsolowerthecostofconsumergoodsandinputs,promotethedevelopmentofnewbusinesses,increaselandvalues,boostagriculturalproduction,andfacilitateknowledgeandtechnologyspillovers.222Morespecifically,Dreheretal.(2021a)findthatoneadditionalChineseODA-orOOF-financedprojectincreaseseconomicgrowthbybetween0.41and1.49percentagepoints(pp)twoyearsafterthefundingfortheprojectisapproved,onaverage.0.95pprepresentsthemidpointofthisrange.221ThisfindingimpliesthatifahostcountrychosetoacceptthreeadditionalChineseODAorOOF-financeddevelopmentprojects,itcouldreasonablyexpecttoboostitseconomic(GDP)growthby2.85percentagepointswithintwoyearsofBeijingagreeingtobankrolltheprojects.FormoreonthesocioeconomicimpactsofChinesegrant-andloan-financeddevelopmentprojects,seeBluhmetal.(2020),Martoranoetal.(2020),Dreheretal.(2022),MandonandWoldemichael(2023),andWellneretal.(forthcoming,2023).220Withthe3.0versionofAidData’sGCDFdataset,wecalculatetheaverageamountoftimeneededtocompleteaChinesegrant-orloan-financedinfrastructureprojectbymeasuringtheaveragenumberofcalendardaysbetweenactualprojectimplementationstartdatesandactualprojectcompletiondates.WeonlyincludeactiveprojectsandcompletedprojectsthatsecuredofficialcommitmentsfromChinainouranalysis.219Between2000and2021,Beijingfinanced735infrastructureprojectswithgrantsandloansworthatleast$250millionin89LICsandMICs,377infrastructureprojectswithgrantsandloansworthatleast$500millionin68LICsandMICs,and163infrastructureprojectswithgrantsandloansworthatleast$1billionin47LICsandMICs.AllofthesefiguresincludeinfrastructureprojectsthatsecuredgrantorloancommitmentsfromChina,whichweresubsequentlysuspendedorcanceled.
120
Beijing’strackrecordofbankrollingandbuildingbig-ticketinfrastructureprojects
withrecordspeedandnear-termeconomicimpactchangedthenatureof
policymakerdemandintheGlobalSouth.ThroughitsListeningtoLeaders
program,AidDatahasrepeatedlysurveyedthousandsofseniorandmid-level
governmentalofficialsacross140LICsandMICs,andtheseleadersnowreporta
strongpreferenceforworkingwithBeijingratherthanitscompetitorson
infrastructureprojects(Custeretal.2021;Horigoshietal.2022;Blairetal.
2022b).
Still,thefactthatChina—seeminglyovernight—becametheGlobalSouth’s
infrastructurefinancieroffirstresortisremarkable.Large-scaleinfrastructure
projectsarenotoriouslydifficulttoimplement.Theyoftenencountermajor
delaysandcostoverrunsandraiseconcernsaboutpollution,biodiversityloss,
thedisplacementofnearbyresidents,andthepreservationofcultural
heritage.224Iftheyarenotcarefullydesignedandimplemented,theycanalso
leadtopublicprotests,lawsuits,laborstrikes,corruptionscandals,and
allegationsofpoliticalfavoritism.225
Beijinginitiallysoughttoovercomethesechallengesbyaskingpoliticalleaders,
ratherthantechnocrats,toproposeinfrastructureprojectsandoverseetheir
implementation(Dreheretal.2019,2022).226Itreliedonpresidentsandprime
ministersandtheirdeputiestoresolvecomplexandcontentiousmattersthat
coulddelayorderailprojects—forexample,thegrievancesandcompensation
demandsofthosedisplacedorharmedbyconstructionactivities(Dreheretal.
2022).Italsodispensedwithmanyofthecostlyandcumbersomerequirements
ofWesterndonorsandmultilateraldevelopmentbanks(MDBs).Ratherthan
insistinguponstrictadherencetointernationalESGsafeguards(suchasthose
226China’sdemand-drivenprojectselectionsystemalsoencouragescollusionbetweenhostcountrypoliticiansandChinesecontractors.AsZhangandSmith(2017:2335)explain,“ChinesefirmsandhostgovernmentsenterintoaninformalallianceinwhichChina’scompaniespersuadehostgovernmentstoraisenewaidprojectswithChinawhilethecontractorspromisetohelpbehindthescenestosecurefinancing.Theprojectsarereverse-engineeredtosuitthepoliticalneedsoflocalpoliticiansandthecommercialstrategiesofChinesecontractors.”Similarly,Bräutigam(2019)arguesthat“[t]heAchillesHeelofChina’sbankfinancingmodelisthatitreliesheavilyonChinesecompaniestodevelopprojectstogetherwithhostcountryofficials.Thiscreatesstrongincentivesforkickbacksandinflatedprojectcosts.Particularlyinelectionyears,companiesandpublicworksministersmaycolludetogetprojectsapproved.”225Onthispoint,see
Isaksson
andKotsadam
(2018a,
2018b),
Isaksson
(2020),Dreheretal.(2019,2022),Anaxagorouetal.(2020),Iacoellaetal.(2021),andBaehretal.(forthcoming).224Onthedifficultyofimplementingthesetypesofprojectsontimeandonbudget,seeFlyvbjergetal.(2002).
121
describedintheIFC’sPerformanceStandardsonEnvironmentalandSocial
Sustainability),Beijing’sstate-ownedbanksrequestedcompliancewithnational
rulesandstandards(Export-ImportBankofChina2017;ChenandLandry2018;
Baehretal.forthcoming).Tofacilitaterapidmobilizationassoonasloanorgrant
applicationswereapproved,theyissuedno-bidcontractstoChinesecompanies
withanestablished,on-the-groundpresence(Bräutigam2019).227Theabsence
ofanyrequirementorexpectationtocoordinatewithotherdonorsandlenders
eliminatedadditionalobstaclestoimplementation(BourguignonandPlatteau
2015;Furukawa2018).
Butspeedandconveniencecameatacost:atleast54%ofChina’soverseas
infrastructureprojectportfoliofrom2000to2017hadsignificantenvironmental,
social,orgovernanceriskexposure(seeFigureA38intheAppendix).228Insome
cases,multi-billiondollarpublicinvestmentprojectsweresuspendedor
rescopedbecauseofinsufficientinattentiontoenvironmentalandsocialimpact
assessments(seeLuetal.2023bandBox3b).Inothercases,journalistsandcivil
societyorganizationsuncoveredevidenceofChinesecompaniesandhost
governmentofficialscolludingbyartificiallyinflatingsole-sourcecontractprices
forconstructionprojectsandsharingtheillicitproceeds(Maliketal.2021:67;
Dreheretal.2022:11-12).Instillothercases,localgrievanceswerenot
addressedinatimelyorthoroughmannerandtheymetastasizedinto
reputationalliabilities(thatweanalyzeatgreaterlengthinChapter4).229
BytheendoftheearlyBRIera,theauthoritiesinBeijingseemedtocoalesce
aroundtheideathatsustainingeliteandpublicsupportforitsflagshipglobal
229Manyoftheseproblemsprobablycouldhavebeenavoidedviaavarietyofmechanisms:environmentalandsocialimpactassessmentstoensurethatindigenouspeoplesaregrantedfree,prior,andinformedconsent(FPIC)andavoidsitingprojectsnearendangeredhabitats;competitivebiddingrulestoensuregoodvalue-for-money;blacklistingprocedurestoavoidhiringcontractorswithatrackrecordofparticipatingincorruptandcollusivebehavior;andgrievancemechanismstomakeiteasiertoidentifyandrespondtotheconcernsoflocalstakeholders(Parks2019;Dreheretal.2022).228ThisfigurerepresentsthecumulativepercentageofChina’sgrant-andloan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)between2000and2017withsignificantenvironmental,social,orgovernanceriskexposure.TheaverageannualESGriskprevalencerate,asdefinedinFigure3.2),was47%between2000and2017.Between2000and2017,1,403infrastructureprojectsinLICsandMICssupportedbygrantsandloansfromChinaworth$383billion(inconstant2021USD)presentedasignificantenvironmental,social,orgovernancerisk(seeFigure1.13).227Sole-sourceprocurementistheruleratherthantheexceptioninmostofCDBandChinaEximbank’soverseasloanagreements.Infact,theseagreementsalmostalwaysreferenceaspecificcommercialcontractwithaspecificChinesefirmandstrictlyinstructtheborrowertoexclusivelyusetheproceedsoftheloantofinancethepre-selectedcommercialcontractthatisreferencedintheloanagreement(Gelpernetal.2021,2022).
122
infrastructureinitiativewouldrequiremoreeffectiveESGriskmanagementand
mitigation.OneofthefirstsignsthatchangewasafootcameinNovember2017
whentheChinaBankingRegulatoryCommission(CBRC)—thecountry'stop
bankingregulator—issuedanewsetofrules,requiringCDBandChina
Eximbanktoputinplacemorerobustenvironmentalandsocialrisk
managementprocedures(CBRC2017a,2017b).230By2018,theauthoritieswere
planningatransition“fromahazilydefinedBRI1.0toamorefine-tunedBRI
2.0”(Ang2019).OnAugust27,2018,intherun-uptothefifthanniversaryof
theBRI,XiJinpingusedaChinesepaintingmetaphortocallfor“aswitchfrom
xieyi,freehandpaintingforoutliningbroadstrokes,togongbi,thecareful
inscriptionofdetails”(Ang2019).Then,inApril2019,hegaveaspeechatthe
SecondBeltandRoadForumforInternationalCooperationwherehe
announcedthatChinawould“adoptwidelyacceptedrulesandstandardsand
encourageparticipatingcompaniestofollowgeneralinternationalrulesand
standardsinprojectdevelopment,operation,procurementandtenderingand
bidding”(MinistryofForeignAffairsofthePeople'sRepublicofChina2019).He
conveyedduringthesamespeechthat“inpursuingBeltandRoadcooperation,
everythingshouldbedoneinatransparentway,andweshouldhavezero
toleranceforcorruption”(MinistryofForeignAffairsofthePeople'sRepublicof
China2019).231
Oneyearlater,Beijingsignaledpreliminaryinterestin“multilateralizing”theBRI
byco-financing,co-designing,andco-implementinginfrastructureprojectswith
Westernandmultilateraldevelopmentfinanceinstitutionsandsubjectingthese
projectstostrongersafeguards.232Itteamedupwitheightmultilateral
institutions—theWorldBank,theInter-AmericanDevelopmentBank,theAsian
DevelopmentBank,theEuropeanBankforReconstructionandDevelopment,
232Thereissomeevidenceofinternationalfinancialinstitution(IFI)leaderstryingtosteerChinainthisdirection(e.g.,Kim2017;Lagarde2019).231InJanuary2021,HuHuaibang,theformerchairmanofCDB,wassentencedtolifeinprisonfortaking$13millioninbribes.Then,inSeptember2021,HeXingxiang,aCDBvicepresident,wasplacedunderinvestigationbyChina’sCentralCommissionforDisciplineInspection(CCDI)for“severedisciplineandlawviolations”(Wilson2022).Oneyearlater,LiLi,theformerPresidentoftheBeijingBranchoftheChinaEximbankwasexpelledfromtheChineseCommunistPartyduetocorruptioncharges.Then,inMarch2023,LiuLiangeresignedfromhispositionaschairmanofBankofChinaandCCDIinvestigatedhimonsuspicionsofcorruptionandgraft(WongandZhai2023).230In2018,BeijingalsofinancedthecreationofaChina-IMFCapacityDevelopmentCentertotraingovernmentofficialsondebtsustainabilityframeworks(DSFs)inlow-incomecountriesandotherBRI-relatedpolicyissues(Morrisetal.2020).
123
theEuropeanInvestmentBank,theInternationalFundforAgricultural
Development(IFAD),CorporaciónAndinadeFomento,andtheAsian
InfrastructureInvestmentBank—toestablishaMultilateralCooperationCenter
forDevelopmentFinance(MCDF).233TheCenter’smandateisto(a)investin
moreupstreamprojectpreparationwork;(b)buildthecapacityoflendersand
borrowerstomoreeffectivelymanageandmitigaterisksrelatedtodebt
sustainability,procurement,corruption,andenvironmentalandsocialissues;and
(c)facilitategreaterinformation-sharingandcoordinationbetweenChineseand
non-Chinesedevelopmentfinanceinstitutions(AIIB2021).234
Then,in2021,SAFE—theultimatesourceoffundingformostofChina’s
state-ownedpolicybanks,commercialbanks,andinvestmentfunds(seeBox2b
inChapter2)—announcedthatitwouldprioritize“adoptingMDB’sESGcriteria”
and“incorporatingESGprinciplesintothewholeprojectinvestmentprocess
fromdecision-makingtopost-investmentmanagement”(SAFE2021:54).235In
parallel,throughamultilateralforumknownastheInternationalPlatformon
SustainableFinance(IPSF),theEUandChinalaunchedajointefforttoassess
thecommonalitiesanddifferencesintheirrespectivetaxonomiesfor
environmentallysustainableinvestments(Moody'sInvestorService2022;IPSF
2022;HKGFAandGuangdongGreenFinanceCommittee2022).236This
collaborationresultedinthe2021publicationofaso-calledCommonGround
Taxonomy(CGT),whichinturnwas“incorporatedintodomesticregulation[in
China]”and“directlyusedbyChinesebanksasstandardsforissuinggreen
bondsintheinternationalmarket”(ChengandZhang2023:10).
236IPSFisamultilateralforumthataimstoenabletheexchangeofpracticesandincreaseinternationalcooperationonsustainablefinancerelatedmatters.ItsmembersincludetheEU,China,Singapore,Japan,andIndia.235InSeptember2021,XiJinpingannouncedattheUNGeneralAssemblythatChinawouldnolongerfinancenewcoal-firedpowerprojectsoverseas.Then,inMarch2022,NDRC,theMinistryofForeignAffairs,theMinistryofCommerce,andtheMinistryofEcologyandEnvironmentpublished“OpinionsonJointlyPromotingGreenDevelopmentoftheBeltandRoad,”clarifyingthatChinawould“stopbuildingnewcoal-firedpowerprojectsabroadandprudentlyproceedwithexistingonesthatareunderconstruction”(NationalDevelopmentandReformCommission,MinistryofForeignAffairs,MinistryofEcologyandEnvironment,andMinistryofCommerceofthePeople’sRepublicofChina2022).234TheMCDF,whichisadministeredbytheAIIB,describesitselfas“amultilateralinitiativetoincreasehigh-qualityinfrastructureandconnectivityinvestmentsindevelopingcountriesincompliancewithInternationalFinancialInstitution(IFI)standards,includingbyencouragingotherinvestorsandfinancialinstitutionstoadoptsuchstandards”(AIIB2021).233TheMOUthatestablishedtheMCDFcanbeaccessedinitsentiretyviahttps://www.ndb.int/wp-content/uploads/2022/11/MCDF-MOU-for-disclosure.pdf.
124
AlloftheseactionsandrhetoricalcommitmentssuggestthatBeijinghassome
levelofinterestinmoreeffectivelymanagingtheESGrisksinitsoverseas
infrastructureprojectportfolio—andpotentiallyevenharmonizingitspolicies
andpracticeswithprevailinginternationaldevelopmentfinancerulesand
standards.However,interestdoesnotnecessarilytranslateintoimplementation,
soouraiminthischapteristodetermineifChinahaslearnedfrompastmistakes
andrecalibratedthewaysthatitfinances,designs,andimplements
infrastructureprojectsintheGlobalSouth.237Morespecifically,wewillusethe
3.0versionofAidData’sGCDFdatasetto(a)documentthescopeandseverityof
theESGrisksinChina’soverseasinfrastructureprojectportfolio;(b)identify
whether,when,andhowithassoughttomitigatetheseprojectimplementation
risks;and(c)determinewhetheritsinfrastructureprojectswithandwithout
strongESGsafeguardshavefareddifferentlyduringimplementation.
Ourfindingsdemonstratethat,althoughtheESGriskprofileofChina’soverseas
infrastructureprojectportfoliodeterioratedduringthepre-BRIperiodandearly
BRIperiod,therearesignsofimprovementduringthelateBRI(“BRI2.0”)
period.ChineselendersanddonorshaverespondedtorisinglevelsofESGrisk
byputtinginplaceincreasinglystringentsafeguardsthatmayultimately
undermineG7andMDBeffortstooutcompeteBeijingon“quality”and“safety”
grounds.Chinesegrant-andloan-financedinfrastructureprojectsthatare
subjectedtostrongESGsafeguardspresentfewerenvironmental,social,and
governancerisksduringimplementation.Theyarealsolesslikelytobe
suspendedorcanceled.Perhapsmostimportantly,Chinesegrant-and
loan-financedinfrastructureprojectsthataresubjectedtostrongESGsafeguards
donotfacesubstantiallylongerimplementationdelaysthanthosesubjectedto
weakESGsafeguards.OurfindingsthereforesuggestBeijingenjoysastronger
positionintheglobalinfrastructurefinancingmarketthanitsbilateraland
multilateralcompetitorsrealize.Developingcountrieshavemadetheir
preferencesveryclear:theywanttoworkwithlendersanddonorsthatare
willingandabletoquicklydesignandimplementbig-ticket,high-impact
237InFebruary2022,YunnanChenoftheOverseasDevelopmentInstitute(ODI)toldEuromoneymagazinethat“Chinaseemstohaveamorespecificandtargetedapproach.Itsfinancialinstitutionsarelearning,recognisingpastmistakesanderrors,andtakingamorerisk-averseapproachtowhatprojectstheyfinance,andhowtheygoaboutfinancingandduediligence”(Wilson2022).
125
infrastructureprojectswithoutunreasonablyhighlevelsofESGrisk.238Beijingis
takingactivemeasurestomeetthischallenge.Whetheritscompetitorswilldo
thesameisanopenquestion.
Section2:MeasuringthescopeandseverityofESGriskexposureinChina'sinfrastructureprojectportfolio
WebeginbymeasuringthenatureandextentofESGriskexposureinChina’s
overseasinfrastructureprojectportfolio—andhowithaschangedovertime.239
Wedosoinfiveways.First,weidentifywhetherChinaislocatinglarge-scale
infrastructureprojectsinenvironmentallysensitiveareas.Second,weanalyze
whetherChinaisplacingsuchprojectsinsociallysensitiveareas—specifically,in
areaswhereindigenouspopulationsareoftendeniedfree,prior,andinformed
consent(FPIC).Third,weassesswhetherChinaislocatinglarge-scale
infrastructureprojectsingeographicalareasthatarevulnerabletopolitical
captureandmanipulationbygoverningelitesinhostcountries.Fourth,we
evaluatetheextenttowhichChinaisrelyingoncontractorssanctionedfor
fraudulentandcorruptbehaviorfortheimplementationofitsoverseas
infrastructureprojects.Fifth,baseduponnarrativeevidencefromthe3.0version
ofAidData’sGCDFdataset,weflagallinfrastructureprojectsforwhichitis
knownthatasignificantenvironmental,social,orgovernancechallengearose
before,during,orafterimplementation.
FigureA39presentsthecumulativenumberofChinesegrant-andloan-financed
infrastructureprojectslocatedinenvironmentallysensitiveareaswithinLICsand
MICsbetween2000and2021.Wedetermineifagiveninfrastructureprojectis
locatedinoneormoreenvironmentallysensitiveareasbyfirstmergingprecisely
geocodeddataonChineseODA-andOOF-financedinfrastructureprojectsites
withtwoseparatedatasets:(1)theboundariesofdesignatedterrestrialand
marineprotectedareasfromtheWorldDatabaseonProtectedAreas(WDPA),
whichweconvertintoa1kmx1kmgridcellraster;and(2)the1kmx1kmgrid
cellrasterofterrestrialandmarinecriticalhabitats(asdefinedbythe
239Ouranalysisnotonlyincludesactiveandcompletedinfrastructureprojects,butalsosuspendedandcanceledprojects.Weincludesuspendedandcanceledprojectstoavoidsampleselectionbias,sincesuchprojectsaremorelikelytopresentsignificantESGrisks(Luetal.2023b).238Onthispoint,seeHumphrey2015;Dollar2016;Swedlund2017;HumphreyandMichaelowa2019;Zeitz2021;Horigoshietal.2022;andBlairetal.2022b.
126
InternationalFinanceCorporation’sPerformanceStandard6).240Wethenidentify
thesubsetofprojectswithoneormoresitesthatoverlapwithaterrestrial
protectedarea,amarineprotectedarea,aterrestrialcriticalhabitat,and/ora
marinecriticalhabitat.241Intotal,wefind1,035infrastructureprojectsin108
countriessupportedbygrantsandloansfromChinaworth$233billionthatare
locatedinenvironmentallysensitiveareas.242
FigureA40presentsthecumulativenumberofChinesegrant-andloan-financed
infrastructureprojectslocatedinsociallysensitiveareaswithinLICsandMICs
between2000and2021.Wedetermineifagiveninfrastructureprojectis
locatedinoneormoreoftheseareasbyfirstmergingpreciselygeocodeddata
onChineseODA-andOOF-financedinfrastructureprojectsiteswitha1kmx1
kmgridcellrasterofindigenouslands.243Wethenidentifythesubsetofprojects
withoneormoresitesthatoverlapwithindigenouslands,whichisauseful
measureofsocialriskbecauseinfrastructureprojectscancauselocalharmby
encroachinguponthetraditionalterritoriesofindigenouscommunitieswithout
free,prior,andinformedconsent(FPIC).244Intotal,wefind547infrastructure
projectsin53countriessupportedbygrantsandloansfromChinaworth$112
billionthatarelocatedinsociallysensitiveareas.245
245Thedecisiontolocateaninfrastructureprojectwithinoroutsidethetraditionalterritoriesofindigenouscommunitiesisonlyonewayofunderstandingthesocialriskprofileofsuchaproject.Consistentwiththeexistingliterature,wefocusonthisdimensionofsocialriskbecauseitcanbeconsistentlymeasuredovergeographicspaceandtimefornearlyallinfrastructureprojects(e.g.,Yangetal.2021).244Free,prior,andinformedconsent(FPIC)referstotherightofIndigenousPeoplestoprovideorwithholdconsent,atanypoint,fordevelopmentprojectsaffectingtheirterritories.ItisarightgrantedtoIndigenousPeoplesintheUNDeclarationontheRightsofIndigenousPeoples(UNDRIP)anditisbasedontheprinciplethat“allpeopleshavetherighttoself-determination.”UNDRIPrequiresstatesto“consultandcooperateingoodfaithwiththeindigenouspeoplesconcernedthroughtheirownrepresentativeinstitutionsinordertoobtaintheirfree,priorandinformedconsentbeforeadoptingandimplementinglegislativeoradministrativemeasuresthatmayaffectthem.”243Garnettetal.2018.242EnvironmentalistshaveexpressedparticularconcernsaboutthesitingofChinesegovernment-financedinfrastructureprojectsingeographicalareasthatmayfacilitatelegalandillegallogging,agriculturalfrontierexpansion,andhumansettlementsinpreviouslyremoteorpristineareas(Lauranceetal.2015;Yangetal.2021;Baehretal.2022).241Morespecifically,weidentifyallChinesegrant-andloan-financedinfrastructureprojectswithlocationsthatphysicallyoverlapwithareasthatweredesignatedasterrestrialormarineprotectedareasor“likely”criticalhabitats(asdefinedbyPS6)atanypointbetween2000and2021.Weexcludeallprojectswithout“precise”or“approximate”geocodesfromtheanalysis.Aprojectwith“precise”geocodesisoneforwhichhavehighlypreciseboundariesoftheproject’sgeofeature(s).Aprojectwith“approximate”geocodeisoneidentifiedwithina5kmradiusofthepreciseboundariesoftheproject’sgeofeature(s).Assuch,allprojectsgeocodedtotheADM8,ADM7,ADM6,ADM5,ADM4,ADM3,ADM2,ADM1,andADM0levelsareexcluded.240Martinetal.(2015);Braunederetal.(2018);andUNEP-WCMC&IUCN(2023).TheInternationalFinanceCorporation’sPerformanceStandard6(PS6)iswidelyusedbyinternationallendersanddonorstoidentify“criticalhabitats,”whichrefertoareasofhighbiodiversityvalue(Narainetal.2020,2022).
127
FigureA41presentsthecumulativenumberofChinesegrant-andloan-financed
infrastructureprojectslocatedingeographicalareaswithinLICsandMICsthat
arevulnerabletopoliticalcaptureandmanipulationbetween2000and2021.
Wedetermineifagiveninfrastructureprojectislocatedinoneormoreofthese
areasbyfirstmergingpreciselygeocodeddataonChineseODA-and
OOF-financedinfrastructureprojectsiteswiththePoliticalLeaders’Affiliation
Database(PLAD),whichidentifiesthehome(birth)districts(ADM2s)ofpolitical
leadersinLICsandMICs.246PreviousresearchhasshownthatChineseaidand
creditisdisproportionatelyallocatedtothehomeprovincesanddistrictsof
politicalleadersinhostcountriesandthatChineselendersanddonorslack
institutionalsafeguardstoreducethelikelihoodthatpoliticallymotivated
projectswillbeapproved(Dreheretal.2019,2022;Anaxagorouetal.2020).247
Intotal,wefind216infrastructureprojectsin69countriessupportedbygrants
andloansfromChinaworth$37billionthatarelocatedingeographicalareas
thatarevulnerabletopoliticalcaptureandmanipulation.248
FigureA42presents2000-2021dataonthecumulativenumberofChinese
grant-andloan-financedinfrastructureprojectsthatreliedoncontractors
sanctionedbyotherinternationalfinanciersforfraudulentandcorruptbehavior.
Wedetermineifagiveninfrastructureprojectreliedoncontractorssanctioned
byotherinternationalfinanciersforfraudulentandcorruptbehaviorwitha
three-stepprocess.First,wecompilealistoffirmshistoricallyorcurrently
debarredbytheWorldBankandfiveothermultilateraldevelopmentbanks
(MDBs)—theAfricanDevelopmentBank,theAsianDevelopmentBank,the
EuropeanBankforReconstructionandDevelopment,theInter-American
248Animportantcaveatisthatweonlyidentifyprojectsasbeinglocatedwithingeographicalareasthatarevulnerabletopoliticalcaptureandmanipulationiftheyfallwithinhomedistricts(ADM2s)ofpoliticalleaders.Therefore,projectsthatfallwithinthehomeregions(ADM1s),butnotthehomedistricts(ADM2s)ofpoliticalleaders,areexcluded.Nordoweconsiderthepresenceofnon-infrastructureprojectsinthehomedistricts(ADM2s)ofpoliticalleaders.247TheWorldBankusesexante,cost-benefitanalysistoscreencandidateprojects.Itemploysasimpleprojectacceptabilityrule—“theexpectedpresentvalueoftheproject'snetbenefitsmustbehigherthanorequaltotheexpectednetpresentvalueofmutuallyexclusiveprojectalternatives”—as“asafeguardagainstprojectchoicesbeingcapturedbynarrowpoliticalorsectionalinterests”(Warner2010:2).Bycontrast,theChinesegrant-givingandlendinginstitutionsdonothaveanalogousinstitutionalsafeguardsinplace(Dreheretal.2019,2022).246PLADprovidesinformationonthebirthplacesof1,109effectivepoliticalleadersfrom177countriesbetween1989and2021(Bomprezzietal.2023).BirthplacesaregeocodedtotheADM2(district)level.Inordertoidentifyprojectsthatarevulnerabletopoliticalcaptureandmanipulation,weidentifyallprojectswithlocationsinthehome(birth)districts(ADM2s)ofeffectivepoliticalleadersthatsecuredChinesegrantorloancommitmentsduringtheperiodsoftimewhentheleadersinquestionheldoffice.
128
DevelopmentBank,andtheAsianInfrastructureInvestmentBank—aswellasthe
datesofeachfirm’sformaldebarmentperiod.Then,wecomparethelistof
debarredfirmnamestothefirmnamesofimplementingagenciesand
contractorsinvolvedintheChinesegrant-andloan-financedinfrastructure
projects.Finally,weidentifythesubsetofprojectsthatreliedupondebarred
firmswhiletheywerestillwithintheirdebarmentperiodsbyidentifyingallcases
inwhichtherewascalendardayoverlapbetweenthestartandenddatesofan
organization’sdebarmentperiodandthecommitment,implementation,or
completiondatesoftheproject(s)itsupported.Projectsthatreliedupona
debarredfirmareidentifiedasposingasignificantgovernancerisk(seeSection
A-6intheAppendixformoredetails).Intotal,wefind296infrastructureprojects
in81countriessupportedbygrantsandloansfromChinaworth$88.8billion
thatrelyonsuchfirms.
FigureA43presentsthecumulativenumberofChinesegrant-andloan-financed
infrastructureprojectsforwhichthereisevidencethatasignificant
environmental,social,orgovernancechallengearosebefore,during,orafter
implementation.AkeyfeatureofAidData’s3.0dataset—thatsetsitapartfrom
otherpubliclyavailableChinesedevelopmentfinancedatasets—istheinclusion
of“cradletograve”narrativesthatprovidedetailedinformationabouthow
projectsweredesignedandimplementedinpracticeandwhytheyfailed,
faltered,orsucceeded.Thesenarrativesconsistof3.48millionwords(roughly
thesamenumberofwordsonewouldfindin34full-lengthbooks)across20,985
projectrecords.Theycapture,amongotherdetails,projectdesignand
implementationchallengesrelatedtolandacquisition;preservationofcultural
heritageandarchaeologicalsites;resettlementandcompensationofindigenous
communities;pollutionofair,water,andsoil;andadherencetoanti-corruption
standards.Tomakeuseofthisvasttroveofqualitativeinformation,weapplya
setofsystematicsearchandcategorizationprocedures(describedinSectionA-7
intheAppendix)toidentifythesubsetofinfrastructureprojectsforwhichthere
isevidencethatasignificantenvironmental,social,orgovernancechallenge
arosebefore,during,orafterimplementation.Intotal,wefindthatatleast356
infrastructureprojectsin131countriessupportedbygrantsandloansfrom
129
Chinaworth$250billionencounteredasignificantenvironmental,social,or
governancechallengebefore,during,orafterimplementation.249
Finally,togainabird’seyeviewofthescopeandseverityofESGriskinChina's
infrastructureprojectportfoliointhedevelopingworld,weuseallfiveofthese
measuresincombination.TodetermineifanygivenChinesegrant-or
loan-financedinfrastructureprojectpresentedasignificantenvironmental,social,
orgovernanceriskbetween2000and2021,weidentifyifitwaslocatedinan
areathatisenvironmentallysensitive,sociallysensitive,orvulnerabletopolitical
captureandmanipulation;reliedoncontractorssanctionedbyother
internationalfinanciersforfraudulentandcorruptbehavior;and/orencountered
asignificantenvironmental,social,orgovernancechallengebefore,during,or
afterimplementation.250Across125LICsandMICsbetween2000and2021,we
findthat1,693infrastructureprojectssupportedbygrantsandloansfromChina
worth$470billionhadsignificantenvironmental,social,orgovernancerisk
exposure(seeFigure1.13).Overthesame22-yeartimeperiod,wefindthat
$265billioninChinesegrantandloancommitmentsfor1,101infrastructure
projectshadsignificantenvironmentalriskexposure,while$192billionin
Chinesegrantandloancommitmentsfor701infrastructureprojectshad
significantsocialriskexposureand$211billioninChinesegrantandloan
commitmentsfor405infrastructureprojectshadsignificantgovernancerisk
exposure(seeFigure3.1).251
251Chinesegrant-andloan-financedinfrastructureprojectscan—andoftendo—facemorethanonetypeofESGrisk.250Werestrictedoursearchestoinfrastructureprojectssupportedbygrantandloancommitmentsworth$20million(inconstant2021USD)ormore.Projectssupportedbylargerfinancialcommitmentsgenerallyhavemoredetailedprojectdescriptions,whichprovideastrongerbasisfortheidentificationofenvironmental,social,andgovernancerisks.Theyalsopresentalowerriskofgenerating“falsenegatives.”249The“atleast”qualifierisimportantbecauseofourinabilitytoaddress“falsenegatives”thatmayaffectourkeyword-searchbasedmeasures(seeSectionA-7intheAppendix).
130
Figure
3.1:
Infrastructure
projects
with
significant
environmental,
social,
and
governance
risk
exposureFigure3.1
Notes:Thepresenceofsignificantenvironmental,social,andgovernancerisk(ESG)exposureisbasedona
project-levelcompositemeasurethatisdescribedinSection2ofChapter3.
TheESGriskprevalencerate—definedastheannualpercentageofChina’s
grant-andloan-financedinfrastructureprojectportfolio(measuredinconstant
2021USD)withsignificantenvironmental,social,orgovernancerisk
exposure—hasfluctuatedovertime(seeFigure3.2).252Duringthepre-BRI
period,itsharplyincreasedfrom12%in2000to65%in2013.Itthenfellto54%
(onaverage)duringtheearlyBRIperiodand47%(onaverage)duringthelate
BRIperiod.253By2021,itfellto33%.
253IftheESGriskprevalencerateisredefinedastheannualpercentageofChina’sgrant-andloan-financedinfrastructureprojectsfacingasignificantenvironmental,social,orgovernancerisk,itrosefrom25%in2000to40%in2013.Itthenfellto36%(onaverage)duringtheearlyBRIperiodand31%(onaverage)duringthelateBRIperiod(seeFigureA44).252AsimilarpatternisobservablewhenonetracksthesheernumberofChinesegrant-andloan-financedinfrastructureprojectsfacingsuchrisks.ThereisanapparentreductioninthenumberofinfrastructureprojectsaffectedbyESGriskin2020and2021(seeFigureA44),butgiventhattheprobabilityofESGrisksmaterializingandbeingdetectedincreasesasaprojectprogressesfromthefinancialcommitmentphasetotheimplementationphaseandthecompletionphase,wethinktheapparentreductionin2020and2021shouldbeinterpretedwithcaution.
131
Figure
3.2:
ESG
risk
prevalence
in
overseas
infrastructure
portfolio
from
China
to
LICs
and
MICsFigure3.2
Notes:ProjectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfromChina,although
theESGrisksthattheyencounteredmayhavematerializedafterthefinancialcommitmentyear.The
presenceofsignificantESGriskexposureisbasedonaproject-levelcompositemeasurethatisdescribed
inSection2ofChapter3.Likewise,environmentalriskexposure,socialriskexposure,andgovernancerisk
exposurearebasedontheproject-levelcompositemeasuresthataredescribedinSection2ofChapter3.
InFigures3.3,3.4,and3.5,weseparatelytracktheenvironmentalrisk
prevalencerate,thesocialriskprevalencerate,andthegovernancerisk
prevalencerate.254Onaverage,overtheentire22-yearperiodofanalysis
(2000-2021),theenvironmentalriskprevalenceratewashigher(27%)thanthe
socialriskprevalencerate(20%)orthegovernanceriskprevalencerate(18%).255
Acrossthesethreemeasures,onecanseeagenerallyconsistentpatternover
time:riskprevalenceratesmostlyincreasedduringthepre-BRIperiod
(2000-2013)andmostlydecreasedduringtheBRIperiod(2014-2021).Someof
255IftheESGriskprevalencerateisredefinedastheannualpercentageofChina’sgrant-andloan-financedinfrastructureprojectswithsignificantenvironmental,social,orgovernanceriskexposure,thesamepatternholds:theenvironmentalriskprevalencerateissubstantiallyhigher(22%)thanthesocialriskprevalencerate(14%)orthegovernanceriskprevalencerate(7%).254InFigures3.3,3.4,and3.5,wedefinetheriskprevalencerateastheannualpercentageofChina’sgrant-andloan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)facingagiventypeofrisk.InFiguresA45,A46,andA47,weredefinetheriskprevalencerateastheannualpercentageofChina’sgrant-andloan-financedinfrastructureprojectsfacingagiventypeofrisk.
132
thelargestdeclinesareobservableduringthelateBRIperiod(2018-2021).256
However,thesedeclinesshouldbeinterpretedwithcaution,astheycouldbe
theresultof(a)newlyapprovedprojectsnothavingprogressedtophasesofthe
projectlifecyclewhenESGriskstypicallymaterialize,(b)actualimprovementsin
theESGriskprofileofChina’soverseasinfrastructureprojectportfolio,or(c)
somecombinationofthesefactors.257
Figure
3.3:
Environmental
risk
prevalence
in
overseas
infrastructure
portfolio
from
China
to
LICs
and
MICsFigure3.3
Notes:ProjectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfromChina,although
theESGrisksthattheyencounteredmayhavematerializedafterthefinancialcommitmentyear.The
presenceofsignificantenvironmentalriskexposureisbasedonaproject-levelcompositemeasurethatis
describedinSection2ofChapter3.
257GiventhattheprobabilityofESGriskdetectionincreasesasaninfrastructureprojectprogressesfromthefinancialcommitmentphasetotheimplementationphaseandthecompletionphase,stillanotherpossibilityisthattheapparentdeclinesinESGriskprevalenceduringthelateBRIperiodreflectmeasurementimprecision.256DuringthelateBRIperiod,thetotalnumberofChinesegrant-andloan-financedinfrastructureprojectswithsignificantenvironmentalriskexposure,socialriskexposure,andgovernanceriskexposurealsoapparentlydeclined(seeFigures3.3,3.4,and3.5).
133
Figure
3.4:
Social
risk
prevalence
in
overseas
infrastructure
portfolio
from
China
to
LICs
and
MICsFigure3.4
Notes:ProjectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfromChina,although
theESGrisksthattheyencounteredmayhavematerializedafterthefinancialcommitmentyear.The
presenceofsignificantsocialriskexposureisbasedonaproject-levelcompositemeasurethatisdescribed
inSection2ofChapter3.
134
Figure
3.5:
Governance
risk
prevalence
in
overseas
infrastructure
portfolio
from
China
to
LICs
and
MICsFigure3.5
Notes:ProjectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfromChina,although
theESGrisksthattheyencounteredmayhavematerializedafterthefinancialcommitmentyear.The
presenceofsignificantgovernanceriskexposureisbasedonaproject-levelcompositemeasurethatis
describedinSection2ofChapter3.
TableA12providescountry-levelsummarystatisticsonthenumberand
monetaryvalueofChinesegrant-andloan-financedinfrastructureprojects
between2000and2021withsignificantenvironmental,social,andgovernance
riskexposure.Italsoprovidesacountry-by-countrybreakdownoftheoverall
ESGriskprevalencerateoverthesametimeperiod.Analysisofthecountry-level
datafromTableA12demonstratesthattheESGriskinChina’soverseas
infrastructureprojectportfolioisdisproportionatelyconcentratedincertain
regions(seeTableA10).Forexample,only16.9%ofChina’sgrant-and
loan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)
waslocatedinSouthandCentralAmericabetween2000and2021,but38%of
itsportfoliowithsignificantgovernanceriskexposurewasconcentratedinthe
135
sameregionduringthesametimeperiod.258ESGriskinBeijing’soverseas
infrastructureprojectportfolioisalsounevenlydistributedacrosscountrieswith
differentpercapitaincomelevels(seeTableA11).Governanceriskisagaina
caseinpoint.37.1%ofChina’sgrant-andloan-financedinfrastructureproject
portfolio(measuredinconstant2021USD)waslocatedinupper-middleincome
countries(UMICs)between2000and2021.259Yetastaggering52.5%ofits
portfoliowithsignificantgovernanceriskexposurewasconcentratedinsuch
countries.260TableA12providesevidencethatasmallsubsetoflargeaidand
creditrecipients—includingVenezuela,Malaysia,andArgentina—contributedto
thedisproportionateconcentrationofgovernanceriskexposureinUMICs.
InFigure3.6,weidentifyglobalhotspotsbyfusingdataontheenvironmental,
social,andgovernanceriskexposureofChinesegrant-andloan-financed
infrastructureprojectswithpoint,polygon,andlinevectordata(describedin
Chapter1)thatcapturethegeographicfootprintsoftheseprojects.Todoso,
wefirstcreatea200kmx200kmgridcoveringeveryLICandMICinthe3.0
versionofAidData’sGCDFdataset.Wethenusethepoint,polygon,andline
vectordatatoassigneverygeocodedinfrastructureprojecttooneofmoreof
thesegridcells.261Wesubsequentlyassigneachgridcellagradationof
color—alonga“heat”spectrum—basedonthecumulativemonetaryvalueof
Chinesegrant-andloan-financedinfrastructureprojectswithenvironmental,
social,orgovernanceriskexposureinthatgeographicalareabetween2000and
261Ifaprojectfallsacrossmultiplegridcells,weassumethemonetaryvalueofthecommitmentfortheprojectisevenlydistributedwithintheproject’slineorpolygon.Thus,thetotalfinancialcommitmentvaluefortheprojectissplitupamonggridcellsbaseduponthepercentageoftheproject’sareathatfallswithineachgridcell.260Only15.8%ofChina’sgrant-andloan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)withsignificantsocialriskexposure—and25.6%withsignificantenvironmentalriskexposure—waslocatedinUMICs(seeTableA11).259AccordingtoTableA11,lower-middleincomecountries(LMICs)andlow-incomecountries(LICs)received26.8%and25.3%,respectively,ofChina’sgrant-andloan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)duringthe2000-2021period.However,35.8%ofitsportfoliowithsignificantsocialriskexposurewasconcentratedinLMICs—and37.8%ofitsportfoliowithsignificantsocialriskexposurewasconcentratedinLICs—between2000and2021.258TableA10demonstratesthatenvironmentalriskexposurewasdisproportionatelyconcentratedinCentralandEasternEurope:whereas13%ofChina’sgrant-andloan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)waslocatedintheregionbetween2000and2021,20.5%ofitsportfoliowithsignificantenvironmentalriskexposurewasconcentratedinCentralandEasternEuropeduringthesame22-yearperiod.Bycontrast,socialriskexposurewasdisproportionatelyconcentratedinAsia:whereas36.4%ofChina’sgrant-andloan-financedinfrastructureprojectportfolio(measuredinconstant2021USD)waslocatedintheregionbetween2000and2021,42.3%ofitsportfoliowithsignificantsocialriskexposurewasconcentratedinAsiaduringthat22-yearperiod(seeTableA10).
136
2021.262WhereaslightpinkgridcellsrepresentareaswhereChinahasa
relativelylowlevelofriskexposureinitsinfrastructureprojectportfolio,dark
purplegridcellsrepresentareaswhereChinahasarelativelyhighlevelofrisk
exposureinitsinfrastructureprojectportfolio.
Themapintheupper-lefthandcornerofFigure3.6demonstratesthatBeijing
hasaparticularlyhighlevelofenvironmentalriskexposureintheTropicalAndes
(includingVenezuela,Ecuador,andPeru),theSouthernCone(including
Argentina),EastAfrica(includingEthiopia,Kenya,andUganda),WestAfrica
(includingGhana,Togo,Benin,Nigeria,andCameroon),CentralAsia(including
Kazakhstan,Kyrgyzstan,andTajikistan),andSoutheastAsia(includingLaos,
Cambodia,Vietnam,andIndonesia).Thegeographicaldistributionofsocialrisk
exposure,asdepictedinthemapintheupper-righthandcornerofFigure3.6,is
broadlysimilar,althoughthehotspotsarelessconcentratedinCentralAsiaand
moreconcentratedinEthiopia,Kenya,Pakistan,andSoutheastAsia.Themapin
thebottom-lefthandcornerofFigure3.6alsodemonstratesthatBeijinghasa
particularlyhighlevelofgovernanceriskexposureintheTropicalAndes,East
Africa,andSouthAsia—includingZambia,Bangladesh,andArgentina(three
countriesforwhichweprovidein-depthcasestudyevidenceinChapter4).
Finally,inthebottom-righthandcornerofFigure3.6,wecollapseallthree
categoriesofriskexposureintoasinglemap,suchthateachgridcellcaptures
theextenttowhichChinesegrant-andloan-financedinfrastructureprojectsin
thatareaencounteredsignificantenvironmental,social,orgovernmentrisks.
262InFigureA62,wereplicateFigure3.6butscalethelevelofriskexposureinagivengridcellaccordingtothecumulativecountofChinesegrant-andloan-financedinfrastructureprojectsratherthanthecumulativemonetaryvalueofChinesegrantandloancommitmentsforthesameprojects.
137
Figure
3.6
A
global
map
of
China’s
infrastructure
project
portfolio
in
LICs
and
MICs
with
significant
environmental,
social,
governance
(ESG)
risk
exposureFigure3.6
A
global
map
of
China’s
infrastructure
project
portfolio
in
LICs
and
MICs
with
significant
environmental,
social,
governance
(ESG)
risk
exposure
Notes:ThismappresentsthegeographicalareaswhereChina’sgrant-andloan-financedinfrastructure
projectportfolio(measuredinconstant2021USD)hassignificantenvironmental,social,orgovernance
(ESG)riskexposure.Darker(purple)colorsrepresentareaswheretheportfoliohashighlevelsofrisk
exposureandlighter(pink)colorsrepresentareaswheretheportfoliohaslowerlevelsofriskexposure.
Environmentalriskexposure,socialriskexposure,andgovernanceriskexposurearebasedonthe
project-levelcompositemeasuresthataredescribedinSection2ofChapter3.
138
Section3:MeasuringthestringencyofESGsafeguardsinChina’sinfrastructureprojectportfoliowithnewsourcesofcontractualevidence
AlthoughBeijingclearlyfacesawidearrayofESGrisksinitsoverseas
infrastructureprojectportfolio,littleisknownaboutthesafeguardsthatithas
putinplacetomanageandmitigatetheserisks.Anotherblindspotiswhether
andhowChinesestate-ownedlendershavestrengthenedorweakenedtheir
ESGsafeguardsovertime.
The3.0versionofAidData’sGCDFdatasetprovidesauniqueopportunitytofill
thisevidentiarygap.263Aspartoftheprimarysourceidentificationworkthatwas
undertakentosupporttheconstructionofthedataset,AidDataobtainedalarge
cacheofunredactedinfrastructurefinancingagreementsviaofficialsourcesin
LICsandMICs,includinggovernmentregistersandgazettes,aidanddebt
informationmanagementsystems,andparliamentaryoversightinstitutions.
Thesegrantandloanagreementsrepresent“high-valuesources,”inthatthey
providedetailedinformationaboutwhetherfinanciers,atthetimethatthey
signedtheagreementswiththeirhostcountrycounterparts,identified
behavioralexpectationsrelatedtoESGriskmanagementandmechanismsto
monitorandenforcecompliancewiththoseexpectations.
Anotherimportantfeatureofthe3.0versionofAidData’sGCDFdataset—and
animprovementoverthe2.0version—isthatitmakestheseunredacted
agreementsavailableforthefullrangeoffinancialinstrumentsthatBeijinguses
tobankrollinfrastructureprojectsinthedevelopingworld,including:
1.Bilateralgrantsandinterest-freeloansissuedbyChina’sMinistryof
Commerce(MOFCOM)
2.BilateralloansissuedbyChinaEximbank
263The3.0versionoftheGCDFdatasetprovidesstableURLstohundredsofunredactedgrant,loan,debtforgiveness,debtrescheduling,andescrowaccountagreements.AidDatapublishedasubsetofthesefinancingagreementsinMarch2021whentheHowChinaLendsreportwasfirstpublished(Gelpernetal.2021,2022).However,the3.0datasetprovidesthefullsetofagreementsretrievedbyAidData.
139
3.BilateralloansissuedbyChinaDevelopmentBank(CDB)
4.BilateralloansissuedbyChinesestate-ownedcommercialbanks,suchas
ICBC,ChinaConstructionBank,andBankofChina
5.SyndicatedloansissuedbyChina’spolicybanks(ChinaEximbankand
CDB)andstate-ownedcommercialbanks
6.SyndicatedloansissuedbyChinesestate-ownedbanksandmultilateral
institutions
7.GrantsandloansthatChinahaschanneledviamultilateralinstitutions
8.Supplier’screditsissuedbyChinesestate-ownedcompanies
Theseeighttypesoffinancingagreements,whichaccountfor90%ofChina’s
grant-andloan-financedinfrastructureprojectportfoliointhedevelopingworld
between2000and2021,includewidelydivergentESGtermsandconditions
(seeTablesA5andA8).264However,variationindejureESGsafeguard
stringencyhasneverbeensystematicallydocumentedacrossagreementtypes.
NorhaspreviousresearchdemonstratedhowBeijing’suseofthesedifferent
typesofagreements—withvaryinglevelsofdejureESGsafeguard
stringency—haschangedwiththepassageoftime.
Inordertoovercometheseobstacles,wedevelopedastandardizedsetof
codingcriteriarelatedtoESGriskmanagementthatcanbeappliedtoanytype
ofChineseloancontractorgrantagreementthatsupportsanoverseas
infrastructureproject.265These26criteria,whicharedescribedinSectionA-8
andTableA3intheAppendix,include8focusedonenvironmentalsafeguards,
265Environmentalandsocialsafeguardsaretypicallyinapplicabletoprojectsthatdonotinvolvetheconstruction,rehabilitation,orexpansionofinfrastructure,althoughtherearesomeexceptionstothisgeneralrule(mostnotably,projectsthatinvolvenaturalresourceextractionwithoutinfrastructurecomponents).264These8financialinstrumenttypeswereusedbyChinatosupport90.2%ofitsgrant-andloan-financedinfrastructureprojectportfolioinLICsandMICsbetween2000and2021.Theremaining9.8%oftheportfolioconsistedofprojectssupportedbymore“exotic”financialinstrumenttypes(e.g.,Engineering,Procurement,ConstructionandFinancing(ECPDF)agreements).The3.0versionoftheGCDFdatasetdoesnotincludemanyunredactedfinancingagreementsfortheseprojects,soweexcludethemfromouranalysis.
140
7focusedonsocialsafeguards,and11focusedongovernancesafeguards.They
arebroadlyalignedwiththeOECDCouncilRecommendationonCommon
ApproachesforOfficiallySupportedExportCreditsandEnvironmentaland
SocialDueDiligence,theIFC'sPerformanceStandardsonEnvironmentaland
SocialSustainability,theUniformFrameworkforPreventingandCombating
FraudandCorruption,theOECDCouncilRecommendationonBriberyand
OfficiallySupportedExportCredits,andtheOECDCouncilRecommendationon
PublicProcurement.266Thecriteriaareorganizedintothreegroups:thosethat
identifythepresenceorabsenceof(1)rulesorstandardstoestablishbehavioral
expectationsrelatedtoESGriskmanagementandmitigation,(2)oversight
mechanismsformonitoringcompliancewiththosebehavioralexpectations;
and/or(3)enforcementmechanismsforsanctioningnoncompliancewiththose
behavioralexpectations(e.g.,indemnification,withholdingdisbursements).
Toconstructourcodingsample,wefirstidentifyalloftherecords(nearly300)in
the3.0versionoftheGCDFdatasetthatincludeunredactedloancontractsand
grantagreements.Wethenremovealloftheloancontractsandgrant
agreementsthatdonotsupportinfrastructureprojects.Wesubsequently
eliminateallloancontractsandgrantagreementsthatdonotcorrespondtoone
ofthe8primaryinfrastructurefinancingagreementtypes.AsshowninSection
A-10,wethenprunetheremainingsampleofloancontractsandgrant
agreementstoidentify3agreementsforeachofthe8financialinstrument
categories267thatprovidebroadgeographicalcoverage(acrossAfrica,Latin
AmericaandtheCaribbean,AsiaandthePacific,CentralandEasternEurope,
andtheMiddleEast)andincomebracketcoverage(acrossupper-middleincome
countries,lower-middleincomecountries,low-incomecountries,andleast
developedcountries),andtemporalcoverage(overour22-yearperiodofstudy).
Foreachfinancialinstrumentcategory,wealsoseektoidentifyagreements
267Foroneoftheeightfinancialinstrumentcategories(“syndicatedloansissuedbyChinesestate-ownedbanksandmultilateralinstitutions”),wewereonlyabletoidentifytwoinfrastructureprojectfinancingagreements.266TheOECDCouncilRecommendationonCommonApproachesforOfficiallySupportedExportCreditsandEnvironmentalandSocialDueDiligencewaspreviouslyknownastheOECDRevisedCouncilRecommendationonCommonApproachesontheEnvironmentandOfficiallySupportedExportCredits.
141
issuedbeforeandafterthelateBRIperiod,268giventhatBeijinghasmademany
rhetoricalcommitmentstostrengthenESGprotectionssincelate2017.
Althoughwedonotselectinfrastructurefinancingagreementsforcoding
purposesthrougharandomsamplingprocedureandourcodingsample
representsasmallpartofChina'sgrant-andloan-financedinfrastructureproject
portfolioinLICsandMICs,ourfindingsdemonstratethatESGtermsand
conditionsarehighlystandardizedbyinfrastructurefinancinginstrument,which
givesusconfidencethatwearecapturingmeaningfuldifferencesindejureESG
safeguardstringencyacrossthemainfinancialinstrumentsthatBeijingusesto
bankrolloverseasinfrastructureprojects.269Wealsofindrelativelylittlevariation
inESGtermsandconditionsacrosscountriesindifferentregionsandincome
brackets(seeTableA4intheAppendix).270Inthisrespect,ourfindingsare
consistentwiththefirststudytoeversystematicallyevaluatethetermsand
conditionsgoverningChina’sloancontractswithoverseasborrowers.Gelpernet
al.(2022:16)concludethat“ouranalysisof[100]contractsshowsthatChinese
lendingtermsarehighlystandardizedbylenderandinstrument,anddonot
exhibitsignificantvariationby[...]regionorincomebracket.”
Thecontract-leveldatafromourcodingsampleareprovidedinTableA8ofthe
Appendix.Inordertoconvertthecontract-leveldataintocategoricalmeasures
ofsafeguardstringencyforeachfinancialinstrumenttype,wefirstmakebinary
determinationsofwhetherthereisanyevidencethateachfinancialinstrument
type(beforeorafterthelateBRIperiod)establishedany(a)rulesorstandards
thatcreatebehavioralexpectationsrelatedtoESGriskmanagementand
mitigation,(b)oversightmechanismsformonitoringcompliancewiththose
270OurcodingsampleunderrepresentsChina’sinfrastructurefinancingtosomeregionsandincomebracketsandoverrepresentsitsinfrastructurefinancingtootherregionsandincomebrackets(seeSectionA-10).TheexternalvalidityofoursamplewouldbeaconcernifChina’sinfrastructurefinancingagreementsvariedsystematicallybyregionorincomebracket.However,wedonotfindmuchevidencethatChina’sinfrastructurefinancingagreementsdiffersignificantlybyregionorincomebracket(seeTableA4intheAppendix).269Theagreement-levelratingsthatarereportedinTableA8demonstratethatmostoftheobservedheterogeneityinESGsafeguardstringencyisacrossfinancialinstrumenttypesratherthanacrossagreementswithinagivenfinancialinstrumenttype.SeealsoTablesA6andA7.268Fortwooftheeightfinancialinstrumentcategories(“supplier’screditsissuedbyChinesestate-ownedcompanies”and“syndicatedloansissuedbyChina’spolicybanksandstate-ownedcommercialbanks”),wereliedoninfrastructurefinancingagreementsthatwereissuedin2022(inlieuofagreementsissuedbetween2018and2021)toensureadequatecoverageduringthelateBRIperiod(seeTablesA5,A6,andA7intheAppendixformoredetails).
142
behavioralexpectations,or(c)enforcementmechanismsforsanctioning
noncompliancewiththosebehavioralexpectations.Baseduponthese
determinations,whicharereportedinTableA8,weassignhigh,medium,orlow
environmental,social,andgovernancesafeguardratingstoeachfinancial
instrumenttypeusingthefollowingcriteria:
●Low:Norulesandstandardsexistandtherearenomechanismsfor
monitoringcomplianceorsanctioningnoncompliance.
●Medium:Rulesandstandardsexist,buttherearenomechanismsfor
monitoringcomplianceorsanctioningnoncompliance.
●High:Thereisamechanismformonitoringcomplianceand/ora
mechanismforsanctioningnoncompliance.
Ourapplicationofthestandardizedcodingcriteriatothesampleofgrantand
loanagreementsproducesasetofsummaryESGratingsforthe8financial
instrumentcategoriesovertwotimeperiods:thepre-BRIandearlyBRIperiod
(2000-2017)andthelateBRIperiod(2018-2021).271Thesesummaryratings,
whichmeasurethestrengthofESGsafeguardsinadejureratherthanadefacto
sense,areprovidedinTable3.1andtheycallattentiontoseveralimportant
patternsandtrends.272First,amongtheinfrastructurefinancinginstrumentsat
Beijing’sdisposal,policybank(ChinaEximbankandCDB)loanagreementsoffer
theweakestESGsafeguards.273ThiswascertainlytruebeforetheBRIwas
273ChinaEximbank’sinfrastructureloanagreementsreceivedlowenvironmental,social,andgovernancesafeguardratings.CDB’sinfrastructureloanagreementsreceivedlowenvironmentalandsocialsafeguardratings,butamediumgovernancesafeguardrating—duetothefactthattwooutofthethreeCDBcontractsinthecodingsampleincludedanti-corruptionandanti-moneylaunderingrequirementsaswellasrequirementstoprepareandsubmitfinancialstatementsinaccordancewithInternationalFinancialReportingStandards(IFRS)ortheGenerallyAcceptedAccountingPrinciples(GAAP)oftheFinancialAccountingStandardsBoard(FASB).272The“High,”“Medium,”and“Low”designationsarenotcomprehensivemeasurementsofESGsafeguardstringencyvis-à-visinternationalstandards,suchasPS6.TheyonlyprovidemeasurementsofwhetherESGrulesandstandardsexistandwhethertherearemechanismsinplaceformonitoringcomplianceorsanctioningnoncompliance.Apotentiallyproductiveavenueforfutureresearchwouldbetoconstruct“distance-to-frontier”safeguardstringencymeasuresthatarebasedonPS6orananalogoussetofinternationalstandardsthatarebroadlyencompassing.271BeyondthefactthatsomeChinesefinanciershavepublishedenvironmentalpoliciesandstandardsontheirwebsitesandothershavenot,Narainetal.(2020)doesnotdocumentanysafeguardvariationacrossChinesestate-ownedcreditorsthatfinanceoverseasinfrastructureprojects.However,Narainetal.(2020)doesnotsystematicallyevaluatethesafeguardprovisionscontainedinthefinancingagreementsofChinesestate-ownedcreditors.NordoesthestudycaptureanyofthechangesthattookplaceduringthelateBRIperiod(2018-2021).
143
launched,anditremainedlargelytrueduringtheearlyandlateBRIperiods.274
Second,China’sstate-ownedcommercialbankshavestrongdejureESG
safeguardsintheiroverseasloanagreements.Theynotonlyapplysuch
safeguardswhentheyissuebilateralloanagreements,butalsowhenthey
participateinsyndicatedloanagreements.Third,strongESGsafeguards
consistentlyapplytothegrantsandloansthatthePBOCandChina’sMinistryof
FinancechanneltoLICsandMICsviamultilateralinstitutions.Theyalsoapplyto
syndicatedloansthatinvolvemultilateralinstitutions,whichhighlightsafourth
(broader)findingfromTable3.1:thefactthatsyndicatedloanshaveconsistently
strongerdejureESGsafeguardsthanbilateralloans.Giventhatallparticipants
inasyndicatedloanagreementforaninfrastructureprojectmustagreetoa
commonsetofcontractualtermsandconditions,includingapplicableESG
safeguards,onemightthinkthata“leastcommondenominator”dynamiccould
gointoeffect.ButTable3.1indicatesthattheoppositeistrue:syndicate
participantsseemtodefertothelendinginstitution(s)withthestrongest
preference(s)forESGriskmitigation.275
Table
3.1:
De
jure
ESG
safeguard
stringency
in
China’s
overseas
infrastructure
portfolio
by
type
of
financing
instrumentTable3.1
De
jure
ESG
safeguard
stringency
in
China’s
overseas
infrastructure
portfolio
by
type
of
financing
instrument
FinancingInstrumentEnvironmental
SafeguardsSocial
SafeguardsGovernance
Safeguards
Pre/Early
BRILateBRIPre/Early
BRILateBRIPre/Early
BRILateBRI
BilateralChinaEximbankloanLowLowLowLowLowLow
BilateralCDBloanLowLowLowLowMediumMedium
BilateralMOFCOMloanorgrantLowMediumLowMediumLowHigh
BilateralChinesestate-owned
commercialbankloanHighHighHighLowHighHigh
SyndicatedloanwithChineseand
multilateralbankparticipantsHighHighHighHighHighHigh
275Theapparentbenefitofincludingamultilateralinstitutionorastate-ownedcommercialbankinalendingsyndicateisthatitcanleadeveryothermemberofthesyndicatetoadopttheir(stronger)safeguards.274ThesefindingsareconsistentwiththoseofNarainetal.(2020,2022).
144
FinancingInstrumentEnvironmental
SafeguardsSocial
SafeguardsGovernance
Safeguards
Pre/Early
BRILateBRIPre/Early
BRILateBRIPre/Early
BRILateBRI
SyndicatedloanwithChinese
state-ownedcommercialbanks
and/orpolicybanksHighHighHighLowHighHigh
PBOC/MOFgrantorloanchanneled
throughmultilateralinstitutionsHighHighHighHighHighHigh
Supplier’screditfromChineseSOELowHighLowLowLowHigh
Notes:Thesafeguardstringencyscoresforeachtypeofgrant-givingandlendinginstrumentarebasedon
theanalysisdescribedinSectionA-9oftheAppendix.
Severalchangesthattookplaceduringourperiodofstudy(2000-2021)also
meritdiscussion.MOFCOM’sgrantandinterest-freeloanagreementshadweak
ESGsafeguardspriortothelateBRIperiod.However,weseeevidenceof
MOFCOMshiftingtowardstrongerdejureESGprotectionsbetween2018and
2021.Thesamepatternisevidentinsupplier’screditagreementsissuedby
Chinesestate-ownedenterprises:formalESGsafeguardstringencyincreased
withthepassageoftime.276Table3.1alsoprovidesevidencethat,duringthe
lateBRIperiod,China’sstate-ownedcommercialbankswatereddowntheir
socialsafeguards.277Duringthepre-BRIandearlyBRIperiods,thesefinancial
institutionshadmechanismsinplacetomonitorcomplianceand/orsanction
noncompliancewithdomesticandinternationalsociallawsandstandards.278
Theirbilateralloanagreementsandsyndicatedloanagreementsmadefinancial
disbursementsconditionaluponcertificationofcompliancewithsociallawsand
standards,orrequiredborrowerstofinanciallycompensate(indemnify)lenders
foranylossesorliabilitiesresultingfromactualorallegedviolationsofsocial
278Inthesampleoffinancingagreementsthatweevaluated,sociallawsandstandardsweregivenexpansivedefinitions,including(a)laws,rules,andregulationsinborrowercountriesrelatedtowork,socialsecurity,industrialrelations,occupationalhealthandsafety,publicparticipation,propertyownership(formalandtraditional),andtheprotectionandempowermentofindigenouspeoplesandethnicgroups;theprojection,restoration,andpromotionofculturalheritageandarchaeologicalartifacts;andtheresettlementoreconomicdisplacementofpersons;(b)theOECDRevisedCouncilRecommendationonCommonApproachesontheEnvironmentandOfficiallySupportedExportCredits;(c)theEquatorPrinciples;(d)UNtreatiesandconventionsonhumanrights;and(e)internationallaboragreements.277Table3.1providesevidencethatChina’sstate-ownedcommercialbanksweakenedthesocialsafeguardsthatapplytotheirbilateralloansandtheirsyndicatedloansduringthelateBRIperiod.276Althoughtheenvironmentalandgovernancesafeguardsthatapplytosupplier’screditsstrengthenedduringthelateBRIperiod,thesocialsafeguardsthatapplytosupplier’screditsdidnot.
145
lawsandstandards.Yet,forreasonsthatwedonotyetunderstand,these
safeguardsvanishedduringthelateBRIperiod.
ThenextstepinouranalysisistoapplytheESGsafeguardstringencyratings
fromTable3.1toChina’sentiregrant-andloan-financedinfrastructureproject
portfoliointhedevelopingworld.Wedosobyfirstmappingallloansand
grantsforactive,completed,suspended,orcanceledinfrastructureprojectsin
the3.0versionoftheGCDFdatasettooneofthe8financialinstrument
categories(wheneverpossible).Then,weassigntheaggregateESGsafeguard
stringencyratings—reportedinTable3.1—totheinfrastructureloansandgrants
inthe3.0versionofGCDFdatasetthatusethesameloanorgrantinstrument.279
279Tomapindividualgrantsandloansforinfrastructureprojectstoourtaxonomyofinfrastructurefinancinginstruments(consistingof8loanandgrant-givinginstruments),weuseacombinationofthefundingagency,implementingagency,co-financingagency,receivingagency,numberoflenders,flowtype,andsupplier’screditsfieldsinthe3.0versionoftheGCDFdataset.These8financialinstrumenttypescover90.2%ofChina’sgrant-andloan-financedinfrastructureprojectportfolioinLICsandMICsbetween2000and2021.WedonotassigndejureESGsafeguardstringencyratingstotheremaining9.8%oftheportfolio,whichrepresentsinfrastructureprojectsfinancedwithothertypesoffinancialinstruments.Assuch,wheneverwereportportfolio-levelsummarystatisticsrelatedtotheapplicationofdejureESGsafeguards,wedisregardprojectsforwhichdejureESGsafeguardstringencyratingscouldnotbereliablyassigned(i.e.,9.8%oftheLICandMICportfolio).
146
Figure
3.7:
De
jure
ESG
safeguard
stringency
in
China’s
overseas
infrastructure
portfolioFigure3.7
Notes:Thesafeguardstringencyratingsforeachgrant-givingandlendinginstrumentarebased
onTable3.1andexplainedinSectionA-9oftheAppendix.
Figure3.7presentstheestimatedpercentageofChina’sgrant-and
loan-financedinfrastructureprojectportfolioinLICandMICswithstrongdejure
environmental,social,andgovernancesafeguardsinplacebetween2000and
2021.OnecanseeamarkedshifttowardstrongerESGprotectionsduringthe
lateBRIperiod(2018-2021).BytheeighthfullyearofBRIproject
implementation(2021),approximately57%ofChina’sgrant-andloan-financed
infrastructureprojectportfolioinLICandMICshadstrongdejure
environmental,social,andgovernancesafeguardsinplace.280Thisrepresentsa
majordeparturefrompastpractice:attheturnofthecentury,China’sentire
grant-andloan-financedinfrastructureprojectportfolioinLICandMICshad
weakdejureenvironmental,social,andgovernancesafeguardsinplace.
280TableA12providescountry-levelsummarystatisticsonthepercentageofChina’sgrant-andloan-financedinfrastructureprojectportfoliowithstrongdejureESGsafeguardsbetween2000and2021.
147
Figure
3.8:
De
jure
safeguard
stringency
in
China’s
overseas
infrastructure
portfolio
by
safeguard
typeFigure3.8.1:Environment
Figure3.8.2:Social
148
Figure3.8.3:Governance
Notes:Thesafeguardstringencyratingsforeachgrant-givingandlendinginstrumentarebasedonTable
3.1andexplainedinSectionA-9oftheAppendix.
However,Beijinghasnotdemonstratedcomparablelevelsofenthusiasmforall
typesofESGsafeguardsduringthelateBRIera.Figure3.8demonstratesthat
37%oftheinfrastructureprojectportfoliowassubjectedtostrongdejure
environmentalsafeguardsfrom2018to2021,ascomparedto20%duringthe
previouseighteen-yearperiod(2000-2017).Similarly,40%oftheinfrastructure
projectportfoliowassubjectedtostrongdejuregovernancesafeguardsfrom
2018to2021,ascomparedto20%duringthepreviouseighteen-yearperiod
(2000-2017).
YetBeijingdemonstratedfarlessinterestinapplyingstringentsocialsafeguards
toitsoverseasinfrastructureprojectportfolioduringthelateBRIera.Between
2018and2021,itshieldedanincreasingproportionofitsgrant-and
loan-financedinfrastructureprojectportfoliofromthesetypesofsafeguards(see
Figure3.8).WhiletheproximateexplanationforthischangeduringthelateBRI
erawastheremovalofsocialsafeguardenforcementmechanismsfromtheloan
contractsofChina’sstate-ownedcommercialbanks(seeTable3.1),the
149
underlyingreasonwhyittookplaceisamystery.Onepossibility—potentially
deservingattentioninfutureresearch—isthatChina’saversiontostrongdejure
socialsafeguardsisrelatedtoitsowndisconcertingexperiencewiththeWorld
Bank’ssocialsafeguardsduringthelate1990sandearly2000s(seeBox3a).281
Box3a:China’sexperiencewiththeapplicationofWorldBanksocialsafeguards
totheWesternPovertyReductionProjectinQinghai
In1997,theWorldBankstartedworkingwithChina’sprovincialgovernmentinQinghaionthedesignofa$40-millionloanfortheWesternPovertyReductionProject.Thepurposeoftheprojectwastoresettleapproximately60,000poorfarmerstoanewirrigationtract.TheresettlementareawaslocatedincentralQinghai,morethan500kilometersfromtheborderoftheTibetAutonomousRegion.However,in1999,atransnationaladvocacynetwork—consistingofTibetNGOs(includingtheTibetInformationNetworkandtheInternationalCampaignforTibet)andmultilateraldevelopmentbankmonitors(suchastheBankInformationCenterandtheCenterforInternationalEnvironmentalLaw)—launchedacampaigntopreventtheWorldBank’sBoardofDirectorsfromgreen-lightingtheproject.Theyclaimed,withsupportfromtheU.S.CongressandU.S.Treasury,thatapprovaloftheprojectwouldbetantamounttobankrollinggenocide(bydilutingTibet’sculturewith60,000ethnicChinese).TheyalsoclaimedthattheWorldBankhadfailedtocomplywithitsownsocialsafeguardspolicy—byclassifyingaprojectas“CategoryB”whenitshouldhavebeenclassifiedas“CategoryA.”282
WhentheBoardofDirectorsvotedtoconditionallyapprovetheprojectinJune1999,agroupofcampaignershunga“WorldBankApprovesChina’sGenocideinTibet”banneroutsideWorldBankheadquarters.RobertWade,whoinvestigatedclaimsabouttheprojectonbehalfoftheWorldBank’sInspectionPanel,recountsthat"[t]heNGOsputtogetheraformidablyeffectivecampaignnetwork.TheyestablishedwebsitestoshareinformationandprovidesampleprotestletterstotheBankwhichcouldbeemaileddirectlyfromthesiteorprintedoutandfaxed.TheTibetlobbysponsoredrockconcertsincitiesaroundtheworld,withpreparedpostcards,faxmachinesandemailfacilitiesonhand.Theresultwasadelugeofletters,postcards,emailsandfaxesthelikeofwhichtheBankhadneverseen,mainlyfromtheU.S.andEurope.TheWesternmedia,bothpressandTV,linedupbehindthecritics.ReportsinleadingnewspaperslikeTheFinancialTimes,TheNewYorkTimesandTheWashingtonPostreadasthoughtakenstraightfromNGOhandouts.TheyrepeatedtheNGOs’portraitoftheprojectinthesamelanguage,oftennotdistinguishingbetweenwhattheNGOsclaimedandwhatthey,thejournalists,reportedasfact.Manyreportedasfact,forinstance,thatthemove-inareawasthebirthplaceoftheDalaiLama,whichissimplyfalse[…].YetforalltheirclaimstospeakforTibetansandforall
282CategoryAprojectsposethemostsevereenvironmentalandsocialrisks;theyofteninvolvelarge-scaleinfrastructure,industrial-scalechemicalmanufacturing,ornaturalresourceextractionactivities.TheWorldBanksubjectstheseprojectstoitsmoststringentESGsafeguards,butitalsoacknowledgesthattheriskstheseprojectsposecanbedifficultorimpossibletofullymitigate.CategoryBprojectsalsoposesignificantenvironmentalandsocialrisks,buttheWorldBankexpectsthatitcanreasonablyandreadilymitigateallormostoftheserisksduringimplementation(Buchananetal.2018).281Itis,however,worthnotingthattherewasnevermuchsupportinBeijingforstrongsocialsafeguardsacrossourentire22-yearperiodofstudy(2000-2021).
150
theirdenunciationoftheconsultationprocess,theNGOsneverproducedevidencethatlocalpeopledidnotwanttheprojectbeyondafewverybriefandanonymousletterssenttotheTibetNGOsbypeopleclaimingtolivenearthemove-inarea”(Wade2009:32).
InJuly2000,theWorldBank’sBoardofDirectorsconvenedtodecideifitwouldapprovetheproject.Thediscussionsdraggedonformultipledays,withdevelopingcountryrepresentativesadvocatingforprojectapprovalandcertaindevelopedcountryrepresentativescallingforprojectcancellation.TheissuewasultimatelyresolvedwhenChina’sExecutiveDirectorwithdrewtheprojectproposalfromconsideration.Beijingannouncedthattheprojectwouldproceedwithanalternativesourceoffunding.
Ofcourse,Beijing’scriticsandrivalsmightquestionwhetheranyofthe“fine
print”initsoverseasinfrastructurefinancingagreementsevenmattersifESG
safeguardsarenotputintopractice.TogaugewhetherChina’sdefacto
applicationofESGsafeguardsmatchesthedejureESGsafeguardsinits
financingagreements,weleveragethedetailedqualitativeinformationthat
AidDatahascollectedabouthowprojectsweredesignedandimplementedin
practice.The“cradletograve”narrativesinthe3.0versionoftheGCDFdataset
includedetaileddescriptionsofeffortstomitigateESGrisksbefore,during,and
afterprojectimplementation—forexample,byadoptingenvironmental
managementplans(EMPs)thatrespondtothefindingsandrecommendationsof
anenvironmentalimpactassessmentorbyprovidingfinancialcompensationto
project-affectedpersons(PAPs).
Tomakeeffectiveuseofthisqualitativeinformation,weuseasetofsystematic
searchandcategorizationprocedures(describedinSectionA-7intheAppendix)
toidentifythesubsetofinfrastructureprojectsforwhichthereisevidenceof
effortsbeingundertakenbyChinesefinanciersorimplementingagenciesto
mitigateenvironmental,social,orgovernancerisksbefore,during,orafter
project’simplementation.Between2000and2021,wefindevidencethatde
factoESGriskmitigationeffortswereundertakentosupportatleast210
infrastructureprojectsin66LICsandMICssupportedbygrantsandloansfrom
China.283TheestimatedcumulativevalueofChina’sgrant-andloan-financed
infrastructureprojectportfoliosupportedbydefactoESGriskmitigationefforts
increasedfrom$55millionin2000to$86billionin2021(seeFigureA48).
283The“atleast”qualifierisimportantbecauseofourinabilitytoaddress“falsenegatives”thatmayaffectourkeyword-searchbasedmeasures(seeSectionA-7intheAppendix).
151
AccordingtoFigure3.9,thepercentageofChinesegrant-andloan-financed
infrastructureprojectsinthedevelopingworldsupportedbyadefactoESGrisk
mitigationeffortsteadilyincreasedfrom2%in2000to18%in2021,whichis
broadlyconsistentwithChina’sincreasinguseofstrongdejureESGsafeguards
overthesametwenty-twoyearperiod.284However,Figure3.9alsohighlightsan
importantshiftthattookplaceovertime:althoughdejureanddefactorisk
mitigationeffortsmostlymovedintandemduringthepre-BRIera(2000-2013),
the“delta”betweendejureanddefactoriskmitigationeffortswidenedduring
theBRIera(2014-2021).285By2021,thegapbetweenhowBeijingappliedESG
safeguardsinprincipleandinpracticewassubstantial:57%ofitsinfrastructure
projectportfolioinLICsandMICsbenefitedfromstrongdejureESG
safeguards,yettherewasevidenceofdefactoESGriskmitigationeffortsbeing
undertakeninonly18%oftheportfolio(seealsoBox3b).286
286AnimportantcaveatisthatourmeasureofwhetheranyeffortwasundertakentomitigateESGrisksbefore,during,orafterproject’simplementationalmostcertainlyunderestimatesthetruelevelofriskmitigationeffort(duetothepreviously-mentioned“falsenegative”challenge).285Figure3.9treatsaninfrastructureproject’sfinancialcommitmentyearastheyearinwhichESGriskmitigationeffortswereundertaken.However,giventhattheprobabilityofESGriskmitigationmeasuresbeingundertakenanddetectedincreasesasaprojectprogressesfromthefinancialcommitmentphasetotheimplementationphase(andthecompletionphase),itmayalsobeusefultotreataninfrastructureproject’scommencement(implementationstart)yearoritscompletion(implementationend)yearastheyearinwhichESGriskmitigationeffortswereundertaken.WedosoinFiguresA51andFigureA52.However,thesetwofiguresdonotshowsubstantiallysmaller(orlarger)gapsbetweendejureanddefactoESGriskmitigationefforts.InFigureA51,theaverageannualpercentagepointdifferencebetweeninfrastructureprojectswithstrongdejureESGsafeguardsandinfrastructureprojectsthatinvolveddefactoESGriskmitigationeffortsbasedonthecompletionyearis10%.InFigureA52,itis13.2%basedonthecommencementyear.InFigure3.9,itis12%.284IntheAppendix,wepresentadifferentversionofthisgraph(FigureA50)thatmeasurestheannualpercentageofChina’sgrant-andloan-financedinfrastructureprojectsinLICsandMICssupportedby(a)oneormoredefactoESGriskmitigationeffortsand(b)strongdejureESGsafeguards.IttooshowsthatBeijing’sdejureriskmitigationeffortsgenerallyoutpaceditsdefactoriskmitigationefforts.
152
Figure
3.9:
Infrastructure
project
portfolio
with
de
jure
vs.
de
facto
ESG
risk
mitigationFigure3.9
Notes:DefactoESGriskmitigationeffortsaremeasuredusingthemethodologythatis
describedinSectionA-8.StrongdejureESGsafeguardsaredefinedinSectionA-9ofthe
Appendix.
Box3b:DejureversusdefactoapplicationofESGsafeguardstotheLahore
OrangeLineMetroTrainProject
TheLahoreOrangeLineisPakistan’sfirst-everurbanmassrailtransitproject.SinceitsinaugurationinOctober2020,theaveragelevelofdailyridership(178,714)onthe27-kmmetrolinehasremainedbelowcapacity,buttransformedthemegacity’spublictransportlandscape(Hasnain2023).DuringPresidentXiJinping’sApril2015visittoPakistan,theprojectwasgrandfatheredintotheChina-PakistanEconomicCorridor(CPEC)asa“giftfromChina”(Khan2018).ButitwasultimatelyfinancedbyChinaEximbankwithamixofconcessionalandnon-concessionalloans,includinga$1.2billionpreferentialbuyer’screditwitha2%interestrate,anRMB1.2billiongovernmentconcessionalloanwitha2%interestrate,anda$203millionbuyer’screditloanwitha5.2%interestrate.287Pakistan’sgovernmentusedtheloanproceedstopartiallyfinancea$1.63billioncommercialcontractbetweenCR-NORINCO—ajointventureofChinaStateRailwayGroupCo.Ltd.(CR)andChinaNorthIndustriesCorporation(NORINCO)—andPunjabMassTransitAuthority.CR-NORINCO,inturn,hiredlocalcontractorstoassistwithavarietyofactivities,includingtheproject'senvironmentalimpactassessment
287Formoredetails,seeProjectID#54420,53820,37280inthe3.0versionofAidData’sGCDFdataset.
153
(EIA),whichwasconductedbyalocalstate-ownedengineeringservicescompany(NESPAK)priortocommencementofconstructioninAugust2015(NESPAK2015b).
NESPAK’s“comprehensiveandcomplete”studies,whichincludedtheEIAanda37-pageenvironmentalmanagementplan(EMP),weredeemedbythird-partyevaluatorstobe“compliantwithinternationalcodesandstandards”(NESPAKv.Mumtaz2017).TheEMPidentified—andsuggestedcorrectivemeasuresfor—aseriesofrisksrelatedtolandacquisitionandresettlement,floraandfauna,airqualityandnoiselevel,publicutilities,seismichazard,andthehealthandsafetyofworkers(NESPAK2015a).Afterrecognizingthatseveralheritagesites,asdefinedbytheAntiquitiesActof1975(“Act”),wouldbeaffectedbyconstruction,NESPAKaffirmedtheneed“toavoidanyinterferencewithculturalheritagesite(s)andpublicpropertyasfaraspossible”(NESPAK2015b).Notingthatheightenednoiselevelscouldaffectthestructuralintegrityofculturalheritagesites,itcalleduponcontractorstoemploy“noisebarriersduringconstruction”(NESPAK2015a).EventhoughtheActprohibitsconstructionactivitywithin200feetofheritagesites,basedontheseplans,theDirectorGeneral(DG)ofArcheologyissuedaNoObjectionCertificate(NOC)inNovember2015,“givingpermissiontocarryonconstructionwithinprohibitedlimitsof200feetofprotectedantiquities”(Mumtazv.Punjab2016).
Lahore’siconic17th-centurymonument,Chauburji,wasbuiltbyMughalEmperorShahjehanforhisbeloved
daughterJahanaraBegumandservedasanentrancetoaroyalgarden,isshownherewiththeOrangeLine
inthebackground.Itisoneof11heritagesitesaffectedbytheproject’sconstructionactivities.
PhotoCredit:AnamHussain/AlJazeera
Atthetime,allenvironmentalandsocialrequirementsunderlocallawsappearedtohavebeenmet,givingCR-NORINCOanditslocalsubcontractorsthego-aheadtoproceedwithimplementation.However,whenconstructioncrewsbeganmarkingsitesfordemolitionand
154
earthworksinOctober2015,itbecameapparenttolocalcommunitiesandcivilsocietygroupsthat“constructionwork[would]becarriedoutwithin95feetofShalimarGardens”andseveralotherheritagesites(Ghani2015).Almostimmediately,agroupofprominentenvironmentallawyers,urbanists,andrightsadvocatespetitionedtheLahoreHighCourt(LHC),arguingthattheissuanceoftheNOCwas“notonlyarbitrary,malafide,patentlyillegal,withoutlawfulauthoritybutalsowithoutapplicationofindependentmind”(Mumtazv.Punjab2016).BeforeissuingtheNOC,theDGofArcheologyallegedlydidnotconsiderPakistan’scommitmentstointernationalconventionsforheritageconservationandwaspressuredbythegovernment“toissueNOCwithintwodaystimewithoutconsultinganyindependentexperts”(Mumtazv.Punjab2016).
Afterthegovernmentfailedtoprovidesatisfactoryresponsestotheseconcerns,inJanuary2016,theLHCorderedanimmediatesuspensionofprojectactivitiesnear11heritagesites.Italsoaskedtheauthoritiestoreportontheiradherencetoalldejurerequirementsrelatedtolandacquisition,noiselevels,andsolidwastemanagement(ShaukatandTanveer2016).Thecourtorderthrewtheprovincialgovernmentintoafrenzy,asitanticipatedlongimplementationdelaysthatcouldpreventtheprojectfromreachingcompletionaheadoftheJuly2018election.ItimmediatelyengagedexpertstoconductseparateStructuralandHeritageImpactAssessments(SIAandHIA)andre-issuedtheNOCinJuly2016—beforetheLHCissueditsfullverdictthefollowingmonth.ThematterwasfinallysettledbytheSupremeCourtofPakistan(SCP)whenitrejectedthegovernment’srevisedNOConthesamegrounds,questioningtheintegrityofthegovernment’sactionsthatclearlysoughttoremovethisroadblock.
Afterseveraladditionalhearingsandengagementswithinternationalexpertstoascertainthetruedangersfromvibrationstotheintegrityofhistoricbuildings,theSCPfinallyauthorizedtheproject’sresumptioninDecember2017ontheconditionthatits31-itemstronglistofrequirementswouldbeimplemented(NESPAKv.Mumtaz2017).Withindays,“Shehbazspeed”wasonfulldisplay,asconstructionaroundthesesitesresumedafteradelayofnearlytwoyears.288Notwithstandingtheseefforts,theSharifadministrationwasunabletocompletetheprojectbytheendofitsterm,ultimatelyallowingitschiefpoliticalrivalsfromthePakistanMovementforJusticepartytocuttheredribboninOctober2020.
ThesagaofthisprojectduringtheearlyBRIperioddemonstratesthatevenwhenstrongdejureESGsafeguardsareinplace,thedefactoimplementationofsuchsafeguardscanfalterorfailforawidevarietyofreasons.Insomecases,localofficialsmaybeincentivizedtoprioritizespeedoversafety.Inothercases,theymaylacktechnicalknowhowtoenforcestandardsormaynotfearpenaltiesfornon-compliance.
288LahoreisthecapitalofthePunjabprovinceandthepoliticalpowerbaseofthethen-incumbentPakistanMuslimLeagueparty.Itsleadership,includingthen-PrimeMinisterNawazSharifandhisyoungerbrother,PunjabChiefMinisterShehbazSharif,belongtothecity’sbusinesselite.Sincefirstcomingtopowerinthemid-1980s,theSharifs’politicalstrategyhashingedonflagshipinfrastructureprojects,suchasmajornewinternationalairportsandinter-citymotorways.Duringhisfirsttenureinofficeafterreturningfromexilein2008,theyoungerSharifdeliveredtheLahoreBusRapidTransit(BRT)at“Shehbazspeed”within10months(Majidetal.2018;ExpressTribune2016).Hispartywasrewardedwitharesoundingelectoralvictoryinthe2013elections,whichitattributedtotheBRT.
155
Whatthencanweconcludebasedupontheavailableevidence?First,itis
increasinglycommonforChinesedonorsandlenderstoincludeESGsafeguard
provisionsintheirinfrastructurefinancingagreementswithLICsandMICs.These
provisionsarebroadlycompatiblewithinternationalESGsafeguards,suchasthe
OECDCouncilRecommendationonCommonApproachesforOfficially
SupportedExportCreditsandEnvironmentalandSocialDueDiligence,theIFC’s
PerformanceStandardsonEnvironmentalandSocialSustainability,theUniform
FrameworkforPreventingandCombatingFraudandCorruption,theOECD
CouncilRecommendationonBriberyandOfficiallySupportedExportCredits,
andtheOECDCouncilRecommendationonPublicProcurement.Second,many
ofthesedejureprovisionsgofarbeyondidentifyingrulesandstandardsrelated
toESGriskmanagement;arapidlyexpandingpercentageofChina’soverseas
infrastructureprojectportfolioisunderpinnedbyfinancingagreementsthat
includemechanismsformonitoringcomplianceand/orsanctioning
noncompliancewiththoseESGrulesandstandards.Third,China’sdefactorisk
mitigationeffortsareontherise.Fourth,thereisagrowinggapbetweenhow
ESGsafeguardsareappliedtoChina’soverseasinfrastructureprojectsin
principle(dejure)andinpractice(defacto),whichisnotunexpectedgiventhat
ESGriskmitigationonlyrecentlybecameapriorityforBeijing.
Section4:DoESGsafeguardsproduceaprojectperformancepayofforpenalty?
GiventhatBeijinghasrecentlytakensignificantmeasurestomitigateESGrisks
initsoverseasinfrastructureprojectportfolio,animportantquestioniswhether
thesecoursecorrectionsarecompatiblewiththecountry’sreputationforspeed
andconvenience.ChinadidnotbecometheGlobalSouth’sgo-tobankerfor
big-ticketinfrastructurebecauseofahappyaccident.Itearnedthepositionby
addressingakeysourceofunmetdemandamongLICsandMICs:financingfor
infrastructurewithoutoverlycomplexandcumbersomeESGsafeguardpolicies
156
andprocedures(Dollar2016;Swedlund2017;Parks2019;Humphreyand
Michaelowa2019;Zeitz2021;Horigoshietal.2022).289
In2015,theG-24—agroupofcountriesthatworktogethertocoordinatethe
positionsofdevelopingcountriesoninternationalmonetaryandfinancial
issues—gavevoicetothefrustrationsofLICsandMICsinareportentitled
InfrastructureFinanceintheDevelopingWorld:
"OneaspectofthebusinesspracticesoftheWorldBankandmajor
[regionalmultilateraldevelopmentbanks]thathasaparticularly
strongimpactoninfrastructureinvestmentisenvironmentaland
socialsafeguardpolicies.Safeguardscompriseproceduresand
restrictionsondifferenttypesoflendingoperationsmeantto
‘safeguard’theprojectfromhavingnegativeimpactsonthe
environmentandsocialgroups.Safeguardswerefirstinstitutedatthe
WorldBankinthe1990s,andtheothermajor[regionalmultilateral
developmentbanks]followedsuitinsubsequentyears.TheWorld
Bank’ssafeguardsarestillconsideredthemostcomprehensiveand
rigorous,butthesafeguardsoftheAsDB,IADB,andAfDBhavebeen
graduallytightenedovertheyearssuchthatthedifferencesbetween
themarerelativelysmall,particularlyonthehot-buttonissuesof
environmentalassessmentandresettlement.Asaprojectundergoes
theinitialscreeningprocess,MDBstaffmembersdeterminewhether
ittriggersanyoftheMDB’sapplicablesafeguards.Shouldthatbethe
case,aseparateseriesofspecialrequirementsmustbefollowed
beforetheloancanbeapprovedanddisbursed.Themostfrequently
triggeredsafeguardsinthecaseoftheWorldBankrelateto
environmentalassessmentandinvoluntaryresettlement,andmost
frequentlyaffectinvestmentprojectsinthetransportation,energy,
289AccordingtoDavidDollar,whoservedastheWorldBank’scountrydirectorforChina(2004-2009)andtheU.S.TreasuryDepartment’seconomicandfinancialemissarytoChina(2009-2013),“[the]proceduresdevelopedbytheWorldBankarethegoldstandardofenvironmentalandsocialsafeguardsininfrastructureprojects.However,theyhavehadanumberofunintendedconsequences.Ithasbecometime-consumingandexpensivetodoinfrastructureprojectswiththeWorldBank,andasaresult,developingcountrieshaveturnedtoothersourcesoffunding.[...]Giventhissituation,theemergenceofChinaasamajorfunderof[...]infrastructureprojectshasbeenwelcomedbymostdevelopingcountries.Chinaisseenasmoreflexibleandlessbureaucratic.Itcompletesinfrastructureprojectsrelativelyquicklysothatthebenefitsareseensooner”(Dollar2016).
157
andurbansectors.Therequiredproceduresareextraordinarily
detailedandspecific,andinmanycases[...]extremelydifficultfor
borrowersandevenstafftofullyunderstand.Requirementsoften
includetime-consuming,lengthystudiestobeundertakenby
third-partyexperts(usuallyatthegovernment’scost),lengthy
consultationswithaffectedparties(sometimesincludingunelected
non-governmentalorganizations),extensivemitigationsmeasures,
andlengthymandatorypriorpublicdisclosureandcommentperiods
duringwhichtimetheprojectcannotmoveahead.These
requirementssupersedewhatevernationallawsmaybeinplacein
theborrowingcountry—aparticularlytroublingpointofprinciplefor
manyborrowingcountries,beyondthepracticalimpactsof
safeguards”(Humphrey2015).
China,whichisamemberoftheG-24andtheWorldBank’slargestborrower,
appreciatedtheseconcerns(seeBox3a)andusedthemasawaytodifferentiate
itsofferingtotheglobalinfrastructurefinancingmarket.Underthebannerof
“South-Southcooperation,”itemphasizeditssolidaritywiththeGlobalSouth
andofferedLICsandMICsanalternativemodelofdevelopmentthatprioritized
therapidinstallationof“hardware”over“software”investmentsthatfocuson
policiesandinstitutions.290Beijing’smessageresonated—somuchsothatit
becamethedevelopingworld’sfinancieroffirstresortforhighways,railroads,
dams,bridges,seaports,airports,powerplants,andelectricitygrids,whilethe
MDBsdownsizedtheirinfrastructuredepartmentsandprogramsduetoalackof
borrowerdemand.291Severalyearsago,ChrisHumphreyofETHZurich’sCenter
forDevelopmentCooperationandKatharinaMichaelowaofUniversityofZurich
publishedinterviewevidencefromthreeAfricancountriesonthechanging
natureofborrowerdemandforinfrastructurefinancing.Theyfoundthat:
“[o]neissuewhichofficialsinallthreecountriesnotedaslimitingtheir
owndemandforinfrastructurelendingfromtheWorldBankandtoa
291In2010,theWorldBank’sIndependentEvaluationGroupfoundthat“[t]wothirdsof[WorldBank]managersinterviewedreportedthatsomeclientshadavoidedorweredroppinga[World]Bankprojectbecauseofsafeguardpolicies”(IEG2010:73).290AttheBeltandRoadForumforInternationalCooperationin2017,XiJinpingdescribedtheBRIas“anewoptionforothercountriesandnationswhowanttospeeduptheirdevelopmentwhilepreservingtheirindependence”(BeltandRoadForumforInternationalCooperation2017,emphasisadded).
158
lesserdegreetheAfDBisthe‘hasslefactor’implicitinthesetypesof
projectsfromprojectdesignrulesandenvironmentalandsocial
safeguards.OfficialsfromallgovernmentsconcurredthattheWorld
Bankisparticularlydifficult.‘Forhydroelectricandrailroads,wedon’t
eventalktothem,wejustgostraighttotheChinese,’saidan
Ethiopianofficial.Discussingamajorgaspipelineproject,aTanzanian
officialsaid,‘TheChineseareabitmoreexpensive,buttheyarealot
easierandalotfasterforthiskindofproject.Wedidn’tevensenda
requesttotheWorldBankforsupport,wewentstraighttothe
Chinese.’EveninMalawi,withonlysmallamountsofChinesefinance,
officialswereexperiencingthesedynamicswithaplannednew
coal-firedpowerplant,tobefundedbytheChineseatmarket-based
interestrates.‘TheWorldBankandAfDBwouldn’tfunditbecausethe
powerfulshareholderswouldnotagreetothatkindofthingfor
environmentalreasons.SowewentwiththeChinese.’”(Humphrey
andMichaelowa2019:23)
InlightofLICandMICdemandforlowlevelsof“hasslefactor,”Beijing’spivot
towardamorestringentESGsafeguardregimeraisesthequestionofwhetherit
willunderminethevaluepropositionthatithastraditionallyusedtodifferentiate
itselffromcompetitorsintheglobalinfrastructurefinancingmarket.Tobetter
understandtheimplicationsofChineselendersanddonorsadoptingstronger
ESGsafeguards,weusethe3.0versionofAidData’sGCDFdatasettocompare
theperformanceofChinesegovernment-financedinfrastructureprojectswith
andwithoutstrongESGsafeguards.Wedosowithsixoutcomemeasures:(1)
thepercentageofprojectsthatrunbehindschedule(2)theaveragelengthof
commencementdelays,(3)theaveragelengthofcompletiondelays,(4)the
averageamountoftimeittakestoreachcompletion,(5)thefrequencyand
valueofprojectsuspensionsandcancellations,and(6)ESGriskprevalencerates.
WebeginbycomparingthepercentageofChina’sinfrastructureproject
portfoliowithandwithoutstrongdejureESGsafeguardsthatranbehind
schedule.Weclassifyaprojectas“behindschedule”ifitsactualimplementation
startdatetookplace3monthsormoreafteritsoriginallyscheduled
implementationstartdate,and/orifitsactualcompletiondatetookplace3
159
months(ormore)afteritsoriginallyscheduledcompletiondate.FigureA58
demonstratesthatChina’sgrant-andloan-financedinfrastructureprojectsinLICs
andMICsareequallyaslikelytorunbehindschedulewhenstrongdejureESG
safeguardsareinplaceastheyarewhenstrongESGsafeguardsarenotinplace:
74%ofinfrastructureprojectswithstrongdejureESGsafeguardsranbehind
schedule,and75%withoutstrongdejureESGsafeguardsranbehindschedule.
Next,weexaminetheaveragelengthofcommencementdelaysforChina’s
overseasinfrastructureprojectsthataresubjecttoESGsafeguardswithvarying
levelsofstringency.FigureA59providesevidencethatChina’soverseas
infrastructureprojectsencounterslightlyshortercommencementdelayswhen
theyareundertakenwithstrongESGsafeguards(47daysshorter,onaverage,
thanChina’soverseasinfrastructureprojectswithoutstrongsafeguards).Thefact
thatinfrastructureprojectswithstrongESGsafeguardstakeslightlylesstimeto
launchthanthosewithoutsuchprotectionsisconsistentwiththeoldadagethat
“anounceofpreventionisworthapoundofcure.”ItalsoimpliesthatBeijing
maybeabletoimplementstrongESGsafeguardswithoutlosingitscompetitive
edge.Asageneralrule,MDB-financedinfrastructureprojectswithstrongESG
safeguardsfacesubstantiallylongercommencementdelays.CharlesKennyof
theCenterforGlobalDevelopmentestimatesthat‘‘CategoryA’’WorldBank
projects—environmentallyandsociallysensitiveprojectssubjectedtothe
organization’smoststringentsafeguards—take7.4years(2,689days),on
average,tomovefromtheproposalstagetothedisbursement(project
commencement)stage.292
Theprincipleof“anounceofpreventionisworthapoundofcure”also
evidentlyappliestocompletiondelaysinChinesegrant-andloan-financed
infrastructureprojects.FigureA60providesevidencethatChina’soverseas
infrastructureprojectsfaceslightlyshortercompletiondelayswhenstrongESG
safeguardsareinplace(91daysless,onaverage,thanChina’soverseas
infrastructureprojectswithoutstrongESGsafeguards).293Aseparate,butclosely
293Thesefindingsarelikelyrelatedtothefindingsoncommencementdelays.ESGsafeguardsoftenrequirethatcontractorsandtheirhostcountrycounterpartstakeaseriesoftime-consumingactions—suchasconductingenvironmentalimpactassessments(EIAs)andpreparingresettlementactionplans292ThefindingsreportedinKenny(2023)arespecifictothe2010-2017timeperiod.In2010,astudybytheWorldBank’sIndependentEvaluationGroup(IEG)alsorevealedthattheaveragecostofsafeguardsforaCategoryAprojectattheWorldBankis$19million(IEG2010).
160
related,insightfromthe3.0versionofAidData’sGCDFdatasetisthatittakesan
averageof3.2years(1,163days)tocompleteaninfrastructureprojectwithout
strongESGsafeguards,andittakes8fewerdays(1,155days)tocompletean
infrastructureprojectwithstrongESGsafeguards(seeFigureA61).294These
findingsdonotsuggestthatChina’sreputationforspeedisinjeopardy.
Whatthencanweconcludeaboutthe“speedofimplementation”differences
betweeninfrastructureprojectswithandwithoutstrongESGsafeguards?The
mostimportantpointisthattheyarenotlarge,whichmeansthatthereisnot
muchevidencetosupporttheideathatESGsafeguardsimposeasignificant
projectperformancepenalty.TheconventionalwisdomisthatESGrisk
mitigationmeasuressubstantiallyimpedeinfrastructureprojectimplementation,
therebyunderminingakeycomponentofthevalueproposition(speed)that
Chinahastraditionallyusedtodifferentiateitselffromitscompetitorsinthe
globalinfrastructurefinancingmarket(Swedlund2017;Parks2019;Humphrey
andMichaelowa2019;Zeitz2021).However,ourfindingsdonotsupportthis
argument.Quitetheopposite:theysuggestthatChinacanreducetheESGrisk
profileofitsoverseasinfrastructureportfolioifitiswillingtoacceptslightly
longerprojectimplementationtimelines(measuredindozensofdaysratherthan
hundredsorthousandsofdays).295
AnotherpotentialwaythattheperformanceofBeijing’soverseasinfrastructure
projectportfoliomightvarybasedonESGsafeguardstringencyisthelikelihood
ofprojectsuspensionorcancellation.ThelatestversionofAidData’sGCDF
datasetdemonstratesthatinfrastructureprojectswithstrongdejureESG
safeguardsaresubstantiallylessvulnerabletosuspensionandcancellationafter
thefinalizationofaChinesegrantorloanagreement.AccordingtoFigure3.10,
while74Chinesegrant-andloan-financedinfrastructureprojects(worth$43
billion)withweakdejureESGsafeguardshavebeensuspendedorcanceled
295ThisfindingisrelevanttotheloanrepaymentchallengesthatwedocumentinChapter2becausethespeedofimplementationcanaffectaproject’srevenuegenerationpotentialandthusaborrower’sabilitytomeetitsloanrepaymentobligations.TheCDB-financedJakarta-BandungHigh-SpeedRailwayConstructionProject,whichisrunningover-budgetandbehindschedule,isacaseinpoint.Sinceitwasfinancedthroughalimitedrecourseprojectfinancetransactionandtherailwayisnotyetinoperation,theborrowerisunabletomakedebtservicepaymentsviarailwayrevenues(MalikandParks2021;Kuo2021).294Bywayofcomparison,ittakesWorldBankandAsianDevelopmentBankprojects,onaverage,6yearstomovefromthecommencementstagetothecompletionstage(seeBulmanetal.2017:362).(RAPs)—duringthepre-implementationphaseofaproject,whichcaneliminateimplementationobstaclesthatwouldotherwisedelaycompletion.
161
since2000,only7Chinesegovernmentgrant-andloan-financedinfrastructure
projects(worth$11billion)withstrongdejureESGsafeguardshavebeen
suspendedorcanceledsince2000.296Thesefindingssuggestthatthe
applicationofmorestringentESGsafeguardsmayhelpratherthanhinder
Beijing’seffortstode-riskitsoverseasinfrastructureprojectportfoliointhe
developingworld.
Figure
3.10:
Monetary
value
of
project
suspensions
and
cancellations
by
de
jure
ESG
safeguard
strengthFigure3.10
Notes:StrongandweakdejureESGsafeguardsaredefinedinSectionA-9oftheAppendix.
Finally,wecanusethe3.0versionofAidData’sGCDFdatasettodetermine
whetherESGriskprevalenceratesinChina’soverseasinfrastructureproject
portfoliovaryaccordingtoESGsafeguardstringency.Figure3.11comparesthe
percentageofChina’sgrant-andloan-financedinfrastructureprojectportfolio
withsignificantESGriskexposureacrosstwocohorts:projectswithandwithout
strongdejureESGsafeguards.Whereas82%ofprojectsthatlackedstrongde
jureESGsafeguardsfacedsignificantESGrisks,only18%ofprojectswithsuch
safeguardsencounteredsimilarrisks.297FiguresA53,A54,andA55inthe
Appendixdemonstratethatthesepatternsareequallyapplicabletoallthree
types(environmental,social,andgovernance)ofESGsafeguards.
297FigureA49tracksthesametwocohortsovertime.NotwithstandingasharpincreaseinthepercentageoftheinfrastructureprojectportfoliosubjectedtostrongdejureESGsafeguardsduringthelateBRIperiod,itshowsthatthesameempiricalpatternisgenerallyconsistentacrossthe2000-2021period.296OurfindingsareconsistentwiththoseofLuetal.(2023b).TheyfindthatChinese-financedpowerplantprojectsposinghigherlevelsofenvironmentalriskaremorelikelytobesuspended.
162
Figure
3.11:
Proportion
of
infrastructure
project
portfolio
facing
significant
ESG
risk
exposure
by
level
of
safeguard
stringencyFigure3.10
Notes:StrongandweakdejureESGsafeguardsaredefinedinSectionA-9oftheAppendix.Thepresence
ofsignificantESGriskexposureisbasedonaproject-levelcompositemeasurethatisdescribedinSection
2ofChapter3.
Section5:DecodingBeijing’sESGriskmitigationstrategy
Intheremainderofthischapter,wewillanalyzethe3.0versionofAidData’s
GCDFdatasettobetterunderstandhowBeijingisseekingtomanageand
163
mitigateESGrisksinitsinfrastructureprojectportfolio.Weseeevidenceof
Chinesestate-ownedfinancierstakingfourESGriskmitigationeffortsto:
1.Defundthebilateraldevelopmentfinanceinstitutionswiththeweakest
safeguards
2.Supportbilaterallendersanddonorswiththestrongestsafeguards
3.Outsourceriskmanagementviasyndicationandmultilateralization
4.Unwindrelationshipswithhigh-riskcountriesanddoubledownon
relationshipswithlow-riskcountries
Riskmitigationstrategy#1:Defundthebilateraldevelopment
financeinstitutionswiththeweakestsafeguards
MostofBeijing’sofficialstatementsandpublicationsaboutde-riskingits
overseasinfrastructureprojectportfolioareanodyneanddifficulttointerpret.
The“GreenDevelopmentGuidelinesforForeignInvestmentandCooperation”
thatChina’sMinistryofCommerceandMinistryofEcologyandEnvironment
issuedinJuly2021statethat“[t]hegreeningofoutboundinvestmentand
cooperationmustbeguidedbyXiJinpingThoughtonSocialismwithChinese
CharacteristicsforaNewEra.Intheprocess,wemustimplementthespiritof
the19thCPCNationalCongressandthe2nd,3rd,4thand5thPlenarySessions
ofthe19thCPCCentralCommittee,XiJinpingThoughtonEcological
CivilizationandthedecisionsoftheCPCCentralCommitteeandtheState
Council.Wemuststaycommittedtothenewdevelopmentconcept,strivingfor
thestrongawarenessofgreendevelopment,efficientuseofresources,strict
protectionoftheenvironmentandeffectivecontrolofcarbonemissions.We
shouldworktoshowcaseChina’sleadershipinglobalendeavortowardgreen
transitionandourcommitmenttobuildingtheworldintoabetterandcleaner
placeandlayingthegroundworkforanewdevelopmentparadigm”(MOFCOM
andMEE2021).
However,whenyoucutthroughthefloweryrhetoricusedbyChinesepoliticians
andbureaucratsbyfollowingthemoney,astarkrealityemerges:Beijingis
164
entrustingashrinkingproportionofitsoverseasinfrastructureprojectportfolio
tothecountry’spolicybanks(CDBandChinaEximbank),whichhaveparticularly
weakdejureESGsafeguards(seeTable3.1andTable3.2).Figure3.12plotstwo
trendsovertime:thepercentageofChinesegrant-andloan-financed
infrastructureprojectswithweakdejureESGsafeguardsandthepercentageof
China’sinfrastructureprojectportfoliofinancedviabilateralloansfromthe
country’spolicybanksfrom2000to2021.Beijing’srelianceuponpolicybank
financingforinfrastructureprojectsinLICsandMICsplummetedfrom86%in
2013to41%in2021.298Theyear-on-yearchangesthattookplaceduringthis
periodalsotrackverycloselywithyear-on-yearchangesinthepercentageof
China’soverseasinfrastructureprojectportfoliobankrolledbyinstitutionswith
weakdejureESGsafeguards(seeFigure3.12).
298This45percentagepointdeclineobscuressomedifferencesacrossthetwopolicybanks.WhereasthepercentageofChina’sinfrastructureprojectportfolioinLICsandMICsfinancedviaCDBdeclinedfrom39.79%inthepre-BRIperiod(2000-2013)to11.61%duringthelateBRIperiod(2018-2021),thepercentagefinancedviaChinaEximbankactuallyincreasedfrom38.06%to48.75%acrossthesetwoperiods(seeTable3.2).However,uponcloserinspection,onecanseethatthepercentagesofChina’sinfrastructureprojectportfolioinLICsandMICsfinancedviaCDBandChinaEximbankdeclined(forthemostpart)overthecourseofthelateBRIperiod.FiguresA56andA57demonstratethatthepercentageoftheportfoliofinancedviaChinaEximbankfellfrom58.7%in2018to18.3%in2021andthepercentageoftheportfoliofinancedviaCDBfellfrom11.9%in2018to4.3%in2020beforetickingbackupto22.5%in2021.
165
Figure
3.12:
Composition
of
infrastructure
project
portfolio:
reliance
upon
the
policy
banks
and
weak
de
jure
ESG
safeguardsFigure3.12
Notes:WeakdejureESGsafeguardsaredefinedinSectionA-9oftheAppendix.
Table
3.2:
China’s
grant-
and
loan-financed
infrastructure
portfolio
by
type
of
financing
instrument
over
timeTable3.2
China’s
grant-
and
loan-financed
infrastructure
portfolio
by
type
of
financing
instrument
over
time
ContractCategoryPre-BRI(2000-2013)EarlyBRI(2014-2017)LateBRI(2018-2021)
BilateralChinaEximbankloan38.06%43.84%48.75%
BilateralCDBloan39.79%23.93%11.61%
BilateralMOFCOMloanorgrant2.64%2.68%3.10%
BilateralChinesestate-ownedcommercialbankloan5.14%9.38%8.84%
SyndicatedloanwithChineseandmultilateralbankparticipants0.80%2.17%1.45%
SyndicatedloanwithChinesestate-ownedcommercialbanksand/orpolicybanks8.86%16.91%20.39%
PBOC/MOFgrantorloanchanneledthroughmultilateralinstitution0%0.12%0.89%
166
ContractCategoryPre-BRI(2000-2013)EarlyBRI(2014-2017)LateBRI(2018-2021)
Supplier'screditfromChineseSOE4.71%0.77%4.97%
Total100%100%100%
Notes:ThistablepresentsthesharesofChina’sgrant-andloan-financedinfrastructureprojectportfolio
(measuredinconstant2021USD)inLICsandMICsdeliveredvia8financialinstrumentcategoriesacross
threedifferenttimeperiods:(1)thepre-BRIperiodfrom2000to2013,(2)theearlyBRIperiodfrom2014to
2017,and(3)thelateBRIperiodfrom2018to2021.
Riskmitigationstrategy#2:Supportbilaterallendersanddonors
withthestrongestsafeguards
AcrossthevariousbilateralinstrumentsthatBeijinghasatitsdisposaltobankroll
infrastructureprojectsinthedevelopingworld,CDBandChinaEximbankloans
offertheweakestdejureESGsafeguards.Ratherthanrelyingonthesepolicy
banks,Beijingisincreasinglyturningtoadifferentsetoffinancial
institutions—withlendingandgrant-givinginstrumentsthatincludeamore
stringentsetofdejureESGsafeguards—tofinanceinfrastructureprojectsin
LICsandMICs.
InSection3ofChapter3,wediscoveredthatbilateralloansfromthecountry’s
state-ownedcommercialbanks,supplier’screditsfromthecountry’sstate-owned
enterprises,andMOFCOMgrantsandinterest-freeloanshavestrongerdejure
ESGsafeguardsthanCDBandChinaEximbankloans(seeTable2.1inChapter
2).Thisfindingbegsthequestion:hasBeijingincreaseditsuseofthesebilateral
infrastructurefinancinginstruments?Table3.2abovedemonstratesthatithasin
factdoneso,albeitinanincrementalway:whereastheproportionofChina’s
infrastructureprojectportfolioinLICsandMICsfinancedthroughthese
instrumentsamountedto12.49%duringthepre-BRIperiod(2000-2013)and
12.83%duringtheearlyBRIperiod(2014-2017),itjumpedupto16.91%during
thelateBRIperiod.
Riskmitigationstrategy#3:Outsourceriskmanagementvia
syndicationandmultilateralization
AnotherstrategythatBeijingcouldpursuetode-riskitsoverseasinfrastructure
projectportfolioisoutsourcingriskmanagementtomultilateralinstitutions,
167
whicharewidelyconsideredtobenorm-setting,industryleadersinthedesign
andimplementationofESGsafeguards(Park2010;Martinetal.2015;Buntaine
2016;Braunederetal.2018;Buchananetal.2018;Narainetal.2020,2022).
Onepathtomultilateralizationistodirectlyentrustthemanagementofgrants
andloanstoanintergovernmentalorganization.Forexample,throughits$2
billiontrustfundattheAfricanDevelopmentBank(knownastheAfricaGrowing
TogetherFund)andits$2billiontrustfundattheInter-AmericanBank(knownas
theChinaCo-FinancingFundforLatinAmericaandtheCaribbean),thePBOC
hasfullydelegatedprojectdesign,preparation,implementation,and
supervisionresponsibilities(includingESGsafeguardapplication)tothe
multilateralinstitutions.
Anothermultilateralizationoptionistoparticipateinsyndicatedloan
agreementsthatarearrangedbymultilateralinstitutions.Manyofthese
agreementshaveanA/Bstructure,wherebyanMDBservesasthe
lender-of-recordandkeepsapartoftheloanforitsownaccount(the“A-loan”)
whilesellingparticipationintheremainderoftheloan(the“B-loan”).Acommon
featureofA/Bsyndicatedloanagreementswithmultilateralarrangersisthatall
B-loanparticipantsyieldauthoritytotheA-loanproviderforriskmitigation
purposes(EstyandMegginson2003;BaeandGoyal2009;HainzandKleimer
2012;Broccolinietal.2021;Luetal.2023a).
AccordingtoTable3.2above,Beijinghasmodestlyincreaseditsrelianceupon
multilateralinstitutionsovertime.TheproportionofChina’sinfrastructureproject
portfolioinLICsandMICsfinancedviasyndicatedloanswithmultilateral
participantsandPBOC/MOFloansandgrantsentrustedtomultilateral
institutionsrosefrom0.8%duringthepre-BRIperiod(2000-2013)to2.34%
duringthelateBRIperiod(2018-2021).Despitethesmallsizeofthisincrease,it
isnoteworthybecausePBOCandMOFgrantsandloansentrustedto
multilateralinstitutionsandsyndicatedloanswithChinesebankandmultilateral
institutionparticipantshavethemoststringentESGsafeguardsinoursampleof
infrastructurefinancingagreements(seeTable3.1inSection3).
168
SyndicatedloanswithChinesepolicybankandstate-ownedcommercialbank
participantsmaybeanalogoustosyndicatedloanswithmultilateralparticipants
ifallmembersofthesyndicategenerallydefertothelendinginstitution(s)with
thestrongestpreference(s)forESGriskmitigation.Consistentwiththis
expectation,Table3.1aboveprovidesevidencethattheESGsafeguardsof
state-ownedcommercialbanksdoindeedprevailoverthoseofthepolicybanks
insyndicatedloanarrangements.299
Beijinghasalsointensifieditsuseofthesebilateralinfrastructurefinancing
instrumentsovertime.Figure3.13presentsthepercentageofChinesegrant-
andloan-financedinfrastructureprojectswithstrongdejureESGsafeguardsin
conjunctionwiththepercentageofChina’sinfrastructureprojectportfolio
financedviasyndicatedloanswithChinesepolicybankandstate-owned
commercialbankparticipantsfrom2000to2021.Beijing’suseofthesetypesof
syndicatedloanarrangementsforinfrastructureprojectsinLICsandMICshas
increaseddramatically—from0%in2000to41%in2021—andintandemwith
theusageofstrongdejureESGsafeguards.300Theyear-on-yearchangesthat
tookplaceoverthistwenty-twoyearperiodtrackcloselywithyear-on-year
changesinthepercentageofChina’soverseasinfrastructureprojectportfolio
bankrolledbyinstitutionswithstrongdejureESGsafeguards.
300AccordingtoTable3.2above,thepercentageofChina’sinfrastructureprojectportfolioinLICsandMICsfinancedviasyndicatedloansinvolvingstate-ownedpolicybanksandcommercialbanksincreasedfrom8.86%duringthepre-BRIperiod(2000-2013)to16.91%duringtheearlyBRIperiod(2014-2017)and20.39%duringthelateBRIperiod(2018-2021).299Sufi(2007)demonstratesthatleadarrangersreducethecostsofduediligenceforallothersyndicateparticipants.
169
Figure
3.13:
Composition
of
infrastructure
project
portfolio:
Use
of
syndicated
loans
and
strong
de
jure
ESG
safeguardsFigure3.13
Notes:StrongdejureESGsafeguardsaredefinedinSectionA-9oftheAppendix.
Riskmitigationstrategy#4:Unwindrelationshipswithhigh-risk
countriesanddoubledownonrelationshipswithlow-riskcountries
OneadditionalwaytoreducetheESGriskprofileofaninfrastructureproject
portfolioistouseinformationaboutthepastperformanceofhostcountriesto
guidefuturelendingandgrant-givingactivities.Thatistosay,onceadonoror
lenderhasidentifiedthesubsetofinfrastructureprojectsinitsportfoliothat
havepresentedsignificantESGrisks,itmayseektoidentifywheretheseprojects
aregeographicallyconcentratedsothatitcanidentifyhostcountriesposing
especiallyhighlevelsofESGriskandredirectfundingforfutureinfrastructure
projectselsewhere.Thisreallocation—or“selectivity”—practiceisfollowedby
severalmajorMDBs,includingtheAsianDevelopmentBankandtheWorldBank
(Buntaine2011,2015,2016).
170
Figure
3.14:
Proportion
of
infrastructure
portfolio
allocated
to
LICs/MICs
with
high
ESG
risk
prevalence
ratesFigure3.14
Notes:ThisfigurecomparestheoverallpercentageofChina’sinfrastructurefinancingtoLICsandMICsthat
wasallocatedtocountrieswithhighESGriskprevalenceratesintwotimeperiods:(1)thepre-BRIandearly
BRIperiod(2000-2017)and(2)thelateBRIperiod(2018-2021).CountrieswithahighESGriskprevalence
ratearedefinedasthosewhereatleast75%ofChina’sgrant-andloan-financedinfrastructureproject
portfoliobetween2000and2017facedsignificantESGrisks.ThepresenceofsignificantESGriskexposure
isbasedonaproject-levelcompositemeasurethatisdescribedinSection2ofChapter3.
InordertogaugewhetherBeijingmovedinthisdirectionduringthelateBRI
era,wefirstcreatetwocohortsofhostcountries:countrieswhereatleast75%of
China’sgrant-andloan-financedinfrastructureprojectportfoliohadsignificant
ESGriskexposurebetween2000and2017andcountrieswherelessthan75%
ofChina’sgrant-andloan-financedinfrastructureprojectshadsignificantESG
riskexposurebetween2000and2017.WethencompareBeijing’sprovisionof
infrastructurefinancingtothesetwocohortsbetween2018and2021.Figure
3.14demonstratesthat2.8%ofinfrastructurefinancingfromBeijingduringthe
lateBRIerawasdirectedto9LICsandMICswhereatleast75%ofChina’sgrant-
andloan-financedinfrastructureprojectportfoliohadsignificantESGrisk
exposurebetween2000and2017.Bywayofcomparison,Beijingallocateda
substantiallylargerproportion(6.83%)ofitsgrant-andloan-financed
infrastructureprojectportfoliotothesame9countriesbetween2000and2017.
ThispatternisconsistentwiththeideathatBeijinghasrebalancedthe
cross-countryallocationofaidandcredittoreducetheESGriskprofileofits
overseasinfrastructureprojectportfolio.
GiventhatChinahasscaledbackinfrastructurespendingincountrieswhereits
projectshavefacedparticularlyhighlevelsofESGriskexposure,anotherway
thatitcouldseektorecalibrateitsportfolioisbyrampingupsupportfor
infrastructureprojectsincountrieswhereitsprojectshavefacedparticularlylow
171
levelsofESGriskexposure.Figure3.15providesevidencethatBeijinghasin
factmovedinthisdirection.Morespecifically,itdemonstratesthatduringthe
lateBRIperiod(2018-2021)7.6%ofinfrastructurefinancingfromChinawas
directedto6LICsandMICswherelessthan10%ofChinesegrant-and
loan-financedinfrastructureprojectshadsignificantESGriskexposurebetween
2000and2017.ThisrepresentedasignificantincreaseinlateBRIeraspending
forlow-riskcountries,asBeijingallocatedonly1.73%ofitsgrant-and
loan-financedinfrastructureprojectportfoliotothesame6countriesbetween
2000and2017.
Figure
3.15:
Proportion
of
infrastructure
portfolio
allocated
to
LICs/MICs
with
low
ESG
risk
prevalence
ratesFigure3.15
Notes:ThisfigurecomparestheoverallpercentageofChina’sinfrastructurefinancingtoLICsandMICsthat
wasallocatedtocountrieswithlowESGriskprevalenceratesintwotimeperiods:(1)thepre-BRIandearly
BRIperiod(2000-2017)and(2)thelateBRIperiod(2018-2021).CountrieswithalowESGriskprevalence
ratearedefinedasthosewherelessthan10%ofChina’sgrant-andloan-financedinfrastructureproject
portfoliobetween2000and2017facedsignificantESGrisks.ThepresenceofsignificantESGriskexposure
isbasedonaproject-levelcompositemeasurethatisdescribedinSection2ofChapter3.
Section6:IsBeijingcourse-correctingorvirtue-signaling?
WhenBeijingfirstsignaledinterestinsubjectingitsoverseasinfrastructure
projectportfoliotomorestringentESGsafeguards,criticsandrivalswerequick
toquestionitssincerity.JonathanE.HillmanoftheCenterforStrategic&
InternationalStudies(CSIS)pennedanop-ed,inwhichhearguedthatChinawas
engagedina“greenwashingeffort.”Hewrotethat“XiknowsthatChinais
vulnerableonenvironmentalissues.AtagatheringofworldleadersinBeijing
lastyear,‘green’and‘sustainable’werethesaltandpepperinhisspeechabout
thenextphaseofBRI.Toaddresscriticism,Chinaunveilednolessthan11new,
greeninitiativesunderBRI,allofwhicharevoluntary,andnoneofwhichare
172
bindingortransparent”(Hillman2020b).KellySimsGallagherandQiQiofTufts
Universitywrotethat“ChinesegovernmentrhetoricaboutgreeningtheBRIis
laudable,butithasyettomakeanysubstantivechangestowardthatgoal”(Sims
GallagherandQi2021).TheU.S.anditsalliesalsorejectedthenotionthat
Beijing’soverseasinfrastructureprojectshadrobustESGprotectionsinplace.In
November2019,theU.S.,Japan,andAustraliaannouncedthattheywere
joiningforcestoestablisha“BlueDotNetwork”thatwould“evaluateand
certifynominatedinfrastructureprojectsbaseduponadherencetocommonly
acceptedprinciplesandstandards”and“promotemarket-driven,transparent,
andfinanciallysustainableinfrastructuredevelopmentintheIndo-Pacificregion
andaroundtheworld.”Morerecently,theU.S.,theU.K.,andtheother
membersoftheG7havepromotedaPartnershipforGlobalInfrastructureand
Investment(PGII)—previouslyknownastheBuildBackBetterWorld(B3W)
initiative—thattheycharacterizeasanalternativetotheBRIandanoptionfor
countriesthatwanttoundertakeinfrastructureprojectsinstrictaccordancewith
internationallyacceptedESGsafeguards.301
However,thischapterdemonstrates—withmanynewsourcesandtypesof
evidence—thatBeijingisnotsimplyengaginginaninternationalvirtue-signaling
exercise.Ithastakenmeaningfulstepstode-riskitsoverseasinfrastructure
projectportfoliobyrampingdowntheinternationallendingactivitiesofbanks
thatlackstrongESGriskmanagementguardrails,rampinguptheprovisionof
infrastructurefinancingviainstitutionsthathavestrongESGsafeguardsinplace,
graduallyunwindingaidandcreditrelationshipswithLICsandMICsthatpresent
highlevelofESGrisk,andredirectingnewinfrastructurefinancingtolower-risk
countries.Atthesametime,itisstillatarelativelyearlystageinitsjourneyfrom
ESGskeptictoadvocateanditwilllikelytakemanyyears—potentiallyeven
decades—toclosethegapbetweenthedejureanddefactoapplicationofESG
safeguards.302
302Itisalsoimportanttokeepinmindthat,asof2021,40%ofChineseinfrastructurefinancingtoLICsandMICswasstillbeingchanneledviabilateralCDBandChinaEximbankgrantsandloans.Thisissignificant,sincethischapterprovidesevidencethatCDBandChinaEximbankhaveforthemostpartnotreformed/modernizedtheirdejureESGsafeguardsinawaythatiscomparabletothepracticesofmultilateralinstitutions,state-ownedcommercialbanks,ortheleadarrangersofsyndicatedloans.301InSeptember2023,theU.S.,France,Germany,Italy,theEU,India,theUAE,andSaudiArabiaalsoannouncedplanstodevelopanIndia-MiddleEast-EuropeEconomicCorridor(IMEC).
173
WedonotseeevidenceofChinesedevelopmentfinanceinstitutionsuniformly
complyingwithinternationalESGsafeguards.Rather,weseeevidencethat
someChineselendersanddonorsaregraduallyandselectivelyharmonizing
theirESGsafeguardpoliciesandpracticeswiththoseoftraditionaldonorsand
lenders.303ThesechangesshouldgivepausetoBeijing’scompetitorsinthe
globalinfrastructurefinancingmarket.TheG7andsomeMDBsarecurrently
tryingtoconvincewould-bepartnersinthedevelopingworldthat(a)theBRIisa
low-qualityinfrastructureoption(privilegingspeedandconvenienceoversafety
andlong-termsustainability)and(b)theycanprovidealternative,high-quality
financingoptionsforcountriesthatwanttoundertakeinfrastructureprojects
basedonstrictadherenceto“internationalbestpractice”ESGsafeguards.
However,thisblack-and-whitebrandingstrategymaylackresonancewithits
targetaudience,asLICsandMICshavealreadymadeitveryclearthattheyhave
lowlevelsofappetitefor“goldstandard”ESGsafeguards.Theywantfinancing
partnersthatcanquicklydesignandimplementbig-ticket,high-impact
infrastructureprojectswithoutunreasonablyhighlevelsofESGrisk.304The
evidenceinthischaptersuggeststhatBeijingmaybebetter-positionedto
answerthiscallthanitscompetitorsrealize.Itisnowdeliveringlarge-scale
infrastructureprojectswithincreasinglyrobustESGsafeguardsbutwithoutthe
lengthyimplementationdelaysthatoftenhobblesimilarprojectsbackedbyG7
membersandMDB
304Onthispoint,seeHumphrey2015;Dollar2016;Swedlund2017;HumphreyandMichaelowa2019;andZeitz2021.303Therearereasonstobelievethatthisapproachofgradualandselectiveharmonizationwillbealonger-termprocess.OnJuly16,2021,China’sMinistryofCommerceandtheMinistryofEcologyandEnvironmentissued“GreenDevelopmentGuidelinesforOverseasInvestmentandCooperation,”whichrecommendthatprojectsponsorsandcontractorscomplywithinternationalstandardsorChinesestandardswhenthelawsandregulationsofhostcountriesarevagueorweak.
174
Chapter4:ReputationalRehabilitationonthe
BeltandRoad
Section1:China’squestforsoftpowerduringtheBRIera
Greatpowersuseaidandcredittoexpandtheirglobalinfluence.305Thesetools
ofeconomicstatecraftcanbeparticularlyusefulforwinningsupportfrom
developingcountries.DuringtheColdWar,theEasternandWesternblocsused
grant-givingandlendinginstrumentstopromotetheircompetingideologies
andstrengthenalliancesaroundtheworld(Heurlin2020).China’sBRIstandsout
asanotherstrikingexampleofthisapproachduringthe21stcentury.Beijinghas
attemptedtobolsteritsglobalinfluencebyofferingdemand-responsive
infrastructurefinancingandrapidprojectdeliverytodevelopingcountries
(Dreheretal.2022).PriortotheunveilingoftheBRI,Beijingtriedtokeepa
relativelylowprofileasaninternationaldonorandlender,adheringtoDeng
Xiaoping’s“hideyourcapabilitiesandbideyourtime”principle.However,the
BRImarkedanimportantshift.Chinabecamemoreproactiveaboutcultivating
itsbrandandbroadcastingpositivemessagesaboutitsoverseasactivities.Itdid
sobyplacingsignageatprojectsites,organizinghigh-profileceremoniesto
celebratethecommencementandcompletionofprojects,cultivating
relationshipswithjournaliststoencouragepositivemediacoverageofproject
accomplishments,andforgingcontent-sharingpartnershipswithradiostations,
televisionchannels,andnewspapers(Custeretal.2018,2019;Wellneretal.
2023).
WashingtonseestheBRIasBeijing’sattemptto“reshapetheinternationalorder
[and][...]therulesoftheroad”by“us[ing]itseconomicpowertocoerce
countries”(WhiteHouse2022).The2022U.S.NationalSecurityStrategy
emphasizestheimportanceof“out-competingChina”andcallsupon
governmentagenciesto“partnerwith,support,andmeettheeconomicand
developmentneedsofpartnercountries”(WhiteHouse2022).However,given
thedifficultyofcompetingdollar-for-dollarwithBeijing,Washingtonis
305Onthispoint,seeGoldsmithetal.(2014),Blairetal.(2022a),Wellneretal.(forthcoming),andAsmusetal.(forthcoming).
175
increasinglyawareofthefactthatitwillneedtocarefullyselectcountries,
sectors,anddeliveryinstrumentsthatofferthegreatestreturnoninvestment.
U.S.policymakershavealsodemonstratedinterestinunderstandingthecauses
andconsequencesofBRIbacklash—andBeijing’sresponsetoit.Onits10th
anniversary,theBRIfacesanarrayofreputationalchallengesinhostcountries
whereenthusiasmfortheinitiativehaswaned.Debtrepaymentchallenges
(discussedinChapter2)andESGproblemsrelatedtoinfrastructureproject
implementation(discussedinChapter3)haveledagrowingnumberofLICsand
MICstoreevaluatetherisksandrewardsofcontinuedparticipationinBeijing’s
“projectofthecentury”(Horigoshietal.2022).
Inthischapter,weseektoexplainwhen,where,andhowChinesegrant-and
loan-financedprojectshavebecomereputationalassetsorliabilitiesforChina.
WealsoexaminehowBeijinghasrecalibrateditsstrategiesandtacticsin
countrieswhereithasencounteredpublicantipathy,unfavorablemedia
sentiment,orinsufficientsupportamonggoverningelites.Withdataonpublic
opinion,mediasentiment,elitesupport,andChinesegrantandloan
commitmentsduringthefirsteightyearsoftheBRIera(2014-2021),weusean
“action-reaction”frameworktoexaminehowsoftpowergainsandlossesinLICs
andMICsduringtheearlyyearsoftheinitiative(2014-2017)influencedBeijing’s
responsesintheso-called“BRI2.0”era(2018-2021).
Thepursuitofsoftpowerforforeignpolicyadvantage
Statesfrequentlyusenon-coercivetools—orso-calledsoftpower
instruments—towinthesupportofforeigncountriesonissuesofmajornational
importance(Nye2004;GoldsmithandHoriuchi2012;Goldsmithetal.2014;
Guisoetal.2009;Rose2016,2019).306ForChina,theseissuesincludeTaiwan,
Tibet,Xinjiang,FalunGong,humanrights,andmaritimeclaimsintheSouth
ChinaSea(Flores-MacíasandKreps2013;Kastner2016;Strüver2016,2017;
Custeretal.2018).
306Apopulardefinitionofsoftpoweris“theabilitytoachievegoalsthroughattractionratherthancoercion”(Nye2004:x).
176
MaintaininginternationalsupportfortheBRIisalsoakeypriority.InOctober
2021,weobservedBeijing’sreadinesstomanagereputationalriskrelatedtoits
flagship,globalinfrastructureinitiative.Withintwodaysofthereleaseofan
AidDatareportonpublicdebtexposuretoChina,aleadingEnglishdailyin
Pakistan,Dawn,wroteahard-hittingeditorialcallingforgreatertransparencyin
theplanningandfinancingoftheChina-PakistanEconomicCorridor(CPEC).
Fivedayslater,inresponse,twoCabinet-levelministersconductedalivepress
conferenceonnationaltelevisioninwhichtheypresentedapoint-by-point
responsetotheeditorialandthereport(Maliketal.2021;Dawn2021).During
subsequentmeetingswithseniorofficialsinPakistan,welearnedthatthepress
conferencewasorganizedinpartduetopressurefromtheChineseembassy,
whichhadexpressedfrustrationatthe“publicrelationsimpacts”resultingfrom
arareinstanceofnegativepresscoverageabouttheCPEC.307
Beijingusesawidearrayofsoftpowerinvestments—includingscholarship
programsthatallowforeignnationalstostudyinChina,trainingprogramsthat
exposeforeigngovernmentofficialstoChina’spolicypositions,junketsfor
foreignjournaliststovisitChinaandlearnaboutitsexperiencesand
perspectives,andpoliticalpartyoutreachandcapacitybuildingactivities—to
rallyinternationalsupportforitsprioritiesandpolicypositions.Accordingto
DongandChapman(2008:162),Chinahasprovidedthousandsofuniversity
scholarshipstoAfricanstudentseachyearwiththeaimof“trainingfuture
leaders[...]whomightserveasopinionleadersoncebackintheirhome
countries.”InananalysisofBeijing’sefforttocultivaterelationshipswithAfrican
politicalparties,Aiping(2015)findsthatAfricancountriesfrequentlydispatch
delegationstoChinatolearnmoreaboutitsmodelofeconomicdevelopment
andsystemofgovernance,andthattheCommunistPartyofChina(CPC)uses
theseexchangestoextolthevirtuesoftherulingpartymaintaininginternal
stabilitytopromotelong-runeconomicgrowthanddevelopment.
307Gelpernetal.(2022:26,emphasisadded)reportasimilarepisodeinvolving“avideoobtainedandreleasedbyinvestigativejournaliststhatrevealedthetermsofEcuador’smulti-billiondollaroil-backeddebttoCDB.Thereleaseofthevideoshortlyafterthedealwassignedpromptedpublicdebateaboutthenewborrowing[...].Inresponse,theheadofCDB’sResidentMissioninEcuadorwrotetohiscounterpartinEcuador’sMinistryofFinance,complainingabouttheborrower’sapparentbreachof[a]confidentialityletter,calledontheEcuadoriangovernmenttolaunchaleakinvestigationanddemandedthatittakemeasurestomitigatethereputationaldamagetoCDBcausedbythevideo.TheCDBletteralsoimplicitlythreatenedtowithholdfuturefinancingiftheborrowerdidnotadequatelyaddresstheincident.”
177
Anecdotalevidencesuggeststhatthesesoftpowerinvestmentsmaybe
effective.Consider,forexample,thecohortofcountriesthatagreedtoabstain
onakeyUNGAvoterelatedtoRussia’sinvasionofUkraine.308OnMarch2,2022,
UNGApassedaresolutioncondemningRussia’saggressionandcallingonitto
“immediately,completelyandunconditionallywithdrawallofitsmilitaryforces
fromtheterritoryofUkraine.”141countriesvotedinfavoroftheresolution,5
countriesvotedagainstit,andanother35countriesabstained.Chinaabstained
andmanyofitslargestaidandcreditrecipients—includingAngola,Bangladesh,
Bolivia,Burundi,CentralAfricanRepublic,Congo-Brazzaville,Cuba,ElSalvador,
Ethiopia,Gabon,Guinea,Iran,Kazakhstan,Kyrgyzstan,Laos,Mongolia,
Mozambique,Namibia,Pakistan,SouthAfrica,SriLanka,Sudan,Tajikistan,Togo,
Uganda,Uzbekistan,Vietnam,andZimbabwe—followedsuit(WhiteandHoltz
2022).
Yetgreatpowersdonotusesoftpowerinstrumentsmerelytosecure
internationalsupportondiscretepolicyissues.Theyalsostrivefordeeper
ideologicalalignmentandwidespreadadmirationoftheireconomic
developmentandgovernancemodels(Repnikova2022).Effective,long-term
softpowerprojectionrequiresattractionalinfluence:resonantideasandvalues
thatshapehowgoverningelitesinothercountriesdiagnoseproblems,think
aboutcause-and-effectrelationships,identifydesirablepolicyoutcomes,assign
priorityamongcompetingobjectives,anddeterminehowpoliciesshouldbe
formulatedandimplemented(Kroenigetal.2010;Atkinson2010).
FormanyofChina’stargetaudiencesintheGlobalSouth,itsmodelofeconomic
developmentisattractivebecauseofitsapparentsuccess:itsuccessfullylifted
680millionpeopleoutofpovertyinageneration(Ravallion2009;LinandWang
2014).309China’smodel,whichprioritizestherapidinstallationoflarge-scale
physicalinfrastructureviastate-ledinvestment,iscentraltothewaythatit
309AccordingtoHalper(2010),“China’sgoverningmodelismoreappealingtothedevelopingworldandsomeofthemiddle-sizedpowersthanAmerica’smarket-democraticmodel.Giventhechoicebetweenmarketdemocracyanditsfreedomsandmarketauthoritarianismanditshighgrowth,stability,improvedlivingstandards,andlimitsonexpression—amajorityinthedevelopingworldandinmanymiddle-sized,non-westernpowersprefertheauthoritarianmodel.”308AnotherinterestingexampleisthepolicypositionthatmanyAfricangovernmentstakeonChina’streatmentofUyghurslivinginXinjiangProvince.RatherthanjumpingontheWesternbandwagonandcriticizingChinaforitsallegedhumanrightsabuses,Africangovernmentsincreasinglypushback,arguingthat“Westernforces[are]hypinguptheso-calledXinjiang-relatedissues[to]servetheirownulteriormotives”(Olewe2021).
178
administersitsoverseasgrant-givingandlendingprogram(Ansaretal.2016).310
Thisapproachsupposedlyallowscountriestoleapfrogtheprocessof
establishingliberalinstitutionsanddemocraticvalues,whichwereprerequisites
foreconomicandsocialprogressinWesterncountries(NaughtonandTsai
2015).
Forpoliticalleadersseekingreelectionwithinfour-tofive-yearcyclesand
autocratsvyingforlegitimacyonthebasisofperformance,China’smodelof
economicdevelopmentcanbeespeciallycompelling.ArthurMutambara,
Zimbabwe’sformerDeputyPrimeMinister,toldtheWallStreetJournalin2011
that“Chinaismyfavoritecountry.[...]China'smodelistellingusyoucanbe
successfulwithoutfollowingtheWesternexample”(Wonacott2011).Several
yearsearlier,AbdoulayeWade,theformerPresidentofSenegal,admonished
Westernaidagenciesandmultilateraldevelopmentbanksfortheirinsensitivity
totheconditionsfacingpolicymakersindevelopingcountriesandpraised
China’smodel:“China’sapproachtoourneedsissimplybetteradaptedthan
theslowandsometimespatronizingpost-colonialapproachofEuropean
investors,donor[agencies]andnongovernmental[organizations].[...]Iamafirm
believeringoodgovernanceandtheruleoflaw.Butwhenbureaucracyand
senselessredtapeimpedeourabilitytoact—andwhenpovertypersistswhile
internationalfunctionariesdragtheirfeet—Africanleadershaveanobligationto
optforswiftersolutions”(Wade2008).
TheeffectivenessofChina’ssoftpowerinstruments
ButChina’ssoftpowerplaybookgoesbeyondbankrollingandbuilding
big-ticketinfrastructureprojects.NotunlikeWesternpowers,itdeploysatoolkit
thatincludesculturaldiplomacythroughConfuciusInstitutes,exchange
diplomacyusingstudentscholarshipsandjournalistvisits,andelite-to-elite
310AnimportantfeatureofChina’smodelisits“portfolioapproach.”ChinandGallagher(2019:256)explainthat“[w]hereasWestern-backed[developmentfinanceinstitutions]and[multilateraldevelopmentbanks]conductindividualprojectfinancing,China’spolicybanks,athomeandabroad,takea[...]portfolioapproachandfinancewhattheyrefertoas‘strategiccreditspaces’wherebundlesofloansorlinesofcreditareissuedforanarrayofcoordinatedandcorrespondingprojects.”Coordinatedpublicinvestmentstrategieshavearichintellectualhistoryrelatedto“bigpush”theory(Rosenstein-Rodan1943)and“growthpole”theory(Perroux1950;Hirschman1958).AccordingtoChinandGallagher(2019:251),“[s]omeintheseniorranksoftheChinesestatepolicybankshavedrawninspirationfrom[“bigpush”theory],includingtheformerchiefeconomistattheChinaDevelopmentBank[CDB],LixingZou,whosawCDBashavingplayedsuchacoordinatingrolewithintheChinesegrowthmiracle.”
179
diplomacyintheformofleadershipvisits(Benabdallah2021;Custeretal.2018).
Beijinghasalsoputinplaceawell-resourcedstrategiccommunicationsstrategy
with“distinctive”Chinesecharacteristics(Shambaugh2015;Snow2022).Ithas
dramaticallyexpandedtheglobalreachofofficialmediaorganizations,suchas
XinhuaNewsAgencyandChinaRadioInternational.311China’smessaging
operationsare,inturn,buttressedbycontent-sharingpartnershipswithlocal
media,whichrepublishstoriescarryingofficialChinesenarratives(Custeretal.
2019).
AgrowingbodyofstatisticalevidencesuggeststhatChinahaseffectively
utilizedsoftpowerinstrumentstoinfluencepublicopinion,mediasentiment,
andtheforeignpolicyprioritiesofgoverningelites,albeitwithvaryingdegrees
ofsuccess(BrazysandDukalskis2019;Eichenaueretal.2021;Blairetal.2022a;
Dreheretal.2022;Wellneretal.forthcoming,2023;Asmusetal.forthcoming).
Wellneretal.(forthcoming)findthat,onaverage,thecompletionofeach
additionalChinesedevelopmentprojectleadstoa3percentagepointincrease
inpublicsupportfortheChinesegovernmentintheshortterm,althoughthis
effecterodesovertime.312BrazysandDukalskis(2019:557)provideevidence
thatgeographicalproximitytoanactiveConfuciusInstitute“significantlyand
substantivelyimprovesthetoneofmediareportingabouteventsrelevantto
China.”
TheauthoritiesinBeijingreportedlyvaluemorefavorablemediasentimentand
publicsentimentbecausetheybelieveitcan“filterupandinfluenceelitepolicy
tobemoreamenableto[theirown]interests”(BrazysandDukalskis2019:567).
However,Beijingalsoseekstodirectlyinfluencetheforeignpolicyprioritiesof
governingelites.Dreheretal.(2022)provideevidencethata10%increasein
votingalignmentwithChinaintheUNGeneralAssemblyyieldsa276%increase
312Buildinguponthesemodelresults,Wellneretal.(2023)estimatethecountry-specificmagnitudesofthesepublicopinioneffectsbyassumingacounterfactualscenarioinwhichtheChinesegovernmentequallydistributeddevelopmentprojectsacrossallcountries.IfCambodiareceived30Chinesedevelopmentprojects(theaveragenumberofprojectsoverthe2006-2017periodofstudy)ratherthanthe91thatitactuallyreceived,theauthorsestimatethatChinawouldhavesuffereda12.55percentagepointlossofpublicsupportinthatcountry.311InAfrica,itsprimaryexport,CGTNAfrica,hasbeenlesssuccessfulinbringingChina’smessagetotheworldthanBBCAfricaduetoalackofqualitystaffing(ZhangandOng’ong’a2021).
180
inaidandcreditfromChina,onaverage.313TheseresultssuggestthatBeijing,
likesomeofitsrivals,usesitsforeignaidandcredittoencourageand/orreward
countriestosupportitsforeignpolicypositions.314
YetmuchremainsunknownaboutChina’ssoftpowerstrategyintheGlobal
South.IsBeijingresponsivetosoftpowergainsorlossesvis-à-vistheU.S.?What
typesofreputationalliabilitiesorassetsleadtocompetitiveresponses?Where,
when,andhowisBeijingusingitsdevelopmentfinanceinstrumentstomakesoft
powergains?
Leveragingthebroadtemporalandgeographicalcoverageofthe3.0versionof
AidData’sGCDFdataset,weexaminethesequestionswithan“action-reaction”
framework.BasedonChina’spriorresponsestochangesinpublicopinion,
mediasentiment,andelitesupportinLICsandMICs,weseektoanticipate
whereandhowChinawillmountcompetitiveresponsesinthefuture.315We
placespecialemphasisonunderstandinghowChinaallocateslimited
competitiveresourcesacross“safebet,”“toss-up,”and“moonshot”countries.
Section2:Vyingforsoftpower:China’splayforpublicopinion,mediasentiment,andelitesupport
Ouraction-reactionframeworkevaluatesBeijing’saidandcreditallocation
decisionsduringthelateBRIperiod(2018-2021)afteritobservedvarious
changesonthreemeasuresofsoftpowerduringtheearlyBRIperiod
(2014-2017):publicsupportfortheChinesegovernment,asmeasuredviathe
315China’ssystemofgovernanceisoftencharacterizedas“fragmentedauthoritarianism,”wheremultilevelpowercentersexist,eachwiththeirownlisteningposts,interests,anddecision-makingprocesses,whichcanmakeitdifficulttogenerateunifiedpositionsandapproaches(Mertha2009).Analystshavealsoarguedthat“fragmentedauthoritarianism”affectstheexecutionofChina’soverseasdevelopmentprogram,wheremultiplestakeholdersassociatedwiththestate(companies,banks,diplomaticmissions,etc.)areadvancingtheirowninterestswithoutnecessarilyfollowingacoherentstrategyorganizedbyacentralcoordinatingagency(Ye2021,Lee2020).AkeyobjectiveofthischapteristoprobetheplausibilityofthehypothesisthatBeijingiscapableofformulatingcoherentresponsestothecompetitive(softpower)challengesthatitfacesintheGlobalSouth.314PreviousresearchdemonstratesthatforeignpowersotherthanChinauseaidandcredittoinfluencetheforeignpolicypositionsofdevelopingcountries(AlesinaandDollar2000;KuziemkoandWerker2006;VreelandandDreher2014;Rose2018).313WethankAxelDreher,AndreasFuchs,AustinStrange,andMikeTierneyforgeneratingandsharingsupplementaryevidencederivedfromastatisticalmodelinthefifthchapteroftheirbookBankingonBeijing.
181
GallupWorldPoll(GWP);thefavorabilityofmediacoveragerelatedtothe
Chinesegovernment,asmeasuredviatheGlobalDatabaseofEvents,
Language,andTone(GDELT);andelitesupportforChina’sforeignpolicy
positions,asmeasuredviaUNGAvotingalignmentwithChina.
InChapter1,wedocumentedthat,intheGlobalSouth,thesethreesoftpower
metricsmovedinvaryingdirectionsoverthefirsteightyearsoftheBRI
(2014-2021).DuringtheearlyBRIperiod,thecontestforinternationalpublic
opinionbetweenWashingtonandBeijingremainedneckandneck.However,
publicsupportforChinafellsharplyby16percentagepointsbetween2019and
2021.Intermsofmediafavorability,afteryearsofmaintainingacompetitive
edge,BeijinglostsignificantgroundtoWashingtonduringthelateBRIperiod;
itsfavorabilityadvantageoverWashingtonshranktoarazor-thinmargin.China
didrelativelywellatmaintainingelitesupportforitsforeignpolicypositions,as
measuredbytheextenttowhichLICsandMICsalignedtheirUNGAvoteswith
China.However,wealsoidentifiedsomegroundsforconcerninBRIparticipant
countries,asgoverningelitesinthiscohorttookforeignpolicypositions(in
UNGA)thatwereincreasinglyoutofalignmentwiththoseofChinaduringthe
lateBRIperiod.
OuranalysisinChapter1alsorevealedthat,whilethebattleforheartsand
mindsbetweenChinaandtheU.S.wasa“toss-up”duringtheearlyBRIperiod,
itbecamesignificantlymorecompetitiveduringthelateBRIperiod.Washington
mademorereputationalgainsatBeijing’sexpense—thanvice-versa—duringthis
period.Thiswasespeciallytrueinthecontestforinternationalpublicopinion,
wherealmost85%ofthecountry-levelchangesinpublicopinionrepresented
relativelossesforChina.
Wenowextendtheanalysisbyfocusingontherelativegainsandlosses
experiencedbybothpowersonanannualbasisacrossthreesoftpower
measuresduringtheearlyBRIperiod.Wedesignedouranalyticapproachin
recognitionoftworealities.First,greatpowercompetitionforsoftpoweris
generallyconsideredtobeazero-sumgamebytheparticipants.316Second,
316Forexample,inanempiricalexaminationofcompetitionbetweenChinaandIndiain2,333provincesacross123countriesbetween2007and2014,onestudyfindsthatan“increaseintheprobabilityofanewIndianEximBankloaninresponsetonewChinesedevelopmentprojectsismorepronouncedwhen
182
events(“shocks”)atlocal,regional,andgloballevelscansimultaneouslyaffect
overseassupportforallgreatpowers.317Toaddresstheserealities,wefirst
measureabsoluteyear-on-year(YoY)changestoourthreemeasuresofsoft
poweratthecountrylevel(singledelta)foreachgreatpower(Chinaandthe
U.S.).Wethenmeasurethedifference(doubledelta)betweenthechangesthat
ChinaexperiencedandthechangesthattheU.S.experienced(seeBox4a).
ThesedoubledeltameasuresallowustoidentifywhenandwhereChina
experienced(a)largerlosses(orsmallergains)thantheU.S.,and(b)largergains
(orsmallerlosses)thantheU.S.318
ThismethodologicalapproachalsomakesiteasiertothinkabouthowBeijing
allocatesscarceresourcesacrossdifferentcategoriesofcountriestoadvanceits
softpowerobjectives.Whenagreatpowerallocatesscarceresourcesinpursuit
ofgrassroots,media,orelitesupport,itmustmakerisk-adjustedreward
calculations—bybalancingthemagnitudeofapotentialsoftpowergainina
givenjurisdictionagainstthelikelihoodofsuccessinthatsetting(Asmusetal.
forthcoming).Asageneralrule,itcanmakesmallergainsinplaceswherethe
probabilityofsuccessishigh(safebetterritory),oritcanventureintomore
challengingplaceswheretheprobabilityofsuccessislow(moonshotterritory)
buttheopportunitytomakealargegain(attheexpenseofarival)ishigh.319In
recognitionofthistradeoff,weorganizeouranalysisaroundfourcohorts:safe
betcountries,toss-upcountriesthatleantowardChina,toss-upcountriesthat
leantowardtheU.S.,andmoonshotcountries(basedonthedatasourcesand
methodsdescribedinBox4a).
319Vadlamannatietal.(2023:16)provideevidencethatthe“BRIpromptsthelargestpositiveU.S.responsewhenChinaisengagedwiththetargetcountrybutnotyetdominant.”TheexistingempiricalliteraturedoesnotprovideconclusiveevidenceaboutwhetherBeijingfollowsastrategyofbalancing,bandwagoning,orhedging.318TableA13providesthecountry-specificdoubledeltascoresforeachsoftpowermeasure.317Forinstance,whenacountrysuffersfromamajornaturaldisasterinducedbyglobalclimatechangeandlevelsofforeignassistancearelacking,publicresentmentcanincreasetowardalloutsideforces,aswellasdomesticinstitutions.popularopinionintherecipientcountryisrelativelymorefavorableaboutIndiathanaboutChina”(Asmusetal.forthcoming:26,emphasisadded).Thesamestudyprovidesevidencethatthiseffect“isdrivenbythedifferenceinpublicsentimenttowardIndiaandChinaratherthanbytheabsolutelevelsofpublicsupportforIndiainthesecountries”(Asmusetal.(forthcoming:26-27).
183
Box4a:HowAidDatameasuresrelativegainsandlossesinChinesesoftpower
ForallthreemeasuresofsoftpowerinLICsandMICs(publicopinion,mediasentiment,andelitesupport),weadoptanidenticalapproachtomeasurerelativegainsandlossesbyChinavis-à-vistheU.S.atthehostcountrylevel.Inordertoexplainourmethodofmeasurement,wedescribehowitisappliedtoonesoftpowerindicator(publicopinion)andonehostcountry(Bangladesh).Wetakethefollowingsteps:
Step1:UseBangladesh’sannualpublicapprovalratings(fromGWP)tocalculatetheyear-on-yearchange(singledelta:t0-t1)forChinaandtheU.S.foreachyearoftheearlyBRIperiod.
Step2:Calculatethedifferencebetweenthe“singledelta”inChina’spublicapprovalratingandthe“singledelta”intheU.S.publicapprovalratingforeachyear.Theresulting“doubledelta”(China-U.S.)measureprovidesanindicationofwhetherandtowhatextentChinagainedorlostgroundattheexpenseoftheU.S.Thelargerthesizeofa(positive)doubledelta,thegreaterthesizeofthegainthatChinamadeattheexpenseoftheU.S.—orviceversa.Figure4.1providesavisualrepresentationofhowthiscalculationworksforonesoftpowerindicator(GWP)inBangladeshbetween2016and2017.
F i g u r e
4 . 1 :
B a n g l a d e s h
e x a m p l e :
I s
C h i n a
l o s i n g
o r
g a i n i n g
g r o u n d
v i s - à - v i s
t h e
U . S .
i n
2 0 1 7 ?Figure4.1
Step3:Calculatetheaverageannualrelativechange(i.e.,averageannualdoubledelta)inpublicapprovalovertheearlyBRIperiod(2014-2017).Averagingoverthe2014-2017periodeffectivelymeasureswhetherChinagainedmoregroundoverallthantheU.S.—orviceversa—duringthe
184
earlyBRIperiod.Table4.1providesthefullcalculationforBangladesh,where1.6isthefinalscoreforBangladeshduringthistimeperiod.
Table
4.1:
Bangladesh
example:
Calculating
an
average
double-delta
score
for
soft
power
measures
Table4.1
Bangladeshexample:Calculatinganaveragedouble-deltascoreforsoftpowermeasures
YearChinaApproval
RateChinaSingleDeltaU.S.ApprovalRateU.S.SingleDeltaChina-U.S.DoubleDelta
201369.6-66.2--
201447.5-22.144.9-21.3-0.817
201555.6+8.155.5+10.6-2.52
201670.4+14.870.5+15.0-0.164
201762.0-8.452.1-18.4+9.99
Average(2014-17)58.8-1.955.7-3.51.6
SafeBets(75-100thpercentile)Toss-Ups-LeaningChina(50-74.9thpercentile)Toss-Ups(25-49.9thpercentile)Moonshots(0-24.9thpercentile)
Step4:Repeatthisprocessforallcountry-yearpairings,rankcountriesbasedontheiraverageannualdoubledeltavaluesduringtheearlyBRIperiod,anddividetheglobaldistributionofLICsandMICsintofourcategories:
●Safebets:Thetopquartile(75thto100thpercentile)ofcountrieswhereChinamadethelargestgainsattheexpenseoftheU.S.duringtheearlyBRIperiod.
185
●Toss-ups(leaningChina):Countriesthatfallbetweenthe50thto74.9thpercentilerepresentcompetitivejurisdictionswhereChinamaderelativelysmallgainsattheexpenseoftheU.S.duringtheearlyBRIperiod.320
●Toss-ups:Countriesthatfallbetweenthe25thto49.9thpercentilerepresentcompetitivejurisdictionswhereChinamadenegligiblegainsattheexpenseoftheU.S.duringtheearlyBRIperiod—ortheU.S.maderelativelysmallgainsattheexpenseofChinaduringtheearlyBRIperiod.
●Moonshots:Thebottomquartile(0to24.9thpercentile)ofcountrieswheretheU.S.madethelargestgainsattheexpenseofChinaduringtheearlyBRIperiod.
Publicopinion
TounderstandChina’spublicopiniongainsandlossesvis-à-vistheU.S.during
theearlyBRIyears,wefirstrank-order104LICsandMICs(forwhichGWPdata
areavailable)accordingtoourdoubledeltameasure(seeFigure4.2).The
observedvaluesonthismeasurerangefrom-18ppinMalaysiato+60ppfor
Libya.Wethencalculateapercentilerankingforeachcountryintheglobal
distribution.Theserankingsaresubsequentlyusedtoassigneachcountryto
oneoffourcategories(describedinBox4a):safebets(75thto100the
percentile),toss-upleaningChina(50thto74.9thpercentile),toss-up(25thto
49.9thpercentile),andmoonshots(0to24.9thpercentile).
320The“Toss-ups(leaningChina)”categorydescriptionisadmittedlyimperfectwhenappliedtotheelite(UNGAvoting)alignmentmeasure,sincetheU.S.madegainsattheexpenseofChinainasubstantialnumberofthecountriesthatfallwithinthecategory.
186
Figure
4.2:
China’s
public
approval
gains
and
losses
vis-à-vis
the
U.S.Figure4.2
Notes:Therelativechangecalculationisbasedonthe“doubledelta”measurethatisdescribedinBox4a.
Basedonitspositionintheglobaldistribution,eachcountryisassignedtooneoffourcategories:safebet,
toss-up(leaningChina),toss-up,ormoonshot(asdescribedinBox4a).
AsFigure4.2demonstrates,duringtheearlyBRIyears,Chinaachievedgainsat
theexpenseoftheU.S.intwo-thirdsofallhostcountrieswithGWPpublic
opiniondata.Theaveragedoubledeltainthiscohortwas+3pp.Thereare
severalreasonswhytheU.S.mayhavesufferedsoftpowerlossesduringthis
period.OneisPresidentTrump’selectionandthe“AmericaFirst”agenda,which
broughtaboutaperiodofglobalretrenchment.AnotherisPresidentXi’s
contrastingapproach,whichsoughttoestablish“win-win”partnershipsandnew
“connectivities”withtheGlobalSouththroughtheBRI(Rolland2017).321
321Beforeandafterhis2016electoralvictory,PresidentTrump’sfiery“AmericaFirst”rhetoricanddivisivepersonalviewsraisedworldwideconcernsabouthiscountry’scommitmenttomaintaininggloballeadership.HiselectoralagendaincludedproposalstoimposetariffsonChineseimportsintotheU.S.for“protectingAmericanjobs”whichraisedconcernsaboutanew“tradewar”thatwouldultimatelyhurtdevelopingcountries(Kucik2017).Atthesametime,XiwasseekingtopositionChinaasaleaderontheworldstage.WhenhefirstintroducedtheBRI,hefocusedoncreating“win-win”partnershipswithhostcountriesthroughbig-ticketinfrastructureprojects.
187
NotwithstandingBeijing’soverallsuccessinmakingpublicopiniongainsatthe
expenseofWashington,thereweresomenotableexceptions.Negativedouble
deltasinMalaysia(-18pp),Vietnam(-10pp)andNiger(-2pp)indicate
significantrelativelossesforChina.Duringthesameperiodoftime,China
registeredlarge,positivedoubledeltasinJamaica(+16pp),Namibia(+10pp),
andEgypt(+8pp).Inatleasttwo-thirdsofcountries,weseeevidenceofrobust
competitionbetweenWashingtonandBeijing,withaveragedoubledeltavalues
inthe-5to+5pprange.
Whileeverycountrycontextisunique,publicapprovallevelsappeartobe
sensitivetocertainfactors,suchaschangesofgovernmentfrompro-China
incumbentstonewpoliticalleaderswhoviewBeijingwithgreaterskepticism
(e.g.,Malaysia),ongoingorworseningterritorialdisputeswithChina(e.g.,
Vietnam),andmajorchangesinlevelsofaidorlendingfromagreatpower(e.g.,
Jamaica).ThecasestudiesinSection4demonstratethattheextenttowhich
policymakersinBeijingusedevelopmentfinanceandpublicdiplomacytoolsto
respondtosuchchangesdependsontheirperceptionsofthestrategicvalueof
agivencountry.
Mediasentiment
Wefollowthesamedoubledeltaapproachtomeasurethemediasentiment
gainsandlossesthatChinaachievedvis-à-vistheU.SduringtheearlyBRI
period.322AccordingtoFigure4.3,Beijingachievedgainsattheexpenseof
Washingtonin79.4%of(104outof131)hostcountriesforwhichGDELTdata
areavailable.Thispatternisconsistentwithpreviousresearchthatsuggeststhe
launchoftheBRI,togetherwithBeijing’sstrategyoframpinguppublic
diplomacyefforts,includinggrassrootsmediamanagement,generatedmore
positivemediacoverageaboutChina’sgovernment(Custeretal.2019;Brazys
andDukalskis2019).
322AswedescribeinBox1binChapter1,werestrictouranalysistomediacoverageaboutChineseandAmericangovernments,ratherthanthecountriesoverall,toensurethatwefocusontheactivitiesoftheofficialsector.Inmosthostcountries,anarrayofChineseandAmericanprivatecompaniesandnonprofitsengageinactivitiesthatarewhollyorpartiallyindependentoftheirhomecountry’sofficialpoliciesandapproaches.
188
Figure4.3alsodemonstratesthatChinaachievedthebiggestrelativegainsin
theDemocraticRepublicoftheCongo(+2.8pp),Vanuatu(+1.8pp),and
Tajikistan(+1.5pp),whileitslargestrelativelossestookplaceinComoros(-2.6
pp),Tunisia(-1.4pp),andDominica(-1pp).Acrosstheentireglobaldistribution,
theaveragedoubledeltasizeduringtheearlyBRIperiodwas+0.54.
Figure
4.3:
China’s
media
sentiment
gains
and
losses
vis-à-vis
the
U.S.Figure4.3
Notes:Therelativechangecalculationisbasedonthe“doubledelta”measurethatisdescribedinBox4a.
Basedonitspositionintheglobaldistribution,eachcountryisassignedtooneoffourcategories:safebet,
toss-up(leaningChina),toss-up,ormoonshot(asdescribedinBox4a).
Elitesupport
AsamemberoftheGlobalSouth,China’svotingpatternsintheUNGAare
closelyalignedoverallwiththoseofdevelopingcountries—especiallyonissues
relatedtohumanrightsandnationalsovereignty(FungandLam2022).323
However,asFigure4.4demonstrates,unlikepublicapprovalandmedia
323Moreoftenthannot,wefindthatChinaandtheU.S.votedifferentlyattheUNGA.Since2000,theyhaveonlytakenthesamepositionon14.5%ofallresolutions.Thismisalignmentbecomesworseincertainissueareas,suchasa7.8%alignmentonhumanrightsresolutionsanda4.6%alignmentonthePalestinianconflict.Thereismuchcloseralignment,19.5%,betweenChinaandtheU.S.onresolutionsconcerningnuclearweaponsandmaterialsandarmscontrolanddisarmament,consistentwiththeirnuclearstatuses.
189
sentimentwhereBeijingachievedsignificantgainsatWashington’sexpense
duringearlyBRIyears,UNGAvotingalignmentisameasureofsoftpowerfor
whichweobservetheoppositetrend:governingelitesinLICsandMICsmoved
theirforeignpolicypositionsintocloseralignmentwithChinainonly28%ofthe
107countries(forwhichUNGAvotingalignmentdataareavailable).324China
experiencedsoftpowerlossesvis-à-vistheU.SinIndia(-0.14pp),Pakistan(-0.12
pp),Brazil(-0.09pp),Venezuela(-0.05pp),Mali(-0.16pp)andRwanda(-0.14
pp).ThecountriesthathandedrelativesoftpowergainstoChinaincludedthe
PacificislandnationsofKiribati(0.4)andMicronesia(0.34),andseveralmid-sized
countriesinLatinAmericaincludingPanama(0.18)andParaguay(0.12).
Atthesametime,thesechangesareimportanttounderstandintheircorrect
context.ThevastmajorityofLICsandMICsremainedfarmorestronglyaligned
withChinathantheU.S.intheUNGAinabsoluteterms.AcrossallUNGAvotes
castbetween2000and2021,wefindthatLICsandMICsonaveragealigned
withChina75.6%ofthetime,ascomparedto23.1%withtheU.S.
Figure
4.4:
China’s
elite
alignment
gains
and
losses
vis-à-vis
the
U.S.Figure4.4
324Acrosstheglobaldistribution,theaveragedoubledelta“idealpoint”distancewas-0.02.
190
Notes:Therelativechangecalculationisbasedonthe“doubledelta”measurethatisdescribedinBox4a.
Basedonitspositionintheglobaldistribution,eachcountryisassignedtooneoffourcategories:safebet,
toss-up(leaningChina),toss-up,ormoonshot(asdescribedinBox4a).
Section3:Fromobservationstoreactions
HavingmeasuredthesoftpowergainsandlossesthatChinaexperiencedduring
theearlyBRIperiod,wenowturnourattentiontothesecondpartofour
action-reactionframework:Beijing’slateBRIresponsestosoftpowerchanges
duringtheearlyBRIperiod.Wedosobyanalyzinganewfeatureofthe3.0
versionofAidData’sGCDFdataset:developmentfinancecommitmentdatafor
allLICsandMICsbetween2018and2021.
QuantifyingBeijing’sstrategicprioritization:Theexpectationversus
realityframework
ThenextstepinouranalysisistodetermineifthereisevidenceofBeijing
strategicallyallocatingaidandcreditduringthelateBRIperiodtopursueitssoft
powerobjectives.WedosobyanalyzingwhetherBeijingfavoredordisfavored
anyspecificsoftpowercountrycohortsbetween2018and2021,relativetoa
referencepointofaidandcreditallocationsthatarebasedonthecross-country
distributionofeconomicneed.325Incaseswhereweobservelargedifferences
betweentheexpectedsizeofChinesedevelopmentfinancecommitments
(basedoneconomicneed)andtheobservedsizeofChinesedevelopment
financecommitments,wearguethatsuchdifferencesmayreflectstrategic
decisionsbyBeijingtoprioritizeorde-prioritizeacountryforsoftpower
purposes.
325Thecross-countryaidallocationliteraturegenerallyusestwovariablestomeasureeconomicneed:incomeandpopulationsize(e.g.,AlesinaandDollar2000;Dreheretal.2022).However,incomeisusedtoaccountforvaryinglevelsofeconomicneedacrossHICsandLICs/MICs.GiventhatouranalysisisexclusivelyfocusedonLICs/MICs(i.e.,onlycountrieswithrelativelylowlevelsofincome),werelyonpopulationsizetocapturecross-countrydifferencesineconomicneed.
191
Table
4.2:
Expected
versus
observed
allocation
of
Chinese
development
finance
by
soft
power
cohortTable4.2
Expected
versus
observed
allocation
of
Chinese
development
finance
by
soft
power
cohort
CategoryPublicOpinionMediaSentimentEliteSupport
Expected(%)Observed(%)Difference(pp)Expected(%)Observed(%)Difference(pp)Expected(%)Observed(%)Difference(pp)
SafeBets1517+2109-1730+22
Toss-Up-LeaningChina1848+291948+291723+6
Toss-Up2215-73635-12816-12
Moonshot4316-27347-274731-16
Notes:Thistableillustrateswhichsoftpowercohorts(safebet,toss-upleaningChina,toss-up,and
moonshot)receivedrelativelymoreorlessdevelopmentfinancefromChinaduringthelateBRIperiod
(2018-2021)comparedtoanexpectedallocation.Theexpectedpercentagesforeachcohortandsoft
powermeasure(publicopinion,mediasentiment,andelitesupport)arebasedonahypotheticalscenario
inwhichChinaallocateddevelopmentfinancetoeachcohortonanon-strategicbasis(i.e.,basedonlyon
thepopulationsizeofeachcohort).TheobservedpercentagesrepresenttheactualpercentagesofChina’s
ODAandOOFportfolioallocatedtoeachsoftpowercohortduringthe2018-2021timeperiod.The
differencesbetweenexpectedandobservedallocationsthereforeprovideanindicationofhowmuchChina
mayhaveprioritizedordeprioritizedeachcohortbasedonstrategicconsiderations.
Toputthisframeworkintopractice,wefollowathree-stepprocess.First,we
assigneachLICandMIC(withavailabledata)toasafebet,toss-up(leaning
China),toss-up,ormoonshotcategorybasedonitspositionwithintheglobal
distributionofrelativegainsandlossesexperiencedbyChinaduringtheearly
BRIperiod.Wefollowthisprocedureforeachofthethreemeasuresofsoft
power,whichresultsineachcountrybeingassignedtoadifferentcategoryfor
eachmeasureofsoftpower.Forexample,Moroccofallswithinthesafebet
categoryforthemeasureofpublicopinion,butitfallswithinthemoonshot
categoryforthemediasentimentmeasureandthetoss-up(leaningChina)
categoryfortheelitesupportmeasure.
Second,wecreatemeasuresoftheexpectedlevelofdevelopmentfinancefrom
Chinabasedonthesizeofeachhostcountry’spopulation.Wefirstaveragethe
192
populationsizeofeachhostcountryoverthe2018-2021timeperiod,andthen
estimateeachhostcountry’sshareofthetotalpopulationacrossallLICsand
MICsthatreceiveddevelopmentfinancecommitmentsfromChinaoverthe
sametimeperiod.326Wethenmultiplythesecountry-specificsharesbythetotal
sizeofChina’sinternationaldevelopmentfinanceportfolio(measuredin
constant2021USD)duringthelateBRIperiod.Forexample,hostcountriesthat
fallwithinthesafebetscategory(onthepublicopinionmetric)arehometoonly
14.9%ofthetotalpopulationacrossallLICsandMICs,soweexpectthesame
percentageofChina’sportfolioofinternationaldevelopmentfinance
commitmentsduringthelateBRIperiod(14.9%of$368.8billion,or$54.9
billion)tobeallocatedtothe26countriesinthesafebetscategory.(SeeTable
A14intheAppendixformoredetails).
Third,foreachofthefourcategoriesacrosseverysoftpowermetric,we
measurethesizeofthegapbetweentheexpectedsizeandactualsizeof
developmentfinancecommitmentsfromChina.Consideragaintheglobal
distributionofpublicopiniongainsandlossesthatChinaexperiencedduring
theearlyBRIperiodandthecountriesthatfallwithinthesafebetscategory.
Withthe3.0versionofAidData’sGCDFdataset,wecalculatetotaldevelopment
financecommitmentsfromChinatothiscohortof26countriesduringthe
2018-2021period.Thisfigureamountsto$62.2billion,whichrepresents17%of
China’sportfolioofinternationaldevelopmentfinancecommitmentsduringthe
lateBRIperiod.The$7.2billiondeltabetweentheactualsizeandexpectedsize
ofChinesedevelopmentfinancecommitmentsrepresentsa+2ppdifferencefor
countriesinthesafebetscategory(seeTable4.2),whichsuggeststhatBeijing
mayhaveprioritizedthisgroupofcountriesforsoftpowerpurposes.However,
onecanalsoseeinTable4.2thatthereissubstantiallystrongerevidenceof
Beijingprioritizingthetoss-up(leaningChina)category(+29ppdifference)and
de-prioritizingthemoonshotscategory(-27ppdifference)forsoftpower
purposes.
326Between2018and2021,the3.0versionofAidData’sdatasetcapturesdevelopmentfinancecommitmentsto138countriesandterritories.
193
DecodingBeijing’ssoftpowerplaybook
Table4.2providesfivekeyinsightsregardinghowBeijingallocatesaidand
creditinresponsetosoftpowergainsandlosses.Together,theseinsightshelp
decodeBeijing’splaybookforcompetingwithWashingtonforexpanded
influenceintheGlobalSouth.
First,acrossallthreemeasuresofsoftpower,Chinadevotednearlytwo-thirdsof
itsentireinternationaldevelopmentfinanceportfolioduringthelateBRIperiod
tocountriesbelongingtothetwotoss-upcategories.Table4.2also
demonstratesthatBeijingprioritizedtoss-upcountriesthatleantowardit.Thisis
trueacrossallthreemeasuresofsoftpower,whichsuggeststhatBeijing’sforeign
policybraintrustismonitoringsoftpowergainsandlossesandredirectingaid
andcreditatpivotalmomentstobattlegroundcountrieswhereithasamodest
advantage.327
Second,acrossallthreemeasuresofsoftpower,wefindlargenegativedeltas
betweenexpectedandobserveddevelopmentfinanceallocationstocountries
inthemoonshotcategory(seeTable4.2andFigure4.5).Thispatternsuggests
thatBeijingisrelativelyrisk-aversewhenitcomestothepursuitofsoftpower.
RatherthanprioritizingaudaciousattemptstolurecountriesoutofWashington’s
orbit,Beijingdevotedfewresourcesbetween2018and2021tocountriesthat
movedinWashington’sdirectionduringtheearlyBRIperiod.ThefactthatChina
deprioritizedcountrieswhereithadrecentlysufferedsoftpowerlossesvis-à-vis
itsprincipalrivalrevealshowkeydecision-makersinBeijingapproach
risk-rewardcalculationsmoregenerally.Moonshotcountriesare“highrisk,high
reward”opportunities,inthattheyrepresentjurisdictionswherethemagnitudes
ofsoftpowergainscanbelarge(duetorelativelylowbaselinelevelsofsupport
forChina)buttheprobabilitiesofsuccessarelow.Figure4.5suggeststhat
Beijing’ssoftpowerinvestmentstrategyisnotfocusedonsuchcountriesbut
ratheroncountrieswheretheoddsofsuccessarehigher—evenifthesizeofthe
gainsthatcanberealizedarelesssubstantial.
327Aswewillsoonexplain,thereissomeevidencethat,inordertosafeguarditsgeopoliticalinterestsinBangladesh(whereitforgedarobustpartnershipwithPrimeMinisterSheikhHasina),Chinesestate-ownedfinanciersandimplementationagenciesacceleratedthepaceofprojectcommencementandcompletionceremoniesduringelectionseason.
194
Figure
4.5:
Beijing’s
aid
and
credit
prioritization
strategyFigure4.5
Notes:Thisfigureprovidesavisualrepresentationofthe“Difference(pp)”columnsfromTable4.2by
focusingonthecategoriesofcountrieswhichBeijingprioritizedordeprioritizedduringthelateBRIperiod,
inapparentresponsetosoftpowergainsandlossesvis-à-vistheU.S.Thelinelabeled0represents
situationswherethereisnodifferencebetweentheexpectationandrealityofBeijing'saidandcredit
allocation.Thepositiveornegativevaluesrepresentcategorieswherespendingexceedsandfailstomeet
expectations,respectively.
Third,inthecontestforinternationalpublicopinion,Chinahasshownalow
tolerancelevelforreputationalrisk.DuringthelateBRIperiod,itdevoted
two-thirdsofitsinternationaldevelopmentfinanceportfoliototoss-upcountries
andassignedspecialprioritytothesubsetoftoss-upcountriesthatleantoward
195
China(seeTable4.2).328Bycontrast,itdirectedonly16%ofitsinternational
developmentfinanceportfolioduringthelateBRIperiodtomoonshot
countries.329TheseresourceallocationpatternssuggestthatChinaisless
interestedincompetingincountrieswhereitsprincipalrivalhasmomentumon
itssideandmoreinterestedinshoringuppublicsupportincountrieswith
favorablebaselineconditions.ZambiaisanexampleofacountrythatBeijing
showeredwithaidandcreditwhenitwasrelativelypro-China.However,aswe
explaininSection4ofthischapter,afterthegovernmentdefaultedonits
externaldebtobligationsinNovember2020,China’spublicapprovalrating
plummeted(by-9.2pp)andtheU.S.publicapprovalratingremainedmostly
unchanged(-0.3pp).Beijing’sprovisionofaidandcredittoZambiaallbut
halted,completingadramaticfull-circleturnsincetheheydayofBeijing’s
engagementinthecountryduringtheearlyBRIperiod.
Fourth,inthequestforfavorablemediacoverage,Chinastronglydisfavors
countriesinthesafebetandmoonshotcategories.Itdevotedonly16%ofits
internationaldevelopmentfinanceportfolioduringthelateBRIperiodtosuch
countries.Instead,itfocused83%ofitsinternationaldevelopmentfinance
portfolioduringthelateBRIperiodoncountriesinthetwotoss-upcategories.
ThesespendingpatternssuggestthatBeijingisneitherinterestedinhigh-risk,
high-rewardopportunitiesnorlow-risk,low-rewardopportunitieswhenitcomes
tothepursuitoffavorablemediacoverage.Instead,itappearstobedoubling
downincompetitivejurisdictionswhereit“hasthewindatitsback.”Figure4.5
providesevidenceofalargepositivedelta(+29pp)betweenexpectedand
observedChinesedevelopmentfinanceallocationstocountriesinthetoss-up
(leaningChina)category.InthecaseofBangladesh,whereChinaexperienced
mediasentimentgainsattheexpenseoftheU.S.,averageannualcommitments
tripledfrom$994millionduringtheearlyBRIperiodtoover$3.3billioninthe
lateBRIperiod,despitechallengesassociatedwithCOVID-19.Inthecontextof
growingstrategiccompetitionintheIndo-Pacific,Beijing’spartnershipwith
329Forcontext,thesamesetofcountriesthatwereclassifiedasmoonshotinouranalysisreceived31%ofChina’stotaldevelopmentfinanceportfoliointhepre-BRIperiod(2000-2013),and12%duringtheearlyBRIperiod(2014-2017).328Forcontext,thesamesetofcountriesthatwereclassifiedastoss-up(leaningChina)ortoss-upinouranalysisreceived21%ofChina’stotaldevelopmentfinanceportfoliointhepre-BRIperiod(2000-2013),and28%duringtheearlyBRIperiod(2014-2017).
196
PrimeMinisterSheikhHasina’s15-yearoldgovernmentstrengthened
significantly.330
Fifth,intheinternationalcontestforelitesupport,Chinashowsrelativelylittle
interestintoss-upcountries.Table4.2}demonstratesthatBeijingallocatedonly
39%ofitsinternationaldevelopmentfinanceportfolioduringthelateBRIperiod
tothesehighlycompetitivejurisdictions.Itdirectedtheremaining61%ofthe
portfoliotocountriesinthemoonshotcategory(i.e.,jurisdictionswheretheU.S.
madesoftpowergainsatitsexpense)andsafebetscategory(i.e.,jurisdictions
whereChinamadesoftpowergainsattheexpenseoftheU.S.).331These
spendingpatternssuggestthatBeijingisconfidentitstried-and-truestrategyof
tradingcashforforeignpolicyconcessionscanworkinanynumberofsettings.
AsweexplaininSection2,thecaseofArgentinaisconsistentwiththenotion
thatBeijingiswillingtoworkwithfriendlygovernments,regardlessoftheir
ideologicalcommitmentsorpoliticalviewpoints.
Section4:Countrycasestudies
InSection3,welearnedthatthereisnoone-size-fits-allapproachinChina’ssoft
powerplaybook.Beijingtailorsitsstrategytolocalconditionsinhostcountries,
whileaccountingforrecentcompetitivegainsandlosses.However,policymakers
andanalystsstilllackaclearunderstandingofhowChinaadaptsitsaidand
creditallocationbasedonlocalcontextualfactors.Togaingreaterinsight,we
turnourattentiontothreecasestudies:Zambia(atoss-upcountry,leaning
towardChinaduringtheearlyBRIperiodonthepublicopinionmeasure),
Bangladesh(atoss-upcountry,leaningtowardChinaduringtheearlyBRIperiod
onthemediasentimentmeasure),andArgentina(asafebetcountryduringthe
earlyBRIperiodontheelitesupportmeasure).332
332Foreachsoftpowermeasure,weselectedacountrythatfellwithinacategory(safebet,toss-up,andmoonshot)thatChinaprioritizedduringthelateBRIperiod.Tomaximizegeneralizability,wealsoselectedcountriesfromdifferentregionsandwithvaryinglevelsofeconomicdevelopment.331AccordingtoTable4.2,thelargestpositivedelta(+22pp)betweenexpectedandobservedChinesedevelopmentfinanceallocationstocountriesisobservedinthesafebetscategory.330Duringthistime,PrimeMinisterSheikhHasinahaspresidedoverbothstellareconomicdevelopmentperformanceandgrowingauthoritarianism.Mostofthelate-BRIfinancinghasfocusedonlarge-scaleinfrastructureprojectsrelatedtotransportationandenergy.
197
Zambiacasestudy:Fromreputationalassetstoreputational
liabilities
IntermsofpublicopinionchangesduringtheearlyBRIperiod,Zambiafallsinto
thetoss-up(learningChina)categoryofcountriesthatmovedclosertoChina
thantheU.S.,butnotenoughtobeclassifiedasasafebet.DuringtheearlyBRI
period(2014-2017),China’saverageannualdevelopmentcommitmentsto
Zambiajumpedto$1.93billion,ascomparedtoonly$230millionduringthe
pre-BRIperiod(2000-2013).WeinvestigatethefactorsthatshapedChina’s
engagementwithZambiaduringtheearlyandlateBRIperiodsinthiscase
study.
Vyingforinfluence:ThelimitsofChinesesoftpowerinZambia
OnherMarch2023triptoZambiaasU.S.VicePresident(VP),whereshelast
visitedasachildtoseeherIndiangrandfather,KamalaHarrisdeclaredher
administration’ssincereandsteadfastcommitmenttoAfrica.“Ourpresencehere
isnotaboutChina,”shesaid.“It’saboutanindependentunderstandingofthe
intertwinedhistoriesofournations,andourmutualcommitmenttodemocratic
principles,andarecognitionandunderstandingofwhatitmeanstoengagein
smartinvestmentsandthepotentialforthefutureoftheentireglobe.”Asproof,
sheannouncedover$7billionincommitmentsfromtheU.S.privatesectorto
variousAfricancountries.Mostoftheinvestmentsfocusedonadvanced
technologies,suchasartificialintelligence(AI)forimprovedweatherprediction,
electricvehicleassemblyplants,andbiofertilizerfacilities(WhiteHouse2023).
IntheearlydaysoftheBidenadministration,aneffortwassetinmotionto
launchamajorcounter-initiativetoChina’sBRI(SangerandLandler2021).Asthe
WhiteHousewasironingoutthedetailsofPGII(rebrandedfromBuildBack
BetterWorld),Africa’slukewarmcondemnationoftheRussianinvasionof
UkraineacceleratedU.S.effortstosecuremoresupportfromAfricanleaders.In
responsetoaMarch2022UNGeneralAssemblyResolution(GA/12407)that
condemnedRussianaggression,only51%ofAfricancountriesvotedwith
Washington(ascomparedto81%fortherestoftheworld).Asignificantnumber
198
ofAfricancountriesabstained,asChinadid,ordecidednottovoteatall(White
andHoltz2022).
Keentoresetforeignrelationsafterhispredecessor'sChinese-backed
infrastructurespendingspree,Zambia’sHichilemaadministrationdecidedto
breakawayfromitsneighbors—ZimbabweandSouthAfrica—andvoteinfavor
oftheUNGAresolution(ObeandVandome2022).Weekslater,Hichilemawas
rewardedwithaninvitationtotheWhiteHousetomeetVPHarris,thefirstbya
Zambianleaderinnearlythreedecades.AtthismeetingandtheDecember
2022U.S.-AfricaLeadersSummit,U.S.officialsdidnotputChinaontheagenda,
focusinginsteadonmessagingaroundtheirofferingstoAfrica.Bycontrast,
duringtheprevious(Trump)administration,seniorofficialsdirectlycriticized
China’sactivitiesinZambia.InDecember2018,nationalsecurityadvisorJohn
Boltonclaimedthat“ChinaisnowpoisedtotakeoverZambia’snationalpower
andutilitycompanyinordertocollectonZambia’sfinancialobligations”(Bolton
2019).Ayearlater,speakingataUNmeetingforAfricanleaders,Secretary
Pompeoadvisedthat“[c]ountriesshouldbewaryofauthoritarianregimesand
theiremptypromises,”because“[t]heybreedcorruption,dependencyand
instability,notprosperity,sovereigntyandprogress”(Paravicini2020).333
ThoughtheBidenadministrationmovedawayfromitsprecedessor’srhetoric,
China’sgrowinginfluenceinAfrica,particularlyinmineral-richregionssuchas
Zambia’scopperbelt,remainedapointofconcernamongU.S.government
officials,anditloomedoverVPHarris’trip.AtthenewsconferencewithVP
HarrisinLusaka,PresidentHichilemaclarifiedhispositionontheU.S.-China
strategiccompetition.“WhenI'minWashington,I'mnotagainstBeijing,”he
said.“AndwhenI'minBeijing,I'mnotagainstWashington”(WhiteHouse
2023).SoonafterhisAugust2021landslidevictoryoverincumbentZambian
PresidentEdgarLungu,Hichilemastruckacordialtonewithbothpowers,
seekingimprovedtieswitheach(ObeandVandome2022).Buthelater
discovered—duringmonthsofcontentiousrestructuringnegotiationswith
creditorsfromtheParisClub,China,andprivatebondholdersfollowingthe
333ManyAfricanleaderscomplainedaboutAmericanneglect,indifference,orevendisdainduringtheTrumpadministration.
199
country’sNovember2020sovereigndefault—thatresolvingZambia’sdebtissues
wouldrequiredelicatemanagementofU.S.-Chinatensions.
Fromasoftpowerperspective,Beijingencounteredthelimitationsofits
state-ledoverseasdevelopmentfinancingmodelbeforeHichilemacameto
power.DespitewarningsfromtheInternationalMonetaryFundthat“thepaceof
borrowingneedstobesloweddownsignificantly,”thePatrioticFront(PF)
presidenciesofMichaelSataandEdwardLunguoversawaperiodof“profligacy
andtheftoftherulingelite”thatincreasedpublicdebtfrom32%ofGDPin
2012to120%by2020(TheEconomist2020).Chinesestate-ownedcreditors
playedamajorroleinbankrollingthecountry’sborrowingbingeforbig-ticket
infrastructureprojects.From2000to2011,China’saverageannuallending
commitmentstoZambiaamountedtoonly$178.3million.However,between
2012and2018(duringtheSataandLunguadministrations),itsaverageannual
lendingcommitmentssoaredto$932million.
Effortstoconstruct,expand,andrehabilitatelarge-scale
infrastructure—includingroads,airports,powerplants,transmissionlines,and
waterandsanitationsystems—withChinesecreditcontinueduntillate2018,
whenthecountry’sdebt-to-GDPratiosurpassed90%.Thisdevelopment
triggeredalarmaroundtheworldandparticularlyinBeijingbecauseofitshigh
levelofexposuretonon-performingloans.Chinaalsowitnessedasharpdecline
initspublicapprovalratingonceitturnedoffthecredittap:whereas79.5%of
ZambiansapprovedoftheChinesegovernmentin2017,thisfiguredroppedby
12.9percentagepoints(to66.6%)in2021.334
Therefore,animportantlessonfromthiscasestudyisthatBeijing’sstrategyof
buyingsoftpowerthroughthesupplyofcredit-drivendevelopmentprojectshas
limitations.ThisislikelytrueofanyBRIparticipantcountry,buttherisksare
especiallyhighinpoorly-governedcountrieswithlonghistoriesofcorruption
andfinancialmismanagement.Beijing’sexperienceinZambiaalsocallsattention
tothefactthatitsrivalsarewell-positionedtomakeinroadswithgoverning
elitesandthegeneralpublicinsuchcountries.
334ApprovalratingsaretakenfromtheGallupWorldPoll.China’slendingcommitmentstoZambiain2019and2020amountedtoonly$143.7million.
200
WhyisZambiarelevanttogreatpowercompetition?
China’sdevelopmentfinanceportfolioinZambiaisamongitsmostwell-known
globally,duetothecountry’swidely-reporteddebtcrisis.Butthereismuch
moretothestoryofChinainZambiathanthelatter’sdebtrepaymentstruggles.
Withtheworld’s7thlargestcopperreserve,amineralthatisneededfor
advancedtechnologiestofuelthegreenenergytransition,Zambiaisa
high-valuetargetinthegreatpowercompetitionbetweenChinaandtheU.S.
(S&PGlobal2022).
WithinyearsofdeclaringindependencefromBritain,Zambiabecamehometo
China’sfirst-everforeignassistancemegaproject,theiconicTAZARARailwaythat
linksZambia’scopperbeltheartlandtoTanzania’sportofDaresSalaam(Dreher
etal.2022).Today,ChinaisZambia’slargesttradingpartner,responsiblefor
buyingover75%ofitsmainexport,copper(TheGrowthLabatHarvard
University2023).ChinaisalsooneofZambia’slargestforeigncreditorsand
donors.Between2000and2021,itissuedgrantsandloanstoZambiaworth
$13.7billion.Beijingfinanced$10.03billionininfrastructureprojectsduringthe
BRIera(2014-2021),includingmajorintercityandurbanhighways,hydropower
stationsandaccompanyingtransmissionlines,internationalairports,
communicationssystems,hospitals,universities,coal-firedpowerplants,and
watersupplysystems.335
335China’sfootprintinZambia’shealthcaresystemgoesbeyondhospitalsandequipment,withmedicalteamsfromHunanprovinceservingon22toursinZambiasince1978.
201
Figure
4.6:
Official
financial
flows
from
China
to
Zambia,
2014-2021Figure4.6
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(as
describedinSectionA-2oftheAppendix).
202
Figure
4.7:
China
versus
the
U.S.
in
Zambia:
Public
approval
ratesFigure4.7
GivenChina’sextensiveengagementswithZambiaformorethanseven
decades,itisnotsurprisingthat72%ofZambiansapprovedoftheChinese
governmentduringtheBRIera(seeFigure4.7above).However,Beijingenjoyed
onlyatwopercentagepointadvantageoveritsprimarycompetitor
(Washington)duringthisperiod,despitethefactthatChinesedevelopment
financecommitments($17.47billion)dwarfedU.S.developmentfinance
commitments($3.40billion).DuringtheBRIera,U.S.developmentfinance
remainedrelativelystable336—bothinvolumeandcomposition—butWashington
sawitspublicapprovalratinginZambiasoarthroughoutearlyandlateBRI
periods,from63.8%in2014to84.7%in2021(seeFigure4.7}).337China,onthe
otherhand,sawitspublicapprovalratingsteadilydeclineduringthelateBRI
period.
337GiventherelativestabilityofaidflowsfromWashington,ZambianpublicopinionappearstobeprimarilydrivenbychangesintheWhiteHouse.Afterpeakingat92.5%duringthefirstyearoftheObamaadministration,itsubsequentlyfelltojust66.7%in2017afterTrumptookoffice,inlinewiththe“Trumpeffect”thathurtU.S.popularityglobally.336Withanaverageannualallocationofonly$423million,U.S.developmentfinancerangedfromalowof$399millionin2019toapeakof$548millionin2017.
203
SupplementarydatafromAfrobarometersuggestthat,duringthelateBRI
period,Zambiansre-evaluatedthewisdomoffollowingChina’sstate-centricand
infrastructure-focuseddevelopmentmodelasanalternativetotheWest’srights,
freedoms,andmarketinstitutions-basedapproach.WhenAfrobarometer
surveyedZambiansin2014-2015aboutthecountrytheyconsideredtobethe
“bestmodelfordevelopment,”32%choseChinaand23%pickedtheU.S.
However,by2019-2021,whenthelinkbetweenpoorly-executedChinese
megaprojectsandZambians’economicwoeshadbeenlaidbare,Beijing’s
nine-pointadvantagehadflippedinfavorofWashington,withonly22%
respondentsexpressingapreferenceforChina’smodeland31%choosingthe
U.S.model.ThisrepresentedareversaloffortuneforChinaattheendofits
decade-longpartnershipwiththerulingparty(thePatrioticFund,orPF),which
startedwithPresidentSata’sMay2013Beijingvisit.
ThePatrioticFund’sinfrastructurebinge
ThePF’srelationshipwithBeijingbeganinoutrighthostilityduringthe
unsuccessful2006presidentialcampaignofafieryformertradeunionist,
MichaelSata,whoseabrasivepoliticsearnedhimthenickname“KingCobra”
(KimenyiandCopley2014).Byorientinghiscampaignaroundanewagendaof
Sino-skepticism,heralliedsupportagainstthe15-yearruleoftheMovementfor
Multi-PartyDemocracy(MMD),arguingthatChineseinvestorswere“comingjust
toinvadeandexploitAfrica”(French2011).Atonepoint,heevensuggested
that,ifelected,hewouldrecognizetheRepublicofChina(Taiwan)asa
“sovereignstate”(Mupuchi2006).China’sAmbassadorinLusakarespondedby
threateningthatChineseinvestmentwouldbe“putonhold[...]untilthe
uncertaintysurroundingourbilateralrelationshipwithZambiaiscleared”—that
is,untilSata’selectorallosswasconfirmed(Shacinda2006).Aftertwo
unsuccessfulpresidentialbids,Satarelaxedhisrhetoricduringasuccessful2011
presidentialcampaignbystatinghisappreciationforthebenefitsofChinese
investmentinZambiawhileinsistingthatforeignminingcompaniesabideby
locallawstoprotectworkers(Mfula2011;Shukla2021).
Sata’sApril2013visittoChinaresetSino-Zambianrelations.Accordingtothe
3.0versionofAidData’sGCDFdataset,Beijingissuedgrantandloan
204
commitmentsworth$810.2millionforninenewprojectsin2013.Inthesame
year,itstarteddeliveringfiveadditionalprojectsworth$111.6million,andit
completeddeliveryofa$7.4-millionequipmentsupplyprojectfortheupkeepof
theTAZARArailway.UnliketheMMD-era,whenChina’slargestfinancial
commitmentswereearmarkedforsportsstadiums,thefocusshiftedtoward
infrastructureprojectsostensiblydesignedtogenerateeconomicgrowth.This
typeofengagementcontinueduntilSata’sdeathinOctober2014.Overthe
courseoftheSatapresidency,Beijingissuedgrantandloancommitmentsworth
$1.6billionforhigh-profileprojects,includingupgradestotheLusakaairport,
improvementstotheLusiwasihydroelectricpowerplant,theconstructionofnew
powertransmissionlines,androadrehabilitationactivities.Foracountrywhere
lessthanhalfthepopulationhaselectricityaccessandpavedroaddensityis
amongtheworld’slowest,China’swillingnesstobankrollinfrastructureprojects
withsignificanteconomicgrowthpotentialwasawelcomedeparturefromthe
OECD-DAC’semphasisonhealth,education,andgovernanceprojects.338
AfterSata’sdeath,hisMinisterofJusticeandDefenceEdwardLunguwona
highlycontestedelectioninJanuary2015tocompletetheremainderofhis
term.Anticipatinganelectionthefollowingyear,LunguvisitedChinainMarch
2015toacceleratethemomentumbehindnewinfrastructureprojects.Intherun
uptoLungu’sAugust2016re-election,Chinesegrantandloancommitments
tripledfrom$910millionin2015to$2.9billionin2016.Asinprioryears,the
bulkofthemoneysupportedphysicalinfrastructure,includingmoreairports,
roads,andwatersupplysystems.However,thistimearound,Chinaalso
committed$468millionforICT(informationandcommunicationstechnology)
sectoractivities,includingtheinstallationofCCTVcamerasandsmartcity
projects.Thesetypesof“AI-enabling”toolsarereservedforBeijing’sclosest
allies,especiallyincountrieswherepoliticalsystemsareturningmore
authoritarian,suchasinKenyaandPakistanwheretheyhavebeendeployed
againstpoliticalopponents(Boueyetal.forthcoming).
338AccordingtotheCIA’sWorldFactbook,withanareaof752,618squarekilometers,Zambiaistheworld’s40thlargestcounty,butitsroadnetworkof67,671kilometersisthe70thmostextensiveintheworld.However,only14,888kilometers,or28%,ofZambia’sroadsarepaved.Thismeansthatitspavedroaddensityisonly0.02kilometerspersquarekilometer,significantlylowerthanotherdevelopingcountrieslikeIndia(0.7)andBrazil(0.6).
205
BeijingdoubleddownonitssupportforZambiaduringtheLungu
administration—perhapsbecauseitthoughtthatastableandpro-China
governmentcouldserveasadefactoinsurerforitsprojectandinvestments.
China’spopularitysoaredduringthisperiod.PublicapprovalforChina,as
measuredbytheGallupWorldPoll,rosefrom66.4%in2015toitshighestever
levelof79.5%in2017.
Afternarrowlywinningre-electionwith50.4%ofthevoteinAugust2016,Lungu
wasquicktoannouncethat“PresidentXiJinpinghasexpressedconfidencein
myleadership.”Hereiteratedhisresolvetofurtherdeepenhisengagements
withChinato“empowerbusinessmenandbenefiteveryZambian”(Shaban
2016).Inthefollowingyear,Chinesegrantandloancommitmentsamountedto
$3.1billion,halfofwhichwasaloanforthe750MWKafueGorgehydropower
project.
Then,inSeptember2018,LunguvisitedChinatoattendtheForumon
China-AfricaCooperation(FOCAC)summitwhereheformallysigneduphis
countrytoparticipateintheBRI.Beijingrespondedbyprovidinganadditional
$2.9billionofaidandcreditforroads($1.4billion),powertransmissionlines
($499million),cementplants($445million),publichousing($229million),and
universityconstruction($208million).339
Whenreputationalassetsbecomeliabilities
AsChinesestate-ownedentitieslentrecordamountstoZambia,theIMF
soundedthealarmaboutthecompositionandsustainabilityofZambia’sexternal
debtburden(InternationalMonetaryFund2017).Between2011and2016,
publicandpubliclyguaranteed(PPG)externaldebtskyrocketedfrom8.4%of
GDPto36.5%.Theshareofcentralgovernmentdebtowedtoprivate
creditors—whichwasprovidedonmostlynon-concessionalterms—increasedto
50%.TheshareofcentralgovernmentdebtowedtoMDBs—whichwas
providedonmostlyconcessionalterms—shrankfrom60%to20.5%.TheIMF
warnedthatthe“paceofborrowingneedstobeslowedsignificantly”toalign
339Throughoutthisperiodofhighinfrastructureactivity,China’ssignaturepublichealthengagementscontinuedthroughmedicalteamvisits,equipmentandmedicalsuppliesdonations,andinterventionstocombatinfectiousdiseases(Dolanetal.2023).
206
withthecountry’s“absorptivecapacity”throughmorestringentstandardsfor
“theselection,procurement,andmonitoringofinfrastructureprojects.”
(InternationalMonetaryFund2017).DespitethesewarningsandZambia’spast
experienceswithdebtdistress,theZambiangovernmentanditsexternal
creditorsfailedtopumpthebrakes(InternationalMonetaryFund2023).
ReportsemergedthatZambiangovernmentagenciespressuredtheMinistryof
Financetosign-offonpoliticallyadvantageousbutcommerciallynon-viable
projects(Mutati2016).Inoneinstance,aseniorgovernmentofficialdealingwith
theMbala-Nakonderoadallegedthat“the[Chinese]contractorhadalready
securedtheagreementwiththeMinistryofFinancewhentheycametous.We
justnegotiatedaboutthedetailsofdesign,nottheamountoftheloan”(Lee
2018).Zambia’srushtolaunchasmanyhighvisibilityprojectsaspossiblewasa
functionofpoliticaleconomyconsiderations.Thinelectoralvictorymarginsin
thecountry’spresidentialelections“createpowerfulincentivestousepublic
workstorewardconstituencies”(Brautigam2021).
AreviewofChina’sODAandOOFportfolioinZambiarevealsthat19outof21
projectsidentifiedashavingESGchallengesfacedgovernance-related
concerns,suchasoverpricing,corruption,andfinancialmismanagement(see
TableA12).340InJuly2019,then-oppositionleaderandfuturepresident
HakaindeHichilmaclaimedthat“thedebtthatwasacquiredforjustoneproject,
Lusaka-Ndola(DualCarriageway)road,whichshouldcostUS$400million[...]is
costingUS$1.2billion.”HearguedthatZambians“aregiving[projects]toChina
inthecorruptway,whichiscostingustoomuchoftaxpayers’money”(Ncube
2019).341In2022,Dr.MbitaChitala,theformerchairmanoftheBoardof
DirectorsofZESCO(Zambia’sstate-ownedpowerutility),publishedabook
entitledCorporateCapture:ThePoliticalEconomyofElectricityManagementin
Zambia2014-2021(HowNottoManageaStateEnterprise).Init,heclaimed
thatlargesegmentsoftheChinaEximbank-financedSmartZambiaNationalICT
DevelopmentProjectwerecompletelyunnecessaryanddesignedtoenrich
seniorgovernmentofficials.Herecalledthat“thewholeprojectwasa
341TheSatagovernmentalsolaunchedaninvestigationofthe750MWKafueGorgehydropowerprojecttodetermineifadequatetransparencyandfinancialcontrolswereinplacetomonitorperformance.340Between2000and2021,theESGriskprevalencerateinChina’sgrant-andloan-financedinfrastructureprojectportfoliowas56%inZambia(seeTableA12).
207
conspiracytodefraudZambia”becauseHuaweiwasproposingsolutionsthat
were“inferiortoZESCOexistingnetwork”andlocatedintownsthatwere
alreadyserved(Chitala2022).342
Bylate2018,theZambiangovernmentwascash-strappedandBeijing’s
developmentprojectportfoliobegantosuffer.TheZambiangovernmentwas
increasinglyunabletomeetitsfinancialcommitmentstoChinesecontractors
andcreditors.Forexample,afterCDBissueda$469-millionloaninDecember
2015,theCopperbeltUrbanRoadsProjectfacedaseriesofimplementation
delaysuntilthestate-ownedengineering,procurement,andconstruction(EPC)
contractor,ChinaHenanInternationalCooperationGroupCompany(CHICO),
abandonedtheprojectduetoinsufficientpaymentsbytheZambian
government.The750MWKafueGorgehydropowerproject,thecountry’ssingle
largestenergysectorproject,facedsevereimplementationchallengeswhenthe
borrower—awhollystate-ownedspecialpurposevehicleknownasKafueGorge
LowerPowerDevelopmentCorporationLimited—defaultedonitsloan
repaymentobligationstoICBCandChinaEximbank.343Inbothcases,senior
governmentofficialsclaimedthatLusakawouldcomeupwiththecashneeded
tocompletetheprojects.The750MWKafueGorgehydropowerprojectwas
notcompleteduntilApril2023—approximately13yearsafterSinohydro,the
EPCcontractor,signedanMOUtocompletetheprojectona
343GiventhattheloanwasinsuredbySinosure,theEPCcontractor(SinohydroCorporationLtd.)soughtcompensationthroughanindemnitypayment.InJanuary2021,SinohydroCorporationLtd.filedaninsuranceclaimwithSinosure,whichinturnmadeanindemnitypayment(worth$57.23million)toSinohydroCorporationLtd.inMay2021.SeeProjectID#92289and#57536inthe3.0versionofAidData’sGCDFdataset.342Dr.ChitalaclaimsinhisbookthatZambia’sSecretarytotheCabinetDr.RolandMsiskaandlaterDr.SimonMiriandhisdeputy,formerDeputyFinanceMinisterandBankofZambiaGovernorPatrickMvunga,placedextraordinarypressureonZESCO(thecountry’sstate-ownedpowerutility)toacceptanon-lendingagreementforPhaseIIoftheSmartZambiaNationalICTDevelopmentProject.Hewritesthat“itwasobviousthatsomepeoplehadeitheralreadybenefitedfromthedealorwerebeingpreventedfrombenefitting.Someofthemembersofthe[ZESCO]managementexpressedgenuinefearofthesituationandrequestedformyprotection.”Healsowritesthat“[t]hePresident[ofZambia]advisedmetocancel[anupcomingZESCOBoard]meetingwhichIgratefullydidasIhadplannedtoresignfromtheBoardiftheBoardmembersresolvedintheaffirmativetoaccepttheon-lentloan.[…]Ihadearliermadeupmymindthatiftheboardacceptedthat[ZESCO]receivetheUSD392millionasonlentto[ZESCO]bytheMinistryofFinanceknowingthat[ZESCOLimitedcouldnotaccommodatetheloanonitsbalancesheetandfurtherknowingthattheloanwasessentiallyawayofdefraudingZambia,IwouldrespectfullyresignaschairmanandboardmemberofZESCOLimited.[…]ItwascommonknowledgethatthemorethanUS$60millionthattheMinistryofFinancereleasedasits15%counterpartcontributionfundingfortheloanwasnotonlywronglyreleasedbutwasalsogrosslyabusedandthatmanypeoplemayhavereceivedbribesonthatdealandhencethepressureonZESCOtoaccepttheUSD392millionon-loanletinstructions.[…]Itwasapparentthatsincegovernmentmoneyhadbeenexpendedwithoutanyworktoshowfor,thiswouldaddtotheotherscandalsthattheAuditorGeneralusedtopublishasexamplesofabuseofpublicofficeoftrustorcommontheftofpublicresourcesbypoliticiansandpublicservants”(Chitala2022).
208
build-operate-transferbasis.TheCopperbeltUrbanRoadsProjectwas
indefinitelysuspended.TheLusaka-Ndola(DualCarriageway)RoadProject
sufferedasimilarfate.
Between2019and2021,newgrantandloancommitmentsfromChinawere
virtuallynon-existent.Afterprovidingaveragecommitmentsworth$3billion
between2016and2018,Beijingsharplyreduceditsaverageannual
commitmentsto$66millionfrom2019to2021.IfChina’sbidtoaccumulatesoft
powerrequiresdemonstratingthatitseconomicdevelopmentmodeliseffective
andsustainable,Zambiarepresentsaspectacularfailure.Projectsthatwere
supposedtobereputationalassets—andspurChinese-ledeconomicgrowthfor
Zambians—becamereputationalliabilities.
Thelimitationsofinfrastructurefinancingforsoftpoweraccumulation
TheZambiancaseillustratesthelimitationsofChina’sstrategyofusing
large-scaleinfrastructureprojectstoexpanditsinfluenceoverseas.Thisoutcome
wasnotcompletelyofChina’smaking.Norwasitinevitable.Rather,itwasdue
tooverborrowingforprojectswhosefeasibilitywasbasedoneitheroptimisticor
“graftridden”projections(TheEconomist2018).Forinstance,thenewterminal
atLusaka’sKennethKaundaAirportwasdesignedtoaccommodate“an
improbableten-foldincreaseinpassengertraffic,”andat$360,000per
kilometer,theaveragepricetagofroadsbuiltbetween2011and2018was
“morethantwicetheAfricanaverage”(TheEconomist2018).HadtheZambian
governmentorChinesecreditorsfollowedmorerobustduediligence
requirements(seerelevantdiscussioninChapters2and3),thesepitfallscould
havebeenminimizedorotherwiseavoided.However,thesefactorsultimately
ledtothedownfalloftheentiremodel,resultinginmassivelossesfortheLungu
administration,China’sreputation,andaboveallelse,thepeopleofZambia.
Goingforward,Beijingmightbeabletoreversethereputationallossesthatit
hassufferedifitfollowsthroughonaJune2023debtreliefcommitmentandif
localeconomicconditionsimprove.Inaddition,severalpreviouslystalled
infrastructureprojects—thosethatweredelayedduetoCOVID-19andfinancial
distress—willsoonreachcompletionandpotentiallyincreasepublicsupportfor
209
China.Beijingcouldalsoconsidersendinghigh-rankingCPCofficialsoreven
organizeastatevisitbyPresidentXitoresetanddeepentieswithLusaka.
However,ascomparedtohispredecessor,PresidentHichilemaappearstobe
playingacarefulbalancingactinhisrelationswithChinaandtheU.S.,andhe
willlikelytreadcarefully.
210
Bangladeshcasestudy:Stabilityinauthoritarianism
Bangladesh,likeZambia,isacompetitivejurisdictionthatweclassifiedasa
toss-up(leaningChina)country.DuringtheearlyBRIperiod,Chinamade
modestmediasentimentgainsthereattheexpenseoftheU.S.Italso
dramaticallyincreasedtheprovisionofaidandcredittoBangladeshduringthe
lateBRIperiod.WhereasaverageannualODAandOOFcommitmentsfrom
ChinatoBangladeshwereonly$994millionbetween2014and2017,thisfigure
soaredto$3.4billionbetween2018and2021.Thisunprecedentedspending
increasecoincidedwithaperiodofpoliticalstability,particularlyafterPrime
MinisterSheikhHasinawonhersecondconsecutiveelectoralvictoryin2018.It
alsocoincidedwithaperiodofrisingauthoritarianism.
China’slateBRIeraforayintoBangladesh
InMay2021,ChineseAmbassadorLiJimingissuedasternwarningto
Bangladesh’selites.InresponsetoreportsthatDhakamightjointhe
QuadrilateralSecurityDialogue(theQuad)—aninformalcoalitionbetweenthe
U.S.,Japan,Australia,andIndia—hesaidthatsuchamovewouldinflict
“substantialdamage”onSino-Banglarelations.Thisuncharacteristicallyblunt
remarkprovokedstrongpushbackfromBangladesh’sForeignMinisterA.K.
AbdulMomen.“Weareanindependentandsovereignstate.Wedecideour
foreignpolicy,”hesaid.HealsoclarifiedthatDhakawasneitherapproachedby
themembersoftheQuadnorwasitsparticipationintheQuadapossibility
underactiveconsideration(Islam2021).
ThisdiplomaticspathighlightsBangladesh’sdesiretodelicatelybalance
relationswiththeU.S.anditsclosestregionalpartner(India),whilebolsteringits
“strategicpartnership”withChina(MinistryofForeignAffairsofthePeople’s
RepublicofChina2016).Bangladeshislocatedinthestrategicallyvital
Indo-Pacificregion,servingasagatewaybetweenSouthAsiaandSoutheast
Asia.ForChina,itisanentrypointtotheBayofBengal,anditsproximityto
MyanmarincreasesitsstrategicvaluetotheU.S.anditsQuadallies.
SincePrimeMinisterSheikhHasina’sassumptionofpowerin2009,Bangladesh
hasmadetremendouseconomicandsocialdevelopmentprogress:extreme
211
povertyhasbeenreducedbyhalf;lifeexpectancyhasrisenby6years;the
averageeconomicgrowthratehasexceeded7%;andpercapitaincomehas
risenby3.5times(UnitedNations2022;Trotsenburg2023).Atthesametime,
SheikhHasinahaseffectivelyturnedBangladeshintoasingle-partysystem,
usheringinaperiodofauthoritarianism.344Between2009and2022,the
country’spositionontheReportersWithoutBorders’MediaFreedomIndexfell
from121to162outof180countries(ReportersWithoutBorders2023).Its
overallperformanceonFreedomHouse’sFreedomintheWorldIndextumbled
byaquarter,withpoliticalrightserodingbynearlyone-third(FreedomHouse
2022,2023).
FortheU.S.anditsallies,Bangladesh’spositiveeconomicdevelopment
trajectory,democraticbacksliding,andgeostrategicpositionpresentsapolicy
conundrum.Bangladesh’sconcentrationofpowerinoneperson,lackof
toleranceforpoliticaldissent,andsingularfocusoninfrastructure-driven,
export-ledeconomicgrowthisstrikinglysimilartoBeijing’sown.Yet,theQuadis
seekingtoexpanditsfootprintintheIndo-Pacificandhasreportedlyreached
outtoBangladeshaboutthepossibilityofaregional“economicpartnership”
(Kishida2023).
BeijingisevidentlyawarethatBangladeshisacompetitivejurisdictionwhereit
cannotaffordtorestonitslaurels.Inarecentlypublishedop-ed,theChinese
AmbassadortoBangladeshannouncedthat“Chinaissettobringmore
developmentopportunitiestoBangladesh,”predictingthatthecountry’s
exportstoChinawouldgrowsignificantlyafterBeijingannouncedduty-free
marketaccesstomostBangladeshiproducts(Wen2023).Inotherstatements,
hehasaccusedtheU.S.ofpressuringDhakatodeviatefromitslongstanding
positionofneutralityingreatpowercompetitions.InApril2023,the
BangladeshiMinistryofForeignAffairsreleaseditsfirst-ever“Indo-Pacific
Outlook,”restatingthe“politicaldictum”ofthecountry’sfoundingfatherSheikh
344Theriseofauthoritarianism,whichbeganduringSheikhHasina’ssecondterm(2009-2014)andwasreinforcedduringherthirdterm(2014-2019)withtheDigitalSecurityActof2018,hascontinuedduringherfourthterm(2019-present).Theoppositionparty(BNP),whichaccusesthegovernmentofusingthejudicialsystemtosilenceitscriticsandhatchingplanstorigtheJanuary2024election,claimsthatthefutureofmulti-partydemocracyisinjeopardy(AmnestyInternational2023;Paul2023).
212
MujiburRehman(‘Friendshiptowardall,malicetowardnone”)asaguiding
principle(MinistryofForeignAffairsofBangladesh2023).
Aswepreviouslynoted,Bangladeshwasselectedasacasestudycountry
becauseitfellinthetoss-up(leaningChina)categoryduringtheearlyBRIperiod
(accordingtothemediasentimentanalysisdescribedinSection2).However,on
multipledimensionsofsoftpower,Bangladeshremainsatoss-upcountry,
leaningtowardneitherChinanortheU.S..Bothpowers—aswellasIndia—are
activelyvyingforinfluencethere(Asmusetal.forthcoming).Beijinghas
respondedtoitsrecentsoftpowergainsinBangladeshbydoublingdownon
theprovisionofaidandcredit.Thisoutcomewaslikelymadepossiblebyseveral
factorsrelatedtoBangladesh’sdomesticpoliticalandeconomicenvironment:a
dominantsinglepartygovernmentin-chargeofacentralizedgovernance
system,astrongfocusonrapiddeliveryofhighlyvisiblephysicalinfrastructure
projects,andproactivedebtmanagement.
SheikhHasina’sriseandtheendofpoliticalinstability
WithinmonthsofindependencefromPakistanin1971,therulingAwamiLeague
(AL)partystartedmovingBangladesh’snascentpoliticalsystemtowarda
single-partysystem.Butacombinationoffactors—includingeconomic
mismanagement,risingpoliticalrepression,andgrowingconcernaboutthe
country’sfuture—ledtomultiplemilitarycoupsanddictatorshipsfrom1975to
1990(Riaz2016).Afterareturntodemocracy,theperiodfrom1991to2008was
marredbypoliticalcrises,dysfunctionalgovernance,andfrequentviolent
clashesbetweensupportersofSheikhHasina’sALandKhaledaZia’sBangladesh
NationalistParty(BNP)(Mannan2018).
Akeysourceofinfightingwasthebitterhistoryofanimositybetweenthetwo
leaders,originatinginthe1975coupd’etatagainstthegovernmentofSheikh
Hasina’sfather,SheikhMujiburRehman(thecountry’sfounder).Militaryofficers
notonlyassassinatedSheikhMujib,butalsoallthreeofSheikhHasina’sbrothers,
twoofhersisters-in-law,andhermother.Afterthecoup,KhaledaZia’shusband
GeneralZiaurRahmanemergedasakeymilitaryandpoliticalleader,eventually
becomingpresidenttwoyearslaterandformingtheBNP,whichheleduntilhis
213
assassinationin1981.Thetremorsfromtheseeventsstillreverberatein
Bangladesh’spolitics,asALseekstofulfillSheikhMujib’sdesiretorunthe
countryasasingle-partydemocracy.
Between1991and2006,ZiaservedtwotermsinofficeandHasinaservedone
terminoffice.Thiserarepresentedanelectorallycompetitiveperiod,albeitone
markedbyfrequentpoliticalandconstitutionalcrises(Vaughn2008).Then,prior
totheelectionsscheduledforearly2007,themilitaryintervenedandappointed
atechnocraticinterimgovernmenttooverseefreshelections.Thesewere
eventuallyheldinDecember2008,leadingtoSheikhHasina’sALreturnto
power,whichshehasheldontoeversince.Shehassteadilysolidifiedhergrip
onthepoliticalsystem,stiflingoppositionforcesandretainingpowerthrough
heavy-handedauthoritarianmeasurestocontrolthemediaandjudiciary(The
Economist2023).
Thesoftpowerbenefitsofdevelopmentfinance
BetweenMarch1971(whenthecountrygaineditsindependence)and
December2008,Bangladesh’slargestforeignaidandcreditproviderswere
Japan,theU.S.,andtheU.K.Chinadidnotmaintainanespeciallylarge
developmentfinanceprogramduringthisperiod.However,duringtheBRIera,
Beijingissuedgrantandloancommitmentsworth$17.5billion,makingit
Dhaka’ssinglelargestdevelopmentpartner.In2006,ChinaalsosurpassedIndia
andtheU.S.tobecomeBangladesh’slargesttradingpartner,with$4.7billionin
totalflows.Onemightassumethatthecombinationofhightradevolumesand
unprecedentedlevelsofaidandcreditforhighlyvisibleinfrastructureprojects
ledtorapidsoftpoweraccumulationforChina.However,inreality,the
relationshipbetweendevelopmentfinanceandsoftpowerhasnotbeen
straightforwardforseveralreasons.
214
Figure
4.8:
Official
financial
flows
from
China
to
Bangladesh,
2014-2021Figure4.8
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(as
describedinSectionA-2oftheAppendix).
Accordingtothe3.0versionofAidData’sGCDFdataset,Chinaissuedgrantand
loancommitmentsworth$20.8billionfor138projectsinBangladeshbetween
2000and2021(seeFigure4.8).345ThatmadeBangladeshoneofChina’s20
largestaidandcreditrecipientsinthedevelopingworldandits7thlargest
recipientinAsia.China’sODAandOOFcommitmentstothecountrysoared
from$3.3billionduringthepre-BRIperiod(2000-2013)to$17.5billionduring
theearlyandlateBRIperiod(2014-2021).DuringtheBRIera,theproject
portfoliowaslargelyfocusedontheconstructionandrehabilitationofpower
plants,transmissionlines,highways,andbridges.
345China’sODAandOOFcommitmentstoBangladeshduringthis22-yearperiodfocusedonfoursectors:energy($10.3billion),transport($6.3billion),industrialdevelopment($1.9billion),andICTinfrastructure($1billion).Beijingalsosupportedalargenumberofeducation,health,andemergencyresponseprojects(withrelativelysmallfinancialcommitments),particularlyincrisisyearssuchastheCOVID-19period.
215
Figure
4.9:
China
versus
the
U.S.
in
Bangladesh:
Public
approval
ratesFigure4.9
Giventheunprecedentedscaleofthesefinancialcommitmentsandthehighly
visiblenatureoftheinfrastructureprojectsundertakenduringtheBRIera,itis
notsurprisingthatBeijinghasmadesomepublicopinionandmediasentiment
gainssince2014.TheaveragelevelofpublicapprovalfortheChinese
governmentwas49%between2006and2013(seeFigure4.9).Then,during
SheikhHasina’ssecondandthirdconsecutivetermsinoffice(from2014to2018
andfrom2019to2022),Beijingsawitsaveragepublicapprovalratingincrease
to60%and56%,respectively.
TheU.S.andChinagenerallyenjoyedsimilarlevelsofpublicsupportin
BangladeshduringtheBRIera(seeFigure4.9).However,shortlyafterDonald
Trumpcametopower,theU.S.sufferedan18percentagepointdeclineinits
publicapprovalrating(from70%in2016to52%in2017),whichgaveChinaa
10%percentagepointadvantageovertheU.S.in2017.Beijing’sadvantagewas
neverthelessshort-lived.By2022,Washingtonsawitspublicapprovalratingrise
to62%,therebyopeningupan11percentagepointadvantageoverBeijingin
2022.
216
Figure
4.10:
China
versus
the
U.S.
in
Bangladesh:
Media
sentiment
scoresFigure4.10
Notes:ThedataaredrawnandprocessedfromtheGDELT1.0EventDatabase(relatedtogovernment
actorsfrommainlandChinaortheU.S.).SeeBox1binChapter1formoredetails.
Intermsofthecompetitionforfavorablemediacoverage,theGDELTdata
indicatethat,duringSheikhHasina’s2014-2018term,theaveragetoneofmedia
sentimentaboutChina(+1.88)wassignificantlymorepositivethantheaverage
toneofmediasentiment(-0.33)abouttheU.S.(seeFigure4.10).However,
between2019and2022,China’saveragemediasentimentscoredeclinedto
+0.48.346OnepotentialreasonwhymediacoverageaboutChinamayhave
becomelessfavorableduringthelateBRIperiodisthat$10billion(ornearly
38%)ofBeijing’sdevelopmentfinanceportfolioinBangladeshencountered
significantESGproblems(seeTableA12),includingsocialandgovernance
challengesassociatedwiththedesignandimplementationofmultiplecoal-fired
346Withscoresof+4.7and+3,China’slargestmediasentimentadvantageovertheU.S.camein2016and2017.Sincethen(2018-2022),theaveragetoneofmediasentimentaboutChinahasfallento+1.1.ThisfindingisconsistentwithempiricalevidencethatsuggeststhesoftpowerbenefitsofChinesedevelopmentprojectsareshort-lived,withasignificantfadingawayofreputationalbenefitsinthemedium-tolong-term(Wellneretal.forthcoming).
217
powergenerationplants.347TheU.S.alsosawitsaveragemediasentimentscore
deteriorate(to-0.6)duringthe2019-2022period,thoughthedeclinethatit
experiencedwassmallerthantheoneexperiencedbyChina.Asaresult,the
mediasentimentgainsthatBeijingmadeattheexpenseofWashingtonduring
theearlyBRIperiodwerepartiallyreversedduringthelateBRIperiod(see
Figure4.10).
ThefoundationsofBeijing’spartnershipwithSheikhHasina
TherapidriseofChinesedevelopmentfinanceduringtheBRIerawasfacilitated
byBeijing’sstrongpartnershipwithSheikhHasina’sALgovernment,whichgrew
asshestrengthenedhergriponpower.SinceDecember2008,thecentralization
ofpoliticalpowerhasresultedinhigherlevelsofadministrativecertaintyand
createdanenablingenvironmentforBeijingtobankrollandbuildbig-ticket
infrastructureprojects.
Table
4.3:
Development
finance
commitments
from
China
to
Bangladesh
by
executive
administrationTable4.3
Development
finance
commitments
from
China
to
Bangladesh
by
executive
administration
PoliticalAdministrationAnnualaveragedevelopmentfinance(ODAandOOF)fromChinaduringeachadministration(USD2021millions)ODA%fromChina
BNP’sKhaledaZia(2002-2006)17572%
Military-appointedFakhruddinAhmed(2007-2008)71100%
SheikhHasinaI(2009-2013)38147%
SheikhHasinaII(2014-2018)2,19120%
SheikhHasinaIII(2019-present)2,2907%
Notes:ThisfigureprovidesasummaryofChineseODAandOOFcommitmentstoBangladeshduringeach
executiveadministrationinBangladeshsince2002.Exactcommitmentdatesinthe3.0versionofAidData’s
347Between2000and2021,theESGriskprevalencerateinChina’sgrant-andloan-financedinfrastructureprojectportfoliowas59%inBangladesh(seeTableA12).
218
GCDF3.0werecomparedtoelectiondatesandpoliticaltransitiondatestocategorizeeach
ODA/OOF-financedprojectaccordingtothechiefexecutivethatwasinpoweratthetimeofthe
commitment.“ODA%fromChina”displaystheproportionofChina’sdevelopmentportfoliothatis
classifiedasODAandthatwascommittedduringeachexecutiveadministration.AidDatarelieson
OECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(asdescribedinSectionA-2of
theAppendix).
AsTable4.3demonstrates,thevolumeandcompositionofChinese
developmentfinanceinBangladeshhavechangedsignificantlyoverthecourse
ofthelastfiveadministrations.Financialcommitmentamountshavesoared
sincethemilitary-backedinterimgovernmentof2007-2008,buttheproportion
ofChinesedevelopmentfinanceprovidedviaODA(grantsandhighly
concessionalloans)hasfallensharply,whichreflectsBeijing’sgrowingfocuson
larger-scale,revenue-generatingprojects,suchaspowerplantsandtolled
bridges.
Between2000and2008,underthegovernmentofBangladeshNationalParty
(BNP)leaderKhaledaZiaandthemilitary-backedtechnocraticgovernmentof
FakhruddinAhmed,averageannualChinesedevelopmentfinancecommitments
amountedtoonly$175millionand$71million,respectively(seeTable4.3).
Duringthisperiod,82%oftotalcommitments,or$1.2billion,supportedjust
sevenlargeprojects,includingatransmissionline,fertilizerplant,andbridge.
Beijingalsosupportedsmallerprojectstopromotepublicgoodwill,suchas
cyclereliefactivities,scholarshipstostudyinChina,andtheChina-Bangladesh
conventioncenterasamonumentoffriendship.
ThenatureofBeijing’sengagementinBangladeshbegantoshiftduringthelate
yearsofSheikhHasina’ssecondterm(2009-2013),withaverageannual
commitments($381million)morethandoublingthoseissuedtothetwo
previoustwogovernments(seeTable4.3).Withagreaterfocuson“bankable”
infrastructure,theshareofChinesedevelopmentfinanceprovidedviaODA
droppedto47%.Projectsizesalsoincreased.Beijingissuedloansworth$697
millionforafertilizerfactoryand$595millionfor3Gmobileandnational
internetbroadbandnetworkupgrades,powerplants,andanurbanwater
treatmentplant.
219
ButthemostdramaticchangestookplaceafterSheikhHasina’s2013visitto
Beijing,atwhichtimeamajornewfocusonpowergenerationandtransport
connectivitywasagreeduponatthehighestlevels.Withgreaterpolitical
stabilityandconfidenceinherpoliticallongevity,Beijingagreedtogreen-lighta
setoftransformativeinvestmentsthatcouldhelpSheikhHasinameetthe
ambitiousgoalsshesetforexport-ledeconomicgrowthasameansoflifting
millionsoutofpoverty.China’sowndevelopmentmodelalignedwithhervision,
whichmayexplainwhyitsaverageannualdevelopmentfinancecommitment
grewbyafactorofseven(to$2.2billion)between2014and2018(seeTable
4.3).
TheconventionalwisdomisthatChinadramaticallyscaledbackitsoverseas
lendingcommitmentsduringthelateBRIperiodduetotheCOVID-19
pandemicandrisingconcernsaboutdebtsustainability(seeChapter2).
However,BeijingbehaviorinBangladeshbeliesthisclaim.Onaverage,between
2019and2021,SheikhHasina’sgovernmentaccepted$2.3billionperyearin
newODAandOOFcommitmentsfromChina,whichenabledtherapidprogress
(orcompletion)ofhigh-profileprojectssuchasthePadmaBridge,theDhaka
elevatedexpressway,andseveralpowerplants(seeTable4.3andTable4.4).
FromSheikhHasina’sperspective,thesearearguablythemostpolitically
advantageousundertakingssupportedbyChina:PadmaBridgesitswithinher
homedistrictandpoliticalconstituency;theDhakaexpresswayimprovesurban
mobilityforthecountry’spoliticalandbusinesselite;andnewpowerplants
couldeliminatecripplingpowershortagesaffectingallBangladeshis.348
Thebestofallworlds:Single-partygovernance,proactivedebt
management,andinsatiableinfrastructuredemand
Despitewidespreadinternationalcondemnationofgrowingauthoritarianismand
allegedhumanrightsabusesduringher15-yearterminoffice,Asia’s“ironlady”
ispoisedtofulfillherlatefather’svisionofturningBangladeshintoaone-party
state(TheEconomist2023).Duringrecentmediainteractions,SheikhHasinahas
claimedthatherALpartyistheonlylegitimatepoliticalforceinthecountrythat
348InChapter3,wediscussthegovernancerisksassociatedwithsitingdevelopmentprojectsinthehomedistrictsofpoliticalleaders.AlsoseeDreheretal.(2019,2022).
220
shouldbeallowedtocontestelectionsbecausethemainoppositionBNPwas
createdbyerstwhilemilitarydictators.Hercriticsarguethat,underherwatch,
Bangladesh’seconomicprogresshasbecomesolelyreliantonnepotistictextile
industrialists,andherteamisoutoffreshideastotackledeeprootedstructural
problems,suchascorruption.349Irrespectiveofthesechallenges,withmajor
oppositionpartiesindisarrayanddissentingvoicesincivilsocietylargelysilent,
todaysheispredictedtowintheJanuary2024election.350
Table
4.4:
Chinese
ODA
and
OOF
infrastructure
project
milestones
during
the
second
half
of
2018
in
BangladeshTable4.4
Chinese
ODA
and
OOF
infrastructure
project
milestones
during
the
second
half
of
2018
in
Bangladesh
ProjectCommitmentAmount(USD2021millions)ActivityTypeMonth
PadmaBridgeRailLink2,900ImplementationStartedJuly2018
DasherkandiSewageTreatmentPlant318ImplementationStartedAugust2018
PayraCoal-FiredPowerPlant2,200ImplementationStartedOctober2018
Bangladesh-ChinaFriendshipBridge94ImplementationStartedNovember2018
BanskhaliCoal-FiredPowerPlant1,300FinancialCommitmentDecember2018
Notes:ThistableidentifiesChineseODA-andOOF-financedinfrastructureprojectmilestonesinthe
6-monthperiodprecedingtheDecember2018nationalelectionsinBangladesh.Thefinancialcommitment
amountsonlyrepresentthosefromofficialsectorinstitutionsinChina.AidDatareliesonOECD-DAC
measurementcriteriatomakeODAandOOFdeterminations(asdescribedinSectionA-2ofthe
Appendix).
Previousresearchhasdemonstratedthatelectionsprovidepowerfulincentives
forpoliticianstoshowcasetheirperformance,andhighlyvisiblephysical
infrastructureprojectsareidealforachievingthisgoal(Marx2018;Anaxagorou
350AccordingtoapollundertakenbytheInternationalRepublicanInstitute’s(IRI)CenterforInsightsinSurveyResearch(CISR)betweenMarch1,2023andApril6,2023,SheikhHasinaenjoysa70%publicapprovalrating.However,approvaloftheoppositionincreasedfrom36%inSeptember2019to63%inMarch/April2023(CISR2023).349Bangladesh’sperceivedlevelofcorruptionisamongtheworstinSouthAsia(TransparencyInternational2021).
221
etal.2020).InZambia,Brautigam(2022)arguesthatthecompetitivenessof
presidentialelectionsheightensthisdesire,butinBangladeshthestakesappear
tobehigheveninrelativelyuncompetitiveelections,suchasin2018whenthe
mainoppositionpartyBNPboycottedthem.Intherun-uptotheDecember
2018election,Beijingalsoseemedtomakespecialeffortstoensurethatits
favoredcandidate,SheikhHasina,wasinthebestpossiblepositiontowin
reelection.351
Inthesix-monthperiodleadinguptotheDecember2018election,Beijing
approved,startedimplementationon,orcompletedprojectsbackedbygrant
andloancommitmentsworth$6.8billion(seeTable4.4).Oneoftheseprojects
includedthe1,320MWBanskhalicoal-firedpowerplant,whichwasfinancedby
fiveChinesecreditors(AgriculturalBankofChina,ChinaDevelopmentBank,
ChinaConstructionBank,andBankofChina,ChinaMinshengBank)andone
localcreditor(RupaliBankLimited)througha$1.78billionsyndicatedloan.
Duringthesamesix-monthperiod,Chinesestate-ownedcompaniesbeganthe
implementationofsixadditionalprojectsbackedbyfinancialcommitments
worth$6.5billion,includingthe1,320MWPayracoal-firedpowerplantandthe
high-profilePadmabridgeraillinkproject.Beijingkeptupthemomentumafter
SheikhHasinawonre-election.Duringthefirstsixmonthsofhernewtermin
office,Chinesedonorsandlendersgreen-litnewprojectsworth$1billion;
Chinesefirmsstartedimplementingnewprojectsbackedbyofficialfinancial
commitmentsworth$1.2billion;andtwoprojects(backedbyChineseloans
worth$300million)werecompleted(seeTable4.4).Asinprioryears,the
projectswereinfrastructuralinnature(forelectricityproductionandnational
internetconnectivity).
SimilartoitsrelationshipwithEdwardLunguinZambiaandtheKirchnersin
Argentina,Beijing’spartnershipwithSheikhHasinaappearstoreflecta
preferenceforworkingwithincumbentswhohavegoodreelection
prospects—orpoliticallongevityforotherreasons.Fromareturn-on-investment
standpoint,thispreferencemayreflectthefactthatChineseinfrastructure
351Chinahaspreviouslydemonstratedthatitiswillingandabletofast-trackinfrastructureprojectstohelpfriendlypoliticalincumbents(e.g.,Jansson2013).
222
financierswanttheirloansrepaidontimeandwithinterest.352However,given
thatinfrastructureprojectcompletionincreasespublicsupportforChina(Wellner
etal.forthcoming2023),Beijing’sdiplomatslikelyalsohaveincentivesto
prioritizecollaborationwithstablegovernmentsthathaveputinplacepolicies
andinstitutionswhichfacilitatethedeliveryoflarge-scaleinfrastructure
projects.353
ThegatewaytoSouthandSoutheastAsia
Givenitsuniquelocationalpositionandrecentgeopoliticaltrends—including
theriseofIndiaandgrowingSino-U.S.tensionsintheSouthChina
Sea—Bangladeshwilllikelyremainakeybattlegroundforgreatpower
competitioninthecomingyears.FromDhaka’svantagepoint,growing
competitionbetweenChinaandtheU.S.intheIndo-Pacificregiononlyoffers
moreopportunitiesforimprovingeconomicgrowth.Itseconomystillneeds
diversification—e.g.,throughnewcompetitiveindustriesadjacenttoits
dominanttextileindustry.Thiswillrequireinvestment,skillstransfer,andfurther
integrationofexistingindustriesintoglobalsupplychains.Adeepeningtrade
relationshipwithChina,potentiallythroughtheplannedraillinkviaMyanmar,
couldincreaseeconomicprosperitythroughjobcreation.
FromtheU.S.perspective,itsrobusttradetieswithBangladeshandstrong
alliancewithIndia(Bangladesh’smostinfluentialregionalpartner)mayprovide
opportunitiestomakeinroads.Yetrecentevents—includingvoterigging,voter
intimidation,theuseofviolence,andthetargetingofNobelPeacePrize
laureateMuhammadYunusthroughthejudicialsystem—haveputWashingtonin
atoughposition(Miller2023).Asadefenderofliberaldemocraciesaroundthe
world,Washingtonmayfeelcompelledtocondemnactionsthatwouldpush
Bangladeshfurtherdownthepathofauthoritarianism.Ontheotherhand,it
cannotignoretherealpolitikconsiderationthatChina’sinfluenceisexpandingin
SouthandSoutheastAsia.
353Effectivepublicdebtmanagementisonesuchexample.Bangladeshcurrentlyhasamanageabledebt-to-GDPratioofunder20%(WorldBank2023a).352Dreheretal.(2022)providestatisticalevidencethatwhenBeijingissuesloansatornearmarketrates,itfavorscountrieswithhighlevelsofpoliticalstability.
223
Argentinacasestudy:‘Afamilyaffair’—China’sgenerosityand
pragmatism
IntermsofchangesinelitesupportforChinaduringtheearlyBRIperiod,
Argentinafallsintothesafebetcategoryofcountries.ItsUNGAvotingpatterns
demonstratethatitmovedintocloseralignmentwithChinathanwiththeU.S.
between2014and2017.WealsochosetoconductacasestudyofArgentina
becauseitwasrichlyrewardedbyBeijingduringthelateBRIperiod—mostlyin
theformofserialemergencyrescuelendingthathashelpedpullthecountry
backfromthebrinkofeconomiccatastrophe.
ItissurprisingthatArgentina—anuppermiddle-incomecountrywithclose,ifat
timesstrained,relationswiththeUnitedStates—isoneofthelargestrecipients
ofChineseaidandcreditinthe3.0versionofAidData’sGCDFdataset.Thiswas
notalwaysthecase.Beijingprovidedloansandgrantsworthonly$4billion
duringthepre-BRIera(2000to2013).Then,duringtheBRIera(2014to2021),
China’sofficialfinancialflowstoArgentinaskyrocketed(seeFigure4.11).
Averageannualfinancialcommitmentsduringthisperiodamountedto$16.8
billion.Beijingbankrolledinfrastructureprojectssuchassolarandhydroelectric
powerplants,irrigationsystems,highways,railwaylinesandlocomotives,and
evenaspacemonitoringstation.However,thevastmajorityofChina’sofficial
sectorfinancialflowstoArgentinacameintheformofcurrencyswapdebtfrom
thePeople’sBankofChina(PBOC).354
354Abilateralcurrencyswap(BCS)agreementisanagreementbetweenthecentralbanksoftwocountriestoexchangecashflowsindifferentcurrenciesatpredeterminedratesoveraspecifiedperiodoftime.ThepartytotheBCSagreementthatdrawsdownontheswaplinebecomestheborroweranditscounterpartybecomeslender.Thecurrencyoftheborrowerisheldascollateralwhilethelenderreceivesinterestontheamountdrawndownbytheborroweruntilrepaymentismade.Inprinciple,swaplineswiththePBOCaredesignedtopromotetheuseofRMBfortradeandinvestmentsettlementpurposes.However,inpractice,theyaremostlyusedtoprovidebalanceofpaymentssupporttoborrowerswithhighlevelsofoutstandingdebttoChinaduringperiodsoffinancialdistress(Hornetal.2023a,2023b).NearlyallPBOCswapborrowingscarrydejurematuritiesoflessthanoneyear.However,PBOCswapdebtisfrequentlyrolledover,resultinginaveragedefactomaturitiesof3.5years(seeBox3ainChapter3).
224
Figure
4.11:
Official
financial
flows
from
China
to
Argentina,
2014-2021Figure4.11
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOF
determinations(asdescribedinSectionA-2oftheAppendix).
Atthesametime,Argentinaisgeographicallyandeconomicallyproximateto
theWest.Ithasbeenamajornon-NATOallytotheUnitedStatessince1998
(CenterforLatinAmericanandLatinoStudies2022).TheU.S.isitsthirdlargest
exportandimportmarket(TheGrowthLabatHarvardUniversity2023).
Argentinahasclose,ifcomplicated,relationswiththeIMF,whichhasprovided
tensofbillionsofdollarsofcredittosupportthecountry.Argentinaisalsoa
vibrantdemocracyandonewithahighlyeducatedpopulace.
ButArgentinahasahistoryofpursuingstrategictieswith
ideologically-confoundingpartnersinpursuitofitsnationalagenda.Duringthe
lateColdWar,theArgentineright-wingmilitaryjuntahadgoodrelationswith
theSovietUniontosupportitsinflation-riddeneconomy(Schumacher1981).
Duringthemoderneraofgreatpowercompetition,Argentina’sstatusasa
regionalleaderinLatinAmericaandaG-20economyhasmadeitparticularly
importanttoboththeUnitedStatesandChina,andithaspursuedtieswithboth
225
nationstobenefititself.Nonetheless,ourUNGA-basedmeasureofelitesupport
suggeststhatArgentinamovedintocloseralignmentwithChinathantheU.S.
duringtheearlyBRIperiod.Butthestorybeginsadecadebeforethelaunchof
theBRI.
TheKirchnersandtheBRI
Duringthe2003-2007and2007-2015PeronistpresidenciesofNéstorKirchner
andhiswifeCristinaFernándezdeKirchner(CFK),U.S.-Argentinarelationswere
shaky,inconsistent,andattimestenseandhostile,althoughnot
cooperation-free(SullivanandNelson2017).TheKirchners’governments
deepenedtieswithAmericanadversaries,includingRussia,Venezuela,Iran,and
Cuba(Filkins2015).
WhilethesetensionswiththeUnitedStatesgrewandU.S.foreignpolicy
focusedontheGlobalWaronTerror,theKirchnersdeepenedArgentina’s
relationshipwithChina,especiallyduringCFK’ssecondterm(Sullivanand
Nelson2016).Chinabecamethesecondlargestexportandimportmarketfor
ArgentinaduringtheKirchners’presidencies(TheGrowthLabatHarvard
University2023).DuringCFK’sfirstterm(December2007toDecember2011),
officialfinancialcommitmentsfromChinatoArgentinaamountedtoonly$1
billion.However,duringhersecondterm(December2011toDecember2015),
officialfinancialcommitmentsfromChinatoArgentinaskyrocketedto$23.9
billion.355
355Thesefigures,aswellasotherfiguresthatrelyonprecisedates,onlyincludeprojectswherethecommitmentdateswithinagivenyearareentirelyunknowniftherelevantmeasure[presidentialterm,monthsbeforeanelection,etc.]makesupmorethanhalfoftheyear;sofor2015,sinceCFKservedaspresidentforallbut21daysoftheyear,AidDataincludescommitmentswithentirelyunknowndatesasunderherpresidency.
226
Figure
4.12:
Argentina’s
U.N.
voting
alignment
with
China
and
the
U.S.Figure4.12
Notes:SeeBox1bformoredetailonthismeasure.
Figure4.12documentstrendsinArgentina’slevelofUNGAvotingalignment
withChinaandtheU.Sbetween2000and2021.TheinfluenceofKirchnerist
presidenciesisevident.ArgentinawasneverfurtherawayfromChina—froma
foreignpolicyalignmentperspective—thanin2003.Then,undertheKirchners’
presidencies,ArgentinasteadilymovedintocloseralignmentwithChina,atrend
thatintensifiedduringtheearlyBRIperiod.
Argentina’stiestoChinastrengthenedduringCFK’ssecondterm,which
coincidedwiththebeginningoftheBRI.Inlate2014,severalChinese
state-ownedbanksgreen-lita$4.71-billionsyndicatedloanforthe1,740MW
NéstorKirchnerandJorgeCepernicHydroelectricPowerPlantConstruction
Project(KCHPProject),atwindamofgreatpersonalsignificancetoCFK,asit
wasnamedinpartafterherlatehusbandwhodiedin2010(Lucci2019).356CFK
356TheunredactedloancontractwasincludedintheHowChinaLendsDataset,Version1.0(Gelpernetal.2022)anditisaccessibleinitsentiretyviahttps://www.documentcloud.org/documents/20484849-arg_2014_435.
227
describedtheprojectas“themostimportanthydroelectricproject”in
Argentina’shistoryandservedasitsmajorpatron(Watts2015).TheEPC
contractwasawardedtoChinaGezhoubaGroupaspartofawiderconsortium
withtwoArgentinecompanies.
ThedamprojectcombinedPeronistpoliticalgoals(infrastructure-driven
nation-building)withCFK’spersonalpoliticalgoals(developingtheSantaCruz
Province,thefamilialandpoliticalcoreofKirchnerism)andtheprinciplesofthe
BRI(infrastructuretospurglobaldevelopmentandtrade).Argentina’sleadership
waswell-positionedtoleveragelarge-scalefinancingfromChinatosupportits
agenda.ForBeijing,theprojectrepresentedanopportunitytopromotetheBRI
andpursueitssoftpowerobjectives.
OnJuly30,2014,twodaysbeforesigningoftheloanagreementfortheKCHP
Project,theGovernmentofArgentinadefaultedonitsforeignbondrepayment
obligations(vonHoffman2014).Atthetime,Argentinawasexperiencinga
recession,acurrencydevaluation,andhighlevelsofinflation.Earlierthatmonth,
onJuly18,2014,duringXiJinping’svisittoBuenosAires,thePeople’sBankof
China(PBOC)andtheCentralBankofArgentina(BCRA)extendedanRMB70
billion(around$11billion)bilateralswapagreementforthepromotionoftrade,
useoftherenminbi,andthebolsteringofArgentina’sforeignexchangereserves
(Arnold2023).Laterthatyear,theBCRAactivatedtheswaparrangement,
drawingdownRMB14.2billionbetweenOctoberandDecember.Then,in2015,
itagaindrewdownRMB70billion.ThePBOCswaplinehelpedrescue
ArgentinawithoutanysignificantinvolvementfromWesternpowers.Thiswas
significantbecausethedefaultoriginatedfromanArgentinerefusaltomeetan
U.S.court-mandatedrepaymenttoa“vulturefund”(Shortell2014).
ThefactthatChinawaswillingtoprovidealargeamountofcreditduringa
periodofcrisisconvincedsomelocalelitesthatChinacouldbeaviable
alternativetotheWest.Afterits2001default,Argentinahadbeenisolatedfrom
internationalcapitalmarkets,sotheemergenceofChinaasamajorinternational
lenderwasamajorboon,onethatwouldallowthecountrytoreduceits
dependenceupontheIMFandWesternsourcesoffunding.Theavailabilityof
228
Chinesecreditcouldalsobeleveragedtosecuremorefavorableoffersfromthe
West.
PartnershipwithChinaalsooffereddomesticbenefitsthatcouldhelpCFK’s
JusticialistPartyremaininpower.Inexchange,Chinabenefitedfroman
administrationinBuenosAiresthatwasmorewillingtoadoptforeignpolicy
positionsthatitfavored.By2015,theUNGAvotingalignmentdatashowsthat
Argentina’sidealdistancefromChinareacheditsclosestpointacrosstheentire
22-yearperiod(at0.0968).
Giventhepro-ChinaorientationofCFK’sJusticialistParty,Beijingwaswillingto
supporttheincumbentbyauthorizingmultipledrawdownsunderthePBOC
swaplinetostabilizetheeconomy.357Nevertheless,theJusticialistParty
candidate,DanielScioli,whohadbeenexpectedtowinbyalargemargin,lost
the2015electionby3percentagepoints(BBCNews2015).Thewinnerwasa
conservativebusinessman,MauricioMacri.Duringhiscampaign,Macri
expressedhisdesiretoimprovetieswiththeU.S.andEuropeanUnion.Macri
neverdirectlyvoicedananti-Chinaposition.Infact,hepubliclystatedthe
importanceofmaintaininggoodrelations.Buthedidsignalthatcontracts
signedbyCFK’sadministrationwithChinesecompaniesrequiredreviewfor
corruptionandinsufficient“technicaldetails.”CFK’spersonalinvolvementinthe
KCHPProjecthadbeenasourceofdomesticpoliticalgrievancethatthenew
administrationwishedtorectify(CenterforLatinAmericanandLatinoStudies
2022).
AfteryearsofheavyfinancingduringCFK’syears,fromBeijing’sperspective,the
prospectofanArgentinianrepositioningtowardtheWestwouldbeapolicy
failure.ThefactthatMacri’saccessioncameinthemidstofawaveofnew
center-rightgovernmentsacrossLatinAmericalikelyamplifiedthisconcern
(CenterforLatinAmericanandLatinoStudies2022).
China’ssuspicionsaboutthismatterwerenotill-founded.UndertheMacri
administration,U.S.-Argentinarelationsweremuchcloserthanwhenhis
357Withthecountry’seconomicproblemsadominantissueduringthecampaign,inflationrunningaround25%,andcurrencyreservesdeclining,Beijing’sfinancialrescuepackagewasusefultogoverningelites(GBH2015).
229
predecessorswereinpower.358Furthermore,in2018,theMacriadministration
negotiatedthelargestIMFloanever—$57.1billion—tosupportthecountry’s
economy(Sen2018).
RelationswithChinawerenotstraightforwardatfirst.InlateDecember2015,
Macrisuspendedthetwindam(KCHP)projectonenvironmentalgrounds(Lucci
2019).359ThistriggeredconcernsfromtheChineselenders,whichhadalready
disbursed$950million.OnMarch10,2016,CDBsentaletteronbehalfofthe
othermembersoftheloansyndicatetoArgentina’sfinanceministryandwarned
oflegalandpoliticalconsequencesofprojectsuspension:"[theKCHPProject
andtheBelgranoRailwayModernizationProjectare]majorprojectspromoted
bytheChinesepartiesinthesameperiodoftimeinArgentinaandeach...[of
theloanagreementsforthetwoprojects]contain'crossdefault'provisions."In
effect,CDBtoldtheArgentineauthoritiesthattheycouldnotcanceltheKCHP
projectwithoutcancelingtheBelgranorailwayproject(Gelpernetal.2022).360
InApril2016,aftermeetingwithXiinWashingtonD.C.,Macriannouncedthat
theKCHPProjectwouldproceed,althoughitwouldbemodifiedtominimize
negativeenvironmentalimpacts(Koop2016).Then,inDecember2016,the
SupremeCourtofArgentinasuspendedconstructionuntilanenvironmental
impactassessmentwascompletedandapublichearingwasheld.Construction
ultimatelyresumedinMarch2018.Macri,afterlearningofChina’sdetermination
toseeaprojectpersonallyendorsedbyXisucceed,decidedthatcancelingthe
projectwasnotworththepotentialconsequencesofalienatingorantagonizing
amajorcreditor(Patey2017).Healsotookstepstodevelopastronger
relationshipwithChinawhilemaintaininggoodrelationswiththeWest.During
theearlyBRIyearsofhispresidency,officialfinancialcommitmentsfromChina
toArgentinaremainedsubstantial:$13.2billionin2016and$14.4billionin
2017.
360Thelettercanbeaccessedinitsentiretyviahttps://www.dropbox.com/s/q6s26ninx4ldnes/Cross-Default%20Letter%20from%20China%20Development%20Bank%20to%20the%20Government%20of%20Argentina%2010%20March%202016.pdf?dl=0.359Between2000and2021,theESGriskprevalencerateinChina’sgrant-andloan-financedinfrastructureprojectportfoliowas44%inArgentina(seeTableA12).358ThiswasbolsteredbytheideologicalsynergybetweenMacriandDonaldTrump.Macri’sArgentinasupportedtheTrumpadministration’scampaignagainstNicolásMaduro’sVenezuela(Wilkinson2019).
230
ThelateBRIperiodunderMacriandFernández
ArgentinasawfurthergrowthinfinancialsupportfromBeijingduringthelateBRI
period(2018-2021).AverageannualfinancialcommitmentsfromChinaduring
thisperiodamountedto$21.4billion.Mostofthisfundingwasprovidedvia
PBOCswaplinedrawdownsforbalanceofpaymentssupport.361Bythetime
MacriexitedofficeinDecember2019,Argentina’sRMBswapdebtrepresented
50.5%ofthecountry’stotalforeignexchangereserves(seeTable4.5).
Table
4.5:
Estimated
percentage
of
Argentina’s
foreign
currency
reserves
derived
from
PBOC
swap
facilityTable4.5
Estimated
percentage
of
Argentina’s
foreign
currency
reserves
derived
from
PBOC
swap
facility
YearTotalreserves(USDbillions)AmountsoutstandingunderPBOCswapfacility(USDbillions)EstimatedpercentageofreservesoriginatingfromPBOCswapfacility
2009$48.01$00%
2010$52.21$00%
2011$46.27$00%
2012$43.22$00%
2013$30.53$00%
2014$31.41$2.327.39%
2015$25.52$10.7842.25%
2016$38.41$10.0726.22%
2017$55.31$10.7519.44%
2018$66.22$18.9628.65%
2019$44.88$18.6041.46%
361InJuly2017,BCRAandPBOCextendedtheswaplinebyanotherthreeyears;then,inDecember2018,duringXi’sstatevisittoArgentina,BCRAandPBOCsignedadealtoincreasetheswapline’slimitfromRMB70billiontoRMB130billion(around$19billion)(Hornetal.2023a).ThisagreementincludedastipulationthatPBOCcouldrejectcurrencyswapdrawdownsifArgentina’sIMFstandbyagreementwassuspendedorcanceled.
231
YearTotalreserves(USDbillions)AmountsoutstandingunderPBOCswapfacility(USDbillions)EstimatedpercentageofreservesoriginatingfromPBOCswapfacility
2020$39.40$19.8950.49%
2021$39.65$20.4051.47%
Notes:ThistableshowstheestimatedpercentageofArgentina’sforeigncurrencyreservesthatoriginated
fromthePBOCswapfacilitybetween2009and2021.ThedataontotalreservesarefromtheWorldBank
andincludegoldreserves.AmountsoutstandingunderthePBOCswapfacilityarefromHornetal(2023a).
AllamountsarereportedinnominalUSD.
Despitetheinitialchallenges,Macrieventuallybecameavaluedpartnerto
China,asdemonstratedbytheincreaseintheswapline,Xi’svisits,andvarious
tradeagreements,includingtheopeningupofChinatoArgentina’ssoymeal
livestockfeedexportsinSeptember2019andlowerbarrierstobeefandsheep
exports(Koop2018,BronsteinandHeath2019).Macritookmeasurestojointhe
China-backedAsianInfrastructureInvestmentBank(AIIB)in2018,whichonly
becameofficialin2020.Also,whileArgentinadidnotofficiallyjointheBRIuntil
thenextadministration,MacriwasoneofonlytwoheadsofstatefromLatin
AmericatoparticipateinthefirstBeltandRoadInitiativeForumin2017(Center
forLatinAmericanandLatinoStudies2022).
BeyondPBOCswaplineborrowings,significantcollaborativeactivitieswith
ChinaduringMacri’spresidencyincludedtheGeneralRocaRailwayEquipment
AcquisitionProject,whichreceiveda$236-millionbuyer’screditfromCDBand
supportedacommercialcontractwithstate-ownedCRRCQingdaoSifang;the
NationalRoadLineB(CorredorBTollRoad)Project,whichwassupportedbya
$1.18-billionsyndicatedloanfromthreebanks,includingtheBankofChinaand
ICBC,toaArgentine-Chinesejointventureresponsiblefortheconstructionand
operationoftheroad;andthe312MWCaucharíPhotovoltaicPowerPlant
Project,financedbya$331.5-millionChinaEximbankloan.Macri’spersonal
pragmatismandtheprospectofwhatChinacouldstillofferhelpedfostera
goodrelationship.
China’sexperiencewithMacrileftBeijingwithanoptimisticoutlookforthe
futureofArgentine-Chineserelations.DespiteMacri’spro-Westernviewsand
actionsandinitialcautiontowardChina,underhisleadershipArgentinagrew
evenclosertoChina.IfChinacouldthriveinArgentinaunderaconservative,
232
pro-Westernpresident,thenitcouldthriveunderalmostanyconceivable
Argentinepresident—especiallyiftheKirchneristsreturnedtopower,whohad
deepenedtherelationshipinthefirstplace.Chinahaddemonstratedduring
bothCFKandMacri’spresidenciesthatitwasareliable,criticalpartnertothe
Argentinegovernmentforthecountry’seconomicstabilityandwillingto
negotiateifnecessary,leavinglittlewillpoweramongArgentineelitestoalienate
orantagonizeavaluablepartner;thus,ChinahadturnedArgentinaintoa“safe
bet”fortheforeseeablefuture.
Macri’scenter-rightgovernmentultimatelyprovedinsufficientlypopularbythe
timeofthe2019Argentinegeneralelections,duetoeconomictroubles.The
KirchneristsreturnedtopowerwithpresidentAlbertoFernándezandformer
president,nowvicepresident,CristinaFernándezdeKirchner(norelation).
Duringhiscampaign,FernándezoftendiscussedChina,arguingthatMacridid
notappreciateChinaenough.Additionally,FernándezmetwithZouXiaoli,the
ChineseAmbassadorinArgentina,andsentseveralofhisforeignpolicyadvisors
totheembassyintheweekspriortotheelection.Observersexpectedthat
undertheFernándezadministration,relationswouldbeclose,withChina’s
projectsinthecountryexpectedtobeboosted(Koop2019).Asa
demonstrationoftheimportanceofrelations,Fernández’sgovernment
appointedSabinoVacaNarvaja,sonofoneofthefoundersoftheMontoneros
guerillasandthebrother-in-lawofCFK’sdaughter,asambassadortoChinaand
assignedageneraltoserveasthedefenseattache,somethingpreviouslyonly
donefortheU.S.(Ellis2021).
Examiningthedataonfinancialcommitmentsfromthe3.0versionofAidData’s
GCDFdatasetduringthelatterhalfofthelateBRIperiod,thereislittleevidence
ofdifferentialtreatmentacrosstheMacriandFernándezadministrations,with
Argentinareceiving$21.2billionin2020and$20.2billionin2021.However,
eventhoughrelationsseemedpoisedtogrow,theFernándezadministration’s
relationswithChinareportedlystagnatedbecauseofissuesontheArgentine
side(GiustoandHarán2023).Numerouspledgedorcommittedprojectswith
Chinaneverreachedimplementationbecauseofcurrencycontrols,
protectionism,bureaucraticbungling,andinconsistentpolicymakingfrom
Argentina’snationalleadership(Economist2023).
233
AnexampleistheAtuchaIIINuclearPlantProject.Firstpledgedin2014during
CFK’ssecondterm,itwentthroughmultipleroundsofnegotiationswiththe
Macriadministrationbutnevermadesolidprogress.Soonafterhisvictory,
Fernándezannouncedthattheprojectwouldproceed.InFebruary2022,
ArgentinasignedanEPCcontractwithChinaNationalNuclearCorporation,only
forArgentinatoaskChinato100%financetheprojectinApril2022(asopposed
tothestandard85%maximum).Morenegotiationsandrequestsfor
modificationsfollowedthroughouttheyear,leavingtheprojectonshaky
grounds(Bernhard2022).Atuchaisnotanoutlier.Between2018and2021,
Chinapledged$664.6millionofadditionalfinancing,butnoneofthesepledges
hadbecomeformalcommitmentsby2023,andanadditionalsetofprojects
worth$146millionweresuspended.362
EvenifFernándezwasnotasinterestedinstrengtheningSino-Argentineties,
Fernández,beingacloseKirchnerally,waspoliticallyalignedwithChina.In
February2022,hevisitedBeijingandsignedanagreementtojointheBRI
(GiustoandHarán2023).ChinaalsoissuedastatementinsupportofArgentine
claimstotheFalklandIslands(knownasLasIslasMalvinasinArgentina)(Rey
2022).AttheUNGA,hecontinuedtheoverallpatternestablishedbyhis
predecessors,bringingArgentinaintocloservotingalignmentwithChina.363
ThelargevolumeofofficialfinancingfromChinatoArgentina—especiallyinthe
formofgovernment-sustainingcurrencyswapdrawdowns,andinadditionto
foreigndirectinvestmentandtradeties—hasmadeChinaacriticalfinancial
partnerforArgentina.DuringtheearlyBRIperiod,whichsawthelastpartof
CFK’sruleandtheinitiallydifficult,butsoonproductive,beginningoftheMacri
administration,Beijingrealizedhowstrongitsfootholdwas,tothepointthatit
couldfeelconfidentthatArgentineeliteswouldnotendangerrelations,turning
itintoa“safebet”amongstallrecipientsforfriendlyelitealignmentwithChina.
363In2021,heachievedoneoftheclosestidealdistances(0.1452)fromChinaduringourentire22-yearperiodofstudy.362InadditiontoprojectsthatsecuredfinancialcommitmentsintheearlyBRIera—whichdonothavetheexcuseofCOVID-19asadelayingfactorandhadmoreopportunitiestobeformallycommitted—over$2.8billionoffinancingthathadbeenpledgedneverreachedimplementationandprojectsworth$11billionweresuspended.
234
BasedonthecloserelationsundertheMacripresidencyandpoliticalalignment
duringtheFernándezpresidency,Beijingappearstohavecorrectlypredicted
thecourseofbilateralrelations.Theoveralltrendhasbeenoneofincreasing
UNGAvotingalignmentbetweenChinaandArgentinaandanexpanding
envelopeofChineseaidandcredit.
Entrenchmentasagoalofsoftpower
DuringtheBRIera,inlightofgoodbilateralrelations,Chinahasprovided
meaningfuleconomicsupporttoArgentina.Becausetheeconomyisthesingle
mostimportantelectoralissueandArgentina’sdependencyonChinaishigh,
majorpoliticalpartieshavestrongincentivestomaintaingoodrelationswith
Beijing.Macriexperiencedthedownsideofbeingseenasuncooperativetoward
Beijing,forcinghimtocoursecorrecttowardafriendlierpath.Butasgreat
powercompetitionbetweenChinaandtheWestintensifies,Argentina’sability
tomaintaingoodtieswithbothsidesisbecominglimited.Acaseinpointisthe
ongoingU.S.pushtosellitsownF-16fighterjetstoArgentina,inorderto
preventitfrompurchasingthecheaperChinese-designedand
Pakistani-manufacturedJF-17jets(BuenosAiresTimes2023).
However,acandidateinthePresidentialrun-offelectionwhoreceived30%of
thepopularvoteduringthefirstroundisfar-rightpopulistandlibertarian
economistJavierMilei.InAugust2023,Mileiannouncedhisintentionstofreeze
relationswithChina,callingitsgovernment“anassassin”anddescribinghis
proposedforeignpolicyasa“fightagainstsocialistsandstatists”(Brandimarte
andTobias2023).AnalystshavecomparedMileitoDonaldTrumpandformer
BrazilianpresidentJairBolsonaro,whobothusedanti-Chinarhetorictobolster
theirelectoralcampaigns(Shi2023).ShouldMileiwintheelection,Argentina’s
foreignpolicywouldlikelyshiftintosubstantiallycloseralignmentwiththe
UnitedStates.BeijingmayhopethataspresidentMileiwouldsoftenhisrhetoric
onChinaonceinofficelikeMacridid,butregardless,Chinawouldlikelyprefer
anothercandidate.364
364Thatbeingsaid,inOctober2023,center-rightcandidatePatriciaBullrichannouncedthatshewouldreversePresidentAlbertoFernández’sdecisiontojointheBRICsblocifelected.ShealsotoldtheFinancialTimesthat“[w]ebelievethatinsomeofthelatest[Chinese]loansthereare
235
Intherunuptothefirst-roundelectionsinOctober2023,Chineseofficialsector
entitiesprovidedadditionalresourcestoArgentina,presumablytobolsterthe
campaignsofcandidatesitdeemsleastproblematic.InJune2023,Sergio
Massa,theincumbenteconomyministerandaleadingcandidate,visited
Beijing,wherehestruckdealsonArgentineexports,$3.05billioninfinancingfor
variousprojects,and,mostimportantly,anextensionandexpansionofthe
PBOCswapline(Alcoba2023).BasedonChina’spastdecisionmakingin
Argentinaandanempiricalpatternofcommitingmorefundsinelectionyears
aroundtheworld,moreagreementsandcommitmentsarelikelyforthcoming
(Dreheretal.2019).
ButthesamelogicofChinabeingtoobigtopushinArgentinamayalsoapply
toitsrelationshipwiththeU.S.Argentinalackstheleverage,orwillingness,to
becomeatrueAmericanadversaryakintoCubaorVenezuela.Factorslike
geographicproximity,democraticaffinity,andculturalsynergyalladvantagethe
U.S.overChina.In2021,59.48%ofArgentiniansapprovedoftheU.S.
government,accordingtoGallup,ascomparedtoonly32.65%forChina.The
economicdimensionisalsonoteworthy.WhilenotnearlyaslargeasChina,the
U.S.isstilloneofthemajortradepartnersofArgentina.Itisalsothesingle
largestshareholderattheIMF,whereArgentinaisthesinglelargestdebtor.
Underitsnewgovernmentin2024,Argentina’sforeignpolicywillhaveto
carefullytreadbetweentieswiththeU.S.andChina,bothofwhicharecritical
foritsfutureeconomicprospects.
Section5:Beijing’sapproachtoreputationalriskmanagement
Inthischapter,wehaveprovidedstatisticalandcasestudyevidencethat
China—likeotherforeignpowers—makesadjustmentstoitsinternational
developmentfinanceportfolioinresponsetosoftpowergainsandlossesonthe
ground.Itistakingreactiveandproactivemeasurestomanagereputationalrisk.
clauseswhichwedon’tknowaboutandwearereadytore-examinethem”(StottandNugent2023).
236
Therearefivekeytakeawaysfromouranalysis.First,onthreedifferentmeasures
ofsoftpower(thatcapturegainsandlossesinpublicopinion,mediasentiment,
andelitesupport),Beijingdevotednearlytwo-thirdsofitsentireinternational
developmentfinanceportfolioduringthelateBRIperiodto“tossup”countries.
ThesecountriesrepresentcompetitivejurisdictionswhereneitherChinanorthe
U.S.openedupaninsurmountableleadvis-à-visitsprincipalrival.
Second,insettingswhereChinarecentlymadereputationalgainsatthe
expenseofitsprincipalcompetitor,itdoubleddownwithadditionalaidand
credit—i.e.,itsoughttomaintainandbuilduponmomentum.365Thefactthat
Beijingisredirectingaidandcredittobattlegroundcountrieswhereithas
recentlygainedadvantagechallengesapopularbeliefamongSinologists:that
thebureaucraticmachineryresponsibleforBeijing’soverseasdevelopment
programisfragmented,uncoordinated,andinattentivetochangingreputational
dynamicsontheground.366
Third,Chinadoesnothavemuchofanappetiteforreputationalrisk.It
consistentlyallocatedalower-than-expectedshareofitsinternational
developmentfinanceportfolioto“moonshot”countries(dedicatingbetween
16%and27%lessthanexpectedforeachsoftpowermeasure,seeFigure4.5).
Thesearecountrieswhereitsprincipalrivalhasmomentumonitsside.China
rarelyseekstowooindifferentorantagonisticcountrieswithaidandcredit,but
insteadprioritizescountriesthatarealreadymovinginitsdirection.Aseparate,
butrelated,pointisthat,whenreputationalassetsturnintoreputational
liabilities(aswesawintheZambiacasestudy),Beijingmostlydisengagesfrom
discussionsaboutnewprojectsandfinancialcommitmentsandrefocuseson
managingriskswithinitsexistingportfolioofgrant-andloan-financedprojects.
Fourth,ascasestudiesofBangladeshandArgentinademonstrate,political
transitionsinhostcountriesarecriticalmomentswhenthenature,level,and
paceofChina’sengagementcanchangesignificantly.Beijingunderstandsthatit
366WethankMarinaRudyakforcallingourattentiontothispoint.365Morespecifically,wefindthatBeiijingprioritizedtheprovisionofaidandcredittocountrieswhereithadexperiencedpublicopinionandmediasentimentgainsattheexpenseoftheU.S.duringtheearlyBRIperiod,whileitdeprioritizedtoss-upcountrieswherethemomentumshiftedinfavoroftheU.S.duringtheearlyBRIperiod.
237
canmakemajorreputationalgainsorsuffermajorreputationallossesduring
thesewindows,andithasdemonstratedawillingnessandabilitytouse
instrumentsofstatepowertoprotectitsinterestswhensuchjuncturesarise.
DuringthetransitionfromtheKirchneradministrationtotheMacri
administration,aconsortiumofChinesestate-ownedpolicyandcommercial
banksinvokedacross-defaultclauseinaloanagreementwiththeArgentine
MinistryofFinancetoblockthenewly-electedpresidentfromfollowingthrough
onhiselectoralpledgetosuspendenvironmentallyriskyprojects,whichcould
havejeopardizeda$5.5billiondamconstructioncontractthatwaspreviously
issuedtoChinaGezhoubaGroupCompanyLimited.Conversely,whennew
leaderscometopowerandtakealessadversarialposture(likeBangladesh’s
SheikhHasinaandZambia’sEdwardLungu),ourfindingssuggestthatBeijing
oftenseekstocementbilateralrelationsbyhelpingincumbentstakecreditfor
high-profileprojects.367Itdoessoby,amongotherthings,approvingnew
financialcommitmentsforprojectsthatwerepreviouslyunderconsideration,
organizinggroundbreakingceremoniesforpreviouslyapprovedprojectsthat
hadnotyetcommenced,andprovidingstate-sponsoredmediacoverageof
recentlycompletedprojects.368Thisstrategyisespeciallyrelevantindemocratic
countrieswhereelectionsaremorecompetitive,asinZambiawhereincumbent
performanceattheballotboxislinkedtoperceivedeffectivenessatdelivering
big-ticketinfrastructureprojectsthatcancreatejobsandstimulateshort-term
economicgrowth.
Finally,forthosewhomakeandshapepolicyinWesterncapitals,akeyinsight
fromthischapteristhatBeijingtendstodisengageratherthandoubledownin
countrieswheretherearestrongindicationsofBRIbuyer’sremorse.Theseare
jurisdictionswhereBeijing’scompetitorsmaybeabletolurecountriesbackinto
theWest’sorbitbyfocusingontheirownareasofcomparativeadvantage.
However,doingsowouldrequirethatWesternpowersactquicklywhenthese
windowsofopportunitiesariseandadapttheirprogrammingtoaddressthe
unmetneedsofpartnercountries.
368Wellneretal.(forthcoming)demonstratethatthecompletionofChinesegrant-andloan-financedprojectsincreasespublicsupportfortheChinesegovernmentandthehostgovernment.367Onthispoint,alsoseeHolslag2011;Jansson2013;DiLorenzoandCheng2019;Dreheretal.2019;Anaxagorouetal.2020;Strange2023;Tang2021;Kernetal.2022.
238
References
Aamir,Adnan,MarwaanMacan-Markar,ShaunTurton,CissyZhouandGraceLi.2022.Roadtonowhere:China’sBeltandRoadInitiativeattippingpoint.10August2022.NikkeiAsia.Accessedon13August2023athttps://asia.nikkei.com/Spotlight/The-Big-Story/Road-to-nowhere-China-s-Belt-and-Road-Initiative-at-tipping-point
Abi-Habib,Maria.2018.HowChinaGotSriLankatoCoughUpaPort.NewYorkTimes.25June2018.Accessedon24July2023athttps://www.nytimes.com/2018/06/25/world/asia/china-sri-lanka-port.html.
Acker,Kevin,DeborahBrautigam,andYufanHuang.2020.DebtReliefwithChineseCharacteristics.SAIS-CARIWorkingPaperNo.2020/39.WashingtonDC:JohnsHopkinsUniversity.
Aiping,Zeng.2015.China-AfricaGovernanceExchangesandExperiences.Beijing:ChinaInstituteofInternationalStudies.
Akhtar,ShayerahI.andNickM.Brown.2021.U.S.InternationalDevelopmentFinanceCorporation:OverviewandIssues.WashingtonDC:CongressionalResearchService.Accessedon16July2023athttps://sgp.fas.org/crs/row/R47006.pdf
AlJazeera.2018.ChinaandArgentinaSignCurrencySwapDealafterG20Summit.AlJazeera.2December2018.Accessedon1October2023athttps://www.aljazeera.com/economy/2018/12/2/china-and-argentina-sign-currency-swap-deal-after-g20-summit.
Alcoba,Natalie.2023.Argentina’sYuanLifelineSignofBrinksmanshipbetweenChina,US.AlJazeera.17July2023.Accessedon1October2023athttps://www.aljazeera.com/economy/2023/7/17/argentinas-yuan-lifeline-sign-of-brinksmanship-between-china-us.
Alesina,Alberto,andDavidDollar.2000.WhoGivesForeignAidtoWhomandWhy?JournalofEconomicGrowth5(1):33–63.
AmnestyInternational.2023.Bangladesh:UnlawfulUseofForceagainstProtestersMustEndImmediately.AmnestyInternational.4August2023.Accessedon1October2023athttps://www.amnesty.org/en/latest/news/2023/08/bangladesh-unlawful-use-of-force-against-protesters-must-end-immediately/.
Anaxagorou,Christiana,GeorgiosEfthyvoulou,andVassilisSarantides.2020.ElectoralMotivesandtheSubnationalAllocationofForeignAidinSub-SaharanAfrica.EuropeanEconomicReview127:103430.
Anderlini,Jamil.2008.BeijingusesforexreservestotargetTaiwan.12September2008.FinancialTimes.Accessedon1August2023athttps://www.ft.com/content/22fe798e-802c-11dd-99a9-000077b07658
Ang,YuenYuen.2019.DemystifyingBeltandRoadTheStruggletoDefineChina’sProjectoftheCentury.ForeignAffairs.22May2019.
239
ANI.2022.FinanceMinisterKamalWarnsDevelopingNationsofChineseLoansStrappedwithDebt-Trap.ANI.10August2022.Accessedon1August2023athttps://en.prothomalo.com/bangladesh/bangladesh-in-world-media/sysadnqpr3.
Ansar,Atif,BentFlyvbjerg,AlexanderBudzierandDanielLunn.2016.Doesinfrastructureinvestmentleadtoeconomicgrowthoreconomicfragility?EvidencefromChina.OxfordReviewofEconomicPolicy32(3):360–390.
Arnold,Vincient.2023.China:CentralBankSwapstoArgentina,2014.JournalofFinancialCrises5(1):158–88.
AsianInfrastructureInvestmentBank(AIIB).2021.AIIB,MCDFPartnershiptoFosterHigh-QualityInfrastructureInvestments.25August2021.Beijing:AIIB.Accessedon15August2023athttps://www.aiib.org/en/news-events/news/2021/AIIB-MCDF-Partnership-to-Foster-High-Quality-Infrastructure-Investments.html
Asmus,Gerda,VeraZ.Eichenauer,AndreasFuchsandBradleyC.Parks.Forthcoming.DoesIndiaUseDevelopmentFinancetoCompetewithChina?ASubnationalAnalysis.JournalofConflictResolution.
Atkinson,Carol.2010.Doessoftpowermatter?Acomparativeanalysisofstudentexchangeprograms.ForeignPolicyAnalysis6:1-22.
Bae,Kee
‐HongandVidhanK.Goyal.2009.CreditorRights,Enforcement,andBankLoans.JournalofFinance64(2):823-860.
Baehr,Christian,ArielBenYishay,andBradleyC.Parks.Forthcoming.HighwaytotheForest?LandGovernanceandtheSitingandEnvironmentalImpactsofChineseGovernment-FundedRoadBuildinginCambodia.JournalofEnvironmentalEconomicsandManagement.
Bahaj,Saleem,andRicardoReis.2022.JumpstartinganInternationalCurrency.CEPRDiscussionPaperDP14793.London:CentreforEconomicPolicyResearch(CEPR).
Bailey,MichaelA.,AntonStrezhnev,andErikVoeten.2017.EstimatingdynamicstatepreferencesfromUnitedNationsvotingdata.JournalofConflictResolution61(2):430-56.
BancoCentraldeBolivia.2016.ChinaEximBank:TheExport-ImportBankofChina[Slidedeck].Retrievedfromhttps://www.dropbox.com/s/ryfvumv7fl85gg3/banco_central_bol_EXIMBANK%20CHINA%202016.pdf?dl=0.
BankforInternationalSettlements(BIS).1995.GuidetotheBISstatisticsoninternationalbanking.Basel,Switzerland:BIS.
BankforInternationalSettlements(BIS).2018.Statisticalrelease:BISinternationalbankingstatisticsatend-June2018.Basel,Switzerland:BIS.Accessedon18July2023athttps://www.bis.org/statistics/rppb1810.pdf
BankforInternationalSettlements(BIS).2022.BISinternationalbankingstatisticsandgloballiquidityindicatorsatend-December2021.Basel,Switzerland:BIS.Accessedon19July2023athttps://www.bis.org/statistics/rppb2204.htm
BankforInternationalSettlements(BIS).2023.Statisticalrelease:BISinternationalbankingstatisticsandgloballiquidityindicatorsatend-December2022.Basel,Switzerland:BIS.Accessedon19July2023athttps://www.bis.org/statistics/rppb2304.htm
240
Barbaro,MichaelandAbdiLatifDahir.2023.AmericaHasaProbleminAfrica:China.NewYorkTimes.6April2023.Accessedon1August2023athttps://www.nytimes.com/2023/04/06/podcasts/the-daily/kamala-harris-us-africa-china.html.
Batson,Andrew.2008.ChinaUsedReservesToSwayCostaRica.13September2008.WallStreetJournal.Accessedon2August2023athttps://www.wsj.com/articles/SB122121919505927749
BBCNews.2015.ConservativeMauricioMacriWinsArgentinaPresidency.BBCNews.23November2015.Accessedon1August2023athttps://www.bbc.com/news/world-latin-america-34897150.
BeltandRoadForumforInternationalCooperation.2017.SpeechbyXiJinpingattheOpeningCeremonyoftheBeltandRoadForumforInternationalCooperationonMay14,2017inBeijing.Accessedon15August2023athttps://web.archive.org/web/20230514064500/http://www.beltandroadforum.org/n100/2017/0514/c24-407.html
Benabdallah,Lina.2020.ShapingtheFutureofPower:KnowledgeProductionandNetwork-BuildinginChina-AfricaRelations.AnnArbor,MI:UniversityofMichiganPress.
Bernhard,Isabel.2022.WhyArgentina’sNuclearPowerProjectWithChinaHasStalled.TheDiplomat.14December2022.Accessedon1August2023athttps://thediplomat.com/2022/12/why-argentinas-nuclear-power-project-with-china-has-stalled/.
Bertaut,Carol,BastianvonBeschwitz,StephanieCurcuru.2023.‘TheInternationalRoleoftheU.S.Dollar’Post-COVIDEdition.FEDSNotes.Washington:BoardofGovernorsoftheFederalReserveSystem.
Blair,RobertA.,RobertMarty,andPhilipRoessler.2022a.ForeignAidandSoftPower:GreatPowerCompetitioninAfricaintheEarlyTwenty-firstCentury.BritishJournalofPoliticalScience52(3):1355-1376.
Blair,RobertA.,SamanthaCuster,PhilipRoessler.2022b.DuelingAidRegimes:AConjointSurveyExperimentonElites’DevelopmentFinancePreferencesin141Low-andMiddle-IncomeCountries.WorkingPaper#119.Williamsburg,VA:AidDataatWilliam&Mary.
Bluhm,Robert,AxelFuchs,BradleyC.Parks,AustinM.Strange,andMichaelJ.Tierney.2020.ConnectiveFinancing:ChineseInfrastructureProjectsandtheDiffusionofEconomicActivityinDevelopingCountries.CEPRDiscussionPaper14818.London,UK:CEPR.
Bolton,JohnR.2019.ANewAfricaStrategy:ExpandingEconomicandSecurityTiesontheBasisofMutualRespect.SpeechattheHeritageFoundationon20February2019.Accessedon1August2023athttps://www.heritage.org/sites/default/files/2019-02/HL1306.pdf.
Bomprezzi,Pietro,AxelDreher,AndreasFuchs,TeresaHailer.AndreasKammerlander,LennartKaplan,SilviaMarchesi,TaniaMasi,CharlotteRobertandKerstinUnfried.2023.LightofTheirLives:CountryLeaders’SpousesandRegionalFavoritism.Mimeo.
Bon,GatienandGongCheng.2021.UnderstandingChina'sroleinrecentdebtreliefoperations:Acasestudyanalysis.InternationalEconomics166(C):23-41.
Bouey,Jennifer,LynnHu,KellerScholl,WilliamMarcellino,RafiqDossani,AmmarA.Malik,KyraSolomon,andShengZhang.Forthcoming.China’sAIExportsattheDawnofanAIAge:
241
Technology,DevelopmentFinance,andDataSafety.WashingtonD.C.:RANDCorporationandAidData.
Bourguignon,FrançoisandJean-PhilippePlatteau.2015.TheHardChallengeofAidCoordination.WorldDevelopment69(C):86-97.
Brandimarte,Walter,andManuelaTobias.2023.Argentina’sMileiSaysHe’dReject‘Assassin’China,LeaveMercosur.Bloomberg.16August2023.Accessedon16August2023athttps://www.bloomberg.com/news/articles/2023-08-16/milei-s-foreign-policy-reject-assassin-china-leave-mercosur.
Brauneder,KerstinM.,ChloeMonte,SimonBlyth,LeonBennun,StuartHMButchart,MichaelHoffmann,NeilDBurgess,AnnabelleCuttelod,MattI.Jones,ValKapos,JohnPilgrim,andMelissaJToll.2018.GlobalscreeningforCriticalHabitatintheterrestrialrealm.PLOSONE13(3):1-16.
Bräutigam,DeborahandKevinAcker.2020.IsChinaHidingitsOverseasLending?Horn,ReinhartandTrebesch'sHiddenLoansandHiddenData.SAIS-CARIBlog.Accessedon19July2023atwww.chinaafricarealstory.com/2020/04/is-china-hiding-its-overseas-lending.html
Bräutigam,Deborah.2019.MisdiagnosingtheChineseInfrastructurePush.TheAmericanInterest.4April2019.Accessedon1August2023atwww.the-american-interest.com/2019/04/04/misdiagnosing-the-chinese-infrastructure-push
Brautigam,Deborah.2021.HowZambiaandChinaCo-CreatedaDebt‘TragedyoftheCommons.’WorkingPaper2021/51.WashingtonDC:SAIS-CARI.
Brazys,Samuel.andAlexanderDukalskis.2019.Risingpowersandgrassrootsimagemanagement:ConfuciusInstitutesandChinainthemedia.ChineseJournalofInternationalPolitics12(4):557–584.
Broccolini,Chiara,GiuliaLotti,AlessandroMaffioli,AndreaF.PresbiteroandRodolfoStucchi.2021.MobilizationEffectsofMultilateralDevelopmentBanks.WorldBankEconomicReview35(2):521–543.
Bronstein,Hugh,andMaximilianHeath.2019.ChinaOpensSoymealMarkettoNo.1ExporterArgentinain‘historic’Deal.Reuters.10September2019.Accessedon1August2023athttps://www.reuters.com/article/us-argentina-soyproducts-idUSKCN1VV2JP.
Brooker,Matthew.2023.Zambia’sDebtDealWithChinaIsaLandmarkMoment.Bloomberg.25June2023.Accessedon1August2023athttps://www.bloomberg.com/opinion/articles/2023-06-25/zambia-debt-restructuring-with-china-is-a-landmark-moment.
Brooks,SarahM.,RaphaelCunha,andLaynaMosley.2015.Categories,Creditworthiness,andContagion:HowInvestors'ShortcutsAffectSovereignDebtMarkets.InternationalStudiesQuarterly59(3):587–601.
Bruni,Jo.2021.EcuadorstrikesfundingagreementwithUS'sDFC.LatinFinance.January15.2021.Accessedon16July2023athttps://www.latinfinance.com/daily-briefs/2021/1/15/ecuador-strikes-funding-agreement-with-uss-dfc
Buchanan,GraemeM.,PaulF.Donald,BradleyParks,BrianO’Donnell,JohnSwaddle,LukaszTracewski,DanielRunfola,andStuartH.M.Butchart.2018.TheLocalImpactsofWorldBank
242
DevelopmentProjectsNearSitesofConservationSignificance.JournalofEnvironmentandDevelopment27(3):299–322.
Buchheit,LeeC.1985.IsSyndicatedLendingaJointVenture?InternationalFinancialLawReview4:12-14.
Buchheit,LeeC.1991.MakingAmendsforAmendments.InternationalFinancialLawReview10:11.
Buchheit,LeeC.andMituGulati.2023.Enforcingcomparabletreatmentinsovereigndebtworkouts.CapitalMarketsLawJournal18(1):71-77.
BuenosAiresTimes.2023.UnitedStatesStepsupPressureonArgentinaoverFighterJetChoice.BuenosAiresTimes.26July2023.Accessedon1August2023athttps://www.batimes.com.ar/news/argentina/the-us-steps-up-pressure-on-argentina-to-choose-danish-f-16s-over-chinese-or-indian-fighter-jets.phtml.
Bulman,David,WalterKolkma,andAartKraay.2017.Goodcountriesorgoodprojects?ComparingmacroandmicrocorrelatesofWorldBankandAsianDevelopmentBankprojectperformance.TheReviewofInternationalOrganizations12(3):335-363.
Buntaine,MarkT.2015.AccountabilityinGlobalGovernance:CivilSocietyClaimsforEnvironmentalPerformanceattheWorldBank.InternationalStudiesQuarterly59(1):99–111.
Buntaine,MarkT.2011.DoestheAsianDevelopmentBankRespondtoPastEnvironmentalPerformancewhenAllocatingEnvironmentallyRiskyFinancing?WorldDevelopment39(3):336-350.
Buntaine,Mark.2016.GivingAidEffectively:ThePoliticsofEnvironmentalPerformanceandSelectivityatMultilateralDevelopmentBanks.Oxford:OxfordUniversityPress.
CenterforLatinAmericanandLatinoStudies.2022.ChangesandContinuitiesthroughSino-ArgentineanRecentHistory.CenterforLatinAmericanandLatinoStudies.https://www.american.edu/centers/latin-american-latino-studies/upload/china-argentina-english-final.pdf.
CenterforInsightsinSurveyResearch(CISR).2023.NationalSurveyofBangladesh:March-April2023.Accessedathttps://www.iri.org/wp-content/uploads/2023/08/April-2023-National-Survey-Public-Release.pdf
CentralBankofSudan(CBOS).2023.SudanExternalDebtPositionFirstQuarterasat31stofMarch2022.Khartoum:CBOS.Accessedon19July2023athttps://web.archive.org/web/20230208195131/https:/cbos.gov.sd/sites/default/files/Quarter%201%20Report%20%202022%20.pdf
CentralIntelligenceAgency.2023.TheWorldFactbook.InCIAWorldFactbook.https://www.cia.gov/the-world-factbook/countries/zambia/.
Cerutti,Eugenio,CatherineCasanova,andSwapan-KumarPradhan.2023.BankingAcrossBorders:AreChineseBanksDifferent?JournalofBanking&Finance154:106920.
Cerutti,Eugenio,CatherineKoch,andSwapan-KumarPradhan.2018.ThegrowingfootprintofEMEbanksintheinternationalbankingsystem.BISQuarterlyReview.Accessedon19July2023athttps://www.bis.org/publ/qtrpdf/r_qt1812e.htm
243
Chen,Xiaoyi.2014.
揭
开
外
储
委
托
贷
款
的
神
秘
面
纱(UnveiltheMysteriesoftheEntrustLoansoftheForeignExchangeReservesinChina).21August2014.Accessedon15September2023athttps://www.dropbox.com/scl/fi/r5idp65hfybugsmc5v7b3/.pdf?rlkey=xr8jqe079moeb50pybrjh90bx&dl=0
Chen,Muyang.2020a.BeyondDonation:China’sPolicyBanksandtheReshapingofDevelopmentFinance.StudiesinComparativeInternationalDevelopment55:436–459.
Chen,Muyang.2020b.StateActors,MarketGames:CreditGuaranteesandtheFundingofChinaDevelopmentBank.NewPoliticalEconomy25(3):453-468.
Chen,Muyang.2023.China’sriseandthereshapingofsovereigndebtrelief.InternationalAffairs99(4):1755-1775.
Chen,YunnanandDavidLandry.2018.Capturingtherains:ComparingChineseandWorldBankhydropowerprojectsinCameroonandpathwaysforSouth-SouthandNorthSouthtechnologytransfer.EnergyPolicy115:561–571.
Cheng,Wenting,andKaiZhang.2023.Setting‘green’boundariesforChinesegreenfinance:Multi-levelgovernanceandregulatorystringency.EnvironmentalPolicyandGovernance1-13.
Chengjun,Zhou.2021.AnewperspectiveontheinternationalapplicationofRMBBonds.ChinaBond11:19.Accessedon1September2023athttps://finance.sina.cn/bond/zsyw/2021-12-03/detail-ikyamrmy6550370.d.html?vt=4&cid=76633&node_id=76633
Chin,GregoryT.,andKevinP.Gallagher.2019.CoordinatedCreditSpaces:TheGlobalizationofChineseDevelopmentFinance.DevelopmentandChange50(1):245–274.
ChinaBankingRegulatoryCommission(CBRC).2017a.MeasuresfortheSupervisionandAdministrationofChinaDevelopmentBank.15November2017.Beijing:CBRC.Accessedon15August2023athttps://www.gov.cn/gongbao/content/2018/content_5260809.htm
ChinaBankingRegulatoryCommission(CBRC).2017b.MeasuresfortheSupervisionandAdministrationoftheExport-ImportBankofChina.15November2017.Beijing:CBRC.Accessedon15August2023athttps://www.gov.cn/gongbao/content/2018/content_5264999.htm
ChinaDaily.2009.HuijintofundExim,Sinosure.17September2009.ChinaDailyAccessedon1August2023atwww.chinadaily.com.cn/business/2009-09/17/content_8701521.htm
ChinaDevelopmentBank(CDB).2020.PricingSupplement.19November2020.Accessedathttps://secure.fundsupermart.com/fsm/bond/relatedBondDocument/XS2263032786/China%20Development%20Bank_Pricing%20Supplement%20_%20Programme%20OC_Issue%20of%20CNY2,800,000,000%203.03%20per%20cent.%20Notes%20due%202023.pdf
Chitala,Mbita.2022.CorporateCapture;ThePoliticalEconomyofElectricityManagementinZambia2014-2021:HowNottoManageaStateEnterprise.Lusaka:ZambiaResearchFoundation.
Chuku,Chuku,PrateekSamal,JoyceSaito,DaliaS.Hakura,MarcosdChamon,MartinD.Cerisola,GuillaumeChabert,andJerominZettelmeyer.2023.AreWeHeadingforAnotherDebtCrisisinLow-IncomeCountries?DebtVulnerabilities:TodayvsthePre-HIPCEra.IMFWorkingPapers2023(079).
244
Cordella,TitoandPowell,Andrew.2021.Preferredandnon-preferredcreditors.JournalofInternationalEconomics132(C):103491
Cruces,JuanJ,andChristophTrebesch.2013.SovereignDefaults:ThePriceofHaircuts.AmericanEconomicJournal:Macroeconomics5(3):85–117.
Custer,Samantha,BrookeRussell,MatthewDiLorenzo,MengfanCheng,SidGhose,JakeSims,JenniferTurner,andHarshDesai.2018.TiesthatBind:QuantifyingChina’sPublicDiplomacyanditsGoodNeighborEffect.Williamsburg,VA:AidDataatWilliam&Mary.
Custer,Samantha,TanyaSethi,JonathanA.Solis,JoyceJiahuiLin,SiddharthaGhose,AnubhavGupta,RodneyKnight,andAustinBaehr.2019.SilkRoadDiplomacy:DeconstructingBeijing’sToolkittoInfluenceSouthandCentralAsia.Williamsburg,VA:AidDataatWilliam&Mary.
Custer,Samantha,TanyaSethi,RodneyKnight,AmberHutchinson,VeraChoo,MengfanCheng.2021.ListeningtoLeaders2021:Areportcardfordevelopmentpartnersinaneraofcontestedcooperation.Williamsburg,VA:AidDataattheCollegeofWilliam&Mary.
Custer,Samantha.2022.(Re)investinginOurReputationalSecurity:AlternativeModelsandOptionsStrengthenU.S.StrategicCommunications.Williamsburg,VA:AidDataatWilliam&Mary.
Custer,Samantha,AxelDreher,Thai-BinhElston,BrookeEscobar,RoryFedorochko,AndreasFuchs,SiddharthaGhose,JoyceJiahuiLin,AmmarMalik,BradleyC.Parks,KyraSolomon,AustinStrange,MichaelJ.Tierney,LydiaVlasto,KatherineWalsh,FeiWang,LincolnZaleski,andShengZhang.2023.TrackingChineseDevelopmentFinance:AnApplicationofAidData’sTUFF3.0Methodology.Williamsburg,VA:AidDataatWilliam&Mary.
Dawn.2021.Pakistandoesnothavea'Chinadebt'problem:AsadUmaronUSreportcriticalofCPEC.Dawn.6October2021.Accessedon20September2023athttps://www.dawn.com/news/1650453
Dennis,StevenA.andDonaldJ.Mullineaux.2000.Syndicatedloans.JournalofFinancialIntermediation9(4):404-426.
DiLorenzo,MatthewandMengfan.2019.PoliticalTurnoverandChineseDevelopmentCooperation.ChineseJournalofInternationalPolitics12(1):123-151.
DoRosario,Jorgelina.2023a.Exclusive:ArgentinatouseIMFmoneytopaybackpartofacurrencyswapwithChina,sourcessay.Reuters.24August2023.Accessedon24August2023athttps://www.reuters.com/world/americas/argentina-use-imf-money-pay-back-part-currency-swap-with-china-sources-2023-08-23/
DoRosario,Jorgelina.2023b.ArgentinausedChinayuanswaptopaytheIMFinOctober-source.Reuters.31October2023.Accessedon31October2023athttps://www.nasdaq.com/articles/argentina-used-china-yuan-swap-to-pay-the-imf-in-october-source
DoRosario,JorgelinaandJorgeOtaola.2023.UPDATE3-Argentina,shortofdollars,touseyuanin$2.7blnIMFpayment.Reuters.29June2023.Accessedon29June2023athttps://www.reuters.com/article/argentina-imf/update-3-argentina-short-of-dollars-to-use-yuan-in-2-7-bln-imf-payment-idINL1N38L21F
DoRosario,JorgelinaandKarinStrohecker.2023.Analysis:ChinaholdsthekeytoavoidingArgentina'sIMFdefault.Thepricetagisunknown.Reuters.2August2023Accessedon2
245
August2023athttps://www.reuters.com/markets/china-holds-key-avoiding-argentinas-imf-default-price-tag-is-unknown-2023-08-02/
DoRosario,JorgelinaandRachelSavage.2023.Chinadevelopmentloanstoemergingeconomieshit13-yearlowin2021.Reuters.24January2023,Accessedon17July2023athttps://www.reuters.com/markets/asia/china-development-loans-emerging-economies-hit-13-year-low-2021-study-2023-01-24/
Dolan,CarrieB.,AmmarA.Malik,ShengZhang,WenhuiMao,KaciKennedyMcDade,EliSvoboda,andJuliusN.Odhiambo.2023.ChineseHealthFundinginAfrica:TheUntoldStory.PLOSGlobalPublicHealth3(6):e0001637.
Dollar,David.2016.ChinaAsaGlobalInvestor.AsiaWorkingGroupPaper4.WashingtonDC:BrookingsInstitution
Dong,LiliandDavidW.Chapman.2008.TheChineseGovernmentScholarshipProgram:AnEffectiveFormofForeignAssistance?InternationalReviewofEducation54(2):155-173.
Douglas,Jason.2022.ChinaPropsUpBelt-and-RoadBorrowersViaUnusualChannel.11December2022.WallStreetJournal.Accessedon1August2023athttps://www.wsj.com/articles/china-props-up-belt-and-road-borrowers-via-unusual-channel-11670727153
Dreher,Axel,AndreasFuchs,BradleyParks,AustinM.Strange,andMichaelJ.Tierney.2018.ApplesandDragonFruits:TheDeterminantsofAidandOtherFormsofStateFinancingfromChinatoAfrica.InternationalStudiesQuarterly62(1):182-194.
Dreher,Axel,AndreasFuchs,BradleyParks,AustinStrange,andMichaelJ.Tierney.2021.Aid,China,andGrowth:EvidencefromaNewGlobalDevelopmentFinanceDataset.AmericanEconomicJournal:EconomicPolicy13(2):135-174.
Dreher,Axel,AndreasFuchs,BradleyParks,AustinStrange,andMichaelJ.Tierney.2022.BankingonBeijing:TheAimsandImpactsofChina’sOverseasDevelopmentProgram.Cambridge,UK:CambridgeUniversityPress.
Dreher,Axel,AndreasFuchs,RolandHodler,BradleyC.Parks,PaulA.Raschky,andMichaelJ.Tierney.2019.AfricanLeadersandtheGeographyofChina’sForeignAssistance.JournalofDevelopmentEconomics140:44–71.
Duggan,Julian,ScottMorris,JustinSandefur,andGeorgeYang.2020.IstheWorldBank’sCOVID-19crisislendingbigenough,fastenough?Newevidenceonloandisbursements.CGDWorkingPaper554.Washington,DC:CenterforGlobalDevelopment.
Dyer,Geoff,JamilAnderlini,andHennySender.2011.China’slendinghitsnewheights.17January2011.FinancialTimes.Accessedon3August2023athttps://www.ft.com/content/488c60f4-2281-11e0-b6a2-00144feab49a
EconomicRelationsDivisionoftheGovernmentofthePeople'sRepublicofBangladesh.2023.Wing-8:Asia&JEC.Accessedon14June2023athttps://erd.gov.bd/site/page/e0ed373a-4243-4dd4-aa10-f402dea4a3cf/e0ed373a-4243-4dd4-aa10-f402dea4a3cf
246
Eichenauer,VeraZ.AndreasFuchs,andLutzBrückner.2021.Theeffectsoftrade,aid,andinvestmentonChina'simageinLatinAmerica.JournalofComparativeEconomics49(2):483-498.
Ellis,Evan.2021.NewDirectionsintheDeepeningChinese-ArgentineEngagement.GlobalAmericans(blog).11February2021.Accessedon20September2023athttps://theglobalamericans.org/2021/02/new-directions-in-the-deepening-chinese-argentine-engagement/.
EmbassyofthePeople'sRepublicofChinaintheRepublicofthePhilippines.2019.BeltandRoadPrimer.Accessedon1July2023atph.china-embassy.gov.cn/eng/sgdt/201904/t20190411_1182334.htm
Esty,BenjaminC.,andWilliamL.Megginson.2003.Creditorrights,enforcement,anddebtownershipstructure:Evidencefromtheglobalsyndicateloanmarket.JournalofFinancialandQuantitativeAnalysis38:37–59.
Euractiv.2023.TheBeltandRoadInitiative--Wherearewe10yearson?EuractivHybridConferenceon19September2023.Accessedon19September2023athttps://events.euractiv.com/event/info/the-belt-and-road-initiative-where-are-we-10-years-on?_ga=2.41858210.1856855957.1695222202-589043501.1692974998
ExpressTribune.2016.ChineseDiplomatCallsCM‘ShahbazSpeed.’ExpressTribune.20December2016.Accessedon20September2023athttps://tribune.com.pk/story/1268314/unprecedented-development-chinesediplomat-calls-cm-shahbaz-speed.
Ferry,LaurenandAlexandraZeitz.2023.China,theIMF,andSovereignDebtCrisis.Paperpresentedatthe6thInterdisciplinarySovereignDebtResearchandManagementConference(DebtCon6)atPrincetonUniversityfrom27-29April2023.Accessedon1August2023athttps://debtcon6.princeton.edu/sites/g/files/toruqf3611/files/documents/ferry.pdf
Filkins,Dexter.2015.ADeadlyConspiracyinBuenosAires?TheNewYorker.13July2015.Accessedon15August2023athttps://www.newyorker.com/magazine/2015/07/20/death-of-a-prosecutor.
FinancialTimes.2016.ChinaRethinksDevelopingWorldLargesseasDealsSour.FinancialTimes.13October2016.Accessedon15August2023atwww.ft.com/content/5bf4d6d8-9073-11e6-a72e-b428cb934b78.
Flores-Macías,GustavoA.,andSarahKreps.2013.TheForeignPolicyConsequencesofTrade:China’sCommercialRelationswithAfricaandLatinAmerica,1992–2006.JournalofPolitics75(2):357–71.
Flyvbjerg,Bent,MetteSkamrisHolm,andSørenBuhl.2002.UnderestimatingCostsinPublicWorksProjects:ErrororLie?JournaloftheAmericanPlanningAssociation68(3):279–295.
Foysal,QuaziOmar.2014.TheChina-BangladeshRelations:FromEternitytoModernEra.Society&ChangeVIII(4).
FreedomHouse2022.FreedomintheWorld2022:TheGlobalExpansionofAuthoritarianRule.NewYork:FreedomHouse.
FreedomHouse.2023.FreedomintheWorld2023.https://freedomhouse.org/country/bangladesh/freedom-world/2023.
247
French,HowardW.2011.InAfrica,anElectionRevealsSkepticismofChineseInvolvement.TheAtlantic.29September2011.Accessedon15August2023athttps://www.theatlantic.com/international/archive/2011/09/in-africa-an-election-reveals-skepticism-of-chinese-involvement/245832/.
Fung,Courtney,andShing-honLam.2022.MixedReportCard:China’sInfluenceattheUnitedNations.LowyInstitute.https://www.lowyinstitute.org/publications/mixed-report-card-china-s-influence-united-nations#heading-7004.
Furukawa,Mitsuaki.2018.ManagementoftheInternationalDevelopmentAidSystem:TheCaseofTanzania.DevelopmentPolicyReview36(1):270–284.
Gallagher,KevinP.andRebeccaRay.2020.ScopeandFindings:China’sOverseasDevelopmentFinanceDatabase.BUGlobalDevelopmentPolicyCenter.Accessedon17July2023athttps://www.bu.edu/gdp/2020/12/13/scope-and-findings-chinas-overseas-development-finance-database/
Gardner,Alysha,JoyceLin,ScottMorris,andBradleyParks.2021.BargainingwithBeijing:ATaleofTwoBorrowers.CenterforGlobalDevelopmentandAidDataatWilliam&Mary.
Garnett,StephenT.,NeilD.Burgess,JohnE.Fa,AlvaroFernandez-Llamazares,ZsoltMolnar,CathyJ.Robinson,JamesE.M.Watson,KerstinK.Zander,BeauAustin,EduardoS.Brondizio,NeilFrenchCollier,TomDuncan,ErleEllis,HayleyGeyle,MichaV.Jackson,HarryJonas,PernillaMalmer,BenMcGowan,AmphoneSivongxay,andIanLeiper.2018.AspatialoverviewoftheglobalimportanceofIndigenouslandsforconservation.NatureSustainability1(7):369-374.
GBH.2015.ArgentineVotersFocusedOnEconomyInPresidentialElection.GBH.24October2015.Accessedon15August2023athttps://www.wgbh.org/news/2015-10-24/argentine-voters-focused-on-economy-in-presidential-election.
Gelpern,Anna,SebastianHorn,ScottMorris,BradParks,ChristophTrebesch.2021.HowChinaLends:ARareLookinto100DebtContractswithForeignGovernments.PetersonInstituteforInternationalEconomics,KielInstitutefortheWorldEconomy,CenterforGlobalDevelopment,andAidDataatWilliam&Mary.
Gelpern,Anna,SebastianHorn,ScottMorris,BradParks,ChristophTrebesch.2022.HowChinaLends:ARareLookinto100DebtContractswithForeignGovernments.EconomicPolicyeiac054.
Gelpern,Anna.2016.Sovereigndebt:nowwhat?YaleJournalofInternationalLaw41(2):45-96.
GeoNews.2018.PTIGovtDecidestoInvestigateMass-TransitProjects.GeoNews.24August2018.Accessedon15August2023athttps://www.geo.tv/latest/208466-pti-govt-decides-to-investigate-mass-transit-projects.
Ghani,Amel.2015.DivisiveProject:LDAHopestoSecureArchaeologyNOCinaWeek.TheExpressTribune.4November2015.Accessedon15August2023athttps://tribune.com.pk/story/985371/divisive-project-lda-hopes-to-secure-archaeology-noc-in-a-week.
Gilchrist,Simon,BinWei,VivianZ.Yue,andEgonZakrajšek.2022.Sovereignriskandfinancialrisk.JournalofInternationalEconomics136(C):103603.
248
Giusto,Patricio,andJuanManuelHarán.2023.Argentina’sFailedChinaPolicy.TheDiplomat.28February2023.Accessedon15August2023athttps://thediplomat.com/2023/02/argentinas-failed-china-policy/
Goldsmith,BenjaminE.,YusakuHoriuchi,andTerenceWood.2014.DoingWellbyDoingGood:TheImpactofForeignAidonForeignPublicOpinion.QuarterlyJournalofPoliticalScience9(1):87–114.
Goldsmith,BenjaminE.andYusakuHoriuchi.2012.InSearchofSoftPower:DoesForeignPublicOpinionMatterforUSForeignPolicy?WorldPolitics64(3):555–585.
Gong,Xue.2023.TheBeltandRoadInitiativeIsStillChina’s‘Gala’butWithoutasMuchLuster.CarnegieChina.WashingtonD.C.:CarnegieEndowmentforInternationalPeace.https://carnegieendowment.org/2023/03/03/belt-and-road-initiative-is-still-china-s-gala-but-without-as-much-luster-pub-89207.
Goodman,Seth,ShengZhang,AmmarMalik,BradleyC.Parks,andJacobHall.2023.AidData'sGeospatialGlobalChineseDevelopmentFinanceDataset.Mimeo.
Gormley,ToddA.2010.TheImpactofForeignBankEntryinEmergingMarkets:EvidencefromIndia.JournalofFinancialIntermediation19(1):26–51.
GovernmentofCostaRica.2007.EstablecimientoderelacionesdiplomáticasconlaRepúblicaPopularChinaharesultadobeneficiosoparaelpaís.10September2007.SanJose:MinisteriodeRelacionesExterioresyCultoDireccióndeComunicaciónInstitucional.Accessedon2August2023athttps://www.dropbox.com/scl/fi/g8kkmicps7o8z9vdxcn5d/1243-RPCH-beneficios.pdf?rlkey=3u5afpnqeeg6myy358wxj2kn9&dl=0
GovernmentofCostaRica.2008.LetterfromCostaRicanMinisterofFinanceGuillermoE.ZúñigatoSAFEDeputyAdministratorFangShangpu.7January2008.Accessedon13August2023athttps://www.dropbox.com/scl/fi/48qdeslzwim9t6dqhyk0m/7-January-2008-Letter-from-Minister-of-Finance-of-Costa-Rica-to-SAFE.pdf?rlkey=pr3p4advi1wfzzwjr5t8gva5m&dl=0
Guiso,Luigi,PaolaSapienza,andLuigiZingales.2009.CulturalBiasesinEconomicExchange?QuarterlyJournalofEconomics124(3):1095–1131.
Gurara,Daniel,AndreaPresbitero,andMiguelSarmiento.2020.Borrowingcostsandtheroleofmultilateraldevelopmentbanks:Evidencefromcross-borderbanksyndicatedlending.JournalofInternationalMoneyandFinance100:102090.
Hainz,C.andS.Kleimeier.2012.PoliticalRisk,ProjectFinance,andtheParticipationofDevelopmentBanksinSyndicatedLending.JournalofFinancialIntermediation21(2):287–314.
Halper,Stefan.2010.TheBeijingConsensus:HowChina’sAuthoritarianModelWillDominatetheTwenty-FirstCentury.Philadelphia:BasicBooks.
Hancock,TomandMatthewHill.2022.DebtDefaultsAreaStressTestforChina’sSoftPowerStrategy.BloombergBusinessweek.26September2022.Accessedon1October2023athttps://www.bloomberg.com/news/articles/2022-09-26/debt-crisis-puts-scrutiny-on-china-soft-power#xj4y7vzkg
Hasnain,Khalid.2023.OrangeLineTouches100mPassengersCount.Dawn.6August2023.Accessedon15August2023athttps://www.dawn.com/news/1768572.
249
Heurlin,Christopher.2020.AuthoritarianAidandRegimeDurability:SovietAidtotheDevelopingWorldandDonor–RecipientInstitutionalComplementarityandCapacity.InternationalStudiesQuarterly64(4):968–79.
Hillman,JonathanE.2020.A‘ChinaModel?’Beijing’sPromotionofAlternativeGlobalNormsandStandards.13March2020.CongressionalTestimony.WashingtonDC:CSIS.Accessedon1October2023athttps://www.csis.org/analysis/china-model-beijings-promotion-alternative-global-norms-and-standards.
Hillman,JonathanE.2020.Don’tLetChinaGreenwashItsBeltandRoad.NikkeiAsia.6October2020.Accessedon22August2023athttps://www.csis.org/analysis/dont-let-china-greenwash-its-belt-and-road
Hirschman,AlbertO.1958.TheStrategyofEconomicDevelopment.NewHaven,CT:YaleUniversityPress.
Hoffman,Constantinevon.2014.HowArgentinaWoundupinAnotherDefaultMess.CBSNews.31July2014.Accessedon15August2023athttps://www.cbsnews.com/news/q-a-the-argentina-bond-default-explained/.
Holslag,Jonathan.2011.ChinaandtheCoups:CopingwithPoliticalInstabilityinAfrica.AfricanAffairs110(440):367–386.
HongKongGreenFinanceAssociation(HKGFA)andGuangdongGreenFinanceCommittee.2022.CGTResearchSeriesPhase2:UnderstandingUseCasesoftheCommonGroundTaxonomy.Accessedon1October2023athttps://www.hkgreenfinance.org/wp-content/uploads/2022/09/CGT_Phase2reportENG_final.pdf
Horigoshi,Ana,SamanthaCuster,BryanBurgess,KelseyMarshall,VeraChoo,KatieAndrzejewski,andEmilyDumont.2022.DeliveringtheBeltandRoad:DecodingthesupplyofanddemandforChineseoverseasdevelopmentprojects.Williamsburg,VA:AidDataatWilliam&Mary.
Horn,Sebastian,BradleyC.Parks,CarmenM.Reinhart,andChristophTrebesch.2023a.ChinaasanInternationalLenderofLastResort.NBERWorkingPaper#31105.Cambridge,MA:NBER.
Horn,Sebastian,BradleyC.Parks,CarmenM.Reinhart,andChristophTrebesch.2023b.DebtDistressonChina’sBeltandRoad.AmericanEconomicAssociationPapers&Proceedings113:131-34.
Horn,Sebastian,CarmenM.Reinhart,andChristophTrebesch.2019.China’sOverseasLending.NBERWorkingPaperNo.26050.Cambridge,MA:NationalBureauofEconomicResearch.
Horn,Sebastian,CarmenM.Reinhart,andChristophTrebesch.2020a.China’sOverseasLending:AResponsetoOurCritics.CGDNote.WashingtonDC:CenterforGlobalDevelopment(CGD).Accessedon19July2023athttps://www.cgdev.org/sites/default/files/chinas-overseas-lending-response-our-critics.pdf
Horn,Sebastian,CarmenM.Reinhart,andChristophTrebesch.2020b.CopingwithDisasters:TwoCenturiesofInternationalOfficialLending.NBERWorkingPaperNo.27343.
Horn,Sebastian,CarmenM.Reinhart,andChristophTrebesch.2021.China’sOverseasLending.JournalofInternationalEconomics133:103539.
250
Horn,Sebastian,CarmenM.Reinhart,andChristophTrebesch.2022a.HiddenDefaults.WorldBankPolicyResearchWorkingPaperSeries9925.WashingtonDC:WorldBank.
Horn,Sebastian,CarmenM.Reinhart,andChristophTrebesch.2022b.HiddenDefaults.AEAPapersandProceedings112:531–35.
HumanRightsWatch.2021.Bangladesh:Eventsof2021.InWorldReport2022.NewYork,NY:HumanRightsWatch.Accessedon15August2023athttps://www.hrw.org/world-report/2022/country-chapters/bangladesh.
Humphrey,Chris,andKatharinaMichaelowa.2019.ChinainAfrica:CompetitionforTraditionalDevelopmentFinanceInstitutions?WorldDevelopment120:15–28.
Humphrey,Christopher.2015.InfrastructureFinanceintheDevelopingWorld:ChallengesandOpportunitiesforMultilateralDevelopmentBank.WashingtonDC:G-24.
Hwang,Jyhjong,,OyintareladoMoses,LucasEngel,andSobiaShadbar.2022.ChineseLoanstoAfricaDuringtheCOVID-19Pandemic.BUGlobalDevelopmentPolicyCenter.GCIPolicyBrief012.Accessedon17July2023athttps://www.bu.edu/gdp/files/2022/11/GCI_PB_012_EN_FIN.pdf
Iacoella,Francesco,BrunoMartorano,LauraMetzger,andMarcoSanfilippo.2021.ChineseOfficialFinanceandPoliticalParticipationinAfrica.EuropeanEconomicReview136:103741.
IndependentEvaluationGroupoftheWorldBank(IEG).2010.SafeguardsandSustainabilityPoliciesinaChangingWorld.WashingtonDC:WorldBank.
InternationalMonetaryFund(IMF).2020.IMFPolicyPaper:TheEvolutionofPublicDebtVulnerabilitiesinLowerIncomeEconomies.WashingtonDC:IMF.
InternationalMonetaryFund(IMF).2021.RequestforanExtendedArrangementUndertheExtendedFundFacility--PressRelease;StaffReport;StaffStatement;andStatementbytheExecutiveDirectorforSuriname.IMFCountryReportNo.21/280.WashingtonDC:IMF.Accessedathttps://www.imf.org/-/media/Files/Publications/CR/2021/English/1SUREA2021001.ashx
InternationalMonetaryFund(IMF).2022.FirstReviewUndertheExtendedArrangementundertheExtendedFundFacility,andFinancingAssurancesReview--PressReleease;StaffReport;StaffStatement;andStatementbytheExecutiveDirectorforSuriname.IMFCountryReportNo.22/90.WashingtonDC:IMF.Accessedon3August2023athttps://www.imf.org/en/Publications/CR/Issues/2022/03/25/Suriname-First-Review-under-the-Extended-Arrangement-under-the-Extended-Fund-Facility-and-515732
InternationalMonetaryFund.2017.Zambia:StaffReportonthe2017ArticleIVConsultation.DebtSustainabilityAnalysis.WashingtonD.C.:InternationalMonetaryFund.https://www.imf.org/en/Publications/Search#sort=%40imfdate%20descending&numberOfResults=50&f:country=[Zambia]&f:series=[COUNTRYREPS,DSAS]:~:text=Zambia%3A%20Debt%20Sustainability%20Analysis%3B%20IMF%20Country%20Report%20No.%2017/327%3B%20September%2025%2C%202017.
InternationalMonetaryFund.2023.InternationalFinancialStatistics.IMFData.https://data.imf.org/?sk=4c514d48-b6ba-49ed-8ab9-52b0c1a0179b.
InternationalPlatformonSustainableFinance(IPSF).2022.CommonGroundTaxonomy–ClimateChangeMitigation.InstructionReportoftheIPSFTaxonomyWorkingGroupCo-chaired
251
bytheEUandChina.3June2022.Accessedon1October2023athttps://finance.ec.europa.eu/system/files/2022-06/220603-international-platform-sustainable-finance-common-ground-taxonomy-instruction-report_en.pdf
Isaksson,Ann-SofieandAndreasKotsadam.2018a.ChineseAidandLocalCorruption.JournalofPublicEconomics159:146–159.
Isaksson,Ann-Sofie,andAndreasKotsadam.2018b.RacingtotheBottom?ChineseDevelopmentProjectsandTradeUnionInvolvementinAfrica.WorldDevelopment106(C):284–298.
Isaksson,Ann-Sofie.2020.ChineseAidandLocalEthnicIdentification.InternationalOrganization74(4):833–852.
Islam,Md.Safiqul,HuangAilian,andZhangJie.2018.MajorChallengesandRemediesinBuildingtheBangladesh-China-India-MyanmarEconomicCorridor.ChinaQuarterlyofInternationalStrategicStudies04(04):613–29.
Islam,Syful.2021.BangladeshHitsBackafterChinaEnvoyWarnsagainstJoiningQuad.NikkeiAsia.11May2021.Accessedon15August2023athttps://asia.nikkei.com/Politics/International-relations/Bangladesh-hits-back-after-China-envoy-warns-against-joining-Quad.
Ivashina,Victoria.2009.Asymmetricinformationeffectsonloanspreads.JournalofFinancialEconomics92(2):300–319.
Jansson,Johanna.2013.TheSicominesAgreementRevisited:PrudentChineseBanksandRisk-TakingChineseCompanies.ReviewofAfricanPoliticalEconomy40(135):152–62.
Jia,Chen.2015.Policybanksget$90bcashinfusion.19August2015.ChinaDaily.Accessedon10July2023athttps://www.chinadaily.com.cn/business/2015-08/19/content_21643373.htm
Jinping,Xi.2017.FullTextofPresidentXi’sSpeechatOpeningofBeltandRoadForum.SpeechpresentedattheTheBeltandRoadForumforInternationalCooperation,Beijing,May14.https://www.fmprc.gov.cn/mfa_eng/wjdt_665385/zyjh_665391/201705/t20170527_678618.html.
Jones,Catherine.2018.China’sChallengetoLiberalNorms:TheDurabilityofInternationalOrder.London:PalgraveMacmillanUK.
Jones,Lee.2020.China’sBeltandRoadInitiativeIsaMess,NotaMasterPlan.ForeignPolicy(blog).October9,2020.https://foreignpolicy.com/2020/10/09/china-belt-and-road-initiative-mess-not-master-plan/.
Joosse,Alexandra,AmmarA.Malik,ShengZhang,andThai-BinhElston.2023.NetworksoftheBelt&Road:TheHiddenRoleofFinancialBrokers.Mimeo.
JusticeEjazAfzalKhan.2017.NESPAK,PunjabMassTransitAuthority,LahoreDevelopmentAuthority,ProvinceofPunjabandCivilSocietyNetworkVersusKamilKhanMumtaz&othersandProvinceofPunjab.SupremeCourtofPakistan.https://www.supremecourt.gov.pk/online-case-status/
Kaminsky,GracielaandCarmenM.Reinhart.2000.OnCrises,ContagionandConfusion.JournalofInternationalEconomics51(1):145-168.
252
Kastner,ScottA.2016.BuyingInfluence?AssessingthePoliticalEffectsofChina’sInternationalTrade.JournalofConflictResolution60(6):980–1007.
Kenny,Charles.2023.WorldBankInvestmentProjectsAren’tDesignedforCrises.15March2023.WashingtonDC:CenterforGlobalDevelopment.Accessedon27August2023athttps://www.cgdev.org/blog/world-bank-investment-projects-arent-designed-crises
Kern,Andreas,BernhardReinsberg,andPatrickShea.2022.IMFPrograms,ChineseLending,andthePoliticalEconomyofLeaderSurvival.WorkingPaper#118.Williamsburg,VA:AidDataatWilliam&Mary.
Khan,Yasir.2018.MetroTrainLahore,aGiftfromChina.ChinaToday.Accessedon28September2023.http://www.chinatoday.com.cn/ctenglish/2018/ii/202011/t20201105_800226035.html.
Kim,JimYong.2017.RemarksofWorldBankGroupPresidentJimYongKimattheBeltandRoadForumforInternationalCooperationon14May2017.Accessedat15August2023athttps://www.worldbank.org/en/news/speech/2017/05/14/remarks-of-world-bank-group-president-jim-yong-kim
Kimenyi,MwangiS.,andAmyCopley.2014.TheDeathofPresidentMichaelSataandIssuesoftheHealthofAfricanLeaders.Brookings(blog).30October2014.Accessedat15August2023athttps://www.brookings.edu/articles/the-death-of-president-michael-sata-and-issues-of-the-health-of-african-leaders/.
Kishida,Fumio.2023.TheFutureoftheIndo-Pacific:Japan’sNewPlanforaFreeandOpenIndo-Pacific.MinistryofForeignAffairsofJapan.https://www.mofa.go.jp/files/100477791.pdf.
Kroenig,Matthew,MelissaMcAdam,andStevenWeber.2010.TakingSoftPowerSeriously.ComparativeStrategy29(5):412-431.
Kondo,IlleninO.,AstghikMkhitaryan,andCesarSosa-Padilla.2022.BorrowingintheShadowofChina.Mimeo.
Kong,Bo,andKevinP.Gallagher.2017.GlobalizingChineseEnergyFinance:TheRoleofPolicyBanks.JournalofContemporaryChina26(108):834–851.
Koop,Fermín.2015.NewArgentinaPresidenttoPutDealswithChinaundertheMicroscope.DialogoChino.27November2015.Accessedon15August2023athttps://dialogochino.net/en/trade-investment/4292-new-argentina-president-to-put-deals-with-china-under-the-microscope/.
Koop,Fermín.2016.ArgentinaConfirmsSantaCruzDamProjects.DialogoChino(blog).15April2016.Accessedon15August2023athttps://dialogochino.net/en/climate-energy/5888-argentina-confirms-santa-cruz-dam-projects/.
Koop,Fermín.2018.MacriExtendsHandtoChinaasTradeDeficitGrows.BuenosAiresTimes.5May2018.Accessedon15August2023athttps://www.batimes.com.ar/news/economy/macri-extends-hand-to-china-as-trade-deficit-grows.phtml.
Koop,Fermín.2019.NewArgentinaPresidentLooksSettoDeepenChinaTies.DialogoChino.31October2019.Accessedon15August2023at
253
https://dialogochino.net/en/trade-investment/31387-new-argentina-president-looks-set-to-deepen-china-ties/.
Kose,M.Ayhan,SergioKurlat,FranziskaOhnsorge,andNaotakaSugawara.2022.ACross-CountryDatabaseofFiscalSpace.JournalofInternationalMoneyandFinance128(November):102682.
Kucik,Jeffrey.2017.Trump’sTradeWarBegins.ForeignAffairs,October10,2017.https://www.foreignaffairs.com/articles/united-states/2017-10-10/trumps-trade-war-begins.
Kuo,MercyA.2021.China’sBRILending:$385Billionin‘HiddenDebts’.TheDiplomat.29November2021.Acccesedon27August2023athttps://thediplomat.com/2021/11/chinas-bri-lending-385-billion-in-hidden-debts/
Kuziemko,Ilyana,andEricWerker.2006.HowMuchIsaSeatontheSecurityCouncilWorth?ForeignAidandBriberyattheUnitedNations.JournalofPoliticalEconomy114(5):905–930.
Lagarda,Guillermo,JenniferLinares,ArnoldoLopezMarmolejoandOsmelManzano.2018.EfficiencyBeforeEffectiveness:TheCaseoftheInter-AmericanDevelopmentBank.GEGIWorkingPaper022.Boston,MA:BostonUniversityGlobalDevelopmentPolicyCenter.
Lagarde,Christine.2019.BRI2.0:StrongerFrameworksintheNewPhaseofBeltandRoad.SpeechdeliveredatBeltandRoadForumon26April2019.Accessedon1August2023athttps://www.imf.org/en/News/Articles/2019/04/25/sp042619-stronger-frameworks-in-the-new-phase-of-belt-and-road
Landers,Clemence,NancyLeeandScottMorris.2021.WhyDoesDFCWanttoPayOffEcuador’sChineseCreditors?19January2021.WashingtonDC:CenterforGlobalDevelopment.Accessedon16July2023athttps://www.cgdev.org/blog/why-does-dfc-want-pay-ecuadors-chinese-creditors
Laurance,WilliamF.,SeanSloan,LingfeiWengandJeffreyA.Sayer.2015.EstimatingtheEnvironmentalCostsofAfrica’sMassiveDevelopmentCorridors.CurrentBiology25(24):3202–3208.
Lee,ChingKwan.2018.TheSpecterofGlobalChina:Politics,Labor,andForeignInvestmentinAfrica.Chicago,IL:UniversityofChicagoPress.https://press.uchicago.edu/ucp/books/book/chicago/S/bo22657847.html.
Lin,JustinYifuandYanWang.2014.China-Africaco-operationinstructuraltransformation:ideas,opportunities,andfinances.UNU-WIDERWorkingPaper2014/046.Helsinki:UnitedNationsUniversity.
Liu,ZongyuanZoe.2023.SovereignFunds:HowtheCommunistPartyofChinaFinancesItsGlobalAmbitions.Cambridge,MA:HarvardUniversityPress.
Longstaff,FrancisA.,Pan,Jun,Pedersen,LasseH.andSingleton,KennethJ.2011.HowSovereignIsSovereignCreditRisk?AmericanEconomicJournal:Macroeconomics3(2):75-103.
Lu,Christina.2023.China’sBeltandRoadtoNowhere.ForeignPolicy.13February2023.Accessedon10July2023athttps://foreignpolicy.com/2023/02/13/china-belt-and-road-initiative-infrastructure-development-geopolitics/
254
Lu,Yangsiyu,CeciliaSpringer,andBjarneSteffen.2023a.CofinancingandInfrastructureProjectOutcomesinChineseLendingandOverseasDevelopmentFinance.GCIWorkingPaper#27.Boston,MA:BostonUniversityGlobalDevelopmentPolicyCenter.
Lu,Yangsiyu,XiaoYanZhou,andB.AlexanderSimmons.2023b.EmpiricalanalysisofChineseoverseaspowerplantinvestments:Likelihoodofsuspensionsandassociatedenvironmentalrisks.iScience26(8):107457.
Lucci,Juan.2019.IsItPossibletoSayNotoChina?TheCaseoftheKirchner-CepernicDamsinArgentinePatagonia.LeadershipAcademyforDevelopment.PaloAlto,CA:StanfordUniversity.
Ma,Guonan.2006.SharingChina’sBankRestructuringBill.China&WorldEconomy14(3):19-37.
Majid,Hadia,AmmarMalik,andKateVyborny.2018.InfrastructureInvestmentsandPublicTransportUse:EvidencefromLahore,Pakistan.IGCWorkingPaperC-89231-PAK-1.London,UK:IGC.Accessedon1October2023athttps://www.dropbox.com/s/2gbnajqo55d14jz/BRT_IGC_WP.pdf?dl=0.
Malik,AmmarA.andBradleyC.Parks.2021.HiddendebtexposuretoChina:Whatisit,whereisit,andshouldwebeconcerned?Williamsburg,VA:AidDataatWilliam&Mary.
Malik,AmmarA.,BradleyC.Parks,BrookeRussell,JoyceLin,KatherineWalsh,KyraSolomon,ShengZhang,Thai-BinhElston,andSethGoodman.2021.BankingontheBeltandRoad:InsightsfromaNewGlobalDatasetof13,427ChineseDevelopmentProjects.Williamsburg,VA:AidDataatWilliam&Mary.
Mandan,Tanvi.2020.EmergingGlobalIssues:TheChina-IndiaBoundaryCrisisandItsImplications.WashingtonD.C.:U.S.BrookingsInstitute.Accessedon18August2023viahttps://www.brookings.edu/articles/emerging-global-issues-the-china-india-boundary-crisis-and-its-implications/.
Mandon,PierreandMarthaTesfayeWoldemichael.2023.HasChineseAidBenefitedRecipientCountries?EvidencefromaMeta-regressionAnalysis.WorldDevelopment166:106211.
Mannan,Md.Abdul.2018.Islamo-Nationalism,DomesticPolitics,andBangladesh’sPolicyofBalancingagainstIndiasincethe1990s.InternationalAreaStudiesReview21(4):340–64.
Martin,C.S.,M.J.Tolley,E.Farmer,C.J.Mcowen,J.L.Geffert,J.P.W.Scharlemann,H.L.Thomas,J.H.vanBochove,D.Stanwell-Smith,J.M.Hutton,B.Lascelles,J.D.Pilgrim,J.M.M.Ekstrom,D.P.Tittensor.2015.Aglobalmaptoaidtheidentificationandscreeningofcriticalhabitatformarineindustries.MarinePolicy53(1):45-53.
Martorano,Bruno,LauraMetzger,andMarcoSanfilippo.2020.ChinesedevelopmentassistanceandhouseholdwelfareinSub-SaharanAfrica.WorldDevelopment129(C):104909.
Marty,Robert,SethGoodman,MichaelLeFew,CarrieB.Dolan,ArielBenYishay,andDanRunfola.2019.AssessingtheCausalImpactofChineseAidonVegetativeLandCoverinBurundiandRwandaUnderConditionsofSpatialImprecision.DevelopmentEngineering4:100038.
Marx,Benjamin.2018.ElectionsasIncentives:ProjectCompletionandVisibilityinAfricanPolitics.Mimeo.Accessedon1October2023athttps://drive.google.com/file/d/1aSDEhq-ZGlUcX5QxWbb-0bAcAdSUXlKc/view
255
Mertha,Andrew.2009.‘FragmentedAuthoritarianism2.0’:PoliticalPluralizationintheChinesePolicyProcess.TheChinaQuarterly200(December):995–1012.
Mfula,Chris.2011.Interview-ZambiaOppositionLeaderSaystoKeepChinaTies.Reuters.16September2011.Accessedon1October2023athttps://www.reuters.com/article/zambia-opposition-idAFL5E7KG14Y20110916.
Mfula,Chris.IMF'sGeorgievaseesbroaddealwithChinaondebtrestructurings.Reuters.24January2023.Accessedon30July2023athttps://www.reuters.com/world/africa/imf-chief-says-its-time-creditors-restructure-zambias-debt-2023-01-24/
Miller,Matthew.2023.U.S.DepartmentofStatePressBriefingbyDepartmentSpokespersonon24May2023.WashingtonDC:U.S.DepartmentofState.Accessedon15September2023athttps://www.state.gov/briefings/department-press-briefing-may-24-2023/
Mingey,MatthewandAgathaKratz.2021.China’sBeltandRoad:DownbutnotOut.RhodiumGroupNote.Accessedon21July2023athttps://rhg.com/research/bri-down-out/
Mingey,MatthewandLoganWright.2023China’sExternalDebtRenegotiationsAfterZambia.RhodiumGroupNote.Accessedon1August2023athttps://rhg.com/research/chinas-external-debt-renegotiations-after-zambia/
MinistryofCommerce(MOFCOM)andMinistryofEcologyandEnvironment(MEE)ofthePeople'sRepublicofChina.2021.GuidlinesforGreenDevelopmentinForeignInvestmentandCooperation.15July2021.Beijing:MOFCOMandMEE.Accessedon27August2023athttps://www.followingthemoney.org/wp-content/uploads/2022/07/2021_MOFCOM-MEE_Guidelines-on-Green-Development-in-Foreign-Investment-and-Cooperation_E.pdf
MinistryofCommerceofthePeople'sRepublicofChina(MOFCOM).2014.MOFCOMHeldaMediaDebriefingontheMeasuresfortheAdministrationofForeignAid.Beijing,China:MinistryofCommerce.Accessedon1August2023atwww.mofcom.gov.cn/article/fbhfn/fbh2014/201412/20141200824824.shtml.
MinistryofFinanceandEconomicDevelopment(MOFED).2022.AnnualPublicDebtBulletin2021FinancialYear.Harare,Zimbabwe:MinistryofFinanceandEconomicDevelopment(MOFED).Accessedon20July2023athttps://veritaszim.net/sites/veritas_d/files/2021%20Annual%20Public%20Debt%20Bulletin.pdf
MinistryofFinanceofthePeople'sRepublicofChina.2020.WrittenInterviewwithFinanceMinisterLiuKunonG20DebtAgenda.Beijing:MinistryofFinanceofthePeople'sRepublicofChina.Accessedon1August2023atwww.mof.gov.cn/en/news/spe/202011/t20201120_3626593.htm
MinistryofForeignAffairsofthePeople'sRepublicofChina.2019.WorkingTogethertoDeliveraBrighterFutureForBeltandRoadCooperation.KeynoteSpeechbyH.E.XiJinpingPresidentofthePeople'sRepublicofChinaAttheOpeningCeremonyoftheSecondBeltandRoadForumforInternationalCooperation.26April2019.Accessedon14August2023athttps://www.fmprc.gov.cn/mfa_eng/wjdt_665385/zyjh_665391/201904/t20190426_678729.html
MinistryofForeignAffairsofthePeople’sRepublicofChina.2014.XiJinpingHoldsTalkswithPresidentCristinaFernándezdeKirchnerofArgentinaandTwoHeadsofStateAnnouncetheEstablishmentofChina-ArgentinaComprehensiveStrategicPartnership.MinistryofForeignAffairsofthePeople’sRepublicofChina.July19,2014.
256
https://www.fmprc.gov.cn/eng/gjhdq_665435/3447_665449/3453_664968/3455_664972/201407/t20140722_594264.html.
MinistryofForeignAffairsofthePeople’sRepublicofChina.2016.XiJinpingHoldsTalkswithPrimeMinisterSheikhHasinaofBangladeshBothSidesAgreetoUpgradeChina-BangladeshRelationstoStrategicPartnershipofCooperationandContinuouslyMoveForwardBilateralRelationshipatHigherLevels,October14,2016.https://www.fmprc.gov.cn/mfa_eng/topics_665678/2016zt/XJPDJPZMJLGJXGSFWBCXZYDGAJXDJZGJLDRDBCHW/201610/t20161018_704267.html.
MinistryofForeignAffairs.2023.Indo-PacificOutlookofBangladesh.MinistryofForeignAffairs,Dhaka.https://mofa.gov.bd/site/press_release/d8d7189a-7695-4ff5-9e2b-903fe0070ec9.
Moody'sInvestorService.2022.ChinaGreenBondPrincipleswillsupportthedomesticsustainablefinancemarket’sdevelopmentandalignwithinternationalstandards.30August2022.Accessedon1October2023athttps://www.moodys.com/web/en/us/capabilities/esg/resources/insights/china-green-bond-principles-will-support-the-domestic-sustainable-finance-markets-development-and-align-with-international-standards.html
Morris,Scott,BradParks,andAlyshaGardner.2020.ChineseandWorldBankLendingTerms:ASystematicComparisonacross157Countriesand15Years.CGDPolicyPaper170.Washington,DC:CenterforGlobalDevelopment.
Moses,OyintareladoJyhjongHwang,LucasEngel,andVictoriaYvonneBien-Aimé.2023.ANewStateofLending:ChineseLoanstoAfrica.GCIPolicyBrief019.Boston,MA:BostonUniversity'sGlobalDevelopmentPolicyCenter.
Mupuchi,Speedwell.2006.Zambia:SataIsMisleadingtheNationOverTaiwan-KK.ThePost.29August2006.Accessedon1October2023athttps://allafrica.com/stories/200608290697.html.
Mutati,Felix.2016.PreparingfortheWorst.AfricaConfidential57(25).16December2016.Accessedon20September2023athttps://www.africa-confidential.com/article-preview/id/11858/Preparing_for_the_worst.
Myers,MargaretandRebeccaRay.2023.AtaCrossroads:ChineseDevelopmentFinancetoLatinAmericaandtheCaribbean,2022.InterAmericanDialogueandBUGlobalDevelopmentPolicyCenter.Accessedon17July2023athttps://www.thedialogue.org/wp-content/uploads/2023/03/Chinese-Development-Finance-2023-FINAL.pdf
NarainD.,MaronM.,TeoH.C.,HusseyK.,LechnerA.M.2020.Best-practicebiodiversitysafeguardsforBeltandRoadInitiative’sfinanciers.NatureSustainability3:650–657.
NarainD.,TeoH.C.,LechnerA.M.,WatsonJ.E.,MaronM.2022.BiodiversityrisksandsafeguardsofChina’shydropowerfinancinginBeltandRoadInitiative(BRI)countries.OneEarth5:1019–1029.
Nasrin,Sharmina.2001.China-BangladeshRelations:NeedforEconomicPartnership.JournalofInternationalAffairs5(3).
NationalBankofSerbia(NBS).2007.FinancialStatementspreparedinaccordancewithInternationalFinancialReportingStandardsfortheyearended31December2006withIndependentAuditor'sReport.Belgrade,Serbia:NBS.Accessedon1July2023at
257
https://nbs.rs/export/sites/NBS_site/documents-eng/onbs-poslovanje/NBS_IAS_Report_2006.pdf
NationalDevelopmentandReformCommission,MinistryofForeignAffairs,MinistryofEcologyandEnvironment,andMinistryofCommerceofthePeople’sRepublicofChina.2022.OpinionsonJointlyPromotingGreenDevelopmentoftheBeltandRoad.16March2022.Accessedon25August2023athttps://www.ndrc.gov.cn/xxgk/zcfb/tz/202203/t20220328_1320629_ext.html
NationalEconomicandDevelopmentAuthority(NEDA).2017.OfficialDevelopmentAssistanceTermsandConditionsofLoanAssistance(asof30June2017).Manila,NEDA.Accessedon13June2023athttps://web.archive.org/web/20210627183145/https://www.neda.gov.ph/wp-content/uploads/2017/07/Q2-2017-ODA-Terms-and-Conditions-Loans-Bilateral-Institutions.pdf
Naughton,Barry,andKelleeS.Tsai.2015.StateCapitalism,InstitutionalAdaptation,andtheChineseMiracle.Cambridge,UK:CambridgeUniversityPress.
Ncube,Sipilisiwe.2019.China’s$22mDebtCancellationaMockery-HH.Zambia:NewsDiggers!(blog).3July2019.Accessedon1October2023athttps://diggers.news/local/2019/07/03/chinas-22m-debt-cancellation-a-mockery-hh/.
NESPAK.2015a.LahoreOrangeLineMetroRailProjectPC-1.Lahore,Pakistan:PunjabMasstransitAuthority.Accessedon1October2023athttps://pma.punjab.gov.pk/system/files/pcI.pdf.
NESPAK.2015b.EnvironmentalImpactAssessment(EIA)LahoreOrangeLineMetroTrainProject.Lahore,Pakistan:PunjabMasstransitAuthority.Accessedon1October2023athttps://epd.punjab.gov.pk/system/files/7%2901%20Tital%2C%20Table%20of%20Contant%2C%20Abbreviations%2C%20Executive%20%20Summary.pdf.
North,DouglassC.1991.Institutions.JournalofEconomicPerspectives5(1):97–112.
Nugent,Clara.2023a.Argentinareacheslast-minutedealwithIMFtoavoiddefault.FinancialTimes.28July2023.Accessedon2August2023athttps://www.ft.com/content/42c509a8-d75e-4350-b448-b6b8db13942a
Nugent,Ciara.2023b.ArgentinatapsChinesecreditlinetokeepeconomyafloatduringelections.FinancialTimes.18October2023.Accessedon19October2023athttps://www.ft.com/content/c92e56b7-e619-4cbb-b615-e401140c10cc
Nye,JosephS.2004.SoftPower:TheMeansToSuccessInWorldPolitics.NewYork,NY:PublicAffairs.
Obe,StephenChan,andChrisopherVandome.2022.Zambia:TheFirstYearofaHichilemaForeignPolicy|ChathamHouse–InternationalAffairsThinkTank.London:ChathamHouse.Accessedon1October2023athttps://www.chathamhouse.org/2022/08/zambia-first-year-hichilema-foreign-policy.
OrganisationforEconomicCo-operationandDevelopment(OECD).2023a.DevelopmentAssistanceCommittee:Totalofficialflows.AccessedonAug10,2023viastats.oecd.org.
OrganisationforEconomicCo-operationandDevelopment(OECD).2023b.ConvergedStatisticalReportingDirectivesfortheCreditorReportingSystem(CRS)andtheAnnualDACQuestionnaire,Chapter1-6.27April2023.Paris:OECD.Accessedon14September2023athttps://one.oecd.org/document/DCD/DAC/STAT(2023)9/FINAL/en/pdf
258
OfficeoftheLeadingGroupforPromotingtheBeltandRoadInitiative.2019.TheBeltandRoadInitiativeProgress,ContributionsandProspects.Beijing:OfficeoftheLeadingGroupforPromotingtheBeltandRoadInitiative.Accessedon1July2023athttps://eng.yidaiyilu.gov.cn/p/86739.html
Olander,Eric.2023.What’sDrivingtheSteadyDeclineinChineseOverseasDevelopmentLending?ChinaGlobalSouthProject.Accesedon01July223athttps://chinaglobalsouth.com/podcasts/whats-driving-the-steady-decline-in-chinese-overseas-development-lending/
Olewe,Dickens.2021.WhyAfricanCountriesBackChinaonHumanRights.BBCNews,1May2021.Accessedon1October2023athttps://www.bbc.com/news/world-africa-56717986.
OrganisationforEconomicCo-operationandDevelopment(OECD).2021.ConvergedStatisticalReportingDirectivesfortheCreditorReportingSystem(CRS)andtheAnnualDACQuestionnaire.Paris:OECD.Accessedon1August2023athttps://one.oecd.org/document/DCD/DAC/STAT(2020)44/FINAL/en/pdf
Paravicini,Giulia.2020.PompeoTakesVeiledSwipeatChinaonFinalLegofAfricaTrip.Reuters.19February2020.Accessedon1October2023athttps://www.reuters.com/article/us-usa-pompeo-ethiopia-idUSKBN20D1PB.
Park,S.2010.TheWorldBank’sglobalsafeguardpolicynorm?inOwningDevelopment:CreatingPolicyNormsintheIMFandtheWorldBank,editedbySusanParkandAntjeVetterlein.Cambridge,UK:CambridgeUniversityPress.
Parks,Bradley.2019.ChineseLeadershipandtheFutureofBRI:WhatKeyDecisionsLieAhead?CGDNote.Washington,DC:CenterforGlobalDevelopment.
Parsons,Nick.2010.CICbonds:Thegood,thebadandtheugly.TheAsset.23April2010.Accessedon15July2023athttps://theasset.com/article/17866/cic-bonds-the-good-the-bad-and-the-ugly
Patey,Luke.2017.ChinaMadeMauricioMacriaDealHeCouldn’tRefuse.ForeignPolicy.24January242017.Accessedon1October2023athttps://foreignpolicy.com/2017/01/24/china-made-mauricio-macri-a-deal-he-couldnt-refuse/.
Paul,Ruma.2023.Bangladeshpromisesfree,fairelectionsinresponsetoUSvisacurbs.25May2023Reuters.Accessedon28September2023athttps://www.reuters.com/world/asia-pacific/bangladesh-says-will-take-steps-address-any-interference-in-elections-2023-05-25/
People’sDaily.2014.XiEyesMoreEnablingIntl’EnvironmentforChina’sPeacefulDevelopment(3).People’sDaily.30November2014.Accessedon1August2023athttp://en.people.cn/n/2014/1130/c90883-8815967-3.html.
Perroux,Francois.1950.EconomicSpace:TheoryandApplications.QuarterlyJournalofEconomics64(1):89–104.
Pilling,David,andEmilyFeng.2018.ChineseInvestmentsinAfricaGoOfftheRails.FinancialTimes.5December2018.Accessedon1August2023athttps://www.ft.com/content/82e77d8a-e716-11e8-8a85-04b8afea6ea3.
Pogliani,PamelaandPhilipWooldridge.2022.Cross-borderfinancialcentres.BISWorkingPaper1035.Basel,Switzerland:BankforInternationalSettlements(BIS).
259
Qing,Shi.2023.DollarHegemonyisthetruecauseofAfrica’sdebtproblems,notChina.19April2023.GlobalTimes.Accessedon1October2023athttps://www.globaltimes.cn/page/202304/1289429.shtml
Ravallion,Martin.2009.ArethereLessonsforAfricafromChina’sSuccessAgainstPoverty?WorldDevelopment37(2):303-313.
Ray,Rebecca.2023.SmallisBeautiful:ANewErainChina’sOverseasDevelopmentFinance?.GCIPolicyBrief017.BUGlobalDevelopmentPolicyCenter.Accessedon17December2023athttps://www.bu.edu/gdp/files/2023/01/GCI_PB_017_CODF_EN_FIN.pdf
Reinhart,CarmenM.andChristophTrebesch.2016.Sovereigndebtreliefanditsaftermath.JournaloftheEuropeanEconomicAssociation14(1):215–251.
Rendon,Moises,andSarahBaumunk.2018.WhenInvestmentHurts:ChineseInfluenceinVenezuela.WashingtonDC:CSIS.Accessedon1October2023athttps://www.csis.org/analysis/when-investment-hurts-chinese-influence-venezuela.
Repnikova,Maria.2022.ChineseSoftPower.Cambridge,UK:CambridgeUniversityPress.https://www.cambridge.org/core/elements/chinese-soft-power/4E9C9E445C5F8A9F81F5AAEEAD2DF21D.
ReportersWithoutBorders.2023.WorldPressFreedomIndex.https://rsf.org/en/index.
RépubliqueduCongo.2018.AnnexeOpérationnelleN°2,DocumentCadredePolitiquesetdeProgrammationMacroéconomiquesetBudgétaires(DPPMB)2018-2022,PlanNationaldeDéveloppement(PND)2018-2022.Brazzaville:RépubliqueduCongo.Accessedon31July2023athttps://web.archive.org/web/20200722130921/https://pnd.plan.gouv.cg/wp-content/uploads/2018/07/Parlement-CMEB-24-07-18.pdf
Reuters.2009.ChinaHuijintorecapitaliseEximBanksoon-media.12November2009.Reuters.Accessedon15July2023athttps://www.reuters.com/article/china-exim-huijin/china-huijin-to-recapitalise-exim-bank-soon-media-idUSPEK13684620091112
Reuters.2010a.SAFEintendstochangepolicy-orientedforeignexchangeloanstoentrustedbankstohandlethem--Media(
外
管
局
拟
将
政
策
性
外
汇
贷
款
改
为
委
托
银
行
办
理
形
式--
媒
体).17May2010.Reuters.Accessedon15July2023https://www.reuters.com/article/idCNCHINA-230572
Reuters.2010b.Huijintoinvest100blnyuaninChinabanks-report.19August2010.Reuters.Accessedon5June2023athttps://www.reuters.com/article/china-economy-huijin/huijin-to-invest-100-bln-yuan-in-china-banks-report-idINTOE67J02720100820
Reuters.2016.Chinaforexregulatorbuys$4.2blninstocksvianewplatform-media.31March2016.Reuters.Accessedon15July2023athttps://www.reuters.com/article/china-markets-regulator-stocks/china-forex-regulator-buys-4-2-bln-in-stocks-via-new-platform-media-idUSL3N1731HH
Rey,Débora.2022.UKRebukesChinaforSupportingArgentina’sFalklandsClaim.APNews.7February2022.Accessedon1October2023athttps://apnews.com/article/europe-china-asia-beijing-xi-jinping-df1ca5eebfd3b988813f22bba0b3a7f0.
260
Riaz,Ali.2016.Bangladesh:APoliticalHistorysinceIndependence.BloomsburyPublishing.
Rolland,Nadège.2017.China’s‘BeltandRoadInitiative’:UnderwhelmingorGame-Changer?TheWashingtonQuarterly40(1):127–42.
Rose,AndrewK.2016.LikeMe,BuyMe:TheEffectofSoftPoweronExports.Economics&Politics28(2):216–232.
Rose,AndrewK.2019.SoftPowerandExports.ReviewofInternationalEconomics27(5):1573–1590.
Rose,Sarah.2018.LinkingUSForeignAidtoUNVotes:WhatAretheImplications.CGDNote.WashingtonDC:CenterforGlobalDevelopment.Accessedon15September2023athttps://www.cgdev.org/publication/linking-us-foreign-aid-un-votes-what-are-implications
Rosenstein-Rodan,PaulN.1943.ProblemsofIndustrializationofEasternandSouth-EasternEurope.EconomicJournal53(210/211):202–211.
S&PGlobal.2022.TheFutureofCopper:WilltheLoomingSupplyGapShort-CircuittheEnergyTransition?https://cdn.ihsmarkit.com/www/pdf/0722/The-Future-of-Copper_Full-Report_14July2022.pdf.
Sahoo,Pravakar,andAbhirupBhunia.2014.BCIMCorridoraGameChangerforSouthAsianTrade.EastAsiaForum.18July2014.Accessedon1October2023athttps://www.eastasiaforum.org/2014/07/18/bcim-corridor-a-game-changer-for-south-asian-trade/.
Sanderson,HenryandMichaelForsythe.2013.China’sSuperbank:Debt,OilandInfluence-HowChinaDevelopmentBankisRewritingtheRulesofFinance.NewYork,NY:Wiley.
Sanger,DavidE.,andMarkLandler.2021.BidenTriestoRallyG7NationstoCounterChina’sInfluence.TheNewYorkTimes.12June2021.Accessedon1October2023athttps://www.nytimes.com/2021/06/12/world/europe/biden-china-g7.html.
Schlegl,Matthias,ChristophTrebesch,andMarkL.J.Wright.2019.TheSeniorityStructureofSovereignDebt.NBERWorkingPaper25793.Cambridge,MA:NationalBureauofEconomicResearch(NBER).
Schumacher,Edward.1981.ArgentinaandSovietarenolongerjustbusinesspartners..TheNewYorkTimes.12July1981.Accessedon1October2023athttps://www.nytimes.com/1981/07/12/weekinreview/argentina-and-soviet-are-no-longer-just-business-partners.html.
Sen,AshishKumar.2018.IMFThrowsArgentinaa$57BillionLifeline.AtlanticCouncil(blog).September27,2018.https://www.atlanticcouncil.org/blogs/new-atlanticist/imf-throws-argentina-a-57-billion-lifeline/.
Séri,ReginaStéphanie.2021.Howgiantdiscoveriesofnaturalresourcesimpactsovereigndebtratingsindevelopingandemergingcountries?ÉtudesetDocumentsn°10.Paris:CERDI.
Setser,BradW.2023a.Shadowreserves—howChinahidestrillionsofdollarsofhardcurrency.TheChinaProject.Accessedon1July2023athttps://thechinaproject.com/2023/06/29/shadow-reserves-how-china-hides-trillions-of-dollars-of-hard-currency/
261
Setser,BradW.2023b.TheCommonFrameworkandItsDiscontents.26March2023.NewYork:CouncilonForeignRelations.Accessedon1August2023athttps://www.cfr.org/blog/common-framework-and-its-discontents
Setser,BradW.2023c.HowtoHideYourForeignExchangeReserves—AUser’sGuide.29June2023.NewYork,NY:CouncilonForeignRelations.Accessedon1September2023athttps://www.cfr.org/blog/how-hide-your-foreign-exchange-reserves-users-guide
Sevastopulo,DemetriandGideonLong.2021.U.S.DevelopmentBankStrikesDealtoHelpEcuadorPayChinaLoans.FinancialTimes.14January2021.Accessedon16July2023athttps://www.ft.com/content/affcc432-03c4-459d-a6b8-922ca8346c14
Shaban,AbdurRahmanAlfa.2016.ChinesePresidentHasConfidenceinMyLeadership–Zambia’sEdgarLungu.Africanews.6September2016.Accessedon1October2023athttps://www.africanews.com/2016/09/06/chinese-president-has-confidence-in-my-leadership-zambia-s-edgar-lungu/.
Shacinda,Shapi.2006.ChinaPutsZambiaMineInvestmentonHoldoverTaiwan.Reuters.5September2006.Accessedon1October2023athttp://www.minesandcommunities.org/article.php?a=5019.
ShahidKarim.2016.KamilKhanMumtazetcVersusGovernmentofPunjabetc.LahoreHighCourt.https://sys.lhc.gov.pk/appjudgments/2016LHC2454.pdf
Shaukat,Aroosa,andRanaTanveer.2016.LHCSuspendsWorkonOrangeLineTrainProject.TheExpressTribune.28January2016.Accessedon1October2023athttps://tribune.com.pk/story/1035831/lhc-suspends-work-on-orange-line-train-project.
Shambaugh,David.2015.China’sSoft-PowerPush.ForeignAffairs.16June2015.Accessedon1October2023athttps://www.foreignaffairs.com/articles/china/2015-06-16/chinas-soft-power-push.
Shan,He.2011.CICinjects20blnyuanintoSinosure.Accessedon1June2023atwww.china.org.cn/business/2011-11/17/content_23944979.htm
Shepherd,Christian,andBenBlanchard.2018.China’sXioffersanother$60billiontoAfrica,butsaysnoto‘vanity’projects.Reuters.3September2018.Accessedon24July2023athttps://www.reuters.com/article/china-africa/chinas-xi-says-funds-for-africa-not-forvanity-projects-idUSL3N1VO018
Shi,Jiangtao.2023.SuddenRiseof‘Argentina’sTrump’PutsEconomicTieswithChinaatRisk.SouthChinaMorningPost.18August2023.Accessedon1October2023athttps://www.scmp.com/news/china/diplomacy/article/3231570/sudden-rise-argentinas-trump-javier-milei-puts-economic-ties-china-risk.
Shortell,Paul.2014.WillDefaultDampenChina-ArgentinaTies?TheDialogue.18September2014.Accessedon1October2023athttps://www.thedialogue.org/blogs/2014/09/will-default-dampen-china-argentina-ties/.
Shukla,Manish.2021.HowAnti-ChinaMovementLaunchedbyLateZambianPresidentMichaelSataIsInfluencingZambianElections.ZeeNews.24January2021.Accessedon1October2023athttps://zeenews.india.com/world/how-anti-china-movement-launched-by-late-zambian-president-michael-sata-is-influencing-zambian-elections-2337420.html.
262
SimsGallagher,KellyandQiQi.2021.ClimateProofingChina’sBeltandRoadInitiative(BRI).7May2021.ClimatePolicyLab.Accessedon25August2023athttps://www.climatepolicylab.org/climatesmart/2021/3/4/climate-proofing-chinas-belt-and-road-initiative-bri
Silver,Laura,KatDevlinandChristineHuang.2020.UnfavorableViewsofChinaReachHistoricHighsinManyCountries.6October2020.WashingtonDC:PewResearchCenter.Accessedon1September2023athttps://www.pewresearch.org/global/2020/10/06/unfavorable-views-of-china-reach-historic-highs-in-many-countries/
Snow,Nancy.2022.AReliableFriendandStrategicPartnerintheIndo-PacificRegion:Japan’sStrategicCommunicationsandPublicDiplomacy.AidDataResearchReport.https://docs.aiddata.org/reports/gf01/gf01-07/A-Reliable-Friend-and-Strategic-Partner-in-the-Indo-Pacific-Region-Japans-Strategic-Communications-and-Public-Diplomacy.html.
StateAdministrationofForeignExchange(SAFE).2004.AnnualReportoftheStateAdministrationofForeignExchange(2004).Beijing:SAFE.Accessedon1August2023athttps://www.safe.gov.cn/en/2004/0726/1272.html
StateAdministrationofForeignExchange(SAFE).2008.LetterfromSAFEDeputyAdministratorFangShangputoCostaRicanMinisterofFinanceGuillermoE.Zúñiga.2January2008.Accessedon13August2023athttps://www.dropbox.com/scl/fi/a20fdxb2lecowt8tp5lym/2-January-2008-SAFE-Letter-to-Minister-of-Finance-of-Costa-Rica.pdf?rlkey=y9ywdrrhap4rdkt0cyqj4mj4y&dl=0
StateAdministrationofForeignExchange(SAFE).2017a.PanGongsheng:ForeignExchangeAdministrationSupportstheBeltandRoadInitiative.Beijing:SAFE.Accessedon1July2023athttps://www.safe.gov.cn/en/2017/0505/1263.html
StateAdministrationofForeignExchange(SAFE).2017b.AnnualReportoftheStateAdministrationofForeignExchange2017.Beijing:SAFE.Accessedon1July2023athttps://www.safe.gov.cn/en/file/file/20180822/2f5f632ccd964e02876b7455883e0449.pdf?n=Annual%20Report%20of%20the%20State%20Administration%20of%20Foreign%20Exchange%20(2017)
StateAdministrationofForeignExchange(SAFE).2021.StateAdministrationofForeignExchange2021AnnualReport.Beijing:SAFE.Accessedon1August2023athttps://www.safe.gov.cn/en/file/file/20221226/9186b372d33e4a6b9d156ab154cc2979.pdf?n=Annual%20Report%20of%20the%20State%20Administration%20of%20Foreign%20Exchange%20(2021)
StateCouncilofthePeople'sRepublicofChina.2021.PartySecretaryGeneralXiJinpingattendedthethird'BeltandRoad'symposium.21November2021.Accessedon1October2023athttps://www.gov.cn/xinwen/2021-11/21/content_5652298.htm
Stott,MichaelandCiaraNugent.2023.ArgentinamustrethinkChinarelationship,sayspresidentialcandidatePatriciaBullrich.FinancialTimes.16October2023.Accessedon16October2023athttps://www.ft.com/content/6aaa091d-8031-493a-9655-9b16bc7aee35
Strange,Austin.2023.SymbolsofState:ExplainingPrestigeProjectsintheGlobalSouth.WorkingPaper.
Strüver,Gregor.2016.WhatFriendsAreMadeOf:BilateralLinkagesandDomesticDriversofForeignPolicyAlignmentwithChina.ForeignPolicyAnalysis12(2):170–91.
263
Strüver,Gregor.2017.China'sPartnershipDiplomacy:InternationalAlignmentBasedonInterestsorIdeology.TheChineseJournalofInternationalPolitics10(1):31–65.
Sufi,Amir,2007.Informationasymmetryandfinancingarrangements:Evidencefromsyndicatedloans.TheJournalofFinance62:629–668.
Sullivan,MarkP.,andRebeccaM.Nelson.2016.Argentina:BackgroundandU.S.Relations.WashingtonDC:CongressionalResearchService.Accessedon1October2023athttps://www.hsdl.org/?view&did=791518.
Sullivan,MarkP.,andRebeccaM.Nelson.2017.Argentina:BackgroundandU.S.Relations.WashingtonDC:CongressionalResearchService.Accessedon1October2023athttps://sgp.fas.org/crs/row/R43816.pdf.
Swedlund,HaleyJ.2017.TheDevelopmentDance:HowDonorsandRecipientsNegotiatetheDeliveryofForeignAid.Ithaca,NY:CornellUniversityPress.
Tang,Keyi.2021.DevelopmentFinanceandDistributivePolitics:ComparingChineseandWorldBankFinanceinsub-SaharanAfrica.SAIS-CARIWorkingPaperNo.2021/45.WashingtonDC:SAIS-CARI.
Tangjun,Song,ZhangJianguang,andWangLin.2016.ResearchonState-ownedAssetsManagementofAdministrativeUnits(
行
政
单
位
国
有
资
产
管
理
研
究).Beijing:ChinaFinancialandEconomicPublishingHouse
Teixeira,JoãoC.A.,FranciscoJ.F.Silva,ManuelB.S.Ferreira,andJoséA.C.Vieira.2018.Sovereigncreditratingdeterminantsunderfinancialcrises.GlobalFinanceJournal36:1-13.
TheControllerandAuditorGeneraloftheNationalAuditOfficeofTanzania.2018.ReportoftheControllerandAuditorGeneralontheConsolidatedandSeparateFinancialStatementsofTanzaniaPetroleumDevelopmentCorporationfortheYearEnded30June2018.DaresSalaam:TheControllerandAuditorGeneraloftheNationalAuditOfficeofTanzania.Accessedathttps://www.dropbox.com/s/dyt43g53m2btrmz/TPDC%20Financial%20Statements%2030%20June%202018.pdf?dl=0
TheEconomist.2011.TryingtoPullTogether.TheEconomist.20April2011.Accessedon1October2023athttps://www.economist.com/briefing/2011/04/20/trying-to-pull-together.
TheEconomist.2018.ZambiaSlumpstowardAnotherDebtCrisis.TheEconomist.13September2018.Accessedon1October2023athttps://www.economist.com/middle-east-and-africa/2018/09/13/zambia-slumps-toward-another-debt-crisis.
TheEconomist.2020.HowtoStopZambiafromTurningintoZimbabwe.TheEconomist.12November2020.Accessedon1October2023athttps://www.economist.com/leaders/2020/11/12/how-to-stop-zambia-from-turning-into-zimbabwe.
TheEconomist.2023.ArgentinaIsWastingtheVastOpportunitiesChinaOffersIt.TheEconomist.7March2023.Accessedon1October2023athttps://www.economist.com/the-americas/2023/03/07/argentina-is-wasting-the-vast-opportunities-china-offers-it.
TheEconomist.2023.SheikhHasina’sIronGriponBangladeshMayyetStrangleItsLaudableProgress.TheEconomist.2June2023.Accessedon1October2023at
264
https://www.economist.com/podcasts/2023/06/02/sheikh-hasinas-iron-grip-on-bangladesh-may-yet-strangle-its-laudable-progress.
TheExport-ImportBankofChina.2017.OfferingMemorandum.7March2017.Accessedon4August2023athttps://links.sgx.com/FileOpen/China%20EximBank_USD%20Bonds%20Offering%20Memorandum%20dated%207%20March%202017.ashx?App=Prospectus&FileID=30816
TheExport-ImportBankofChina.2021.ThePartySecretaryandPresidentoftheExport-ImportBankofChinaHuXiaolianattendseventsatthe2021BoaoAsiaForum.20April2021.Accessedon1October2023atwww.eximbank.gov.cn/info/news/202104/t20210420_30419.html
TheGrowthLabatHarvardUniversity.2023.TheAtlasofEconomicComplexity.Accessedon1August2023athttp://www.atlas.cid.harvard.edu.
ThePeople'sBankofChina(PBOC).2007.ChinaDevelopmentBanktoReceiveUS$20billionasCapitalInjection.31December2007.Beijing:PBOC.Accessedon2June2023atwww.pbc.gov.cn/en/3688110/3688172/4048344/3715513/index.html
ThePeople'sBankofChina(PBOC).2012.ChinaFinancialStabilityReport2012.eijing:PBOC.Accessedon1June2023atwww.pbc.gov.cn/eportal/fileDir/image_public/UserFiles/english/upload/File/China%20Financial%20%20Stability%20Report(1).pdf
TheWhiteHouse.2022.NationalSecurityStrategy.WashingtonDC:TheWhiteHouse.Accessedon2June2023athttps://www.whitehouse.gov/wp-content/uploads/2022/10/Biden-Harris-Administrations-National-Security-Strategy-10.2022.pdf.
TheWhiteHouse.2023.FactSheet:VicePresidentHarrisAnnouncesOver$7BillioninPrivateSectorandU.S.GovernmentCommitmentstoPromoteClimateResilience,Adaptation,andMitigationacrossAfrica.WashingtonDC:TheWhiteHouse.Accessedon2June2023athttps://www.whitehouse.gov/briefing-room/statements-releases/2023/03/31/fact-sheet-vice-president-harris-announces-over-7-billion-in-private-sector-and-u-s-government-commitments-to-promote-climate-resilience-adaptation-and-mitigation-across-africa/.
TransparencyInternational.2021.CorruptionPerceptionsIndex.https://www.transparency.org/en/cpi/2021.
Tripathi,Salil.2021.Bangladesh’sLongJourneyFrom‘BasketCase’toRisingStar.ForeignPolicy(blog).10April2021.Accessedon2June2023athttps://foreignpolicy.com/2021/04/10/bangladesh-independence-anniversary-basket-case-rising-star/.
Trotsenburg,Axelvan.2023.DefyingtheOdds:Bangladesh’sJourneyofTransformationandResilience.1March2023.WashingtonDC:WorldBank.Accessedon2June2023athttps://www.worldbank.org/en/news/opinion/2023/03/01/defying-the-odds-bangladesh-s-journey-of-transformation-and-resilience.
U.S.DepartmentofDefense.2018.USNationalDefenseStrategy2018:SharpeningtheAmericanMilitary’sCompetitiveEdge.11August2018.Washington,DC:U.S.DepartmentofDefense.Accessedon1October2023athttps://dod.defense.gov/Portals/1/Documents/pubs/2018-National-Defense-Strategy-Summary.pdf
265
U.S.InternationalDevelopmentFinanceCorporation(DFC).2021a.DFCSurpasses$6.7BillioninCommitmentsin2021DuringRecordFiscalYear.October12,2021.WashingtonDC:DFC.Accessedon16July2023athttps://www.dfc.gov/media/press-releases/dfc-surpasses-67-billion-commitments-2021-during-record-fiscal-year
U.S.InternationalDevelopmentFinanceCorporation(DFC).2021b.DFCandEcuadorSignFrameworkAgreementtoSupportDevelopment,PrivateSectorInvestment.January14,2021.WashingtonDC:DFC.Accessedon16July2023athttps://www.dfc.gov/media/press-releases/dfc-and-ecuador-sign-framework-agreement-support-development-private-sector
U.S.InternationalDevelopmentFinanceCorporation(DFC).2022.DFCMarksLargestYearwithMoreThan$7.4Billion
CommittedinFY2022.November15,2022.WashingtonDC:DFC.Accessedon16July2023athttps://www.dfc.gov/media/press-releases/dfc-marks-largest-year-more-74-billion-committed-fy-2022
UNEP-WCMC&IUCN.2023.ProtectedPlanet:TheWorldDatabaseonProtectedAreas(WDPA).August2023.Cambridge,UK:UNEP-WCMC&IUCN.
UnitedNations.2022.WorldPopulationProspects2022.Dataset.https://population.un.org/wpp/.
Vadlamannati,KrishnaChaitanya,SamuelBrazys,AlexanderDukalskis,andYuanxinLi.2023.BuildingBridgesorBreakingBonds?TheBeltandRoadInitiativeandForeignAidCompetition.ForeignPolicyAnalysis19(3):orad015.
Vaughn,Bruce.2008.Bangladesh:PoliticalTurmoilandTransition.CRSReportforCongress.WashingtonDC:CongressionalResearchService.https://www.everycrsreport.com/files/20080530_RL34522_27d61b3aea89509df1d24e4d57895904992d8120.pdf.
Vreeland,JamesRaymond,andAxelDreher.2014.ThePoliticalEconomyoftheUnitedNationsSecurityCouncil.MoneyandInfluence.NewYork:CambridgeUniversityPress.
Wade,Abdoulaye.2008.TimefortheWesttoPracticeWhatItPreaches.FinancialTimes.23January2008.Accessedon15August2023athttps://www.ft.com/content/5d347f88-c897-11dc-94a6-0000779fd2ac.
Wade,RobertH.2009.AccountabilityGoneWrong:TheWorldBank,NongovernmentalOrganisationsandtheUSGovernmentinaFightoverChina.NewPoliticalEconomy14(1):25-48.
Warner,Andrew.2010.Cost-BenefitAnalysisinWorldBankProjects.Washington,DC:IndependentEvaluationGroupoftheWorldBank.
Watts,Jonathan.2015.ArgentinaLeaderLeavesControversialLegacywithPatagoniaDamsProject.TheGuardian.1December2015.Accessedon2June2023athttps://www.theguardian.com/world/2015/dec/01/argentina-president-cristina-fernandez-de-kirchner-patagonia-hydroelectric-dam-project.
Wei,Lingling.2013.China'sZhuChanghongHelpsSteerNation'sCurrencyReserves.16July2013.WallStreetJournal.Accessedon1August2023athttps://www.wsj.com/articles/SB10001424127887323664204578606301739504368
266
Wellner,Lukas,AxelDreher,AndreasFuchs,BradleyC.Parks,andAustinStrange.2023.Chinesedevelopmentfinanceandpublicopinion.2October2023.VoxEUcolumn.Accessedon2October2023athttps://cepr.org/voxeu/columns/chinese-development-finance-and-public-opinion
Wellner,Lukas,AxelDreher,AndreasFuchs,BradleyC.Parks,andAustinStrange.Forthcoming.CanAidBuyForeignPublicSupport?EvidencefromChineseDevelopmentFinance.EconomicDevelopmentandCulturalChange.
Wen,Yao.2023.BDanImportantStrategicPartnerofChina.TheFinancialExpress(blog).2August2023.Accessedon2August2023athttps://thefinancialexpress.com.bd/views/views/bd-an-important-strategic-partner-of-china.
White,Abraham,andLeoHoltz.2022.FigureoftheWeek:AfricanCountries’VotesontheUNResolutionCondemningRussia’sInvasionofUkraine.WashingtonDC:BrookingsInstitution.Accessedon2October2023athttps://www.brookings.edu/articles/figure-of-the-week-african-countries-votes-on-the-un-resolution-condemning-russias-invasion-of-ukraine/.
Wigglesworth,RobinandSunYu.2023.HowChinachangedthegameforcountriesindefault.13April2023.FinancialTimes.Accessedon14August2023athttps://www.ft.com/content/19add278-aa83-45f8-a84f-12750f32258f
Wilkinson,Tracy.2019.ArgentinaElectionRaisesDoubtsaboutTrump’sBetonRight-WingLeaders.LosAngelesTimes.10September2019.Accessedon2June2023athttps://www.latimes.com/world-nation/story/2019-09-09/argentina-election-trump-right-wing-latin-america-support.
Wilson,Elliot.2022.BRI:HavewepassedpeakChina?16February2022.Euromoney.Accessedon3August2023athttps://www.euromoney.com/article/29ppq3xrh5p5bxgoh9dz4/opinion/bri-have-we-passed-peak-china
Wonacott,Peter.2011.InAfrica,U.S.WatchesChina’sRise.WallStreetJournal.2September2011.Accessedon2June2023athttps://www.wsj.com/articles/SB10001424053111903392904576510271838147248
Wong,ChunHanandKeithZhai.2023.CourtPapersOpenRareWindowIntoRoleofGraftinChina’sOverseasLending.9June2023.WallStreetJournal.Accessedon7August2023athttps://www.wsj.com/articles/court-papers-open-rare-window-into-role-of-graft-in-chinas-overseas-lending-69f7ab81
Wooley,Alex.2023.AidData’sChinadatawins2023BestDatasetAwardfromIPES.TheFirstTranche.29August2023.Accessedon10September2023athttps://www.aiddata.org/blog/aiddatas-china-data-wins-2023-best-dataset-award-from-ipes
WorldBankGroup.2020.WorldBankGroupApproachestoMobilizePrivateCapitalforDevelopment:AnIndependentEvaluation.WashingtonDC:WorldBankhttps://ieg.worldbankgroup.org/sites/default/files/Data/Evaluation/files/PrivateCapitalMobilization.pdf
WorldBank.2000.WorldBankDebtorReportingSystemManual.WashingtonDC:WorldBank.Accessedon20February2023viahttps://web.archive.org/web/20220901043448/https://databankfiles.worldbank.org/data/download/debt/DRS%20Manual%202013.pdf
267
WorldBank.2012.WorldBankSanctionsProcedures.WashingtonDC:WorldBank.Accessedon20August2023viahttps://www.worldbank.org/content/dam/documents/sanctions/other-documents/osd/WBGSanctions_Procedures_April2012_Final.pdf
WorldBank.2020a.InternationalDebtStatistics2021.WashingtonDC:WorldBank.Accessedon20February2023viahttps://openknowledge.worldbank.org/handle/10986/34588
WorldBank.2020b.WorldBankDebtorReportingSystem–WhatitMeasures.WorldBankDevelopmentEconomicsDataGroup.WashingtonDC:WorldBank.Accessedon15September2023athttps://pubdocs.worldbank.org/en/951771586884732835/pdf/DRS-What-it-Measures.pdf
WorldBank.2021.InternationalDebtStatistics2022.WashingtonDC:WorldBank.Accessedon20February2023viahttps://openknowledge.worldbank.org/handle/10986/36289
WorldBank.2023a.WorldDevelopmentIndicators.https://databank.worldbank.org/source/world-development-indicators.
WorldBank.2023b.Procurement-WorldBankListingofIneligibleFirmsandIndividuals.WashingtonDC:WorldBank.AccessedonJune1,2023viahttps://www.worldbank.org/en/projects-operations/procurement/debarred-firms
WorldBank.2023c.RemarksbyWorldBankGroupPresidentAjayBangaatthe2023AnnualMeetingsPlenaryon13October2023.WashingtonDC:WorldBank.Accessedon13October2023athttps://www.worldbank.org/en/news/speech/2023/10/13/remarks-by-world-bank-group-president-ajay-banga-at-the-2023-annual-meetings-plenary
Xiao,Cai.2016.SAFEbuysintoA-sharemarket.1April2016.ChinaDaily.Accessedon1June2023athttps://www.chinadaily.com.cn/cndy/2016-04/01/content_24224169.htm
Xie,YifanandMiaLamar.2016.ChinaCurrencyRegulatorInvestsinBankStocks.4April2016.WallStreetJournal.Accessedon5July2023athttps://www.wsj.com/articles/chinas-foreign-exchange-regulator-dips-into-stocks-1459523466
Xinhua.2015Chinaraisescapitalbaseofdevelopmentbank.18August2015.Xinhua.Accessedon5June2023athttps://www.chinadaily.com.cn/business/2015-08/18/content_21640311.htm
Xining,Wang.2021.Whole-ProcessPeople’sDemocracyIsAHighQualityDemocracy.2021.Beijing:MinistryofForeignAffairsofthePeople’sRepublicofChina.Accessedon5September2023athttps://www.fmprc.gov.cn/eng/wjb_663304/zwjg_665342/zwbd_665378/202112/t20211213_10467431.html.
Xueqing,Jiang.2017.CBRCTellsPolicyBankstoBeefupRiskManagement.16November2017.ChinaDaily.Accessedon1August2023athttps://www.chinadaily.com.cn/business/2017-11/16/content_34591397.htm
Xueqing,Jiang.2022.TsinghuareportexposesAfrica'sbondcrises.5August2022.ChinaDaily.Accessedon5August2022athttps://www.chinadaily.com.cn/a/202208/05/WS62ec70d1a310fd2b29e706f6.html
268
Yang,Hongbo,B.AlexanderSimmons,RebeccaRay,ChristophNolte,SuchiGopal,YaxiongMa,XinyueMa,KevinP.Gallagher,2021.RiskstoglobalbiodiversityandIndigenouslandsfromChina’soverseasdevelopmentfinance.NatureEcology&Evolution11:1520-1529.
Ye,Min.2021.FragmentedMotivesandPolicies:TheBeltandRoadInitiativeinChina.JournalofEastAsianStudies21(2):193–217.
Ying,Diao.2008.Policybanktograduallygocommercial.3March2008.ChinaDaily.Accessedon15June2023athttps://www.chinadaily.com.cn/china/2008-03/03/content_6500728.htm
Yuan,Yao.2014.GlobalInvestmentM&ANewsletter:RevealingtheMysteryofForeignExchangeReserveEntrustedLoans.Accessedon1July2023athttps://m.gaodun.com/caiwu/379542.html
Yuzhe,Zhang.2010.Chinafindspolicyjobforforexreserves.23May2010.CaixinOnline.Accessedon1August2023athttps://web.archive.org/web/20100523082852/http://english.caing.com/2010-05-20/100145743.html
Zeitz,AlexandraO.2021.Emulateordifferentiate?Chinesedevelopmentfinance,competition,andWorldBankinfrastructurefunding.TheReviewofInternationalOrganizations16:265–292.
Zhang,Yanqiu,andDanielOlooOng’ong’a.2022.UnveilingChina’sDigitalDiplomacy:AComparativeAnalysisofCGTNAfricaandBBCNewsAfricaonFacebook.AsianJournalofComparativePolitics7(3):661–83.
Zhang,Denghua,andGraemeSmith.2017.ForeignAidSystem:Structure,Agencies,andIdentities.ThirdWorldQuarterly38(10):2330–2346.
Zhou,HaonanandEugenioCerutti2018.TheChinesebankingsystem:Muchmorethanadomesticgiant.VoxEUBlog.Accessedon19July2023athttps://cepr.org/voxeu/columns/chinese-banking-system-much-more-domestic-giant
Zhu,Yanling.2022.China’s‘NewCulturalDiplomacy’inInternational
Broadcasting:BrandingtheNationthroughCGTNDocumentary.International
JournalofCulturalPolicy28(6):671–83.
269
Appendix:SupplementaryMaterial
SectionA-1:Figuresandtablesreferencedinthereport
Table
A1:
AidData’s
Global
Chinese
Development
Finance
Dataset,
Version
3.0
at
a
glance
TableA.1
AidData ’s
Global
Chinese
Development
Finance
Dataset,
Version
3.0
at
a
glance
AidDataDatasetGlobalChineseDevelopmentFinanceDataset,Version2.0(PublishedSeptember2021)GlobalChineseDevelopmentFinanceDataset,Version3.0(PublishedNovember2023)
Scope&CoverageSectorsAllAll
CountryCoverage165countriesglobally(including145countrieswithprojectsidentified)165countriesglobally(including146countrieswithprojectsidentified)
Financiers334Chineseofficialsectordonorsandlenders791Chineseofficialsectordonorsandlenders
FinancialInstrumentLoans,grants,scholarships,technicalassistance,debtrescheduling,debtforgivenessLoans(withcategorizationof23distinctloaninstruments),grants,scholarships,technicalassistance,debtrescheduling,debtforgiveness
DatasetSummaryNumberofRecords13,42720,985
NumberofFields70133
SourcesPubliclyAvailable91,125(including62,750uniquesources)147,703(including99,393uniquesources)
TotalFinancialValue$851billion(2017prices)$1.34trillion(2021prices)(excludingshort-term“rollover”facilities)
Timeframe2000-2017(withimplementationdetailsthrough2021)2000-2021(withimplementationdetailsthrough2023)
ProjectDetailsFinancialDetailsTransactionamount,collateral,interestrate,graceperiod,maturity,commitmentfee,managementfeeTransactionamount,collateral,interestrate,defaultinterestrate,graceperiod,maturity,commitmentfee,managementfee,insurancefee,firstandlastloanrepaymentdates,levelofpublicliability
ParticipatingAgenciesFundingagencies,co-financingagencies,receivingagencies,implementingagencies,accountableagenciesFundingagencies,co-financingagencies,directreceivingagencies,indirectreceivingagencies,implementingagencies,guarantor,insuranceprovider,collateralprovider,securityagent/collateralagent
ImplementationDetailsCommitmentyear,status,plannedandactualstartandcompletiondatesCommitmentdate,status,plannedandactualstartandenddates,deviationfromplannedstartandcompletiondates,
270
AidDataDatasetGlobalChineseDevelopmentFinanceDataset,Version2.0(PublishedSeptember2021)GlobalChineseDevelopmentFinanceDataset,Version3.0(PublishedNovember2023)
infrastructureprojectflag
DescriptionAverageof142wordsperprojectAverageof166wordsperproject
OECDClassificationsSector,flowclassSector,flowclass,recipientcountryincomeclassification,grant-equivalentmeasure
Sub-nationalDetails3,285physicallocations9,497physicallocations
Figure
A1:
Composition
of
official
financial
flows
from
China
to
the
developing
world,
2000-2021
FigureA.1
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(see
SectionA-2oftheAppendixfordetails).TheVague(OfficialFinance)isaresidualcategoryforofficial
financialcommitmentsfromChinathatcouldnotbereliablycategorizedasODAorOOFbecauseof
insufficientlydetailedinformation.
271
Figure
A2:
Stock
of
official
financial
flows
from
China
and
the
G7
to
the
developing
world
during
the
early
and
late
BRI
periods
FigureA.2
Notes:AidDatareliesonOECD-DACmeasurementcriteriatomakeODAandOOFdeterminations(as
describedinSectionA-2oftheAppendix).TheVague(OfficialFinance)isaresidualcategoryforofficial
financialcommitmentsfromChinathatcouldnotbereliablycategorizedasODAorOOFbecauseof
insufficientlydetailedinformation.G7ODAandOOFdatarepresentgrossdisbursementsfromthe
272
OECD-DAC.Thisfigureexcludesshort-term“rollover”facilitiesfromthetallyofofficialfinancial
commitments(seeBox2candSectionA-3intheAppendix).
Figure
A3:
Sectoral
composition
of
official
financial
flows
from
China
to
developing
world,
2014
vs.
2021
FigureA.3
AidData’sGlobalChineseDevelopmentFinanceDataset,Version3.0ata
glance
Notes:ThisfigureshowsthesectoralcompositionofChineseODAandOOFcommitments(measuredin
constant2021USD)in2014(leftpanel)and2021(rightpanel).
273
Figure
A4:
Number
of
financially-distressed
low-
and
middle-income
countries
with
outstanding
debt
to
China
FigureA.4
Notes:Todetermineifacountryhas“outstandingdebttoChina,”weusethe3.0versionofAidData’s
GCDFdatasettomeasurewhetherinagivenyearifithadatleastoneofficialsectorloanfromChinawithin
itsoriginallyscheduledrepaymentperiod(i.e.,aftertheexpirationofthegraceperiodbutbeforethefinal
maturitydate).Todetermineifacountrywasexperiencingfinancialdistressinagivenyear,weusethe
measurethatisdescribedinBox1a.Since2000isthefirstyearinwhichwemeasureloancommitmentsin
the3.0versionoftheGCDFdataset,wedonotcaptureanyoutstandingdebttoChina(loanswithintheir
originallyscheduledrepaymentperiods)inthatyear.
274
Figure
A5:
Sovereign
overdue
repayments
owed
to
official
sector
creditors
in
China
FigureA.5
Notes:Thisfiguremeasurestheaveragepercountrymonetaryvalue(inconstant2021USDmillions)of
sovereignarrearstoasubsetofofficialsectorcreditorsinChina(usingdatafromtheWorldBank’s
InternationalDebtStatistics).Sovereignarrearscaptureprincipalandinterestarrears(i.e.,overdue
repayments)onPPGdebttoChinaEximbank,ChinaDevelopmentBank,andChina’sMinistryofCommerce
contractedbyasubsetofLICsandMICsthatparticipateintheWorldBank’sDebtorReportingSystem
(DRS).YearsinwhichacountrymaintaineddiplomaticrelationswithTaiwanareexcluded.Eachcountry-year
observationisgivenequalweightinagivenyeartogenerateglobalaverages.
275
Figure
A6:
Project
life
cycle
FigureA.6
Project
life
cycle
276
Figure
A7:
Average
length
of
completion
delays
across
all
Chinese
ODA-
and
OOF-financed
infrastructure
projects
FigureA.7
Notes:Thisgraphshowstheaveragelengthofcompletiondelays(incalendardays)acrossallChinese
ODA-andOOF-financedinfrastructureprojectsbyyearthatwereformallycommitted,undergoing
implementation,orcompleted.Completiondelaysarecalculatedbytakingthedifference(incalendardays)
betweentheoriginallyscheduledprojectcompletiondateandtheactualprojectcompletiondate.The
dataaredrawnfromthe3.0versionofAidData’sGCDFdataset.
277
Figure
A8:
China’s
soft
power
gains
and
losses
vis-à-vis
the
U.S.
FigureA.8
Notes:FigureA8presentstheproportionofLICsandMICsinagivenyear(from2014-2021)inwhichChina
experiencedrelativegainsorlossesinpublicapproval,mediasentiment,andUNGAvotingalignment
278
vis-à-vistheU.S.Tomeasuretherelativegainsorlossesinpopularsupport,wefollowathree-step
calculationforeachcountry:(1)calculatethedifferencebetweenthepublicapprovalratingforChinaina
givenyearandtheprioryear;(2)calculatethedifferencebetweenpublicapprovalratingfortheU.S.ina
givenyearandtheprioryear;and(3)calculatethe“doubledifference”between(1)and(2)todetermineif
ChinaexperiencedagreatergainorlossinpublicsupportthantheU.S.inthesamecountry-year.For
relativegainsandlossesinmediasentimentandUNGAvotingalignment,thesamethree-stepcalculation
wasfollowedusingtheaveragemediasentimentscoreforeachcountry-yearfromtheGDELT1.0Event
Database(relatedtogovernmentactorsfrommainlandChinaortheU.S.)andtheaverage
“idealpointdistance”estimatebetweeneachcountryandChina(ortheU.S.)inagivenyear.
Figure
A9:
China
versus
the
U.S.:
Public
disapproval
rates
FigureA.9
Notes:ThisgraphshowstheannualweightedaveragedisapprovalrateforChinaandtheU.S.between
2014-2021fromGallupWorldPoll.Theaveragedisapprovalrateisweightedbypopulationsizeforeach
country.Whensurveyingrespondents,respondentscananswer“approve”,“disapprove”,or“unsure”
whenaskediftheyapproveofChina’sortheU.S’sleadership.Thisgraphrepresentstheaveragethat
answered“disapprove”foreachcountry’sleadership.Theconstructionofthisvariableisdescribedin
greaterdetailinBox1b.
279
Figure
A10:
Composition
of
China’s
overseas
lending
portfolio
by
level
of
public
liability
FigureA.10
Notes:ThisgraphshowstheannualcompositionofChina’soverseaslendingportfolio(asmeasuredin2021
constantUSD)inLICsandMICsaccordingtotheextenttowhichthehostgovernmentsmayeventuallybe
liablefordebtrepayment.Centralgovernmentdebtandotherpublicsectordebtrepresentloanswhere
theborrowerisagovernmentagencyorawholly-ormajority-ownedstateentity.Centralgovernmentdebt
representsloansthathaveasovereignguaranteefromthehostgovernment.Potentialpublicdebt
representsloanstoentities(includingspecialpurposevehiclesorjointventures)wherethehost
governmenthasaminoritystake.Privatedebtcapturesloanstoprivateentities.
280
Figure
A11:
Proportion
of
China’s
overseas
lending
portfolio
provided
via
buyer’s
credits
FigureA.11
Figure
A12:
Proportion
of
China’s
overseas
lending
portfolio
provided
via
Government
Concessional
Loans
(GCLs)
and
Preferential
Buyer’s
Credits
(PBCs)
FigureA.12
281
Notes:ThisgraphshowstheannualproportionofofficialsectorloancommitmentsfromChina(measured
inconstant2021USD)toLICsandMICsprovidedviaGovernmentConcessionalLoans(GCLs)and
PreferentialBuyer’sCredits(PBCsfromChinaEximbank.
Figure
A13:
Proportion
of
China’s
overseas
lending
portfolio
provided
via
infrastructure
project
loans
FigureA.13
Notes:ThisgraphshowstheannualproportionofofficialsectorloancommitmentsfromChina(measured
inconstant2021USD)toLICsandMICsthatwereissuedtobuild,maintain,orrenovateinfrastructurein
theborrowingcountry.Infrastructureprojectloansinthe3.0versionofAidData’sGCDFdatasetarethose
(a)categorizedasinvestmentprojectloans(IPLs),and(b)explicitlydesignatedassupportinginfrastructure
(throughthe“infrastructure”marker).
282
Figure
A14:
Composition
of
China’s
overseas
lending
portfolio
by
emergency
and
non-emergency
lending
instrument
FigureA.14
Notes:ThisfiguremeasuresthepercentagesofChina’soverseaslending(inconstant2021USD)toLICs
andMICsthatconsistsofemergencyrescueloans(ERLs)thatarerollovers,ERLsthatarenotrollovers,and
allothertypesofloans.The“rescue”variableinthe3.0versionoftheGCDFdatasetisusedtoidentify
emergencyrescueloans.RolloverERLamountsarecalculatedbysubtractingthevaluesintheAdjusted
Amount(ConstantUSD2021)fieldfromthevaluesintheAmount(ConstantUSD2021)field.Non-rollover
ERLamountsaredirectlydrawnfromtheAdjustedAmount(ConstantUSD2021)field.
283
Figure
A15:
Proportion
of
RMB-denominated
rescue
lending
to
countries
in
and
not
in
financial
distress
FigureA.15
Notes:ThisgraphpresentstheannualproportionofChina’sRMB-denominatedloancommitments(as
measuredin2021constantUSD)totwocountrycohortsbetween2000and2021:(1)countriesinfinancial
distress,and(2)countriesnotinfinancialdistress.Todetermineifacountryexperiencedfinancialdistressin
agivenyear,weusethebinarymeasurethatisdescribedinBox1ainChapter1.
284
Figure
A16:
Composition
of
China’s
overseas
lending
portfolio
by
currency
of
denomination
for
countries
in
financial
distress
FigureA.16
Notes:ThisfigurepresentsthecompositionofChina’slendingportfolioinLICsandMICs(asmeasuredin
2021constantUSD)bythecurrenciesinwhichtheloansweredenominatedforcountry-yearsthatwere
designatedas“infinancialdistress.”Todetermineifacountryexperiencedfinancialdistressinagivenyear,
weusethebinarymeasurethatisdescribedinBox1ainChapter1.
285
Figure
A17:
Composition
of
China’s
overseas
loan
portfolio
by
currency
of
denomination
for
countries
not
in
financial
distress
FigureA.17
Notes:ThisfigureshowsthecompositionofChina’slendingportfolioinLICsandMICs(asmeasuredin
2021constantUSD)bythecurrenciesinwhichtheloansweredenominatedforcountry-yearsthatwerenot
designatedas“infinancialdistress.”Todetermineifacountryexperiencedfinancialdistressinagivenyear,
weusethebinarymeasurethatisdescribedinBox1ainchapter1.
Figure
A18:
Early
versus
late
BRI:
weighted
average
interest
rates
FigureA.18
286
Figure
A19:
Early
versus
late
BRI:
weighted
maturity
lengths
FigureA.19
Figure
A20:
Early
versus
late
BRI:
weighted
average
grant
element
FigureA.20
287
Figure
A21:
Weighted
average
grant
element
of
overseas
lending
from
China
FigureA.21
Notes:ThisgraphshowstheaveragegrantelementofofficialsectorlendingfromChinatoLICandMICs
between2000and2021.ThisgrantelementiscalculatedusingtheGrantElement(OECDcash-flow)
variableinthe3.0versionofAidData’sGCDF3.0,whichusesa10%discountrateforallborrowing
countriesbasedontheOECD’scash-flowgrantelementcalculation.Theannualaveragesareweightedby
theconstant2021USDvaluesoftheloancommitmentsineachrespectiveyear.
288
Figure
A22:
Weighted
average
maturity
lengths
for
countries
in
and
not
in
financial
distress
FigureA.22
Notes:ThisgraphshowstheaveragematuritylengthacrossallofficialsectorloansfromChinatoLICsand
MICsbetween2000and2021acrosstwocohorts:(1)countriesinfinancialdistressand(2)countriesnotin
financialdistress.Theannualaveragesareweightedbytheconstant2021USDcommitmentvaluesofthe
loansineachrespectiveyear.Todetermineifacountryexperiencedfinancialdistressinagivenyear,we
usethebinarymeasurethatisdescribedinBox1ainChapter1.
289
Figure
A23:
Weighted
average
interest
rates
for
countries
in
and
not
in
financial
distress
FigureA.23
Notes:ThisgraphshowstheaverageinterestrateacrossallofficialsectorloansfromChinatoLICsand
MICsbetween2000and2021fortwocohorts:(1)countriesinfinancialdistressand(2)countriesnotin
financialdistress.Theannualaveragesareweightedbytheconstant2021USDcommitmentvaluesofthe
loansineachrespectiveyear.Todetermineifacountryexperiencedfinancialdistressinagivenyear,we
usethebinarymeasurethatisdescribedinBox1ainchapter1.
290
Figure
A24:
Weighted
average
grant
elements
for
countries
in
and
not
in
financial
distress,
2014-2021
FigureA.24
Notes:ThisgraphshowstheaveragegrantelementacrossallofficialsectorloansfromChinatoLICsand
MICsbetween2000and2021fortwocohorts:(1)countriesinfinancialdistressand(2)countriesnotin
financialdistress.Todetermineifacountryexperiencedfinancialdistressinagivenyear,weusethebinary
measurethatisdescribedinBox1ainchapter1.ThegrantelementiscalculatedusingtheGrantElement
(OECDcash-flow)variableinthe3.0versionofAidData’sGCDFdataset,whichusesa10%discountratefor
allborrowingcountriesbasedontheOECD’scash-flowgrantelementcalculation.Theannualgrant
elementaveragesareweightedbytheconstant2021USDcommitmentvaluesoftheloansineach
respectiveyear.
Figure
A25:
Grant
element
across
China’s
portfolio
to
countries
in
and
not
in
financial
distress
FigureA.25
Notes:ThisgraphshowstheaveragegrantelementacrossChina’sportfolioofofficialsectorloanstoLICs
andMICsacrosstwocohorts:countriesexperiencingfinancialdistressandcountriesnotexperiencing
financialdistress.ThegrantelementiscalculatedusingtheGrantElement(OECDcash-flow)variableinthe
291
3.0versionofAidData’sGCDFdataset,whichusesa10%discountrateforallborrowingcountriesbased
ontheOECD’scash-flowgrantelementcalculation.Thegrantelementaverageisweightedbyconstant
2021USDcommitmentvalues.
Figure
A26:
Composition
of
China’s
overseas
lending
portfolio
by
type
of
lending
instrument
FigureA.26
Notes:Thisfigureshowstheannualpercentageofofficiallendingcommitments(inconstant2021USD)
fromChinatoLICsandMICsthatwasprovidedvia(1)investmentprojectloans,(2)emergencyrescue
loans,and(3)allothertypesofloans.Projectloansaredefinedasthoseinthe3.0versionofAidData’s
GCDFdatasetthatarecategorizedasinvestmentprojectloans(IPLs).
292
Figure
A27:
Composition
of
China’s
overseas
lending
portfolio
by
type
by
creditor
category
FigureA.27
Figure
A28:
Proportion
of
China’s
overseas
lending
portfolio
involving
a
multilateral
institution
FigureA.28
293
Notes:ThisgraphshowstheannualpercentageofprojectloansfromofficialsectorcreditorsinChinato
borrowersinLICsandMICsthatinvolvemultilateralinstitutionsasco-financiers,implementingagencies,or
areceivingagencybetween2000and2021.
Figure
A29:
Composition
of
China’s
overseas
lending
with
and
without
credit
enhancements
to
countries
in
and
not
in
financial
distress
FigureA.29
Notes:ThisgraphshowstheproportionofChina’soverseaslendingcommitments(measuredasconstant
2021USD)toLICsandMICsbackedbyacreditinsurancepolicy,athird-partyrepaymentguarantee,
and/orcollateraltotwocohorts:(1)countriesinfinancialdistressand(2)countriesnotinfinancialdistress.
Todetermineifacountryexperiencedfinancialdistressinagivenyear,weusethebinarymeasurethatis
describedinBox1ainChapter1.
Figure
A30:
Composition
of
China’s
non-emergency
overseas
loan
portfolio
by
repayment
risk
FigureA.30
Notes:ThisgraphshowsthepercentageofChina’snon-emergencyloancommitmentsbetween2000and
294
2021thatsupportedtwocohorts:(1)LICsandMICswithscoresof5orlessonthesovrateindex;and(2)
LICsandMICswithscoresabove5onthesovrateindex.TheWorldBank’ssovrateindexisameasureof
repaymentriskthatvariesfrom0to21,withhigherscoresindicatinglowerlevelsofsovereigncreditrisk
(Koseetal.2022).Countrieswithscoresof0-5areindefaultoratahighriskofdefault(seeBox1a).Official
sectorChineselendingcommitmentsaremeasuredinconstant2021USD.Country-yearobservations
withoutofficialsectorChineselendingcommitmentsorsovratescoresareexcluded.
295
Figure
A31:
Annual
loan
commitments
to
LICs
and
MICs
by
financial
institution
and
capital
injections
from
SAFE,
2000-2021
FigureA.31
Annual
loan
commitments
to
LICs
and
MICs
by
financial
institution
and
capital
injections
from
SAFE,
2000-2021
Notes:ThistablepresentsannuallendingcommitmentstoLICsandMICs(inconstant2021USD)from
selectedChinesestate-ownedpolicybanks,state-ownedcommercialbanks,andstate-ownedfunds.Italso
presentscumulativelendingcommitmentstoLICsandMICs(inconstant2021USD)thatarebackedby
creditinsurancefromSinosure.TheverticaldashedlinesrepresentyearsinwhichaSAFEcapitalinjectionis
knowntohavetakenplace.
296
Figure
A32:
Proportion
of
China’s
syndicated
overseas
lending
that
involves
non-Chinese
creditors
FigureA.32
Notes:ThisgraphshowstheannualpercentageofsyndicatedloansinvolvingChinesestate-owned
creditors(bythenumberofloans)thatalsoinvolvenon-Chinesecreditors.
297
Figure
A33:
Proportion
of
Chinese
overseas
syndicated
lending
involving
a
multilateral
institution
FigureA.33
Notes:ThisgraphshowstheannualpercentageofsyndicatedloansinvolvingChinesestate-owned
creditors(bythenumberofloans)thatalsoinvolvemultilateralinstitutions.
Figure
A34:
Average
financial
commitments
from
Chinese
state-owned
creditors
to
different
types
of
syndicated
loans
FigureA.34
Notes:Thisfigurerepresentstheaveragemonetarycommitmentvaluesize(inconstant2021USD)from
officialsectorcreditorsinChinathatparticipatedinsyndicatedloanstoLICsandMICsacrosstwocohorts:
(1)syndicatedloanswithChineseandnon-Chineseparticipants,and(2)syndicatedloanswithonlyChinese
participants.
298
Figure
A35:
Chinese
state-owned
bank
contributions
to
syndicated
loans
in
LICs
and
MICs
by
participant
cohort
FigureA.35
Notes:Thisfigurepresentsallindividualsyndicatedloanscapturedinthe3.0versionofAidData’sGCDF
datasetbasedonthesizeofthemonetarycommitments(inconstant2021USD)fromtheofficialsector
creditorsinChinathatcontributedtotheloans.Thecolorcodescorrespondtotwoseparatetypesof
syndicatedloans:(1)thosewithChineseandnon-Chineseparticipants,and(2)thosewithonlyChinese
participants.
Figure
A36:
Composition
of
China’s
overseas
project
lending
portfolio
by
channel
of
delivery
FigureA.36
299
Figure
A37:
Official
lending
commitments
by
lending
institution
type,
2000-2021
FigureA.37
300
Figure
A38:
Cumulative
percentage
of
infrastructure
portfolio
with
significant
environmental,
social,
or
governance
risk
exposure
FigureA.38
301
Figure
A39:
Infrastructure
projects
located
within
environmentally
sensitive
areas
FigureA.39
Notes:Thisfigureshowsthecumulativenumberandmonetaryvalue(inconstant2021USD)ofChinese
grant-andloan-financedinfrastructureprojectsinLICsandMICsbetween2000and2021thatarelocated
inenvironmentallysensitiveareas.Section2ofChapter3describesthemethodsthatwereusedtoidentify
infrastructureprojectsinenvironmentallysensitiveareas.
302
Figure
A40:
Infrastructure
projects
located
within
socially
sensitive
areas
FigureA.40
Notes:Thisfigureshowsthecumulativenumberandmonetaryvalue(inconstant2021USD)ofChinese
grant-andloan-financedinfrastructureprojectsinLICsandMICsbetween2000and2021thatarelocated
insociallysensitiveareas.Section2ofChapter3describesthemethodsthatwereusedtoidentify
infrastructureprojectsinsocialllysensitiveareas.
303
Figure
A41:
Infrastructure
projects
located
in
geographical
areas
vulnerable
to
political
capture
and
manipulation
FigureA.41
Notes:Thisfigureshowsthecumulativenumberandmonetaryvalue(inconstant2021USD)ofChinese
grant-andloan-financedinfrastructureprojectsinLICsandMICsbetween2000and2021thatarelocated
inareasthatarevulnerabletopoliticalcaptureandmanipulation.Section2ofChapter3describesthe
methodsthatwereusedtoidentifyinfrastructureprojectsingeographicalareasthatarevulnerableto
politicalcaptureandmanipulation.
304
Figure
A42:
Infrastructure
projects
involving
contractors
sanctioned
for
fraudulent
and
corrupt
behavior
FigureA.42
Notes:Thisfigureshowsthecumulativenumberandmonetaryvalue(inconstant2021USD)ofChinese
government-financedinfrastructureprojectsinLICsandMICsthatreliedoncontractorssanctionedbyother
internationalfinanciersforfraudulentandcorruptbehaviorbetween2000and2021.Section2ofChapter3
describesthemethodsusedtoidentifyinfrastructureprojectsthatreliedoncontractorssanctionedbyother
internationalfinanciersforfraudulentandcorruptbehavior.
305
Figure
A43:
Infrastructure
projects
that
encountered
significant
environmental,
social,
or
governance
challenges
over
time
FigureA.43
Notes:Thisfigureshowsthecumulativenumberandmonetaryvalue(inconstant2021USD)ofChinese
government-financedinfrastructureprojectsinLICsandMICsthatencounteredsignificantenvironmental,
social,orgovernancechallengesovertime.Whetheraprojectexperiencedasignificantenvironmental,
social,orgovernancechallengeismeasuredwithabinaryindicatorfromAidDatathatusestheproject’s
descriptionfieldtodetermineiftherewasevidenceofasignificantenvironmental,social,orgovernance
challengebefore,during,orafterimplementation.Formoredetails,seeSection2ofChapter3.
306
Figure
A44:
ESG
risk
prevalence
in
overseas
infrastructure
portfolio
from
China
to
LICs
and
MICs
FigureA.44
Notes:Thisfigureshowstheannualnumber(bars)and(withasolidline)theannualpercentageofthe
Chinesegrant-andloan-financedinfrastructureprojects(count)inLICsandMICsthatencountered
significantenvironmental,social,orgovernancerisksbetween2000-2021.Projectsarerecordedinthe
yearswhentheysecuredfinancialcommitmentsfromChina,althoughtheESGrisksthattheyencountered
mayhavematerializedafterthefinancialcommitmentyear.ThepresenceofsignificantESGriskexposureis
basedonaproject-levelcompositemeasurethatisdescribedinSection2ofChapter3.
307
Figure
A45:
Environmental
risk
prevalence
in
overseas
infrastructure
FigureA.45
Notes:Thisfigureshowstheannualnumber(bars)andpercentage(ofprojectcount)ofChinese
government-financedinfrastructureprojectsinLICsandMICsthatencounteredsignificantenvironmental
risksbetween2000-2021.Projectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfrom
China,althoughtheenvironmentalrisksthattheyencounteredmayhavematerializedafterthefinancial
commitmentyear.Thepresenceofsignificantenvironmentalriskexposureisbasedonaproject-level
compositemeasurethatisdescribedinSection2ofChapter3.
308
Figure
A46:
Social
risk
prevalence
in
overseas
infrastructure
portfolio
FigureA.46
Notes:Thisfigureshowstheannualnumber(bars)andpercentage(ofprojectcount)ofChinese
government-financedinfrastructureprojectsinLICsandMICsthatencounteredsignificantsocialrisks
between2000-2021.Projectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfrom
China,althoughthesocialrisksthattheyencounteredmayhavematerializedafterthefinancial
commitmentyear.Thepresenceofsignificantsocialriskexposureisbasedonaproject-levelcomposite
measurethatisdescribedinSection2ofChapter3.
309
Figure
A47:
Governance
risk
prevalence
in
overseas
infrastructure
portfolio
FigureA.47
Notes:Thisfigureshowstheannualnumber(bars)andpercentage(ofprojectcount)ofChinese
government-financedinfrastructureprojectsinLICsandMICsthatencounteredsignificantgovernancerisks
between2000-2021.Projectsarerecordedintheyearswhentheysecuredfinancialcommitmentsfrom
China,althoughthegovernancerisksthattheyencounteredmayhavematerializedafterthefinancial
commitmentyear.Thepresenceofsignificantgovernanceriskexposureisbasedonaproject-level
compositemeasurethatisdescribedinSection2ofChapter3.
310
Figure
A48:
Infrastructure
projects
supported
by
one
or
more
de
facto
environmental,
social,
or
governance
risk
mitigation
efforts
FigureA.48
Notes:ThisfigureshowsthecumulativemonetaryvalueofChinesegrant-andloan-financedinfrastructure
projects(inconstant2021USD)supportedbyoneormoredefactoenvironmental,social,orgovernance
riskmitigationefforts(asmeasuredviathekeywordsearch-basedmethod)from2000to2021.
311
Figure
A49:
De
jure
ESG
safeguard
stringency
in
China’s
overseas
infrastructure
portfolio
with
ESG
risk
exposure
FigureA.49
Notes:ThisfigureshowstheannualpercentageofChina’sgrant-andloan-financedinfrastructureproject
portfolioinLICsandMICs(inconstant2021USD)thatpresentedsignificantenvironmental,social,or
governancerisksacrosstwocohorts:thosewithstrongdejureESGsafeguardsandthoseweakdejureESG
safeguards.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwooutof
three(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”Projectsthatdo
notmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.Thepresenceofsignificant
ESGriskexposureisbasedonaproject-levelcompositemeasurethatisdescribedinSection2ofChapter
3.
312
Figure
A50:
Infrastructure
project
portfolio
with
de
jure
vs.
de
facto
ESG
risk
mitigation
FigureA.50
Notes:Thisfigurepresents(inbars)theannualpercentageofChinesegrant-andloan-financed
infrastructureprojects(byprojectcount)inLICsandMICssupportedbyoneormoredefacto
environmental,social,orgovernanceriskmitigationefforts(asmeasuredviathekeywordsearch-based
method)from2000to2021.Italsopresents(withaline)theannualpercentageofChinesegrant-and
loan-financedinfrastructureprojects(byprojectcount)inLICsandMICswithstrongdejureESG
safeguards.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwooutof
three(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”
313
Figure
A51:
Infrastructure
project
portfolio
with
de
jure
vs.
de
facto
ESG
risk
mitigation,
by
year
of
project
completion
FigureA.51
Notes:Thisfigurepresents(inbars)theannualpercentageofChinesegrant-andloan-financed
infrastructureprojectportfolio(inconstant2021USD)inLICsandMICssupportedbyoneormoredefacto
environmental,social,orgovernanceriskmitigationefforts(asmeasuredviathekeywordsearch-based
method)from2000to2021.Italsopresents(withaline)theannualpercentageofChinesegrant-and
loan-financedinfrastructureprojectportfolio(inconstant2021USD)inLICsandMICswithstrongdejure
ESGsafeguards.Bothmeasuresarebasedontheyeartheprojectwascompleted(whereknown)insteadof
thecommitmentyear.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwo
outofthree(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”
314
Figure
A52:
Infrastructure
project
portfolio
with
de
jure
vs.
de
facto
ESG
risk
mitigation,
by
year
of
project
commencement
FigureA.52
Notes:Thisfigurepresents(inbars)theannualpercentageofChinesegrant-andloan-financed
infrastructureprojectportfolio(inconstant2021USD)inLICsandMICssupportedbyoneormoredefacto
environmental,social,orgovernanceriskmitigationefforts(asmeasuredviathekeywordsearch-based
method)from2000to2023.Italsopresents(withaline)theannualpercentageofChinesegrant-and
loan-financedinfrastructureprojectportfolio(inconstant2021USD))inLICsandMICswithstrongdejure
ESGsafeguardsoverthesametimeperiod.Bothmeasuresarebasedontheyeartheprojectbegan
implementation(whereknown)insteadofthecommitmentyear.Projectswith“strong”dejureESG
safeguardsaredefinedasthosewithatleasttwooutofthree(environmental,socialandgovernance)
safeguardstringencymetricsthatare“high.”
315
Figure
A53:
Proportion
of
infrastructure
project
financing
facing
significant
environmental
risks
by
whether
the
project
financing
had
strong
de
jure
environmental
safeguards
versus
those
with
weak
de
jure
ESG
safeguards
FigureA.53
Notes:ThisfigureshowsthepercentageofChinesegrant-andloan-financedinfrastructureprojectportfolio
(inconstant2021USD)inLICsandMICsfrom2000to2021thatpresentedsignificantenvironmentalrisk
acrosstwocohorts:(1)projectswithstrongdejureenvironmentalsafeguards,and(2)projectswithout
strongdejureenvironmentalsafeguards.Projectswith“strong”dejureenvironmentalsafeguardsare
definedasthosewithascoreof“high”onAidData’ssafeguardstringencyscale.Projectsthatpresent
significantenvironmentalriskaremeasuredwith1/0project-levelcompositemeasure(basedontwoofthe
1/0inputindicatorsdescribedinSection2ofChapter3:whethertheprojectwaslocatedinan
environmentallysensitiveareaand/orwhethertheproject’sdescriptionfieldprovidesevidenceofa
significantenvironmentalchallengebefore,during,orafterimplementation).
316
Figure
A54:
Proportion
of
infrastructure
project
financing
facing
significant
social
risks
by
whether
the
project
financing
had
strong
de
jure
environmental
safeguards
versus
those
with
weak
de
jure
ESG
safeguards
FigureA.54
Notes:ThisfigureshowsthepercentageofChinesegrant-andloan-financedinfrastructureprojectportfolio
(inconstant2021USD)inLICsandMICsfrom2000to2021thatpresentedsignificantsocialriskacrosstwo
cohorts:(1)projectswithstrongdejuresocialsafeguards,and(2)projectswithoutstrongdejuresocial
safeguards.Projectswith“strong”dejuresocialsafeguardsaredefinedasthosewithascoreof“high”on
AidData’ssafeguardstringencyscale.Projectsthatpresentsignificantsocialriskaremeasuredwith1/0
project-levelcompositemeasure(basedontwoofthe1/0inputindicatorsdescribedinSection2of
Chapter3:whethertheprojectwaslocatedinasociallysensitiveareaand/orwhethertheproject’s
descriptionfieldprovidesevidenceofasignificantsocialchallengebefore,during,orafter
implementation).
317
Figure
A55:
Proportion
of
infrastructure
project
financing
facing
significant
governance
risks
by
whether
the
project
financing
had
strong
de
jure
environmental
safeguards
versus
those
with
weak
de
jure
ESG
safeguards
FigureA.55
Notes:ThisfigureshowsthepercentageofChinesegrant-andloan-financedinfrastructureprojectportfolio
(inconstant2021USD)inLICsandMICsfrom2000to2021thatpresentedsignificantgovernancerisk
acrosstwocohorts:(1)projectswithstrongdejuregovernancesafeguards,and(2)projectswithoutstrong
dejuregovernancesafeguards.Projectswith“strong”dejuregovernancesafeguardsaredefinedasthose
withascoreof“high”onAidData’ssafeguardstringencyscale.Projectsthatpresentsignificantgovernance
riskaremeasuredwith1/0project-levelcompositemeasure(basedonthreeofthe1/0inputindicators
describedinSection2ofChapter3:whethertheprojectwaslocatedinanareavulnerabletopolitical
captureandmanipulation,whethertheprojectreliedoncontractorssanctionedbyotherinternational
financiersforfraudulentandcorruptbehavior;and/orwhethertheproject’sdescriptionfieldprovides
evidenceofasignificantgovernancechallengebefore,during,orafterimplementation).
318
Figure
A56:
Composition
of
infrastructure
project
portfolio:
Reliance
on
China
Eximbank
and
weak
de
jure
ESG
safeguards
FigureA.56
Notes:ThisfigureshowsthepercentofChinesegrant-andloan-financedinfrastructureprojects(measured
asconstant2021USD)withweakdejureESGsafeguardsaswellasthepercentageofChinesegrant-and
loan-financedinfrastructureprojects(measuredbyconstant2021USD)financedviabilateralloansfrom
ChinaEximbank.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwoout
ofthree(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”Projectsthat
donotmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.
319
Figure
A57:
Composition
of
infrastructure
project
portfolio:
Reliance
on
China
Development
Bank
and
weak
de
jure
ESG
safeguards
FigureA.57
Notes:ThisfigureshowsthepercentageofChinesegrant-andloan-financedinfrastructureprojects
(measuredasconstant2021USD)withweakdejureESGsafeguardsaswellasthepercentageofChinese
grant-andloan-financedinfrastructureprojects(measuredbyconstant2021USD)financedviabilateral
loansfromCDB.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwooutof
three(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”Projectsthatdo
notmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.
Figure
A58:
Proportion
of
China’s
infrastructure
project
portfolio
behind
schedule
by
de
jure
ESG
safeguard
strength
FigureA.58
Notes:ThisfigurepresentsthepercentageofChina’sgrant-andloan-financedinfrastructureproject
portfolio(inconstant2021USD)thatranbehindscheduleacrosstwocohorts:projectswithstrongdejure
environmental,social,andgovernancesafeguardsandprojectswithweakdejureenvironmental,social,
andgovernancesafeguards.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithat
320
leasttwooutofthree(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”
Projectsthatdonotmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.“Behind
schedule”isdefinedasprojectswheretheactualimplementationstartdatetookplace3monthsormore
afteritsoriginallyscheduledimplementationstartdateaswellasprojectswheretheactualcompletiondate
tookplace3months(ormore)afteritsoriginallyscheduledcompletiondate.Onlyactiveprojectsand
completedprojectsthatsecuredofficialcommitmentsfromChinaareincludedintheanalysis.
Figure
A59:
Average
length
of
commencement
delays
in
infrastructure
projects
by
de
jure
ESG
safeguard
strength
FigureA.59
Notes:Thisfigurecomparestheaveragelengthofcommencementdelays(indays)inChina’sgrant-and
loan-financedinfrastructureprojectportfolioacrosstwocohorts:projectswithstrongdejureenvironmental,
social,andgovernancesafeguardsandprojectswithweakdejureenvironmental,social,andgovernance
safeguards.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwooutof
three(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”Projectsthatdo
notmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.Commencementdelaysare
calculatedbytakingthedifference(incalendardays)betweentheoriginallyscheduledproject
implementationstartdateandtheactualprojectimplementationstartdate.Onlyactiveprojectsand
completedprojectsthatsecuredofficialcommitmentsfromChinaareincludedintheanalysis.
Figure
A60:
Average
length
of
completion
delays
in
infrastructure
projects
by
de
jure
ESG
safeguard
strength
FigureA.60
Notes:Thisfigurecomparestheaveragelengthofcompletiondelays(indays)inChina’sgrant-and
loan-financedinfrastructureprojectportfolioacrosstwocohorts:projectswithstrongdejureenvironmental,
social,andgovernancesafeguardsandprojectswithweakdejureenvironmental,social,andgovernance
safeguards.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwooutof
three(environmental,socialandgovernance)safeguardstringencymetricsthatare“high.”Projectsthatdo
notmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguardsCompletiondelaysare
calculatedbytakingthedifference(incalendardays)betweentheoriginallyscheduledprojectcompletion
321
dateandtheactualprojectcompletiondate.Onlyactiveprojectsandcompletedprojectsthatsecured
officialcommitmentsfromChinaareincludedintheanalysis.
Figure
A61:
Average
time
to
complete
an
infrastructure
project
by
de
jure
ESG
safeguard
strength
FigureA.61
Notes:ThisfigurecomparestheaveragenumberofcalendardaysthatithastakentocompleteChinese
grant-andloan-financedinfrastructureprojectsinLICsandMICsacrosstwocohorts:projectswithstrong
dejureenvironmental,social,andgovernancesafeguardsandprojectswithweakdejureenvironmental,
social,andgovernancesafeguards.Projectswith“strong”dejureESGsafeguardsaredefinedasthosewith
atleasttwooutofthree(environmental,socialandgovernance)safeguardstringencymetricsthatare
“high.”Projectsthatdonotmeetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.The
amountoftimeneededtocompleteaprojectiscalculatedbymeasuringthenumberofcalendardays
betweentheactualprojectimplementationstartdateandtheactualprojectcompletiondate.Onlyactive
projectsandcompletedprojectsthatsecuredofficialcommitmentsfromChinaareincludedintheanalysis.
322
Figure
A62:
ESG
map
-
project
count
FigureA.62
ESG
map
-
project
count
Notes:ThismappresentsthegeographicalareaswhereChina’sgrant-andloan-financedinfrastructure
projectportfolio(measuredinprojectcounts)hassignificantenvironmental,social,orgovernance(ESG)risk
exposure.Darker(purple)colorsrepresentareaswheretheportfoliohashighlevelsofriskexposureand
lighter(pink)colorsrepresentareaswheretheportfoliohaslowerlevelsofriskexposure.Environmentalrisk
exposure,socialriskexposure,andgovernanceriskexposurearebasedontheproject-levelcomposite
measuresthataredescribedinSection2ofChapter3.Ifaprojectfallsacrossmultiplegridcells,itis
countedineverygridcellthatitintersects.Inotherwords,foreverygridcell,wecountthenumberof
projectpoints,linesorpolygonsthatintersectit.
SectionA-2:HowAidDatameasuresconcessionalityandintent
Aspartofitsdatacollectionandclassificationsystem,AidDatadesignateseachfinancialandin-kindtransfer(“flow”)fromanofficialsectorinstitutionasOfficialDevelopmentAssistance(ODA)orOtherOfficialFlows(OOF).TheOECD’sDevelopmentAssistanceCommittee(DAC)hasusedthesedesignationssince1972todistinguishbetweenflowsfromofficialsectorinstitutionsthat(a)areprovidedonconcessionaltermsandthatpromoteandspecificallytargettheeconomicdevelopmentandwelfareofdevelopingcountries(ODA),and(b)areprovidedonnon-concessionaltermsordonotspecificallytarget
323
theeconomicdevelopmentandwelfareofdevelopingcountries(OOF).ThesumofODAandOOFissometimesreferredtoasOfficialFinancialFlows,OfficialFinancing,orOverseasDevelopmentFinance.ManyDACcountries,non-DACcountries,andmultilateralinstitutionsreportthevolumeandcompositionoftheirofficialsectorflowsaccordingtothesecategoriesandcriteria.InalignmentwiththeOECD-DAC’sowndefinitions,AidDataclassifieseachprojectinthe3.0datasetaseither“ODA-like”or“OOF-like.”Thisuniquefeatureofthe3.0datasetsetsitapartfromotherpubliclyavailabledatasetsthatmeasureChinesedevelopmentfinanceinthatitallowsanalyststomake“apples-to-apples”comparisonsofChinesedevelopmentfinanceandotherinternationalsourcesofdevelopmentfinance(thatreporttheirODAandOOFflowdatatotheOECD-DAC).
ThecriteriaforwhetheraflowqualifiesasODAorOOFisdeterminedbytheOECD-DAC.Itisbasedon(1)theintentoftheflow(whetheritsprimaryintentwasdevelopmentornot),(2)theincomeclassificationofthereceivingcountry,and(3)theconcessionalityleveloftheflow.369Allgrantsandin-kindtransfersaretreatedasconcessional.However,a“grantelement”measureisusedtocalculatetheconcessionalitylevelofallloans.Thismeasure,whichvariesfrom0percentto100percent,seekstocapturethegenerosityofaloan—ortheextenttowhichitispricedbelowmarketrates.Inprinciple,anyloanprovidedonentirelynon-concessionaltermsshouldhaveagrantelementof0percent.
WhilethefirsttwocriteriahaveremainedconsistentsincetheconceptofODAwasintroducedmorethanfivedecadesago,theOECD-DACrecentlymadechangestothethird(concessionality)criterion.Until2017,aloanfromanofficialsectorinstitutiontoalow-incomeormiddle-incomecountryhadtomeetaconcessionality(grantelement)thresholdof25%toqualifyasODA.However,in2018,theOECD-DACintroducedatieredsystemofdiscountratesandconcessionalitythresholdsbasedontheincomeclassificationsofborrowercountriesandwhetherborrowinginstitutionsareofficialsectororprivatesectorinstitutions.The2018definitionofconcessionalityisbasedonthefollowingcriteria:
●Forloanstoofficialsectorinstitutions,thefollowingconcessionalitythresholdsapply:
○Least-developedcountriesandlow-incomecountries:aminimumgrantelementof45%(calculatedusinga9%discountrate).
○Lower-middleincomecountries:aminimumgrantelementof15%(calculatedusinga7%discountrate).
○Upper-middleincomecountries:aminimumgrantelementof10%(calculatedusingadiscountrateof6%).
●Forloanstoprivatesectorinstitutions,theOECD-DACmaintainsthepre-2018definitionofconcessionalityandrequiresagrantelementofatleast25%(thatiscalculatedusinga10%discountrate).370
Toensurecomparabilitybetweentheflowsdocumentedinthe3.0versionoftheGCDFdatasetandtheflowdatapublishedbytheOECD-DAC,AidDatahasappliedthesedefinitionsinthefollowingmanner:
Intent:AidDatacodestheintentofeachfinancialandin-kindtransfer(“flow”).Flowswith“developmentintent”arethosethatareprimarilyorientedtowardthepromotionofeconomicdevelopmentandwelfareintherecipientcountry.Flowswith“commercialintent”arethosethatprimarilyseektopromotethecommercialinterestsofthecountryfromwhichthefinancialtransferhasoriginated(e.g.,encouragingthe
370AccordingtotheOECD,themethodforcalculatingtheODAgrantequivalentforloanstoprivatesectorinstitutionshasnotyetbeenformalized,anddiscussionstodosoarecurrentlyongoingattheOECD-DAC.Untilanagreementhasbeenformalized,thepre-2018concessionalitydefinitionstillapplies.369Anadditionalcriteriaisthattheflowmustbeprovidedbyofficialagencies,includingstateandlocalgovernmentsortheirexecutiveagencies.AidData’sGCDF3.0onlytracksofficialChineseagencies,sothiscriteriaisalwaysmet.
324
exportofChinesegoodsandservices).Flowswith“representationalintent”arethosethatprimarilyseektopromoteabilateralrelationshipwithanothercountryorotherwisepromotethelanguage,culture,orvaluesofthecountryfromwhichthefinancialtransferhasoriginated(e.g.,theestablishmentofaConfuciusInstituteorChineseculturalcenter).Flowswith“militaryintent”arethosethatseektopromotethesecurityinterestsofthecountryfromwhichthefinancialtransferoriginatesorstrengthenthelethalforcecapabilitiesofmilitaryinstitutionsintherecipientcountry.
ODAIncomeClassification:AidDatareportstheincomeclassificationgroupoftheborrowingcountry.FlowstocountriesnoteligibleforODAareautomaticallyassignedtothe“OOF-like”category.
Concessionality:
●Forflowscommittedbetween2000and2017,aflowisclassifiedas“ODA-like”whenit(1)hasdevelopmentintent,(2)hasagrantelementofatleast25%(usinga10%discountrate),and(3)supportsacountrythatisODA-eligibleaccordingtotheOECD-DAC’sODAincomeclassificationlist.
●Forflowscommittedbetween2018and2021,aflowisclassifiedas“ODA-like”whenit(1)hasdevelopmentintent,(2)hasaconcessionalitylevelthatmeetsthenewcriteria(establishedin2018definition),and(3)supportscountrythatisODA-eligibleaccordingtotheOECD-DAC’sODAincomeclassificationlist.
Bydefinition,anyinternationalofficialsectorflowsnotclassifiedasODA-likeareclassifiedasOOF-like.TheOOF-likeflowsinthe3.0versionofAidData’sGCDFdatasetlargelyconsistofexportcreditsandnon-concessionalloans.
Insomecases,wearenotabletodetermineifaninternationalofficialsectorflowwouldqualifyasODAor
OOFbecauseofinsufficientlydetailedinformationinsourcedocumentation.Insuchcases,theflowin
questioniscategorizedasVague(OfficialFinance).
SectionA-3:HowdoesAidDatameasurethecumulativestockofofficialfinancialflowsfromChinatoLICsandMICs?
AsweexplainatgreaterlengthinChapter2,short-termemergencyrescueloansrepresentanincreasingly
importantpartofChina’soverseasportfolioofloanstoLICsandMICs.Nearlyalloftheseborrowings,which
aretypicallyusedtorefinancematuringdebts,carrydejurematuritiesofoneyearorless(i.e.,theyare
initiallyscheduledforrepaymentin12monthsorless).However,itisnotunusualforfinancially-distressed
LICsandMICstoreceiveshort-termemergencyrescueloansfromthesameChinesecreditorinaseriesof
consecutiveyears.So-called“rollover”emergencyrescueloanscomeintwovarieties:(1)thosethatreach
theiroriginalcontractualmaturitydatesandsecurefinalmaturitydateextensions;and(2)thosethatare
repaidontheiroriginalcontractualmaturitydatesandreissued(withsimilarordifferentfacevaluesand
borrowingterms)andassignednewmaturitydates.However,amongserialrecipientsofshort-term
emergencyrescueloans,itisseldompossible—withpubliclyavailablesourcesofinformation—to
differentiatebetweenthosewhohadtheirfinalmaturitydatesextendedandthosewhofullyrepaidontheir
originalcontractualmaturitydatesbutwerereissuednewloans.
ThisrelativelynewfeatureofChina’soverseaslendingprogramraisesanimportantquestionabouthowto
accuratelyestimatethecumulativestockofofficialfinancialflows—orlendingcommitments—fromChinato
325
LICsandMICs.NeithertheOECD'sCreditorReportingSystem(CRS)northeWorldBank’sDebtor
ReportingSystem(DRS)asklendersorborrowerstodiscloseloanswithmaturitiesofoneyearorless.
However,mostofChina’sshort-termemergencyrescueloanshavedefactomaturitiesthatsubstantially
exceedoneyear(Hornetal.2023a),whichmakesitdifficulttojustifytheexclusionofallemergencyrescue
loansfromstock-orflow-basedmeasuresofofficialfinancialcommitments(orlendingcommitments)from
ChinatoLICsandMICs(seeBox2c).Atthesametime,rolloverdebtpresentsanovercountingriskbecause
itstraddlesafinelinebetweennewlendingcommitmentsandmaturityextensionsofexistinglending
commitments.Thisriskisparticularlyrelevanttoestimationsofthecumulativestockofofficialfinancial
flows(orlendingcommitments)fromChina.
Inordertoaddressthischallenge,the3.0versionofAidData’sGCDFdatasetincludesthreenewvariables
(fields)thatmeasuretransactionamountswithoutincludinganyrolloveramountsfromPBOCswapline
borrowingsoremergencyrescueloansfromothercreditors(withmaturitiesofoneyearorless).371These
amountsarereportedintheiroriginalcurrenciesofdenomination,nominalUSD,andconstant2021USD
viathe"AdjustedAmount(OriginalCurrency),”"AdjustedAmount(ConstantUSD2021),"and"Adjusted
Amount(NominalUSD)"variables.372
Figures1.3,1.4,1.8,2.8,andA14—aswellasTables2.1,A15,andA16—usethe“AdjustedAmount
(ConstantUSD2021)”variabletorepresentthecumulativestockofofficialfinancialflowsfromChinato
countriesorborrowinginstitutions.Theotherfiguresandtablesinthisreportusethe“Amount(Constant
USD2021)”fieldforfinancialcalculations,unlessotherwisenoted.
SectionA-4:AdditionaldetailsontheBIS-basedestimatesofChina’sinternationallendingportfolio
Inthesection,weprovideastep-by-stepdescriptionofhowtheBIS-basedmeasuresoftotaloutstanding
creditfromChinesebankstooverseasborrowersarederived.However,beforewedoso,therearefivekey
caveatsandconsiderationsthatreadersshouldkeepinmindregardingtheBIScross-borderlending
data.373First,thedataarerepresentedasamountsoutstanding,whichiseffectivelyequivalentto
cumulativedisbursementsminuscumulativerepayments(i.e.,creditstocksratherthancreditflows).A
separate,butrelated,pointisthatcumulativelendingcommitmentsusuallyexceedamountsoutstanding.
373WearegratefultoHaonanZhouofPrincetonUniversityforhisadviceonhowtoutilizetheLBSdatafromtheBIS.372Usersofthe3.0versionofAidData’sGCDFdatasetcanestimate“rollover”loanamounts(intheiroriginalcurrenciesofdenomination)bysubtractingthevaluesintheAdjustedAmount(OriginalCurrency)fieldfromthevaluesintheAmount(OriginalCurrency)field.NominalUSD“rollover”loanamountscanbeestimatedbysubtractingthevaluesintheAdjustedAmount(NominalUSD)fieldfromthevaluesintheAmount(NominalUSD)field.Constant2021USD“rollover”loanamountscanbeestimatedbysubtractingthevaluesintheAdjustedAmount(ConstantUSD2021)fieldfromthevaluesintheAmount(ConstantUSD2021)field.371Wheneverpossible,foreachemergencyrescueloan(PBOCswapborrowing)oftherollovervariety,wecalculateatransactionamountthatexcludestherolloveramountbytakingthedifferencebetweenthelevelofoutstandingdebtinthecurrentyearandthepreviousyear.ThisapproachisconsistentwiththeonetakenbyHornetal.(2023a)toderivenet(new)PBOCswapborrowings.Incaseswhenthisapproachcannotbeappliedbutthereisevidenceofthesamelenderprovidingaseriesofshort-termemergencyrescueloans(withidenticalfacevaluesanddejurematuritiesof1yearorless)tothesameborrowerthatarerepaidontheiroriginalcontractualmaturitydatesandsubsequentlyreissuedinconsecutiveyears,werecordthefacevalueoftheoriginalloancommitmentinthefirstyearbutnotthefacevaluesoftheloancommitmentsinsubsequentyears.
326
Second,theBISreliesonself-reporteddatafrom“internationallyactivebanks,”sounlikeAidData’sGCDF
datasetitdoesnotcaptureoverseascreditextendedbynon-bankinstitutions(e.g.,suppliercreditsfrom
Chinesecompanies,loansfromChina’sMinistryofCommerce).Third,theBISdataarereportedby
state-ownedandprivately-ownedbanks,butChina’sbankingsectorisdominatedbystate-ownedbanks,
whichareresponsiblefornearlyallofChina’soverseaslendingactivities(Hornetal.2021).Fourth,BISdata
canbeaccessedandanalyzedaccordingtothenationalityofthereportinginstitutionortheresidenceof
thecounterparty,andsinceasignificantproportionofChinesebanklendingtoLICsandMICsischanneled
throughoffshorefinancialcentersandforeignaffiliatesofChinesebanks(Ceruttietal.2018),itisgenerally
advisableforanalystsofChina’soverseaslendingactivitiestousetheBISdatathatareorganizedaccording
tothenationalityofthereportinginstitution.Fifth,BISdatarelyonthevoluntarydisclosureeffortsofbanks,
andHornetal.(2021)provideevidenceofsomeunderreportingtotheBIS.374
WenowprovideadetailedexplanationofhowwederivedtheBIS-basedmeasuresoftotaloutstanding
creditfromChinesebankstooverseasborrowers.BIS(2022)reportsthefollowingbreakdownoftotal
outstandingcross-bordercreditduringthesecondquarterof2018:$20.27trilliontoHICborrowers,$4.02
trilliontoLICandMICborrowers,and$4.76trilliontooffshorefinancialcenters(OFC)borrowers.375The
LBSdatafromtheBISalsoindicatethattotaloutstandingcreditfromChinesebankstoalloverseas
borrowers—includingthoseinLICs,MICs,HICs,andOFCs—was$2.15trillionduringthesecondquarterof
2018.376Ceruttietal.(2023)providecomplementaryevidencefromtheLBS;theyfindthat,duringthe
secondquarterof2018,Chinesebankswereresponsiblefor2.4%oftotaloutstandingcross-bordercredit
toHICborrowers,23.7%oftotaloutstandingcross-bordercredittoLICandMICborrowers,and13.7%of
totaloutstandingcross-bordercredittoOFCborrowers.
Thesefiguresimplythat,duringthesecondquarterof2018,22.6%ofoverseascreditfromChinesebanks
wasdirectedtoHICborrowers,while44.3%wasdirectedtoLICandMICborrowersand30.3%was
directedtoOFCborrowers.377GiventhatthetotalamountsoutstandingunderChinesebankloansto
overseasborrowersareavailablefromtheLBSdata,onecanusetheseratios(22.6%,44.31%,and30.33%)
toestimatethethetotalamountsoutstandingunderChinesebankloanstoHICborrowers,LICandMIC
borrowers,andOFCborrowersinpreviousyears(2015-2017)andsubsequentyears(2019-2021).378Wedo
sointhethird,fifth,andeighthcolumnsofTable2.1.
However,somethingimportanthappenedinDecember2022:theBISremovedOFCsasacountrygrouping
fromthestatisticaltablesonitspublicwebsiteandreassignedthecountriesthatwerepreviouslyassigned
totheOFCgroupingtotheHICcountrygroupingortheLIC/MICcountrygrouping.379Thelatestvintageof
379OFCsarelegaljurisdictionsthatserveasintermediariesofcross-borderfinancialflowsandspecializeintheprovisionofbankingservicestonon-residents(PoglianiandWooldridge2022).TheBISdefinesOFCsas“countrieswithbankingsectorsdealingwithnon-residentsand/orinforeigncurrencyonascaleoutofproportionrelativetothesizeofthehosteconomy”(BIS1995).TheBISidentified21OFCsasof2011:378Theseincomebracket-levelratiosareveryslow-movingovertime.Seehttps://www.bis.org/statistics/rppb2304/intgraphs/ch1graphA2.htm.377TheestimateforAE(HIC)borrowersisbasedonthefollowingcalculation:($20.27trillion*0.024)/$2.15trillion.TheestimateforEMDE(LICandMIC)borrowersisbasedonthefollowingcalculation:($4.02trillion*0.237)/$2.15trillion.TheestimateforOFCborrowersisbasedonthefollowingcalculation:($4.76trillion*0.137)/$2.15trillion.376Seehttps://stats.bis.org/statx/srs/table/a7?c=CN&p=20182.375Seehttps://www.bis.org/statistics/rppb2207/intgraphs/ch1graphA2.htm.374Morespecifically,theyfindthattheirowndebtstockestimates(basedonAidDataandothersources)“significantlyexceedBISimplieddebtstocksforsomeoftheriskiestandmostvolatiledebtorcountriesworldwide,suchasAngola,EquatorialGuinea,VenezuelaorZimbabwe”(Hornetal.2021:30).
327
theLBSdataontotaloutstandingcross-bordercredit,whichreliesupontheseupdatedcountrygroupings,
isthereforeusefulinthatithascreatedanalternativewayofestimatingtotaloutstandingcross-border
creditfromChinesebankstoLIC/MICandHICborrowers.
BIS(2023)providesevidencethattotalcross-borderlendingtoHICborrowerswas$20.5trillion,total
cross-borderlendingtoLICandMICborrowerswas$5.48trillion,andtotalcross-borderlendingtoother
borrowers(thatcouldnotbeassignedtoeithertheHICorLIC/MICcategory)was$3.4trillionduringthe
secondquarterof2018.380Thesefigures—inconjunctionwiththeaforementionedCeruttietal.(2023)
estimates—implythatduringthesecondquarterof2018:22.8%oftotalcross-borderChinesebanklending
wasdirectedtoHICborrowers,while60.4%wasdirectedtoLICandMICborrowers,and21.6%was
directedto“other”borrowers(i.e.,overseasborrowersthatcannotbeallocatedbycounterparty
residence).381
Hereagain,sincetotalamountsoutstandingundercross-borderChinesebankloansareavailablefromthe
LBSdata,onecanusethese(22.8%,60.4%,and21.6%)ratiostoestimatethethetotalamounts
outstandingunderChinesebankloanstoHICborrowers,LICandMICborrowers,and“other”borrowersin
previousyears(2015-2017)andsubsequentyears(2019-2021).Table2.1providesalowerboundestimate
($1.16trillion)andanupperboundestimate($1.58trillion)oftotaloutstandingcross-bordercreditfrom
ChinesebankstoLICandMICborrowersasof2021aswellasalowerboundestimate($594.6billion)and
anupperboundestimate($599.9billion)oftotaloutstandingcross-bordercreditfromChinesebanksto
HICborrowersin2021.382Regardlessofwhetheroneuseslowerboundorupperboundestimates,Table
2.1demonstratesthattotaloutstandingcreditfromChinesebankstoLICandMICborrowerseffectively
doubledbetween2015and2021(eitherfrom$644billionto$1.16trillionorfrom$878billionto$1.58
trillion).ThistoplinepatternisremarkablyconsistentwiththeGCDFdataonChina’scumulativeoverseas
lendingcommitmentsbetween2015and2021,whichincreased(innominalterms)from$620billionin
2015to$1.03trillionin2021.383
383Forcomparability’ssake,thesefiguresarereportedincurrent(nominal)USD.However,the3.0versionofAidData’sGCDFdatasetdemonstratesthatChina’scumulativeoverseaslendingcommitmentsincreasedfrom$823billionin2015to$1.3trillionin2021inconstant2021USD.382Theseestimatesarereportedinthesecond,third,fifth,andsixthcolumnsofTable2.1.TheEMDE(LICandMIC)estimatesarebasedonalowerboundassumption(that44.31%oftotalcross-borderChinesebanklendingwasdirectedtoEMDEborrowers)andanupperboundassumption(that60.4%oftotalcross-borderChinesebanklendingwasdirectedtoEMDEborrowers).TheAE(HIC)estimatesarebasedonalowerboundassumption(that22.6%oftotalcross-borderChinesebanklendingwasdirectedtoAEborrowers)andanupperboundassumption(that22.8%oftotalcross-borderChinesebanklendingwasdirectedtoAEborrowers).Table2.1alsoindicatesthattotaloutstandingcross-bordercreditfromChinesebankstoOFCborrowersin2021wasapproximately$798billionandtotaloutstandingcross-bordercreditfromChinesebanksto“other”borrowersin2021wasapproximately$568.3billion.AllofthesefiguresarebasedonaggregateBISdatathatmeasuretotaloutstandingcross-bordercreditfromChinesebanksinthefinalquarterofeachyearbetween2015and2021.Seehttps://stats.bis.org/statx/srs/table/a7?c=CN&p=20154381TheestimateforAE(HIC)borrowersisbasedonthefollowingcalculation:($20.5trillion*0.024)/$2.15trillion.TheestimateforEMDE(LICandMIC)borrowersisbasedonthefollowingcalculation:($5.48trillion*0.237)/$2.15trillion.Theestimatefor“other”borrowersisbasedonthefollowingcalculation:($3.4trillion*0.137)/$2.15trillion.380Seehttps://www.bis.org/statistics/rppb2304/intgraphs/ch1graphA2.htm.Aruba,Bahamas,Bahrain,Barbados,Bermuda,BritishVirginIslands(VG)andWestIndiesUK,CaymanIslands,Gibraltar,Guernsey,HongKongSAR,IsleofMan,Jersey,Lebanon,Liberia,MacaoSAR,Mauritius,NetherlandsAntilles(CuraçaoandSintMaarten),Panama,Samoa,Singapore,andVanuatu.TheBISdidnotidentifyVGasastand-alonejurisdiction;instead,itisincludedinalarger,BIS-definedcountrygroupingcalled“WestIndiesUK”thatincludesAnguilla,AntiguaandBarbuda,Montserrat,andStKittsandNevis.
328
Tobeclear,thetime-varying,BIS-basedestimatesrestuponanimportantassumption:thatthegeographic
allocationofoutstandingcreditfromChinesebanksacrossincomegroups(22.8%toborrowersinHICs,
60.4%toborrowersinLICs/MICsand21.6%to“other”borrowers)wasstablebetween2015and2021.If
thisassumptionisnottrue(oratleastnotforthemostparttrue),analternativeexplanationforthemajor
increaseinChina’stotaloutstandingcredittoborrowersinLICs,MICs,HICs,and“other”overseas
jurisdictions(capturedintheninthcolumnofTable2.1)—from$1.45trillionin2015to$2.63trillionin
2021—couldbethatlendingtoborrowersinHICsand/or“other”overseasjurisdictionsoutpacedlending
toLICs/MICsbetween2015and2021.However,wethinkthatthisalternativeexplanationisimplausible.
AidDatarecentlycollecteddataonallofficialsectorlendingcommitmentsfromChinatoborrowersin26
HICsandthepreliminarysummarystatisticsshowthatannuallendingcommitmenttoHICsdeclinedfrom
2015to2021(seeFigureA63).Thedataalsodemonstratethatchangesinofficialsectorlending
commitmentsfromChinatoHICsandLICs/MICshavegenerallymovedintandemovertime.Between
2015and2021,China’sofficialsectorlendingcommitmentstoLICs/MICsdeclinedby38.5%.China’s
officialsectorlendingcommitmentsto26HICsdeclinedby36.7%overthesamesix-yearperiod.384
Figure
A63:
Official
sector
PRC
lending
to
LICs
and
MICs
versus
Y1
HICs
FigureA.63
Notes:ThisfigurepresentsannuallendingcommitmentsfromofficialsectorinstitutionsinChinatotwo
cohortsbetween2000and2021:(1)therecipientcountriescapturedinthe3.0versionofAidData’sGCDF
dataset,whichcovers165low-incomeandmiddle-incomecountries,and(2)twenty-sevenhigh-income
countries.385Thedataonhigh-incomecountriesrepresentspreliminarydatacollectedbyAidDataaspartof
385The3.0versionofAidData’sGCDFdatasetcoversatotalof165countriesandterritories.Itspecificallycoversallcountriesthatareorhavebeenclassifiedaslow-ormiddle-incomecountriesduringthe2000-2021timeperiod.384Overthenexttwoyears,AidDataexpectstoachievecomprehensivecoverageofallofficialsectorlendingandgrant-givingcommitmentsfromChinatoHICsbetween2000-2021.GiventheimportantbutunderappreciatedrolethatOFCsplayinChina’soverseaslendingprogram,AidDataalsointendstoallowusersofitsdatatodifferentiatebetweenofficialsectorfinancialflowsfromChinathattraveltoversusthroughcountries.
329
anongoingdatacollectionefforttocaptureallofficialsectorlendingcommitmentsfromChinatoall
high-incomecountriesbetween2000and2021.
SectionA-5:HowdoesAidDatacategorizeChineselendingtodifferenttypesofborrowers?
The“LevelofPublicLiability”fieldinthe3.0versionofAidData’sGCDFdatasetcapturestheextenttowhichthehostgovernmentmayeventuallybeliablefordebtrepayment.Itishierarchicallyandautomaticallydeterminedbasedonthefollowingcriteria:
1.Theloanrecordisclassifiedas"Centralgovernmentdebt"ifitisanofficialsectorloantoacentralgovernmentinstitutionintherecipientcountry,measuredbywhetherthereisatleastonereceivingagency(directorindirect)fromtherecipientcountrythatisclassifiedasagovernmentagency;2.Iftheloanrecorddoesnotmeetthefirst(1)criterion,itisclassifiedas"Centralgovernment-guaranteeddebt"ifitisanofficialsectorloantoastate-ownedentity(e.g.,state-ownedenterpriseandstate-ownedbank)orprivately-ownedentityintherecipientcountrythatbenefitsfromasovereign(centralgovernment)repaymentguarantee;3.Iftheloanrecorddoesnotmeetthefirst(1)criterionorthesecond(2)criterion,itisclassifiedas"Otherpublicsectordebt"if(a)itisanofficialsectorloantoastate-ownedentity(suchasacity/municipalgovernment,astate-ownedbank,orastate-ownedenterprise)intherecipientcountrythatdoesnotbenefitfromasovereign(centralgovernment)repaymentguarantee;(b)itisanofficialsectorloantoaprivateentityorstate-ownedentityintherecipientcountrythatisbackedbyarepaymentguaranteefromastate-ownedentityotherthanthecentralgovernmentintherecipientcountry(suchasacity/municipalgovernment,astate-ownedbank,orastate-ownedenterprise),OR(c)itisanofficialsectorloantoaspecialpurposevehicle(SPV)orjointventure(JV)thatismajority-ownedbyoneormorepublicsectorinstitutionsintherecipientcountryandthatdoesnotbenefitfromasovereign(centralgovernment)repaymentguaranteeorarepaymentguaranteefromastate-ownedentityotherthanthecentralgovernmentintherecipientcountry(suchasacity/municipalgovernment,astate-ownedbank,orastate-ownedenterprise).4.Iftheloanrecorddoesnotmeetthefirst(1)criterion,thesecond(2)criterion,orthethird(3)criterion,itisclassifiedas"Potentialpublicsectordebt"ifitisanofficialsectorloantoaspecialpurposevehicle(SPV)orjointventure(JV)borrowerthatisminority-ownedbyoneormorepublicsectorinstitutionsintherecipientcountryandthatdoesnotbenefitfromasovereign(centralgovernment)repaymentguaranteeorarepaymentguaranteefromastate-ownedentityotherthanthecentralgovernmentintherecipientcountry(suchasacity/municipalgovernment,astate-ownedbank,orastate-ownedenterprise).5.Iftheloanrecorddoesnotmeetthefirst(1)criterion,thesecond(2)criterion,thethird(3)criterion,andthefourth(4)criterion,itisclassifiedas"Privatedebt"ifitisanofficialsectorloantoaprivately-ownedentitythatdoesnotbenefitfromarepaymentguaranteefromapublicsectorinstitutionintherecipientcountry(thisincludeslendingtoaprivateentity,orlendingtoaJointVentureorSpecialPurposeVehiclewithnolevelofhostgovernmentownership(i.e.,the"JV/SPVHostGovernmentOwnership"variableissetto"NoHostGovernmentOwnership";6.Iftheloanrecorddoesnotmeetthefirst(1)criterion,thesecond(2)criterion,thethird(3)criterion,thefourth(4)criterion,orthefifth(5)criterion,thenitisclassifiedas"Unallocable"duetoalackofinformation.
330
SectionA-6:Howdidweidentifyprojectsthatrelyoncontractorssanctionedforfraudulentandcorruptbehavior?
AspartofourbroaderefforttomeasurethenatureandextentofgovernanceriskinChina’soverseasinfrastructureprojectportfolio,wesoughttoidentifytheextenttowhichChineselendersanddonorsreliedoncontractorsformallydebarredforfraudulentandcorruptbehaviortoimplementtheirgrant-andloan-financedinfrastructureprojectsinLICsandMICs.Debarmentisatypeofsanctionimposedonfirmsorindividuals,whichprohibitstheentityinquestionfromparticipatinginfutureorcurrentprojectpreparationorimplementation.MDBsregularlyinvestigatereportsofwrongdoing—suchasfraud,corruption,andcollusion—byentitiesinvolvedintheirprojects.Iftheinvestigationconfirmsthefirmorindividualparticipatedintheallegedwrongdoing,asanctionsboardorcommitteedebarsthatfirmorindividual,renderingthemineligibleto“beawardedorotherwisebenefitfromaBank-financedcontract.”386Eachdebarmentisusuallyenforcedforaspecificperiodoftime,withthelengthofdebarmentproportionaltotheseverityofthewrongdoinginwhichthefirmorindividualparticipated.Inextremecases,afirmorindividualmaybedebarredindefinitely.WhileMDBsmaintaintheirownlistofdebarredinstitutions,agrowinggroupofMDBshaveagreedtoupholdthedebarmentsfromeachother’slists(anapproachthatisknownas“cross-debarment”).387
ToidentifythesubsetofChinesegrant-andloan-financedinfrastructureprojectsthatposeelevatedlevelsofgovernancerisk,webeganbycollectingcurrentandhistoricalinformationonfirmsthathavebeendebarredbyMDBssince2004(whentheearliestdataisavailable).WecollectedthenameofeachfirmdebarredandthelengthofeachdebarmentfromWorldBankSanctionsSystemsAnnualReportsfrom2004-2022.TheseannualreportsprovideinformationonWorldBankhistoricaldebarmentbeginningin2004aswellasanycross-debarmentsfromtheotherMDBs(theAfricanDevelopmentBank,theAsianDevelopmentBank,theEuropeanBankforReconstructionandDevelopment,theInter-AmericanDevelopmentBank,andtheAsianInfrastructureInvestmentBank)beginningin2010.Wethencombinedthislistwithalistofcurrentdebarmentsandcross-debarmentsmaintainedbytheWorldBankonitswebsite(WorldBank2023).388Wesubsequentlymatchedthelistofdebarredfirmswithalistofallthefirmsthatare(orwere)directlyinvolvedinChinesegrant-andloan-financedinfrastructureprojects—specifically,anyfirmlistedasanimplementingagencyorareceivingagencyforainfrastructureprojectinthe3.0versionofAidData’sGCDFdataset.Intotal,wefound21formallydebarredorganizationsthatwereinvolvedin453Chinesegrant-andloan-financedinfrastructureprojects.
Afterisolatingtheprojectsthatinvolveddebarredorganizations,wesoughttoidentifythesubsetofprojectsthatreliedupondebarredorganizationswhiletheywerestillwithintheirdebarmentperiods.Wedidsobyidentifyingallcasesinwhichtherewascalendardayoverlapbetweenthestartandenddatesofanorganization’sdebarmentperiodandthecommitment,implementation,orcompletiondatesoftheproject(s)itsupported.Aftereliminatingallprojectsthatinvolvedthedebarredagencybeforeoraftertheirofficialdebarmentperiods,weidentified324Chinesegrant-andloan-financedinfrastructureprojectsthat
388WecomparedsimilaronlineresourcesfromtheotherMDBsparticipatingincross-debarmentandfoundtheWorldBank’stobethemostcomprehensive.387TheWorldBank,theAfricanDevelopmentBank,theAsianDevelopmentBank,theEuropeanBankforReconstructionandDevelopment,theInter-AmericanDevelopmentBank,andmostrecentlytheAsianInfrastructureInvestmentBankhavesignedthe"AgreementonMutualEnforcementofDebarmentDecisions.”386AccordingtotheWorldBank,debarmentisdefinedas“[t]hesanctionedpartyisdeclaredineligible,eitherindefinitelyorforastatedperiodoftime,(1)tobeawardedorotherwisebenefitfromaBank-financedcontract,financiallyorinanyothermanner;(2)tobeanominatedsub-contractor,consultant,manufacturerorsupplier,orserviceproviderofanotherwiseeligiblefirmbeingawardedaBank-financedcontract;and(3)toreceivetheproceedsofanyloanmadebytheBankorotherwisetoparticipatefurtherinthepreparationorimplementationofanyBank-FinancedProject”(WorldBank2012).
331
involvedthedebarredagencyduringtheactivedebarmentperiod(eitherasanimplementingagencyorareceivingagency).TableA2providesalistofthedebarredagenciesandassociateddetails.
Table
A2:
Contractors
sanctioned
for
fraudulent
and
corrupt
behavior
that
were
involved
in
Chinese
ODA-
and
OOF-financed
infrastructure
projects
during
their
debarment
periods
TableA.2
Annual
loan
commitments
to
LICs
and
MICs
by
financial
institution
and
capital
injections
from
SAFE,
2000-2021
FirmNameGroundsfordebarmentBeginningofDebarmentEndofDebarmentDebarringInstitution
ChinaCAMCEngineeringCo.,LTD.Fraudulentpractices03/09/202210/26/2026AsianDevelopmentBank
ChinaEnergyEngineeringGroupHunanElectricPowerDesignInstituteCo.,Ltd.Fraudulentpractices07/01/201905/10/2021WorldBank
ChinaFirstHighwayEngineeringCo.LtdFraudulentandcollusivepractices07/01/201406/30/2018AfricanDevelopmentBank
ChinaFirstMetallurgicalConstructionCorporation(CFMCC)Fraudulentpractices09/28/201109/27/2014WorldBank
ChinaGeo-EngineeringCorporationFraudulentpractices01/28/20091/27/2014WorldBank
ChinaInternationalWater&ElectricCorp.Fraudulentpractices09/24/201409/24/2017WorldBank
ChinaJiangsuInternationalEconomicAndTechnicalCooperationGroupLtd.Fraudulentpractices02/14/20142/13/2017WorldBank
ChinaJiangxiCorporationForInternationalEconomicAndTechnicalCooperationFraudulentpractices04/15/20144/15/2015WorldBank
ChinaMachineryIndustryConstructionGroupInc.(alsoknownasSINOCONSTRUCTION)Fraudulentpractices07/18/201807/17/2022WorldBank
ChinaNuclearIndustryFifthConstructionCo.Ltd.Fraudulentpractices07/19/201807/18/2020WorldBank
ChinaRailway20BureauGroupCo.Fraudulentpractices06/26/201712/31/2017WorldBank
ChinaRailwayConstructionCorporation(International)LimitedFraudulentpractices06/05/201902/04/2020WorldBank
ChinaRailwayFirstGroupCo.Ltd.Fraudulentpractices09/18/201908/18/2021WorldBank
ChinaStateConstructionEngineeringCorporation(CSCEC)Fraudulentpractices01/14/20091/13/2015WorldBank
ChinaWuyiCo.Ltd.Fraudulentandcollusivepractices01/14/20091/13/2015WorldBank
ShandongTaikaiPowerEngineeringCompanyLimitedFraudulentpractices08/18/201502/17/2017WorldBank
SNC-LavalinGroupInc.Fraudulentandcorruptpractices08/18/201502/17/2017WorldBank
332
FirmNameGroundsfordebarmentBeginningofDebarmentEndofDebarmentDebarringInstitution
TractebelEngineeringS.A.Fraudulentandcorruptpractices12/23/202109/28/2025Inter-AmericanDevelopmentBank
ZhengtaiGroupCo.,Ltd.Fraudulentpractices05/05/201708/05/2018WorldBank
ZhonghaoOverseasConstructionEng.Co.,Ltd.Fraudulentpractices06/02/201206/01/2014WorldBank
NorconsultCorruptpractices07/01/201312/31/2014WorldBank
Notes:Thistablelistscontractorsthat(1)weresanctionedbymultilateraldevelopmentbanksforfraudulent
andcorruptbehavior,and(2)duringthetimeofthecontractor’sdebarment,theywereidentifiedasa
receivingorimplementingagencyonaChinese-financedinfrastructureproject.SeeSection2ofChapter3
formoredetailsonhowthesefirmswereidentified.TheWorldBank,theAfricanDevelopmentBank,the
AsianDevelopmentBank,theEuropeanBankforReconstructionandDevelopment,theInter-American
DevelopmentBank,andtheAsianInfrastructureInvestmentBankhaveaformalagreementtocross-debar
firms.
SectionA-7:HowdidweidentifyESGrisksandriskmitigationmeasuresinAidDataprojectdescriptions?
Inordertosupporttheanalysisinchapter3,weconstructedthreebinaryvariablesthatidentifywhether
thereisanynarrativeevidencefromthe3.0versionofAidData’sGCDFdatasetthatagiveninfrastructure
projectencounteredasignificantenvironmentalproblem,socialproblem,orgovernanceproblembefore,
during,orafterimplementation.Todoso,weidentifiedasetofkeywordsforeachtypeofchallenge
(environmental,social,orgovernance)that,ifpresentinaproject’s“description”fieldwithinthe3.0
dataset,mayindicatetheexistenceofsuchaproblem.Weappliedthesekeywordsearchestoallprojects
backedbyofficialcommitments(includingthosethatweresubsequentlysuspendedorcanceled)andall
implementedandcompletedprojects.389Afteridentifyingthesubsetofprojectswith“description”fields
thatcontainoneormoreofthepre-specifiedkeywords,wemanuallyreviewedeachofthecorresponding
projectdescriptionstoeliminate“falsepositives”—e.g.,adescriptionthatreferencesthekeyword“lawsuit”
todescribeenvironmentallitigationaffectinganearlierphaseofaprojectbackedbyanon-Chinese
financierbutnotthesubsequent,Chinesegovernment-financedphaseoftheproject.Morespecifically,we
confirmedordisconfirmedthattheprojectsinquestionencounteredthetypesofproblemsthatwe
intendedtoidentifywiththeprespecifiedsetofkeywords.However,atthemanualreviewstage,wedidnot
seektodifferentiatebetweenproblemsthatwereeffectivelyorineffectivelymanaged.Nordidweseekto
independentlyevaluatetheveracityofanyreportedinstancesofharm,wrongdoing,orpoorperformance.
Ourgoalwastoidentifywhichtypesofproblemsaroseduringimplementation(regardlessofwhethera
mitigationresponsewasundertaken).However,wewerenotabletoaccountfor“falsenegatives”(i.e.,
casesinwhichaspecifictypeofproblemaroseduringprojectimplementation,buttheproblemwasnot
389Wealsorestrictedoursearchestoinfrastructureprojectssupportedbygrantandloancommitmentsworth$20million(inconstant2021USD)ormore.Projectssupportedbylargerfinancialcommitmentsgenerallyhavemoredetailedprojectdescriptions,whichprovideastrongerbasisfortheidentificationofenvironmental,social,andgovernancerisksandriskmitigationefforts.Theyalsopresentalowerriskofgenerating“falsenegatives.”
333
capturedinthe“description”fieldwithinAidData’s3.0dataset),sothesummarystatisticsthatwereport
shouldbetreatedaslower-boundestimates.
Weidentifiedthesubsetofinfrastructureprojectsthatencounteredsignificantenvironmental
problems—before,during,orafterimplementation—byfirstapplyingthefollowingkeywordsearchterms
totheinfrastructureprojectdescriptionsinthe3.0dataset:disaster,destroy*,destruction,ruin*,despoil,
degrad*,pollut*,dispos*,displace*,danger*,hazard*,exploit*,involuntary,unsafe,crack*,substandard,
low-quality,deficien*,defect*,ecology*,ecosystem,habitat,biodiverse*,flora,fauna,species,non-native,
invasive,pesticide*,chemical*,cancer*,discharge*,emission*,emit*,subcritical,odor,smell.Wethen
reviewedeachoftheidentifiedprojectdescriptions“byhand”toremovefalsepositives.
Wefollowedananalogoussetofprocedurestoidentifythoseinfrastructureprojectsthatencountered
significantsocialchallengesandgovernanceproblems,respectively.Toconstructabinary,project-level
measureofexposuretosocialproblems,weappliedthefollowingkeywordsearchterms:strike*,protest*,
riot*,violat*,noncomplian*,non-complian*,involuntary,forced,evict*injur*,noise,vibrat*,nuisance,dead,
death,died,kill*,harm*,unsafe,crack*,substandard,low-quality,defect*,deficien*,danger*,indigenous,
burial,sacred,spiritual,ritual*,religio*,heritage,overcrowd*,displace*,grievanc*,underserve*,
disadvantage*,minorit*,aboriginal,tribe,vulnerab*,marginalize*,ethnic,archaeol*,custom*,manipul*,
interfer*,coerc*,discriminat*,intimidat*,workplace,layoff,fired,underpaid,unpaid,wages,“working
conditions”,abuse*,resettle*,aggrieved,sexual.Toconstructabinary,project-levelmeasureofexposure
togovernanceproblems,weappliedthefollowingkeywordsearchterms:misuse*,abuse*,bid-rig*,
misappropriat*,mismanage*,steal,theft,stole*,corrupt*,bribe*,graft,fraud*,kickback,siphon*,
embezzle*,illicit,illegal,inflat*,overprice*,over-price*,wrongdoing,collusion,collusive,collude*,loot*,
plunder*,abuse*,obstructive,defraud*,fictitious,launder*.Wethenmanuallyreviewedeachofthe
identifiedprojectdescriptionstoremovefalsepositives.
Inadditiontokeywordsthatarespecifictoenvironmental,social,orgovernancechallenges,weidentifieda
setofkeywordsthatcouldindicateachallengeinanyofthethreecategories.Weusedthesecross-cutting
keywordstoidentifyadditionalprojectsrequiringmanualreviewforapotentialenvironmental,social,or
governancechallenge.Forthis“cross-cutting”keywordsearch,weappliedthefollowingsearchterms:
allege*,allegation,criticiz*,alarm,criticism,complain*,controvers*,fiasco,turmoil,breach*,probe*,
irresponsib*,audit*,inquiry,scrutin*,uncover*,scandal,dispute*,fined,sued,lawsuit,arbitrat*,
adjudicate*,court,litigat*,rescope*,jeopard*,rescind*,revoke*,failure,delay*,threat*,renege*,
renegotiat*,“conditionsprecedent”,*indemnif*,adverse,“liable”.
Wealsoconstructedthreebinaryvariablesthatidentifywhetherthereisanynarrativeevidencefromthe3.0
versionofAidData’sGCDFdatasetthateffortswereundertakenbyChinesefinanciersorimplementing
agenciestomitigateenvironmental,social,orgovernancerisksbefore,during,orafterproject
implementation.390ThesevariablesseektomeasurewhetherESGriskmitigationmeasureswereundertaken
byChinesefinanciersorimplementingagencies,irrespectiveofwhetherornotsuchmeasureswere
successful.391Consistentwiththeapproachthatwasusedtoidentifyinfrastructureprojectsthat
391WeonlysoughttoidentifyESGriskmitigationmeasuresthatinvolvedchoiceorconsentfromtheChineseside.Therefore,ifaChinesefinancierorimplementingagencywascompelledbyanentityoutsideofChinatotakeanESGriskmitigationmeasure(i.e.,aChinesecompanydoesnotimplementan390Heretoowerestrictedoursearchestoinfrastructureprojectssupportedbygrantandloancommitmentsworth$20million(inconstant2021USD)ormore.
334
encounteredsignificantESGchallenges,weconstructthreebinary,project-levelmeasuresofenvironmental
riskmitigation,socialriskmitigation,andgovernanceriskmitigationbyapplyingthefollowingkeyword
searchterms:
●Environmentalriskmitigation:remedia*,compensat*;resettle*,reclaim*,reclamation,restor*,
restitution,certify*,permit*,feedback,disclosure,amend*,minimize,consent*,FPIC,safeguard,
mandat*,mitigat*,heritage,consult*,dialogue,“publichearing”,“EnvironmentalAssessment”,
“EA”,“EnvironmentalImpactAssessment”,“EIA”,“EnvironmentalandSocialImpact
Assessment”,“ESIA”,“Resettlementactionplan”,“RAP”,“EnvironmentalManagementPlan”,
“EMP”,“ProjectAffectedPersons”,“PAPs”,“Project-Affected”,“EnvironmentalCategorization”,
“FullResettlementPlan”,“FRP”,“compliancecertificate”,“ECC”,“EnvironmentalandSocial
Policy”,“E&S”,“StakeholderEngagementPlan”,“SEP”,“DueDiligence”,“ESDD”,
“Environmental,Health,SafetyandSocialManagementSystem”,“EHSSMS”,“Environmentaland
SocialActionPlan”,“ESAP”,“ESCP,“protectedarea”,"EquatorPrinciples",desulphurization,
desulphurization,“FGD”,super-critical,“InitialEnvironmentalExamination”,“IEE”,
“Environmentalindemnity”,“Environmentalclaim”
●Socialriskmitigation:compensat*;resettle*,reclaim*,reclamation,restor*,restitution,consent*,
FPIC,safeguard,mandat*,mitigat*,certify*,permit*,feedback,disclosure,amend*,minimize,
heritage,consult*,dialogue,“publichearing”,“healthandsafety”,“Resettlementactionplan”,
“RAP”,“FullResettlementPlan”,“FRP”,“EnvironmentalandSocialImpactAssessment”,“ESIA”,
“compliancecertificate”,“EnvironmentalandSocialPolicy”,“E&S”,“StakeholderEngagement
Plan”,“SEP”,“DueDiligence”,“ESDD”,“Environmental,Health,SafetyandSocialManagement
System”,“EHSSMS”,“EnvironmentalandSocialActionPlan”,“ESAP”,“ESCP”,occupational
health,“OHS”,"EquatorPrinciples",,“GenderActionPlan”,“collectivebargaining”,“Safeguards
MonitoringReport”
●Governmentriskmitigation:debar*,cross-debarment,sanction*,blacklist*,prosecut*,arrest,
convict*,criminal,extradit*,imprison,anticorruption,anti-corruption,value-for-money,freeze,
froze*,halt*,suspend*,suspension,cancel*,withheld,withhold,abandon*,postpon*,abeyance,
mothball*,forensic,prohibit*,“OpenCompetitiveBidding”,OCB,“InternationalCompetitive
Bidding”,ICB,“CompetitiveBidding”,IFRS,“InternationalFinancialReportingStandards”,
“GAAP”,“knowyourcustomer”,“generallyacceptedaccountingprinciples”,“anti-money
laundering”
Wethenmanuallyreviewedeachoftheidentifiedprojectdescriptionstoremovefalsepositives.
SectionA-8:HowdidAidDataidentifyESGriskmitigationmeasuresininfrastructurefinancingagreements?
InordertobenchmarktheESGriskmanagementprovisionsintheinfrastructurefinancingagreementsof
Chinesestate-ownedlendersanddonors,wefirstdevelopedastandardizedsetof26evaluationcriteria(or
EnvironmentalImpactAssessmentuntilitiscompelledtodosobyacourtorderissuedbyajudgeintherecipientcountry),itdoesnotcountasanenvironmentalriskmitigationmeasurebecauseoftheabsenceofchoice/consentbytheChinesecompany.Similarly,ifaChinesecompanyemployeeisconvictedoncorruptionchargesintherecipientcountry,thisdoesnotcountasagovernanceriskmitigationmeasurebecauseoftheabsenceofchoice/consentontheChineseside.
335
diagnosticquestions).Thecriteriaareorganizedintothreegroups:thosethatidentifythepresenceor
absenceof(1)rulesorstandardstoestablishbehavioralexpectationsrelatedtoESGriskmanagementand
mitigation,(2)oversightmechanismsformonitoringcompliancewiththosebehavioralexpectations;and/or
(3)enforcementmechanismsforsanctioningnoncompliancewiththosebehavioralexpectations(e.g.,
indemnification,withholdingdisbursements).8ofthecriteriaarespecifictoenvironmentalsafeguards,
while7arespecifictosocialsafeguards,and12arespecifictogovernancesafeguards.
Thecriteriathatweusedtoevaluatethe23contractsinthecodingsampleareprovidedbelowinTableA3
(andorganizedaccordingtosafeguardtypeandwhethertheevaluationcriteriarelatestostandards,
monitoring,orenforcement).
Table
A3:
Criteria
to
evaluate
environmental,
social,
and
governance
safeguards
in
Chinese
grant
and
loan
contracts
TableA.3
Criteria
to
evaluate
envir onmental,
social,
and
governance
safeguar ds
in
Chinese
grant
and
loan
contr acts
QuestionQuestionTypeSafeguardType
Areanyenvironmentalclausesorconditionsincludedinthe
agreement?StandardsEnvironmental
IsanEnvironmentalImpactAssessment(EIA)identifiedasa
requirementintheagreement?StandardsEnvironmental
IsanEnvironmentalManagementPlan(EMP)–orafunctional
equivalentofanEMP–identifiedasarequirementinthe
agreement?StandardsEnvironmental
Istheborrowerrequiredtoreporttothelenderonits
implementationoftheEIArecommendationsorEMP?MonitoringEnvironmental
Doestheagreementincludeanyenvironmentalconditions
precedent(forentryintoforceorinitialorongoing
disbursement)?EnforcementEnvironmental
Doestheagreementspecifythatnoncompliancewith
environmentalrules,standardsorlawsisasufficientbasisforthe
lendertodemandearlyrepayment(“prepayment”)oftheloan
orcanceltheloan?EnforcementEnvironmental
Doestheagreementrequirecompliancewithinternational
environmentalstandards,suchastheEquatorPrinciplesorthe
OECDRevisedCouncilRecommendationonCommon
ApproachesontheEnvironmentandOfficiallySupportedExport
Credits?StandardsEnvironmental
Istheborrowerrequiredtoindemnifythelenders,guarantors,or
insurers("FinanceParties")againstanylossorliabilityincurred
bytheFinancePartiesasaresultofanyactualorallegedbreach
ofanenvironmentallaworstandard?EnforcementEnvironmental
Areanysocialclausesorconditionsincluded?StandardsSocial
IsanEnvironmentalandSocialImpactAssessment(ESIA)
identifiedasarequirement?StandardsSocial
336
IsanEnvironmentalandSocialActionPlan(ESAP),a
ResettlementActionPlan(RAP),orthefunctionalequivalentof
anESAPorRAPidentifiedasarequirement?StandardsSocial
Istheborrowerrequiredtoreporttothelenderonits
implementationoftheESIArecommendationsorRAP?MonitoringSocial
Doestheagreementincludeanysocialconditionsprecedent(for
entryintoforceorinitialorongoingdisbursement)?EnforcementSocial
Doestheagreementindicatethepenaltyofviolatingsocial
standards/lawsleadstodemandingearlyprepaymenttoclose
theloan?EnforcementSocial
Istheborrowerrequiredtoindemnifythelenders,guarantors,or
insurers("FinanceParties")againstanylossorliabilityincurred
bytheFinancePartiesasaresultofanyactualorallegedbreach
ofasociallaworstandard?EnforcementSocial
Areanygovernance(e.g.,anti-corruption,competitivebidding,
auditedfinancialstatements,and/oranti-moneylaundering)
clausesorconditionspresentintheagreement?StandardsGovernance
Areindependentlyauditedfinancialstatementsrequiredofthe
borrower?StandardsGovernance
Musttheborrower’sfinancialstatementscomplywith
InternationalFinancialReportingStandards(IFRS)standards?StandardsGovernance
Doestheagreementincludeanycompetitivebidding
requirements?StandardsGovernance
Arebiddingdocumentsorbidevaluationreportssubjecttoprior
approvalbythelender?MonitoringGovernance
Doestheagreementincludeanyanti-corruptionoranti-money
launderingrequirements?StandardsGovernance
Istheborrowerrequiredtoindemnifythelenders,guarantors,or
insurers("FinanceParties")againstanylossorliabilityincurred
bytheFinancePartiesasaresultofanyactualorallegedbreach
ofananti-corruptionoranti-moneylaunderinglaworstandard?EnforcementGovernance
Doesthelenderexplicitlyreservetherighttopreventor
investigateanti-corruptionoranti-moneylaunderingcrimes?MonitoringGovernance
Doestheborroweraffirmthatitwillseektoensurethatthe
proceedsfromtheloanorgrantarenotusedtofinanceor
benefitanyinternationallysanctionedentity?MonitoringGovernance
Doestheagreementspecifythatnoncompliancewith
governancerules,standards,orlawsisasufficientbasisforthe
lendertodemandearlyrepayment(“prepayment”)oftheloan
orcanceltheloan?EnforcementGovernance
Doestheagreementincludeanycompetitivebidding,
anti-corruption,oranti-moneylaunderingconditionsprecedent
(forentryintoforce,initialdisbursement,orongoing
disbursements)?EnforcementGovernance
337
SectionA-9:AssigningESGSafeguardStringencyRatingstoInfrastructureFinancingAgreementsandthe8InfrastructureFinancingInstrumentCategories
Toassessthepresenceorabsenceofeachsafeguardtype(standards,monitoring,andenforcement)within
eachinfrastructurefinancingagreement(“contract”),weexaminethebinary(yes/no)responsesto26
diagnosticquestionsthatarecategorizedbysafeguardtype(seeTableA8).392Foreachcontract,ifwe
identify“yes”responsestoanyquestionsrelatedtostandardsinagivenESGdomain(environmental,
social,orgovernance),thenwedeterminerulesandstandardstobepresentinthatparticularESG
domain-contractdyad.Conversely,ifweidentify“no”responsestoallquestionsrelatedtostandardsina
givenESGdomain(environmental,social,orgovernance),thenwedeterminerulesandstandardstobe
absentinthatparticularESGdomain-contractdyad.Thesamecriteriaareappliedtothequestionsrelated
tomonitoringandthequestionsrelatedtoenforcementinagivenESGdomain.TableA4recordsallofthe
yes/nodeterminationsatthelevelofESGdomain-contractdyads.
Then,foreachESGdomain-contractdyad,weassignhigh,medium,orlowratingsbaseduponthe
followingcriteria:
●Low:Contractslackingestablishedstandards,monitoringmechanisms,andenforcement
mechanisms.
●Medium:Contractsthatstipulatestandards,butlackmonitoringmechanismsorenforcement
mechanisms.
●High:Contractsthatstipulatestandardsandincorporateeithermonitoringmechanismsor
enforcementmechanisms.
TableA4recordsalloftheESGdomain-contractdyadratings.Next,toaccountforchangesinESG
safeguardstringencywithineachfinancialinstrumentcategoryovertime,wesegmentourESG
domain-contractdyadratingsintotwodistinctperiods:2000-2017(pre-andearlyBRIperiod)and
2018-2022(lateBRIperiod).Wegeneratesummary(high/medium/low)ratingsfortheenvironmental
safeguards,socialsafeguards,andgovernancesafeguardsthatappliedtoeachfinancialinstrument
categoryovereachtimeperiod(2000-2017and2018-2021).Wedosobyfollowinga"mostfrequent
designation"decisionrule.Ifthemajorityofcontractswithinafinancialinstrumentcategorysharethesame
ratingforagivenESGdomain,themostfrequentratingdesignationisappliedtotheentirefinancial
instrumentcategory.However,incaseswheretwocontractswithinthesametimeperiodandfinancial
instrumentcategoryareassigneddifferentratingsforagivenESGdomain,thehigherratingisappliedto
thefinancialinstrumentcategoryasawhole.
Forvarioustypesofanalysis,weusethedistinctionbetween“strong”and“weak”dejureESGsafeguards.
Projectswith“strong”dejureESGsafeguardsaredefinedasthosewithatleasttwooutofthree
392The“standards”measuresseektoidentifywhetherthecontractidentifiesrulesorstandardsthatcreatebehavioralexpectationsrelatedtoESGriskmanagementandmitigation.The“monitoring”measuresseektoidentifywhetherthecontractidentifiesoversightmechanismsformonitoringcompliancewiththosebehavioralexpectations.The“enforcement”measuresseektoidentifywhetherthecontractidentifiesenforcementmechanismsforsanctioningnoncompliancewiththosebehavioralexpectations.
338
(environmental,socialandgovernance)safeguardstringencyratingsthatare“high.”Projectsthatdonot
meetthisstandardareclassifiedashaving“weak”dejureESGsafeguards.
Table
A4:
Safeguard
stringency
ratings
for
infrastructure
financing
agreements
at
the
ESG
domain-contract
dyad
level
TableA.4
Safeguar d
stringency
ratings
for
infrastructur e
financing
agreements
at
the
ESG
domain-contr act
dyad
level
Financial
Instrument
CategoryContractESGDomainStandardsMonitoringEnforcementLevel
BilateralChina
EximbankloanMombasa-NairobiStandardGaugeRailwayProject
(Kenya)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceNoNoNoLow
NewPowerPlantProject(AntiguaandBarbuda)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceNoNoNoLow
NewCentennialWaterSource-KaliwaDamProject
(Philippines)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceNoNoNoLow
BilateralCDB
loanTermFacilityforInfrastructureandSocial
DevelopmentProjects(CostaRica)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceYesNoYesHigh
ShipConstructionProject(MarshallIslands)EnvironmentalYesNoYesHigh
SocialNoNoNoLow
GovernanceYesNoNoMedium
TermFacilityAgreementforInfrastructureProjects
(Ecuador)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceNoNoNoLow
Bilateral
MOFCOMloan
orgrantProjectfortheRealizationoftheBridgeRenovation
ontheRiverTara(Montenegro)EnvironmentalYesNoNoMedium
SocialYesNoNoMedium
GovernanceYesNoNoMedium
NationalEmergencyOperationsCenter(COEN)
Project(Peru)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceNoNoNoLow
PortInfrastructureConstructionProject(Mongolia)EnvironmentalNoNoNoLow
339
Financial
Instrument
CategoryContractESGDomainStandardsMonitoringEnforcementLevel
SocialNoNoNoLow
GovernanceYesYesNoHigh
Bilateral
Chinese
state-owned
commercial
bankloanMasindi-Biiso,Kibaale-Kiziranfumbiand
Hohwa-Nyairongo-Kyarusesa-ButooleRoad
UpgradingProject(Uganda)EnvironmentalYesNoYesHigh
SocialNoNoNoLow
GovernanceYesYesYesHigh
ElectrificationProject(Ghana)EnvironmentalYesNoNoMedium
SocialNoNoNoLow
GovernanceYesNoYesHigh
PatucaIIIHydroelectricPlantProject(Piedras
Amarillas)PhaseII(Honduras)EnvironmentalYesNoYesHigh
SocialYesNoYesHigh
GovernanceYesNoNoMedium
Syndicated
loanwith
Chineseand
multilateral
bank
participantsItuangoHydroelectricProject(Colombia)EnvironmentalYesYesYesHigh
SocialYesYesYesHigh
GovernanceYesYesYesHigh
CastellanaWindPowerProject(Argentina)EnvironmentalYesYesYesHigh
SocialYesYesYesHigh
GovernanceYesYesYesHigh
Syndicated
loanwith
Chinese
state-owned
commercial
banksand/or
policybanksWorksonConstructionoftheMunicipal(Sewage)
InfrastructureProject(Serbia)EnvironmentalYesNoYesHigh
SocialNoNoNoLow
GovernanceYesYesYesHigh
PresidentDrNéstorCarlosKirchnerandGobernador
JorgeCepernicHydropowerProject(Argentina)EnvironmentalYesNoYesHigh
SocialNoNoNoLow
GovernanceYesNoNoMedium
QueenElizabethllQuayatFreetownPortProject
(SierraLeone)EnvironmentalYesNoYesHigh
SocialYesNoYesHigh
GovernanceYesYesYesHigh
PBOC/MOF
grantorloan
channeled
through
multilateral
institutionsInfrastructureTransportProgram(PIT)(CostaRica)EnvironmentalYesYesYesHigh
SocialYesYesYesHigh
GovernanceYesYesYesHigh
SustainableWaterSupplyandSanitationProgram
(Rwanda)EnvironmentalYesYesNoHigh
SocialYesYesYesHigh
340
Financial
Instrument
CategoryContractESGDomainStandardsMonitoringEnforcementLevel
GovernanceYesYesYesHigh
WaterSupplySchemeforTeteSettlementProject
(PapuaNewGuinea)EnvironmentalYesYesYesHigh
SocialYesNoYesHigh
GovernanceYesYesYesHigh
Supplier's
creditfrom
ChineseSOERegionalHospitalsProject–Lot3(Guyana)EnvironmentalYesNoYesHigh
SocialNoNoNoLow
GovernanceYesNoYesHigh
Phase1Lot5Adenta-DodowaDualCarriageway
Project(Ghana)EnvironmentalYesNoYesHigh
SocialNoNoNoLow
GovernanceYesNoYesHigh
EnergyTransmissionNetworkConstructionProject
AssociatedwiththeImboulouPowerPlant(Congo)EnvironmentalNoNoNoLow
SocialNoNoNoLow
GovernanceNoNoNoLow
Table
A5:
Sample
of
infrastructure
financing
agreements
used
to
code
the
de
jure
stringency
of
ESG
safeguards
TableA.5
Sample
of
infrastructur e
financing
agreements
used
to
code
the
de
jure
stringency
of
ESG
safeguar ds
Financial
Instrument
CategoryCommitm
entYearProjectTitleLender/DonorBorrower/Recipient
BilateralChina
Eximbankloan2014Mombasa-NairobiStandardGauge
RailwayProject[hyperlink]Export-ImportBankofChinaGovernmentof
Kenya
2008NewPowerPlantProject[hyperlink]Export-ImportBankofChinaMinistryofFinance
andEconomyof
Antiguaand
Barbuda
2018TheNewCentennialWater
Source-KaliwaDamProject[hyperlink]Export-ImportBankofChinaMetropolitan
Waterworksand
SewerageSystemof
thePhilippines
341
Financial
Instrument
CategoryCommitm
entYearProjectTitleLender/DonorBorrower/Recipient
BilateralCDBloan2008TermFacilityforInfrastructureandSocial
DevelopmentProjects[hyperlink]ChinaDevelopmentBankNationalBankof
CostaRica
2013ShipConstructionProject[hyperlink]ChinaDevelopmentBankNereusNavigation
Ltd.andIrises
ShippingLtd
2016TermFacilityAgreementfor
InfrastructureProjects[hyperlink]ChinaDevelopmentBankGovernmentof
Ecuador
BilateralMOFCOM
loanorgrant2020ProjectfortheRealizationoftheBridge
RenovationontheRiverTara[hyperlink]ChinaMinistryofCommerceGovernmentof
Montenegro
2014NationalEmergencyOperationsCenter
(COEN)Project[hyperlink]ChinaMinistryofCommerceGovernmentofPeru
2019PortInfrastructureConstructionProject
[hyperlink]ChinaMinistryofCommerceGovernmentof
Mongolia
BilateralChinese
state-owned
commercialbank
loan2021Masindi-Biiso,Kibaale-Kiziranfumbiand
Hohwa-Nyairongo-Kyarusesa-Butoole
RoadUpgradingProject[hyperlink]ChinaConstructionBankMinistryofFinance,
Planningand
Economic
Developmentofthe
RepublicofUganda
2016ElectrificationProject[hyperlink]IndustrialandCommercialBank
ofChinaMinistryofFinance
ofGhana
2013PatucaIIIHydroelectricPlantProject
(PiedrasAmarillas)PhaseII[hyperlink]
[hyperlink]IndustrialandCommercialBank
ofChinaEmpresaNacional
deEnergíaEléctrica
ofHonduras
Syndicatedloan
withChineseand
multilateralbank
participants2017ItuangoHydroelectricProject[hyperlink]People'sBankofChinavia
Inter-AmericanDevelopment
Bank,Inter-American
InvestmentCorporation,
Inter-AmericanDevelopment
BankEmpresasPúblicas
deMedellínE.S.P.
2017CastellanaWindPowerProject
[hyperlink]StateAdministrationon
ForeignExchange(SAFE)via
InternationalFinance
Corporation,International
FinanceCorporation,
Inter-AmericanInvestment
Corporation,Inter-American
DevelopmentBank,Canadian
ClimateFundforthePrivateCPLaCastellana
S.A.U.
342
Financial
Instrument
CategoryCommitm
entYearProjectTitleLender/DonorBorrower/Recipient
SectorintheAmericas
Syndicatedloan
withChinese
state-owned
commercialbanks
and/orpolicy
banks2022WorksonConstructionoftheMunicipal
(Sewage)InfrastructureProject
[hyperlink]BankofChinaHungarian
Branch,BankofChinaSrbija
A.D.BeogradMinistryofFinance
ofSerbia
2014PresidentDrNéstorCarlosKirchnerand
GobernadorJorgeCepernic
HydropowerProject
[hyperlink]ChinaDevelopmentBank,
IndustrialandCommercialBank
ofChina,andBankofChinaMinistryofEconomy
andPublicFinance
ofArgentina
2017QueenElizabethllQuayatFreetown
PortProject
[hyperlink]IndustrialandCommercialBank
ofChina,Export-ImportBank
ofChinaNationalPort
Development(SL)
Limited
PBOC/MOFgrant
orloanchanneled
through
multilateral
institutions2014InfrastructureTransportProgram(PIT)
[hyperlink]People'sBankofChinavia
Inter-AmericanDevelopment
BankPeople'sBankof
Chinavia
Inter-American
DevelopmentBank
2017SustainableWaterSupplyandSanitation
Program[hyperlink][hyperlink]People'sBankofChinavia
AfricanDevelopmentBankGovernmentof
Rwanda
2019WaterSupplySchemeforTete
SettlementProject[hyperlink]PRCPovertyReductionFund
viaAsianDevelopmentBankGovernmentof
PapuaNewGuinea
Supplier'scredit
fromChineseSOE2022RegionalHospitalsProject–Lot3
[hyperlink]ChinaCAMCEngineeringMinistryofFinance
ofGuyana
2018Phase1Lot5Adenta-DodowaDual
CarriagewayProject[hyperlink]SinohydroCorporationLimitedGovernmentof
Ghana
2005EnergyTransmissionNetwork
ConstructionProjectAssociatedwiththe
ImboulouPowerPlant[hyperlink]ChinaNationalMachinery&
Equipmentlmport&Export
Corporation(CMEC)Governmentof
Republicofthe
Congo
343
Table
A6:
Summary
of
environmental,
social,
and
governance
safeguard
clauses
in
the
contract
sample
for
the
pre-BRI
and
early
BRI
period
(2000-2017)
TableA.6
Summar y
of
envir onmental,
social,
and
governance
safeguar d
clauses
in
the
contr act
sample
for
the
pre-BRI
and
early
BRI
period
(2000-2017)
EnvironmentalSocialGovernance
StandardsMonitoringEnforcementStandardsMonitoringEnforcementStandardsMonitoringEnforcement
BilateralChinaEximbankloanNoneNoneNoneNoneNoneNoneNoneNoneNone
BilateralCDBloan1of3contractsincludescomplianceandnoenvironmentalclaimsrequirementNone1of3contractsincludesEnvironmentalIndemnityclauseNoneNoneNoneMostcontractsincludeanti-moneylaunderingandIFRSrequirementsNone1of3contractstreatbreachingtheserequirementsaseventofdefault,thusrequiresindemnification
BilateralMOFCOMloanorgrantNoneNoneNoneNoneNoneNoneNoneNoneNone
BilateralChinesestate-ownedcommercialbankloanCompliancewithEnvironmentalLawsandnoEnvironmentalClaimsrequirementsNone1of2contractsincludesissuanceoftheCertificateofEnvironmentalComplianceasaconditionsprecedent1of2contractsincludesCompliancewithSocialLawsrequirementNone1of2contractsincludesCompliancewithSocialLawsasconditionsprecedentAnti-corruptionandanti-moneylaunderingrequirementsNone1of2contractsrequirescancellationandmandatoryprepaymentintheventofbreach
SyndicatedloanwithChineseandmultilateralbankparticipantsRequirementstoComplywithenvironmentallawandinternationalenvironmentalstandards;conductEIAandEMPRequiresImplementationreportofEIA/EMPRequiresenvironmentalconditionsprecedent;requirescancelation,mandatorypre-payment,orindemnificationIneventsofbreachRequirementstoComplywithSociallaw;conductESIAandESAP;RequiresImplementationreportofESIA/ESAPRequiresSocialconditionsprecedent;requirescancelation,mandatorypre-payment,orindemnificationineventsofbreachAnti-corruptionandanti-moneylaunderingrequirements;requiresindependentauditorcompliancewithIFRSReserverightstoinvestigatepotentialbreach;requirestheproceedingswillnotbenefitsanctionedentitiesCancellation,mandatoryprepayment,orindemnificationineventofbreach
Syndicatedloan2of2:EMP,None1of21of2noNone1of2nosocial1of2requires1of21of2no
344
EnvironmentalSocialGovernance
StandardsMonitoringEnforcementStandardsMonitoringEnforcementStandardsMonitoringEnforcement
withChinesestate-ownedcommercialbanksand/orpolicybanksEnvironmentallawcompliancerequired
1of2:internationalenvironmentalstandards(EquatorPrinciples,OECD)IndemnificationandaccelerationviaEventofDefaultifenvironmentallawsbroken
1of2indemnificationifenvironmentallawsbrokensocialstandards
1of2sociallawcompliancerequiredenforcement
1of2indemnificationforbreachofsociallawsorstandardsprovidinginformationforknowyourcustomerchecks
1of2requiresanticorruptionlawcompliance,independentlyauditedfinancialstatementscompliantwithIFRSrepresentationmadethatneitherborrowernorassociatesareinternationallysanctionedgovernanceenforcement
1of2accelerationviaEventofDefaultformisrepresentationaboutcompliancewithanticorruptionlawsorsanctionsbyguarantor
PBOC/MOFgrantorloanchanneledthroughmultilateralinstitutionsESMPandEIA(orequivalents)requiredpresentationofEMAP(orequivalent)tolender,ongoingreportingonsocialmanagement1of2conditionsprecedent,mandatoryprepayment/accelerationRAPandESMP(orequivalents)required,1of2requiresESIApresentationofRAPandESAP(orequivalents)tolender,ongoingreportingonsocialmanagement1of2acceleration/mandatoryprepaymentforviolations,2of2conditionsprecedentCompetitivebiddingrequirements,anti-corruptionlawsBiddingdocumentssubjecttolenderreview,1of2investigationbylenderAcceleration,mandatoryprepaymentforanti-corruption/anti-moneylaunderingviolations,1of2indemnificationforviolations
Supplier'screditfromChineseSOEnonenonenonenonenonenonenonenonenone
345
Table
A7:
Summary
of
environmental,
social,
and
governance
safeguard
clauses
in
the
contract
sample
for
the
late
BRI
period
(2018-2021)
TableA.7
Summar y
of
envir onmental,
social,
and
governance
safeguar d
clauses
in
the
contr act
sample
for
the
late
BRI
period
(2018-2021)
EnvironmentalSocialGovernance
StandardsMonitoringEnforcementStandardsMonitoringEnforcementStandardsMonitoringEnforcement
BilateralChinaEximbankloanNoneNoneNoneNoneNoneNoneNoneNoneNone
BilateralCDBloan1of3contractsincludescomplianceandnoenvironmentalclaimsrequirementNone1of3contractsincludesEnvironmentalIndemnityclauseNoneNoneNoneMostcontractsincludeanti-moneylaunderingandIFRSrequirementsNone1of3contractstreatbreachingtheserequirementsaseventofdefault,thusrequiresindemnification
BilateralMOFCOMloanorgrant1of2contractsrequiresEnvironmentalImpactAssessment(EIA)NoneNone1of2contractsrequireequaltreatmentprincipleNoneNone1of2contractsincludesCompetitiveBiddingrequirements;Anti-corruptionrequirementBiddingdocumentssubjecttolenderapprovalNone
BilateralChinesestate-ownedcommercialbankloanEnvironmentalLawsComplianceandnoEnvironmentalClaimsRequirementNoneBreachofEnvironmentalLawsistreatedaseventofdefault,thusrequiresindemnificationandaccelerationNoneNoneNoneAnti-corruptionandanti-moneylaunderingrequirementsReserverightstoinvestigatepotentialbreach;requirestheproceedingswillnotbenefitsanctionedentitiesBreachtreatedasaneventofdefault,conditionprecedentthattheborrowersshallprovideallrequestedevidencetoprovecompliance
SyndicatedloanwithChineseandmultilateralbankparticipantsRequirementstoComplywithenvironmentallawandinternationalenvironmentalstandards;conductEIAandEMP;RequiresImplementationreportofEIA/EMPRequiresenvironmentalconditionsprecedent;requirescancelation,mandatoryprepayment,orindemnificationineventsofbreachRequirementstoComplywithSocialLaw;conductESIAandESAP;RequiresImplementationreportofESIA/ESAPRequiresSocialconditionsprecedent;requirescancelation,mandatoryprepayment,orindemnificationinAnti-corruptionandanti-moneylaunderingrequirements;requiresindependentauditorcompliancewithIFRSReserverightstoinvestigatepotentialbreach;requirestheproceedingswillnotbenefitsanctionedentitiesRequirescancellation,mandatoryprepayment,orindemnificationineventofbreach
346
EnvironmentalSocialGovernance
StandardsMonitoringEnforcementStandardsMonitoringEnforcementStandardsMonitoringEnforcement
eventsofbreach
SyndicatedloanwithChinesestate-ownedcommercialbanksand/orpolicybanksCompliancewithenvironmentallawsandpermitsNoneNoneNoneNoneNoneCompliancewithAnti-CorruptionandAnti-MoneyLaunderingLawsProceedsnottobenefitsanctionedentitiesMandatoryprepaymentforbusinesswithsanctionedentities,non-compliancewithAnti-CorruptionorAnti-MoneyLaunderingLaws
PBOC/MOFgrantorloanchanneledthroughmultilateralinstitutionsEMPrequiredReportingonEMPprogressrequiredWithdrawalofgrantfundsfornoncomplianceGenderActionPlan(GAP),preventionofinvoluntaryresettlement,laborprotectionsinbiddingdocumentsandcontractsrequiredNoneWithdrawalofgrantfundsfornoncomplianceCompliancewithanti-corruption,anti-moneylaundering,anti-terrorismfinancelaws;competitivebiddingrequiredLenderreservesrighttoinvestigatecorruptpracticesWithdrawgrantfundsfornoncompliance
Supplier'screditfromChineseSOECompliancewithenvironmentallawsandpermitsNoneIndemnificationandaccelerationclausesNoneNoneNoneAnti-moneylaundering,anti-corruptionandcounterterrorismfinancinglawcomplianceNoneIndemnificationandaccelerationclauses
347
Table
A8:
Environmental
Social
Governance
Safeguard
Questions
TableA.8
Envir onmental
Safeguar d
Questions
Page1of6(Environment)GrantandLoanInstrumentCategory
①BilateralChinaEximbankloan
②BilateralCDBloan
③BilateralMOFCOMloanorgrant
④BilateralChinesestate-ownedcommercialbankloan
①Kenya(2014)
①Antigua(2008)
①Philippines(2018)
②CostaRica(2008)
②MarshallIslands(2013)
②Ecuador(2016)
③Montenegro(2020)
③Peru(2014)
③Mongolia(2019)
④Uganda(2021)
④Ghana(2016)
④Honduras(2013)
Areanyenvironmentalclausesorconditionsincludedintheagreement?NoNoNoNoYesNoYesNoNoYesYesYes
IsanEnvironmentalImpactAssessment(EIA)identifiedasarequirementintheagreement?NoNoNoNoNANoYesNoNoNoNoNo
IsanEnvironmentalManagementPlan(EMP)–orafunctionalequivalentofanEMP–identifiedasarequirementintheagreement?NoNoNoNoNANoNoNoNoNoNoNo
IstheborrowerrequiredtoreporttothelenderonitsimplementationoftheEIArecommendationsorEMP?NoNoNoNoNANoNoNoNoNoNoNo
Doestheagreementincludeanyenvironmentalconditionsprecedent(forentryintoforceorinitialorongoingdisbursement)?NoNoNoNoNoNoNoNoNoYesNoYes
Doestheagreementspecifythatnoncompliancewithenvironmentalrules,standardsorlawsisasufficientbasisforthelendertodemandearlyrepayment(“prepayment”)oftheloanorcanceltheloan?NoNoNoNoNoNoNoNoNoYesNoNo
Doestheagreementrequirecompliancewithinternationalenvironmentalstandards,suchastheEquatorPrinciplesortheOECDRevisedCouncilRecommendationonCommonApproachesontheEnvironmentandOfficiallySupportedExportCredits?NoNoNoNoNoNoNoNoNoNoNoNo
Istheborrowerrequiredtoindemnifythelenders,guarantors,orinsurers("FinanceParties")againstanylossorliabilityincurredbytheFinancePartiesasaresultofanyactualorallegedbreachofanenvironmentallaworstandard?NoNoNoNoYesNoNoNoNoYesNoNo
348
Page2of6(Environment)GrantandLoanInstrumentCategory
⑤SyndicatedloanwithChineseandmultilateralbankparticipants
⑥SyndicatedloanwithChinesestate-ownedcommercialbanksand/orpolicybanks
⑦PBOC/MOFgrantorloanchanneledthroughmultilateralinstitutions
⑧Supplier'screditfromChineseSOE
⑤Colombia(2017)
⑤Argentina(2017)
⑥Serbia(2022)
⑥Argentina(2014)
⑥SienaLeone(2017)
⑦CostaRica(2014)
⑦Rwanda(2017)
⑦PapuaNewGuinea(2019)
⑧Guyana(2022)
⑧Ghana(2018)
⑧Congo(2005)
Areanyenvironmentalclausesorconditionsincludedintheagreement?YesYesYesYesYesYesYesYesYesYesNo
IsanEnvironmentalImpactAssessment(EIA)identifiedasarequirementintheagreement?YesYesNoNoYesYesYesNoNoNoNo
IsanEnvironmentalManagementPlan(EMP)–orafunctionalequivalentofanEMP–identifiedasarequirementintheagreement?YesYesNoYesYesYesNoYesNoNoNo
IstheborrowerrequiredtoreporttothelenderonitsimplementationoftheEIArecommendationsorEMP?YesYesNoNoNoYesYesYesNoNoNo
Doestheagreementincludeanyenvironmentalconditionsprecedent(forentryintoforceorinitialorongoingdisbursement)?YesNoNoYesNoYesNoNoNoNoNo
Doestheagreementspecifythatnoncompliancewithenvironmentalrules,standardsorlawsisasufficientbasisforthelendertodemandearlyrepayment(“prepayment”)oftheloanorcanceltheloan?YesNoYesYesNoYesNoYesYesYesNo
Doestheagreementrequirecompliancewithinternationalenvironmentalstandards,suchastheEquatorPrinciplesortheOECDRevisedCouncilRecommendationonCommonApproachesontheEnvironmentandOfficiallySupportedExportCredits?NoYesNoNoYesNoNoNoNoNoNo
Istheborrowerrequiredtoindemnifythelenders,guarantors,orinsurers("FinanceParties")againstanylossorliabilityincurredbytheFinancePartiesasaresultofanyactualorallegedbreachofanenvironmentallaworstandard?NoYesNoYesYesNoNoNoYesYesNo
349
Page3of6(Social)GrantandLoanInstrumentCategory
①BilateralChinaEximbankloan
②BilateralCDBloan
③BilateralMOFCOMloanorgrant
④BilateralChinesestate-ownedcommercialbankloan
①Kenya(2014)
①Antigua(2008)
①Philippines(2018)
②CostaRica(2008)
②MarshallIslands(2013)
②Ecuador(2016)
③Montenegro(2020)
③Peru(2014)
③Mongolia(2019)
④Uganda(2021)
④Ghana(2016)
④Honduras(2013)
Areanysocialclausesorconditionsincluded?NoNoNoNoNoNoYesNoNoNoNoYes
IsanEnvironmentalandSocialImpactAssessment(ESIA)identifiedasarequirement?NoNoNoNoNANoNoNoNoNoNoNo
IsanEnvironmentalandSocialActionPlan(ESAP),aResettlementActionPlan(RAP),orthefunctionalequivalentofanESAPorRAPidentifiedasarequirement?NoNoNoNoNANoNoNoNoNoNoNo
IstheborrowerrequiredtoreporttothelenderonitsimplementationoftheESIArecommendationsorRAP?NoNoNoNoNANoNoNoNoNoNoNo
Doestheagreementincludeanysocialconditionsprecedent(forentryintoforceorinitialorongoingdisbursement)?NoNoNoNoNoNoNoNoNoNoNoYes
DoestheagreementindicatethepenaltyofviolatingSocialstandards/lawsleadstodemandingearlyprepaymenttoclosetheloan?NoNoNoNoNoNoNoNoNoNoNoNo
Istheborrowerrequiredtoindemnifythelenders,guarantors,orinsurers("FinanceParties")againstanylossorliabilityincurredbytheFinancePartiesasaresultofanyactualorallegedbreachofansociallaworstandard?NoNoNoNoNoNoNoNoNoNoNoNo
350
Page4of6(Social)GrantandLoanInstrumentCategory
⑤SyndicatedloanwithChineseandmultilateralbankparticipants
⑥SyndicatedloanwithChinesestate-ownedcommercialbanksand/orpolicybanks
⑦PBOC/MOFgrantorloanchanneledthroughmultilateralinstitutions
⑧Supplier'screditfromChineseSOE
⑤Colombia(2017)
⑤Argentina(2017)
⑥Serbia(2022)
⑥Argentina(2014)
⑥SienaLeone(2017)
⑦CostaRica(2014)
⑦Rwanda(2017)
⑦PapuaNewGuinea(2019)
⑧Guyana(2022)
⑧Ghana(2018)
⑧Congo(2005)
Areanysocialclausesorconditionsincluded?YesYesNoNoYesYesYesYesNoNoNo
IsanEnvironmentalandSocialImpactAssessment(ESIA)identifiedasarequirement?YesYesNoNoNoNoYesNoNoNoNo
IsanEnvironmentalandSocialActionPlan(ESAP),aResettlementActionPlan(RAP),orthefunctionalequivalentofanESAPorRAPidentifiedasarequirement?YesYesNoNoNoYesYesNoNoNoNo
IstheborrowerrequiredtoreporttothelenderonitsimplementationoftheESIArecommendationsorRAP?YesYesNoNoNoYesYesNoNoNoNo
Doestheagreementincludeanysocialconditionsprecedent(forentryintoforceorinitialorongoingdisbursement)?YesNoNoNoNoYesYesNoNoNoNo
DoestheagreementindicatethepenaltyofviolatingSocialstandards/lawsleadstodemandingearlyprepaymenttoclosetheloan?YesNoNoNoNoYesNoYesNoNoNo
Istheborrowerrequiredtoindemnifythelenders,guarantors,orinsurers("FinanceParties")againstanylossorliabilityincurredbytheFinancePartiesasaresultofanyactualorallegedbreachofansociallaworstandard?NoYesNoNoYesNoNoNoNoNoNo
351
Page5of6(Governance)GrantandLoanInstrumentCategory
①BilateralChinaEximbankloan
②BilateralCDBloan
③BilateralMOFCOMloanorgrant
④BilateralChinesestate-ownedcommercialbankloan
①Kenya(2014)
①Antigua(2008)
①Philippines(2018)
②CostaRica(2008)
②MarshallIslands(2013)
②Ecuador(2016)
③Montenegro(2020)
③Peru(2014)
③Mongolia(2019)
④Uganda(2021)
④Ghana(2016)
④Honduras(2013)
Areanygovernance(eg.anti-corruption,competitivebidding,auditedfinancialstatements,and/oranti-moneylaundering)clausesorconditionspresentintheagreement?NoNoNoYesYesNoYesNoYesYesYesYes
Areindependentlyauditedfinancialstatementsrequiredoftheborrower?NoNoNoNoNoNoNoNoNoNoNoNo
Musttheborrower’sfinancialstatementscomplywithInternationalFinancialReportingStandards(IFRS)standards?NoNoNoYesYesNoNoNoNoNoNoNo
Doestheagreementincludeanycompetitivebiddingrequirements?NoNoNoNoNoNoYesNoYesNoNoNo
Arebiddingdocumentsorbidevaluationreportssubjecttopriorapprovalbythelender?NoNoNoNoNoNoNoNoYesNoNoNo
Doestheagreementincludeanyanti-corruptionoranti-moneylaunderingrequirements?NoNoNoYesYesNoYesNoNoYesYesYes
Istheborrowerrequiredtoindemnifythelenders,guarantors,orinsurers("FinanceParties")againstanylossorliabilityincurredbytheFinancePartiesasaresultofanyactualorallegedbreachofananti-corruptionoranti-moneylaunderinglaworstandard?NoNoNoYesNoNoNoNoNoYesNoNo
Doesthelenderexplicitlyreservetherighttopreventorinvestigateanti-corruptionoranti-moneylaunderingcrimes?NoNoNoNoNoNoNoNoNoYesNoNo
Doestheborroweraffirmthatitwillseektoensurethattheproceedsfromtheloanorgrantarenotusedtofinanceorbenefitanyinternationallysanctionedentity?NoNoNoNoNoNoNoNoNoYesNoNo
Doestheagreementspecifythatnoncompliancewithgovernancerules,standards,orlawsisasufficientbasisforthelendertodemandearlyrepayment(“prepayment”)oftheloanorcanceltheloan?NoNoNoYesNoNoNoNoNoYesYesNo
Doestheagreementincludeanycompetitivebidding,anti-corruption,oranti-moneylaunderingconditionsprecedent(forentryintoforce,initialdisbursement,orongoingdisbursements)?NoNoNoNoNoNoNoNoNoYesNoNo
Page6of6(Governance)
352
GrantandLoanInstrumentCategory
⑤SyndicatedloanwithChineseandmultilateralbankparticipants
⑥SyndicatedloanwithChinesestate-ownedcommercialbanksand/orpolicybanks
⑦PBOC/MOFgrantorloanchanneledthroughmultilateralinstitutions
⑧Supplier'screditfromChineseSOE
⑤Colombia(2017)
⑤Argentina(2017)
⑥Serbia(2022)
⑥Argentina(2014)
⑥SienaLeone(2017)
⑦CostaRica(2014)
⑦Rwanda(2017)
⑦PapuaNewGuinea(2019)
⑧Guyana(2022)
⑧Ghana(2018)
⑧Congo(2005)
Areanygovernance(eg.anti-corruption,competitivebidding,auditedfinancialstatements,and/oranti-moneylaundering)clausesorconditionspresentintheagreement?YesYesYesYesYesYesYesYesYesYesNo
Areindependentlyauditedfinancialstatementsrequiredoftheborrower?YesYesNoNoYesYesNoNoNoNoNo
Musttheborrower’sfinancialstatementscomplywithInternationalFinancialReportingStandards(IFRS)standards?NoYesNoNoYesNoNoNoNoNoNo
Doestheagreementincludeanycompetitivebiddingrequirements?NoNoNoNoNoYesYesYesNoNoNo
Arebiddingdocumentsorbidevaluationreportssubjecttopriorapprovalbythelender?NoNoNoNoNoYesYesNoNoNoNo
Doestheagreementincludeanyanti-corruptionoranti-moneylaunderingrequirements?YesYesYesYesYesYesYesYesYesYesNo
Istheborrowerrequiredtoindemnifythelenders,guarantors,orinsurers("FinanceParties")againstanylossorliabilityincurredbytheFinancePartiesasaresultofanyactualorallegedbreachofananti-corruptionoranti-moneylaunderinglaworstandard?NoYesNoNoNoNoYesNoYesYesNo
Doesthelenderexplicitlyreservetherighttopreventorinvestigateanti-corruptionoranti-moneylaunderingcrimes?YesYesNoNoNoYesNoYesNoNoNo
Doestheborroweraffirmthatitwillseektoensurethattheproceedsfromtheloanorgrantarenotusedtofinanceorbenefitanyinternationallysanctionedentity?YesYesYesNoYesNoNoNoNoNoNo
Doestheagreementspecifythatnoncompliancewithgovernancerules,standards,orlawsisasufficientbasisforthelendertodemandearlyrepayment(“prepayment”)oftheloanorcanceltheloan?YesYesYesNoYesYesYesYesYesYesNo
Doestheagreementincludeanycompetitivebidding,anti-corruption,oranti-moneylaunderingconditionsprecedent(forentryintoforce,initialdisbursement,orongoingdisbursements)?NoYesNoNoNoYesNoNoNoNoNo
353
SectionA-10:HowdoesthecompositionofthecodingsampleofinfrastructurefinancingagreementscomparetothecompositionofChina’sentiregrant-andloan-financedinfrastructureprojectportfolioinLICsandMICs?
InChapter3,weconstructacodingsampleof23infrastructurefinancingagreementstoanalyzetrendsinESGsafeguardsinthefull
grant-andloan-financedinfrastructureprojectsportfoliofromChinatoLICsandMICs.These23financingagreementswereidentified
outofasetofnearly300unredactedloancontractsandgrantagreementsincludedinthe3.0versionoftheGCDFdataset.All23
contractsrepresentinfrastructurefinancingagreementsthatcorrespondtothe8primaryinfrastructureagreementtypeslistedin
Chapter3Section3.These8financialinstrumenttypeswereusedbyChinatosupport90.2%ofitsgrant-andloan-financed
infrastructureprojectportfolioinLICsandMICsbetween2000and2021.Theremaining9.8%oftheportfolioconsistedofprojects
supportedbymore“exotic”financialinstrumenttypes(e.g.,EPCFagreements).The3.0versionoftheGCDFdatasetdoesnotinclude
manyunredactedfinancingagreementsfortheseprojects,soweexcludethemfromouranalysis.
Thiscontractcodingsamplehasbroadgeographicalcoverage,incomebracketcoverage,andtemporalcoverage(forboththepre-
andearlyBRIperiodaswellasthelateBRIperiod).TableA9describesthecompositionofthecontractcodingsampleandthe
compositionofChina’sentiregrant-andloan-financedinfrastructureprojectportfolio(asmeasuredinthe3.0versionofAidData’s
GCDFdataset)onthreedimensions:region,incomebracket,andtimeperiod.
Table
A9:
Composition
of
coding
sample
of
infrastructure
financing
agreements
versus
China’s
entire
grant-
and
loan-financed
infrastructure
project
portfolio
TableA.9
Composition
of
coding
sample
of
infrastructur e
financing
agreements
versus
China ’s
entir e
grant-
and
loan-financed
infrastructur e
project
portfolio)
ContractCodingSample(count)Fullinfrastructureprojectportfolio(%ofconstantUSD2021value)
Region
Africa34.7%(8)29%
LatinAmericaandtheCaribbean43.4%(10)18%
AsiaandthePacific13%(3)38%
CentralandEasternEurope8.2%(2)11%
MiddleEast0%1%
IncomeBracket
Upper-middleincomecountries52.2%(12)34%
Lower-middleincomecountries34.8%(8)24%
354
ContractCodingSample(count)Fullinfrastructureprojectportfolio(%ofconstantUSD2021value)
Lowincome&leastdevelopedcountries13%(3)34%
TimePeriod
Pre-andearlyBRI(2000-2017)65%(15)81%
LateBRI(2018-2022)35%(8)19%
SectionA-11:ESGriskexposureanddejureESGsafeguardprotectioninChina’soverseasinfrastructureprojectportfoliotables
Table
A10:
Regional
distribution
of
environmental,
social,
and
governance
risk
exposure
in
China’s
overseas
infrastructure
project
portfolio
TableA.10
Regional
distribution
of
envir onmental,
social,
and
governance
risk
exposur e
in
China ’s
overseas
infrastructur e
project
portfolio
RegionInfrastructureprojectswithESGriskexposure(%)Infrastructureprojectswithenvironmentalriskexposure(%)Infrastructureprojectswithsocialriskexposure(%)Infrastructureprojectswithgovernanceriskexposure(%)Infrastructureprojects(%)
Africa24.66%26.31%31.24%23.29%27.61%
South&CentralAmerica20.90%13.87%12.45%38.05%16.92%
Asia34.26%35.46%42.39%24.74%36.40%
Europe16.60%20.54%10.13%11.01%13.12%
MiddleEast2.95%2.98%2.75%2.58%4.53%
Oceania0.63%0.84%1.04%0.33%1.42%
Total100.00%100.00%100.00%100.00%100.00%
Notes:ThistableprovidesaregionalbreakdownofChina’sgrant-andloan-financedinfrastructureprojectportfolioinLICsandMICs
withenvironmental,social,orgovernancerisk(ESG)exposure(columns2-5).ItalsoprovidesaregionalbreakdownofChina’sgrant-
andloan-financedinfrastructureprojectportfolioinLICsandMICs(column6).ESGriskexposure(column2)isbasedonthe
project-levelcompositemeasurethatisdescribedinSection2ofChapter3.Likewise,environmentalriskexposure,socialrisk
exposure,andgovernanceriskexposure(column3-5)arebasedontheproject-levelcompositemeasuresthataredescribedinSection
2ofChapter3.AllpercentagesrepresentpercentagesofChina’sgrant-andloan-financedinfrastructureprojectportfolio(inconstant
2021USD)between2000and2021(exposedtooneormoretypesofESGrisk).TheregionalgroupingsaredefinedbytheOECD’s
regionalclassifications.The“Other”categorycapturesgrantandloancommitmentsforinfrastructureprojectsincountriesthatwere
classifiedashigh-incomecountriesasof2021andcommitmentsthatcouldnotbeassignedtoonespecificcountry.
355
Table
A11:
Distribution
of
environmental,
social,
and
governance
risk
exposure
in
China’s
overseas
infrastructure
project
portfolio
by
host
country
income
level
TableA.11
Distribution
of
envir onmental,
social,
and
governance
risk
exposur e
in
China ’s
overseas
infrastructur e
project
portfolio
by
host
countr y
income
level
Notes:ThistableprovidesanincomebracketbreakdownofChina’sgrant-andloan-financedinfrastructureprojectportfolioinLICs
andMICswithenvironmental,social,orgovernancerisk(ESG)exposure(columns2-5).Italsoprovidesanincomebracketbreakdown
ofChina’sgrant-andloan-financedinfrastructureprojectportfolioinLICsandMICs(column6).ESGriskexposure(column2)isbased
ontheproject-levelcompositemeasurethatisdescribedinSection2ofChapter3.Likewise,environmentalriskexposure,socialrisk
exposure,andgovernanceriskexposure(column3-5)arebasedontheproject-levelcompositemeasuresthataredescribedinSection
2ofChapter3.AllpercentagesrepresentpercentagesofChina’sgrant-andloan-financedinfrastructureprojectportfolio(inconstant
2021USD)between2000and2021(exposedtooneormoretypesofESGrisk).TheincomebracketsaredefinedbytheOECD’sODA
incomecategoriesasof2021.The“Other”categorycapturesgrantandloancommitmentsforinfrastructureprojectsincountriesthat
wereclassifiedashigh-incomecountriesasof2021andcommitmentsthatcouldnotbeassignedtoonespecificcountry.
356IncomelevelofhostcountryInfrastructureprojectswithESGriskexposure(%)Infrastructureprojectswithenvironmentalriskexposure(%)Infrastructureprojectswithsocialriskexposure(%)Infrastructureprojectswithgovernanceriskexposure(%)Infrastructureprojects(%)
Upper-middleincomecountries(UMIC)34.30%25.66%15.88%52.51%37.17%
Lower-middleincomecountries(LMIC)23.54%26.00%35.89%12.88%26.85%
Low-incomecountries(LIC)26.61%28.44%37.87%24.74%25.37%
Other15.55%19.91%10.36%9.87%10.61%
GrandTotal100.00%100.00%100.00%100.00%100.00%
Table
A12:
Country-by-country
distribution
of
ESG
risk
exposure
and
de
jure
ESG
safeguard
protection
in
China’s
overseas
infrastructure
project
portfolio,
2000-2021
TableA.12
Countr y-by-countr y
distribution
of
ESG
risk
exposur e
and
de
jure
ESG
safeguar d
protection
in
China ’s
overseas
infrastructur e
project
portfolio,
2000-2021
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Albania51%333%0000530%0%
Angola17,67044%4820%11,691364,602712,0833027%8%
AntiguaandBarbuda24649%933%24694816610%0%
Argentina10,04644%2653%9,020139,180228,090790%82%
Azerbaijan27434%330%001561118298%60%
Bahamas3,596100%583%3,53633,41216120%0%
Bangladesh12,07559%2745%6,542158,423156,0281122%24%
Barbados32%110%3100000%0%
Belarus2,36721%1523%1,8631122211,08657%3%
Benin1,65170%2456%1,57419917657131%13%
Bolivia3,28598%1169%1,85081,988455310%9%
BosniaandHerzegovina1,38164%327%1,1362798124510%0%
Botswana16911%842%1698002183%7%
Brazil1,29216%1135%1,0221000270134%63%
357
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Brunei2,10299%686%2,1026000089%83%
Bulgaria16231%450%162200020%0%
BurkinaFaso11831%225%118211820124%20%
Burundi10%313%0000130%0%
CaboVerde14335%614%1530013440%0%
Cambodia8,66559%5036%7,258456,00134351420%13%
Cameroon3,31739%1526%85572,5037172310%19%
CentralAfricanRepublic9625%729%00967010%0%
Chad87751%932%71188779000%0%
Chile16410%528%71171193415%60%
Colombia1,10274%1169%55775795244242%57%
Comoros6827%735%68700000%0%
Congo(Republic)3,87452%3038%3,18323864852360%0%
CookIslands3645%450%36400000%0%
CostaRica1,77299%1091%7578001,02843%10%
Coted'Ivoire2,76742%1828%1,235102,34412102245%33%
Cuba2188%313%21830018420%0%
NorthKorea578%638%02552220%0%
358
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Congo(DRC)2,51928%3130%1,260171,401201,17140%0%
Djibouti1,04846%411%582258221,02133%5%
Dominica11140%623%815301000%0%
Ecuador6,44765%1435%5,627112,54343,700418%20%
Egypt1833%37%183315020060%27%
ElSalvador18598%338%821124261123%20%
EquatorialGuinea2,20426%48%2,204400000%0%
Eritrea1,92775%730%001,895432365%13%
Ethiopia17,84586%7674%9,5103714,507737,4821319%4%
Fiji33866%2168%233201051000%0%
Gabon2,03567%2662%1,180101,060210122%31%
Gambia182100%4100%18249919910%0%
Georgia4917%233%49200000%0%
Ghana1,67019%2127%1,67021000022%36%
Grenada30567%1644%3051600000%0%
Guinea1,76043%1554%479141,28111,281116%28%
Guinea-Bissau12067%1255%12012006240%0%
Guyana35832%1260%7522102269100%0%
359
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Honduras33690%150%336133613361100%100%
India35214%213%352200000%0%
Indonesia20,91250%6043%14,4234916,586391,285333%39%
Iran9,30937%1327%5,13353,46274,24346%14%
Iraq3145%317%3122002193%79%
Israel2,43573%990%2,4359001,181260%75%
Jamaica1,85281%1357%1,378120047410%0%
Jordan1,80885%629%1311,80850191%35%
Kazakhstan12,19625%1321%10,400122,77011,796145%31%
Kenya10,76182%4147%9,658288,266185,4591413%7%
Kiribati00%229%0200000%0%
Kyrgyzstan2,83177%2142%1,726952761,369121%3%
Laos13,53171%4935%11,6994113,2094612117%16%
Lesotho34276%832%00193234280%0%
Liberia22559%1642%66200159140%0%
Libya00%00%000000100%100%
Madagascar32244%414%3224003210%0%
Malawi20923%512%107300103313%10%
360
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Malaysia24,10594%1959%2,733101,357520,014410%36%
Maldives1,51674%1862%1,5161800210230%29%
Mali1,02365%2158%61041,02321000%0%
MarshallIslands00%00%00000065%22%
Mauritania50930%1130%170350911010%0%
Mauritius808%13%00008010%0%
Mexico28932%427%28126130017%80%
Micronesia2121%932%21900000%0%
Mongolia1,09541%1542%9331190210043%3%
Montenegro1,09090%250%004511,09020%0%
Morocco33230%110%332133210042%38%
Mozambique2,37443%925%1,22551,3364694329%12%
Myanmar8,39565%2123%3,03183,211124,90774%11%
Namibia7315%39%733551000%0%
Nauru00%125%0100000%0%
Nepal1,38889%4087%832181,3884025640%0%
Niger1,98851%1856%33721,988181,34210%0%
Nigeria6,32652%2247%3,022104,49216787328%35%
361
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Niue1683%125%16100000%0%
NorthMacedonia1,34699%754%1,0326001,086317%20%
Oman00%00%000000100%100%
Pakistan20,44852%6349%6,3731617,887557,3801625%31%
Panama20781%571%1233001404100%100%
PapuaNewGuinea2,27033%1122%1,63851,8946633631%6%
Paraguay00%00%000000100%100%
Peru45458%873%0042113470%0%
Philippines2,33537%1636%1,978144917736455%27%
Russia63,12881%1542%42,875515,939719,197813%32%
Rwanda262%13%00002617%10%
St.Lucia5276%133%52100000%0%
Samoa14834%822%1488006110%0%
SaoTomeandPrincipe00%120%0100000%0%
Senegal2,43667%1636%2,436161,681675646%6%
Serbia3,44373%956%3,4439591005%14%
Seychelles7440%1442%741400000%0%
SierraLeone1,22733%2552%1,2272500954321%8%
362
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
SouthAfrica4,40450%1045%2,6759001,72916%33%
SouthSudan1,81792%1441%50610001,31040%0%
SriLanka10,62070%5054%7,8933485036,476251%3%
Sudan4,23028%98%1,72742,00851,97610%0%
Suriname66448%724%00844579310%19%
Syria94%330%0000930%0%
Tajikistan2,05736%2029%1,6121500988610%15%
Tanzania1,85764%2334%1,75318159620738%16%
Thailand1102%210%001102002%17%
Timor-Leste2930%436%29400000%0%
Togo72150%2147%666191624000%0%
Tonga7023%411%70400000%0%
TrinidadandTobago94365%655%943639124620%0%
Tunisia00%212%01000126%14%
Türkiye4,59043%827%3,76862,339434183%95%
Turkmenistan110%214%00001120%0%
Uganda3,27474%1232%2,84791,69932,68466%5%
Ukraine56211%440%10420045828%43%
363
CountryESGRisk($)ESGRisk(%of$)ESGrisk(#)ESGRisk(%of#)Environmentalrisk($)Environmentalrisk(#)Socialrisk($)Socialrisk(#)Governancerisk($)Governancerisk(#)StrongdejureESGsafeguards(%of$)StrongdejureESGsafeguards(%of#)
Uruguay2049%133%2010000100%100%
Uzbekistan00%00%00000096%36%
Vanuatu6323%117%63100000%0%
Venezuela64,32177%3544%10,390254,529363,743136%3%
Vietnam15,96762%4849%12,418318,0863288351%34%
Zambia8,10156%4541%2,045262,26165,8291936%29%
Zimbabwe55913%714%38760034520%0%
Africa,regional40769%667%407637952810%0%
Asia,regional00%00%000000100%100%
Notes:ESGriskexposureisbasedonaproject-levelcompositemeasurethatisdescribedinSection2ofChapter3.Likewise,environmentalriskexposure,socialriskexposure,andgovernanceriskexposurearebasedonproject-levelcompositemeasuresthataredescribedinSection2ofChapter3.StrongdejureESGsafeguardsaredefinedinSectionA-9oftheAppendix.Allmonetaryamountsareprovidedinconstant2021USDmillions.Allvaluesinthe"%of$"columnsrepresentpercentagesofChina’sgrant-andloan-financedinfrastructureprojectportfolio(inconstant2021USD)between2000and2021inaspecificrecipientcountry.Allvaluesinthe"%of#"columnsrepresentpercentagesofChina’sgrant-andloan-financedinfrastructureprojectsbetween2000and2021inaspecificrecipientcountry.
364
SectionA-12:SoftpowergainsandlossesduringearlyandlateBRIperiods
TableA13showstherelativechange(doubledelta)scoresforeachsoftpowermeasuredescribedinBox
4a,includingsteps1-3forboththeearlyBRIperiod(2014-2017)andthelateBRIperiod(2018-2021).
PublicopinionisbasedonGallupWorldPolldata.Mediasentimentisbasedontheaveragetoneofmedia
coveragerelatedtogovernmentactors,asreportedbyGDELT.ElitesupportisbasedonUNGAvoting
alignmentscores.SeeBox1bforadditionalinformationonthedatasourcesthatweusetoconstructthese
measures.Withrespecttotherelativechange(doubledelta)scoresforthepublicopinionandmedia
sentimentmetrics(reportedinTableA13),thehigherthescore,thelargerthemagnitudeofthegain
achievedbyChinaattheexpenseoftheU.S.Therelativechange(doubledelta)scoresfortheelitesupport
(idealpoint)metric(reportedinTableA13)reflectrawscoresthathavenotbeeninverted—sothelowerthe
idealpointscore(reportedinTableA13),thelargerthesizeofChina’sgainattheexpenseoftheU.S.
Table
A13:
China’s
soft
power
gains
and
losses
vis-à-vis
the
U.
S.
during
early
and
late
BRI
periods
TableA.13
China ’s
soft
power
gains
and
losses
vis-à-vis
the
U.
S.
during
early
and
late
BRI
periods
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
Afghanistan3.0%0.5320.0502.1%(0.170)0.031
Albania0.1%(0.031)0.003-4.5%0.8270.020
Algeria-3.3%0.5480.0500.7%(0.650)0.031
Angola1.9%0.0410.050(0.266)0.031
AntiguaandBarbuda4.2190.050(3.747)0.031
Argentina3.0%0.921(0.018)-12.4%(0.626)(0.023)
Armenia3.8%0.7430.020-0.2%0.843(0.032)
Azerbaijan5.2%0.8490.050-3.4%(0.607)0.031
Bahamas0.424(0.059)0.2330.031
Bangladesh1.6%0.6630.050-2.8%(0.583)0.031
Barbados0.6370.028(1.020)0.031
Belarus4.1%0.3950.023-7.7%1.232(0.039)
365
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
Benin-1.1%(1.410)0.050-0.2%(0.429)0.031
Bolivia5.7%0.6750.050-12.4%(0.193)0.031
BosniaandHerzegovina3.2%(0.047)-4.5%(0.086)
Botswana4.9%(0.401)(0.005)0.7%1.3500.031
Brazil3.0%0.1940.092-8.8%(0.159)0.159
BruneiDarussalam0.0360.0500.2180.031
Bulgaria4.9%0.4420.003-6.8%0.2840.024
BurkinaFaso0.0%0.3320.050-0.6%0.6760.031
Burundi-15.1%0.5360.050(1.229)0.031
CaboVerde(1.005)0.0500.031
Cambodia0.4%1.3210.0500.0%(0.799)0.031
Cameroon2.1%0.983(0.070)-3.6%(1.027)(0.092)
CentralAfricanRepublic4.4%(0.460)0.141(0.550)(0.112)
Chad6.8%(1.028)0.050-0.9%1.2900.031
Chile8.1%0.288(0.113)-11.1%0.1970.053
Colombia6.5%(0.181)(0.090)-9.0%1.1640.061
Comoros(2.648)0.050-0.5%(5.837)0.031
Congo0.4%1.6560.050-2.4%(0.214)0.031
CookIslands(1.629)(1.198)
CostaRica10.0%0.644(0.040)-11.5%(0.336)(0.004)
Coted'Ivoire-1.3%0.363(0.075)-1.2%1.0090.080
Cuba0.5410.0500.5910.031
NorthKorea0.1600.1310.180(0.032)
366
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
DRC0.0%2.7840.050(1.029)0.151
Djibouti1.1370.050(1.032)0.031
Dominica(1.015)(0.057)(2.684)0.031
DominicanRepublic5.7%0.309(0.015)-5.7%1.3190.053
Ecuador7.2%0.8100.050-8.6%0.0590.031
Egypt7.8%0.5370.050-36.7%0.3380.031
ElSalvador10.9%(0.649)(0.030)-6.8%(0.926)0.031
EquatorialGuinea1.972(0.037)1.2410.031
Eritrea2.5310.0500.7100.031
Ethiopia-0.1%1.0540.0509.4%0.0740.031
Fiji0.6650.027(0.202)0.031
Gabon0.1%1.5190.050-0.5%(2.304)0.031
Gambia(0.236)0.050(0.248)0.031
Georgia2.3%0.015-3.5%(14.520)(0.080)
Ghana2.8%(0.152)0.050-1.9%0.3820.031
Grenada0.498(0.088)0.0620.031
Guinea1.5%0.7280.050-3.6%(0.611)0.031
Guinea-Bissau2.3700.050(3.296)0.031
Guyana(0.265)0.0500.4010.031
Haiti4.3%0.9450.000-5.1%(2.191)0.056
Honduras5.8%0.040-11.2%(3.421)(0.088)
India-2.3%(0.184)0.145-5.4%(0.230)0.015
Indonesia2.9%0.1910.050-3.5%(0.426)0.031
367
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
Iran7.2%0.7230.0502.3%(0.109)0.031
Iraq-1.0%0.6300.050-5.4%(0.551)0.031
Israel-1.2%0.454(0.062)4.7%(0.309)(0.064)
Jamaica16.1%0.3770.027-12.1%(0.290)0.031
Jordan0.3600.050-26.8%0.1080.031
Kazakhstan5.0%0.4250.037-9.4%(0.774)0.031
Kenya2.3%0.3500.039-5.0%(0.593)0.031
Kiribati(0.397)(1.334)0.031
KyrgyzRepublic2.5%(0.041)0.050-7.2%0.2170.031
Laos0.8390.050(0.253)0.031
Lebanon-1.1%1.4000.050-4.6%0.3750.031
Lesotho13.5%(0.155)0.0500.6%1.0630.031
Liberia0.0%0.4720.0011.8%(0.072)0.235
Libya59.7%0.5200.050-3.7%0.1680.031
Madagascar0.1%1.3770.0461.4%(0.174)0.052
Malawi1.1%0.086(0.162)-0.4%0.9360.037
Malaysia-18.2%0.2330.0508.2%(0.517)0.031
Maldives0.9430.023(0.791)0.031
Mali0.2%1.2350.159-0.3%(1.396)(0.078)
MarshallIslands(0.064)(8.304)(0.068)
Mauritania5.2%1.1990.050-0.6%(0.630)0.031
Mauritius5.1%1.5270.040-4.9%(0.270)0.031
Mexico8.4%0.3230.031-8.7%0.100(0.080)
368
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
Micronesia1.101(0.341)(1.693)0.062
Moldova6.8%(0.430)(0.118)-2.2%(1.467)(0.088)
Mongolia4.1%(0.180)0.006-1.1%0.2280.031
Montenegro-0.6%0.078-0.3%(0.006)
Morocco9.5%(0.588)0.050-7.5%(0.655)0.031
Mozambique9.4%0.4070.050-3.3%(0.604)0.031
Myanmar3.3%0.4290.050-8.5%(0.345)0.031
Namibia9.5%(0.215)0.050-6.5%(0.605)0.031
Nauru0.029(8.454)(0.124)
Nepal0.5%0.0150.0502.0%(0.059)0.031
Nicaragua5.5%0.5150.050-5.2%1.1610.031
Niger-3.7%0.5230.050-5.2%(0.024)0.031
Nigeria1.7%0.4140.0500.5%0.2010.031
NorthMacedonia3.5%1.412-0.4%(1.283)
Oman2.0680.050(1.293)0.031
Pakistan4.6%0.4090.125-6.4%(0.182)0.042
Palau0.055
Panama6.7%0.135(0.180)-11.3%0.179(0.046)
PapuaNewGuinea(0.071)(0.156)0.009(0.001)
Paraguay4.6%4.208(0.122)-1.0%1.445(0.126)
Peru7.3%0.997(0.067)-5.2%0.224(0.066)
Philippines3.0%0.4960.040-2.9%(0.674)0.031
Romania5.3%0.051-6.6%(0.018)
369
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
Russia5.7%0.4110.0320.5%(0.085)(0.021)
Rwanda2.5%0.0680.142-1.6%0.143(0.013)
SaintLucia(5.200)
Samoa(0.502)(0.170)(0.795)(0.007)
SaoTomeandPrincipe4.7260.0500.031
Senegal4.4%0.3700.050-0.9%0.7960.031
Serbia6.3%0.535(0.080)-3.1%(0.674)(0.087)
Seychelles1.3170.008(0.996)0.031
SierraLeone2.4%1.1040.038-2.6%(0.203)0.031
SolomonIslands0.495(0.189)
Somalia-2.9%1.8370.0500.2020.031
SouthAfrica3.1%0.8580.022-3.9%(0.698)0.031
SouthSudan-1.2%0.1240.1851.337(0.490)
SriLanka-6.5%0.1920.050-3.1%(0.379)0.031
Sudan0.0%0.5330.050(0.092)0.031
Suriname(1.848)0.0502.8280.031
SyrianArabRepublic8.8%0.3270.613
Tajikistan-2.9%1.5270.050(2.183)0.031
Tanzania2.1%0.4220.050-0.9%(0.036)0.031
Thailand4.2%0.164(0.003)-4.2%(0.959)0.031
Timor-Leste0.0110.031
Togo-0.2%1.9580.019-2.6%(2.328)(0.088)
Tonga0.153(0.123)0.422(0.002)
370
EarlyBRI(2014-2017)LateBRI(2018-2021)
CountryPublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)PublicOpinion(%)MediaSentiment(score)EliteSupport(idealpoint)
TrinidadandTobago0.1070.0060.8550.031
Tunisia3.6%(1.372)0.050-7.8%(0.233)0.031
Turkey-0.6%0.7130.0861.7%(0.590)(0.022)
Turkmenistan0.2%0.5820.0502.8%(1.278)0.031
Uganda2.8%(0.290)0.050-3.0%1.3430.031
Ukraine-0.2%1.1010.141-0.8%(0.607)(0.025)
Uruguay8.8%0.660(0.000)-11.1%(0.254)0.094
Uzbekistan5.2%1.3400.022-7.6%(1.164)(0.038)
Vanuatu1.750(0.160)1.448(0.060)
Venezuela-0.8%0.2110.050-3.4%0.8490.031
VietNam-9.8%0.3550.0504.5%(0.070)0.031
WestBankandGazaStrip1.7%0.453-0.1%(0.309)
Yemen-0.3%(0.063)0.0508.1%(1.016)0.031
Zambia4.0%0.7450.050-7.7%(0.126)0.031
Zimbabwe2.7%0.1580.050-2.1%(0.157)0.031
TableA14presentsthesoftpowercohorts(safebet,toss-upleaningChina,toss-up,andmoonshot)to
whicheachLIC/MICbelongsbasedonitsrelativechange(doubledelta)scoresduringtheearlyBRI
(2014-2017)period.TableA14alsoreportsobservedandexpecteddevelopmentfinancecommitments
fromBeijingduringthelateBRIperiod(2018-2021)toeachLIC/MIC.Expecteddevelopmentfinance
commitmentsarebasedonahypotheticalscenarioinwhichChinaallocateddevelopmentfinancetoeach
countryonanon-strategicbasis(i.e.,basedonlyonthepopulationsizeofeachcountry).Thedifferences
betweenexpectedandobservedallocations(inconstant2021USDmillionsbetween2018and2021and
constant2021USDmillionsasapercentageoftotalODAandOOFfromChinabetween2018and2021)
thereforeprovideanindicationofhowmuchChinamayhaveprioritizedordeprioritizedeachcohortbased
onstrategicconsiderations.
371
Table
A14:
Country-by-country
breakdown
of
expected
versus
observed
development
finance
allocations
during
late
BRI
period
TableA.14
Countr y-by-countr y
breakdown
of
expected
versus
obser ved
development
finance
allocations
during
late
BRI
period
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)
AfghanistanToss-upToss-up(China)Toss-up2,741.782.42,659.30.7%
AlbaniaToss-upMoonshotToss-up(China)203.01.6201.50.1%
AlgeriaMoonshotToss-up(China)Toss-up3,076.368.93,007.30.8%
AngolaToss-upMoonshotToss-up2,349.44,658.9(2,309.5)-0.6%
AntiguaandBarbudaSafeBetsMoonshot6.6108.9(102.3)0.0%
ArgentinaToss-up(China)Toss-up(China)SafeBets3,225.5(85,906.1(82,680.6)-22.4%
ArmeniaToss-up(China)Toss-up(China)Toss-up(China)201.02.0199.00.1%
AzerbaijanToss-up(China)Toss-up(China)Toss-up717.8492.6225.20.1%
BahamasToss-upSafeBets28.90.928.00.0%
BangladeshToss-upToss-up(China)Toss-up(11,893.1(13,488.6(1,595.4)-0.4%
BarbadosToss-up(China)Toss-up(China)20.01.318.80.0%
BelarusToss-up(China)Toss-upToss-up(China)671.11,837.4(1,166.3)-0.3%
BeninMoonshotMoonshotMoonshot890.61,240.0(349.4)-0.1%
BoliviaSafeBetsToss-up(China)Moonshot846.5131.1715.40.2%
BosniaandHerzegovinaToss-up(China)SafeBets238.4469.5(231.1)-0.1%
BotswanaToss-up(China)MoonshotSafeBets180.15.0175.10.0%
BrazilToss-up(China)Toss-upMoonshot(15,170.07,384.97,785.12.1%
BruneiDarussalamMoonshotToss-up31.41,926.4(1,895.0)-0.5%
BulgariaToss-up(China)Toss-upSafeBets496.7583.4(86.7)0.0%
372
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)
BurkinaFasoMoonshotToss-upToss-up1,517.4446.91,070.50.3%
BurundiMoonshotToss-up(China)Toss-up859.734.4825.30.2%
CaboVerdeMoonshotMoonshot41.42.538.90.0%
CambodiaToss-upSafeBetsToss-up(China)1,164.76,476.1(5,311.4)-1.4%
CameroonToss-upSafeBetsSafeBets1,867.0504.51,362.50.4%
CentralAfricanRepublicToss-up(China)MoonshotMoonshot376.9286.790.20.0%
ChadSafeBetsMoonshotMoonshot1,170.7381.8788.90.2%
ChileSafeBetsToss-upSafeBets1,366.85,050.4(3,683.6)-1.0%
ColombiaSafeBetsMoonshotSafeBets3,605.8954.42,651.40.7%
ComorosMoonshotMoonshot57.1125.0(67.9)0.0%
CongoToss-upSafeBetsToss-up402.7106.8295.90.1%
CostaRicaSafeBetsToss-up(China)SafeBets364.379.8284.50.1%
Coted'IvoireMoonshotToss-upSafeBets1,891.73,938.2(2,046.4)-0.6%
CubaToss-up(China)Toss-up807.2151.5655.80.2%
NorthKoreaToss-upMoonshot1,843.51,163.9679.60.2%
DRCMoonshotSafeBetsToss-up6,531.51,005.75,525.81.5%
DjiboutiSafeBetsMoonshot77.3301.5(224.2)-0.1%
DominicaMoonshotSafeBets5.136.3(31.2)0.0%
DominicanRepublicSafeBetsToss-upSafeBets781.655.0726.50.2%
EcuadorSafeBetsToss-up(China)Toss-up1,245.51,388.4(142.9)0.0%
EgyptSafeBetsToss-up(China)Moonshot7,609.1(19,114.8(11,505.7)-3.1%
ElSalvadorSafeBetsMoonshotSafeBets449.4287.3162.10.0%
373
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)EquatorialGuineaSafeBetsSafeBets112.21,257.7(1,145.4)-0.3%
EritreaSafeBetsMoonshot252.21,762.9(1,510.8)-0.4%
EthiopiaMoonshotSafeBetsToss-up8,263.43,367.44,896.01.3%
FijiToss-up(China)Toss-up(China)65.835.330.40.0%
GabonToss-upSafeBetsMoonshot161.9187.2(25.2)0.0%
GambiaMoonshotMoonshot181.616.4165.20.0%
GeorgiaToss-upToss-up(China)265.794.4171.20.0%
GhanaToss-upMoonshotMoonshot2,275.23,298.0(1,022.8)-0.3%
GrenadaToss-up(China)SafeBets8.857.8(49.0)0.0%
GuineaToss-upToss-up(China)Toss-up931.63,196.5(2,264.8)-0.6%
Guinea-BissauSafeBetsToss-up(China)142.459.383.00.0%
GuyanaMoonshotToss-up56.9340.1(283.2)-0.1%
HondurasSafeBetsToss-up(China)717.036.0681.10.2%
IndiaMoonshotMoonshotMoonshot(99,221.0214.699,006.426.8%
IndonesiaToss-upToss-upToss-up(19,327.27,265.812,061.43.3%
IranSafeBetsToss-up(China)Toss-up(China)6,203.86,689.4(485.5)-0.1%
IraqMoonshotToss-up(China)Toss-up3,004.55,700.4(2,695.9)-0.7%
IsraelMoonshotToss-up(China)SafeBets652.1140.0512.10.1%
JamaicaSafeBetsToss-upToss-up(China)201.376.6124.80.0%
JordanToss-upToss-up772.132.2739.90.2%
KazakhstanToss-up(China)Toss-upToss-up(China)1,331.35,501.6(4,170.4)-1.1%
KenyaToss-upToss-upToss-up(China)3,676.91,310.82,366.20.6%
374
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)
KiribatiSafeBets9.0161.4(152.5)0.0%
KyrgyzRepublicToss-upMoonshotMoonshot465.2230.1235.10.1%
LaosToss-up(China)Moonshot519.05,734.9(5,216.0)-1.4%
LebanonMoonshotSafeBetsToss-up410.5334.176.40.0%
LesothoSafeBetsMoonshotToss-up160.0279.7(119.7)0.0%
LiberiaMoonshotToss-up(China)SafeBets359.9139.8220.10.1%
LibyaSafeBetsToss-up(China)Toss-up472.110.7461.40.1%
MadagascarToss-upSafeBetsToss-up(China)1,991.5352.21,639.40.4%
MalawiToss-upMoonshotSafeBets1,366.2422.3943.90.3%
MalaysiaMoonshotToss-upToss-up(China)2,356.9(10,235.3(7,878.4)-2.1%
MaldivesSafeBetsToss-up(China)36.3195.5(159.2)0.0%
MaliToss-upSafeBetsMoonshot1,493.5203.51,290.00.3%
MarshallIslandsSafeBets3.11,379.5(1,376.3)-0.4%
MauritaniaToss-up(China)SafeBetsToss-up317.3292.924.40.0%
MauritiusToss-up(China)SafeBetsToss-up(China)90.4205.3(114.9)0.0%
MexicoSafeBetsToss-upToss-up(China)8,961.31,255.27,706.12.1%
MicronesiaSafeBetsSafeBets8.081.0(73.0)0.0%
MoldovaSafeBetsMoonshotSafeBets189.72.8186.90.1%
MongoliaToss-up(China)MoonshotToss-up(China)232.88,698.4(8,465.6)-2.3%
MontenegroMoonshotMoonshot44.470.3(26.0)0.0%
MoroccoSafeBetsMoonshotToss-up(China)2,607.2334.62,272.60.6%
MozambiqueSafeBetsToss-upToss-up2,195.9416.51,779.50.5%
375
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)
MyanmarToss-up(China)Toss-upMoonshot3,802.5679.93,122.60.8%
NamibiaSafeBetsMoonshotToss-up(China)176.3146.629.70.0%
NauruToss-up(China)0.90.00.90.0%
NepalToss-upMoonshotToss-up2,084.5550.91,533.60.4%
NicaraguaSafeBetsToss-up(China)Toss-up479.432.6446.80.1%
NigerMoonshotToss-up(China)Moonshot1,707.41,201.0506.40.1%
NigeriaToss-upToss-upToss-up(14,704.86,488.88,216.02.2%
NorthMacedoniaToss-up(China)SafeBets148.1239.6(91.6)0.0%
OmanSafeBetsToss-up326.2732.5(406.3)-0.1%
PakistanToss-up(China)Toss-upMoonshot(16,101.3(43,114.6(27,013.3)-7.3%
PanamaSafeBetsToss-upSafeBets304.4385.6(81.2)0.0%
PapuaNewGuineaMoonshotSafeBets688.8276.6412.20.1%
ParaguayToss-up(China)SafeBetsSafeBets469.63.3466.30.1%
PeruSafeBetsSafeBetsSafeBets2,358.24,566.7(2,208.5)-0.6%
PhilippinesToss-up(China)Toss-up(China)Toss-up(China)7,947.22,223.95,723.31.6%
RomaniaToss-up(China)Moonshot1,379.2181.91,197.30.3%
RussiaSafeBetsToss-upToss-up(China)(10,293.5(18,081.5(7,788.0)-2.1%
RwandaToss-upMoonshotMoonshot928.2480.0448.20.1%
SamoaMoonshotSafeBets15.32.013.30.0%
SaoTomeandPrincipeSafeBetsToss-up(China)15.595.8(80.3)0.0%
SenegalToss-up(China)Toss-upMoonshot1,158.81,410.4(251.6)-0.1%
SerbiaSafeBetsToss-up(China)SafeBets494.02,403.6(1,909.6)-0.5%
376
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)
SeychellesSafeBetsToss-up(China)7.07.3(0.3)0.0%
SierraLeoneToss-upSafeBetsToss-up(China)581.52,213.5(1,632.0)-0.4%
SolomonIslandsToss-up(China)48.8162.0(113.2)0.0%
SomaliaMoonshotSafeBetsMoonshot1,160.726.91,133.80.3%
SouthAfricaToss-up(China)Toss-up(China)Toss-up(China)4,172.07,312.3(3,140.3)-0.9%
SouthSudanMoonshotToss-upMoonshot753.61,312.3(558.8)-0.2%
SriLankaMoonshotToss-upToss-up(China)1,563.46,793.4(5,230.0)-1.4%
SudanMoonshotToss-up(China)Toss-up3,131.0101.43,029.60.8%
SurinameMoonshotToss-up43.1947.6(904.5)-0.2%
SyrianArabRepublicSafeBetsToss-up1,456.034.51,421.50.4%
TajikistanMoonshotSafeBetsMoonshot674.31,479.9(805.6)-0.2%
TanzaniaToss-upToss-upToss-up(China)4,344.1578.33,765.81.0%
ThailandToss-up(China)Toss-upSafeBets5,098.7319.24,779.61.3%
Timor-LesteToss-up(China)92.25.586.70.0%
TogoMoonshotSafeBetsToss-up(China)596.135.1560.90.2%
TongaToss-upSafeBets7.543.4(35.9)0.0%
TrinidadandTobagoMoonshotToss-up(China)108.4142.4(34.0)0.0%
TunisiaToss-up(China)MoonshotToss-up864.5112.5751.90.2%
TurkeyMoonshotToss-up(China)Moonshot5,986.1(17,859.1(11,873.0)-3.2%
TurkmenistanToss-upToss-up(China)Toss-up443.21.6441.50.1%
UgandaToss-upMoonshotToss-up3,120.2948.92,171.30.6%
UkraineMoonshotSafeBetsMoonshot3,159.7598.52,561.20.7%
377
EarlyBRI(2014-2017)(ObservedSoftPowerChanges)LateBRI(2018-2021)(ExpectedandObservedAllocations)
CountryPublicOpinionMediaSentimentEliteSupportExpectedODA/OOFAllocation(USDmillions)ObservedODA/OOFAllocation(USDmillions)Difference(USDmillions)Difference(percentoftotalChineseODA/OOF)
UruguaySafeBetsToss-up(China)SafeBets244.836.4208.40.1%
UzbekistanToss-up(China)SafeBetsToss-up(China)2,423.15,152.0(2,729.0)-0.7%
VanuatuSafeBetsSafeBets22.0176.1(154.1)0.0%
VenezuelaMoonshotToss-upMoonshot2,062.4319.11,743.30.5%
VietNamMoonshotToss-upToss-up6,872.03,313.03,559.01.0%
WestBankandGazaStripToss-upToss-up339.012.4326.60.1%
YemenMoonshotMoonshotToss-up2,278.827.32,251.40.6%
ZambiaToss-up(China)Toss-up(China)Moonshot1,332.53,060.8(1,728.3)-0.5%
ZimbabweToss-upToss-upMoonshot1,108.5429.5679.00.2%
378
SectionA-13:China’sofficialsectorlendingportfolioinLICsandMICs:acomparisonofAidData,IDS,andCODF
TableA15presentstheaggregatemonetaryvalue(inconstant2021USD)ofloancommitmentsfromofficial
sectorcreditorsinChinatoborrowersinLICsandMICs,asmeasuredbythreedifferentsources:the3.0
versionofAidData’sGCDFdataset,theWorldBank’sInternationalDebtStatistics(IDS),andBoston
University’sChina’sOverseasDevelopmentFinance(CODF)dataset.393Thecountry-levelsummarystatistics
areorganizedbylevelofpublicliability.394IDSandCODFprovidedataonpublicandpublicly-guaranteed
debt(PPG)forseveralofficialsectorcreditorsinChina.IDSprovidescoveragefor89LICsandMICs
(excludingChina)from2000to2021andCODFprovidescoveragefor96LICsandMICsfrom2008to
2021.The3.0versionofAidData’sGCDFdatasetcoversPPGandnon-PPGdebtfrom180officialsector
creditorsinChinato165LICsandMICs(ofwhich126contractedloansfromofficialsectorcreditorsin
China)from2000to2021.395
AidDatameasuresPPGdebtbyaggregatinglendingcommitmentstogovernmentandmajority
state-ownedinstitutionsaswellasotherinstitutionsthatsecuredcentralgovernmentrepaymentguarantees
orrepaymentguaranteesfromstate-ownedentitiesotherthanthecentralgovernmentinthehostcountry
(i.e.,byusingthe“LevelofPublicLiability”variableinthe3.0versiontotheGCDFdatasettoidentifyall
loancommitmentsassignedtothe“centralgovernmentdebt,”“centralgovernment-guaranteeddebt,”
and“otherpublicsectordebt"categories).InadditiontoPPGdebt,the“LevelofPublicLiability”variable
inthe3.0versionofAidData’sGCDFdatasetcaptures“potentialpublicsectordebt”(loanstominority
state-ownedinstitutionswithoutpublicsectorrepaymentguarantees),“privatedebt,”anddebtthatcannot
beeasilycategorizedbasedonthelevelofpublicliability(referredtoas“unallocable”debtinTableA15).
The“TotalDebttoChina”columninTableA15)representsthesumofallloancommitmentsfromofficial
sectorcreditorsinChinatoborrowersinLICsandMICsbetween2000and2021thatqualifyasPPGor
non-PPGdebt.
395InTableA15,theloancommitmenttotalsfromthe3.0versionofAidData’sGCDFdatasetexcludetheshort-term"rollover"facilitiesdescribedinBox2candSectionA-3.394SectionA-5intheAppendixprovidesmoredetailsonhowAidDatausesthe“levelofpublicliability”measuretocategorizelendingtodifferenttypesofborrowers.393AsofAugust2023,theChineseLoanstoAfrica(CLA)databasecaptured$203billioninChineselendingcommitmentstogovernmentandmajoritystate-ownedinstitutionsinAfricafrom2000to2020(deflatedtoconstantUSD2021).Overthesametimeperiod,the3.0versionofAidData’sGCDFdatasetcaptures$266billioninChineselendingcommitmentstogovernmentandmajoritystate-ownedinstitutionsaswellasotherinstitutionsthatsecuredcentralgovernmentrepaymentguaranteesorrepaymentguaranteesfromstate-ownedentitiesotherthanthecentralgovernmentinthehostcountry(i.e.,“centralgovernmentdebt,”“centralgovernment-guaranteeddebt,”and“otherpublicsectordebt")inAfrica.Additionally,in2021,AidDatarecords$5.3billioninChineselendingcommitmentstoAfricathatqualifyas“centralgovernmentdebt,”“centralgovernment-guaranteeddebt,”and“otherpublicsectordebt.”AidDataalsocapturesChineselendingcommitmentstoAfricathatareexcludedfromtheCLAdatabasebydefinition:$31billiontoprivatesectorborrowinginstitutions(“privatedebt”)inAfricaand$8.6billiontospecialpurposevehicleandjointventureborrowersthatareminority-ownedbypublicsectorinstitutionsinAfricawithoutrepaymentguaranteesfrompublicsectorinstitutionsinhostcountries(“potentialpublicsectordebt”)between2000and2021.
379
TheAidData,CODFandIDSestimatesofPPGandnon-PPGdebttoChinainTableA15shouldbe
interpretedwithcautionsincetheydonotrepresentamountsoutstandingandassuchdonotaccountfor
disbursementsorrepayments.396
Table
A15:
Official
sector
lending
commitments
from
China
to
LICs
and
MICs:
Country-by-country
comparison
of
AidData,
IDS,
and
CODF
in
constant
2021
USD
millions
TableA.15
Official
sect or
lending
commitments
from
China
to
LICs
and
MICs:
Countr y-by-countr y
comparison
of
AidData,
IDS,
and
CODF
in
constant
2021
USD
millions
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
Albania100NANA000100
Algeria1142NA000114
Angola63,02238,59038,7151,422936264,816
AntiguaandBarbuda310NA221009320
Argentina32,8708,58121,723383,3311,48437,724
Armenia2828NA00028
Azerbaijan288NANA77611801,182
Bahamas132NA11903,41203,544
Bangladesh17,16811,84517,30102,58431820,070
Barbados216NA191000216
Belarus10,82212,6338,491124118011,064
Benin1,3971,0373902484002,260
Bolivia2,3822,9233,7440602,388
396AidDatarecentlylaunchedanewdatacollectioninitiativetotrackdisbursements,repayments,andamountsoutstandingonaloan-by-loanbasis.Inthe3.0versionoftheGCDFdataset,wehavedocumenteddisbursements,repayments,andamountsoutstandinginthe“description”fieldforasubsetofcountries.However,inthefuture,weintendtopublishloan-leveldataondisbursements,repayments,andamountsoutstandinginauser-friendlyformatforamorecompletesetofcountries.
380
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
BosniaandHerzegovina1,359NA1,016077302,132
Botswana1,660194NA0001,660
Brazil39,24023,82436,180013,5971,50554,341
Brunei0NANA2,1022402,126
Bulgaria9115026801023571,371
BurkinaFaso24217889000242
Burundi48198NA38024109
CaboVerde19863NA000198
Cambodia6,1736,7764,3061,2788,7925116,293
Cameroon7,4446,8996,4544494338,433
CentralAfricanRepublic26693NA02000466
Chad1,073603771860561,214
Chile312NANA7195,1732456,449
Colombia296NANA761,169561,597
Comoros138138134000138
Congo(Republic)7,1492,8924,502059507,258
CookIslands38NANA00038
CostaRica753265504000753
Coted'Ivoire7,0145,4043,8880007,014
Cuba4,086NA42708404,171
NorthKorea76NANA00076
381
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
Congo(DRC)4,0538,1302,3561,2177,22764113,137
Djibouti2,1711,5981,4230002,171
Dominica76586000076
DominicanRepublic0NA65400139139
Ecuador25,67215,45122,471012345826,254
Egypt14,3606,5038,593048913714,986
ElSalvador0NANA05379132
EquatorialGuinea9,009NA2,5810009,009
Eritrea1,05327062612101,7122,886
Ethiopia17,44012,1549,72802,8879120,419
Fiji444390664000444
Gabon2,8682,1231,58712002,880
Gambia30302700030
Georgia99NA0127330466
Ghana7,8547,9113,0845431,41229,812
Grenada84797600084
Guinea3,0162,1812,26309278244,768
Guinea-Bissau27NANA00027
Guyana408493342021082700
Honduras336NANA01140450
HungaryNANA2,122NANANANA
382
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
India968NANA09,6721410,654
Indonesia22,6204,50317,2673,25027,2131,91754,999
Iran24,8252,31033,39904302,69627,951
Iraq6,782NANA06373937,812
Israel0NANA02,7872,8665,653
Jamaica1,7041,7092,43805172012,422
Jordan3672NA01,94801,984
Kazakhstan27,1134,85622,12627,5083,7415,81864,180
Kenya11,79311,47210,5680860012,653
Kiribati0NANA11500115
Kyrgyzstan2,6982,5462,717068903,386
Laos10,4409,25310,8706,5262,5861,02420,577
Lebanon317NANA000317
Lesotho317295278000317
Liberia62725800564626
Libya451NANA000451
Madagascar5575644610150572
Malawi6414643180490689
Malaysia12,584NA5,27702,56378915,936
Maldives1,9161,2531,1060001,916
Mali1,2571,2076870001,257
383
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
MarshallIslands144NANA04,98905,133
Mauritania9631,223817016501,129
Mauritius8878335850283711,241
Mexico832NA1,12802,7594744,064
Micronesia4NANA0004
Moldova16NANA00016
Mongolia5,4781,8062,24103297516,558
Montenegro1,1631,1661,163045241,233
Morocco1,5861,5021,422051802,104
Mozambique3,1773,7582,5811,962381935,613
Myanmar10,6403,4695,7602432,18063813,701
Namibia2,020NA139001552,175
Nepal49841843803460844
Nicaragua0NANA020020
Niger1,7191,7463871,39725003,366
Nigeria12,4728,7677,7158361,1447514,528
NorthMacedonia1,31989688400271,346
Oman6,944NA3,687186007,130
Pakistan67,22145,99926,001091078768,918
Panama105NANA05340640
PapuaNewGuinea1,6501,5531,5812,9891,9863006,924
384
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
Paraguay56NANA020076
Peru421NA75014,1772,29316,890
Philippines2,5043,1801,2471,1814,09007,774
Romania292NANA0285146722
Russia129,191NA75,822037,1842,883169,258
Rwanda7196526470260745
Samoa288272233000288
Senegal3,6102,9382,07100493,660
Serbia4,7764,4062,7040688695,712
Seychelles2NANA0002
SierraLeone878113341952,0511,4094,533
SouthAfrica11,6784,8125,13309,08856121,326
SouthSudan5,283NA4460005,283
SriLanka17,35613,09714,4204301,11856719,472
Sudan18,0015,9262,15000018,001
Suriname1,514NA9670001,514
Syria15077550492,2282,427
Tajikistan3,5952,4131,1656835992315,108
Tanzania2,8593,0172,4220002,859
Thailand3,267NANA01,5572415,065
Timor-Leste0NA560000
385
CountryTotalPPGDebttoChinaTotalPPGDebttoChinaTotalPPGDebttoChinaPotentialPublicSectorDebttoChinaPrivateSectorDebttoChinaUnallocableDebttoChinaTotalDebttoChinaSourceAidDataIDSCODFAidDataAidDataAidDataAidData
YearRange2000-20212000-20212008-20212000-20212000-20212000-20212000-2021
Togo1,12889065505701,185
Tonga193176126000193
TrinidadandTobago646NA761000646
Tunisia314189120000314
Türkiye17,6643,0931,3897899,8344528,331
Turkmenistan12,21711,7279,25700012,217
Uganda4,1694,1373,933002694,438
Ukraine3,0071,7612,57102301013,339
Uruguay130NANA010923262
Uzbekistan10,2323,9209,1367,20936720518,013
Vanuatu323270260000323
Venezuela112,782NA71,341000112,782
Vietnam18,5955,36511,6132,3627,73810228,797
Yemen357448NA000357
Zambia12,3177,1417,37769631014713,469
Zimbabwe3,8993,8732,869106374444,990
Regional57NA2,04801,3228,4499,828
Total1,109,793378,195605,29467,184216,36249,9541,280,125
SectionA-14:ComparisonofAidDataandIDSestimatesofPPGandnon-PPGdebtexposuretoChina
TableA16comparestotalloancommitments(inconstant2021USDmillionsandasapercentageofGDP)
fromofficialsectorcreditorsinChinatoborrowersinLICsandMICsfrom2000-2021,ascapturedbythe3.0
386
versionofAidData’sGCDFdatasetandtheWorldBank’sIDS.397Thecountry-levelsummarystatisticsare
organizedbylevelofpublicliability.398
89LICsandMICs(excludingChina)voluntarilyreportonpublicandpublicly-guaranteeddebt(PPG)tothe
WorldBankthroughitsDebtorReportingSystem(DRS).TheWorldBank’sdefinitionofPPGdebtincludes
(a)long-termexternalobligationsofpublicdebtors,includingthenationalgovernment,apolitical
subdivision(oranagencyofeither),andautonomouspublicbodies;and(b)long-termexternalobligations
ofprivatedebtorsthatareguaranteedforrepaymentbyapublicentity(WorldBank2000).“Public
debtors”includeentitiesinwhichthehostgovernmentholdsatleastfiftypercentownership.
TomaximizecomparabilitywithIDS,the3.0versionofAidData’sGCDFdatasetclassifiesallloansaccording
totheirlevelsofpublicliability.AidDatameasuresPPGdebtbyaggregatinglendingcommitmentsto
governmentandmajoritystate-ownedinstitutionsaswellasotherinstitutionsthatsecuredcentral
governmentrepaymentguaranteesorrepaymentguaranteesfromstate-ownedentitiesotherthanthe
centralgovernmentinthehostcountry(i.e.,byusingthe“LevelofPublicLiability”variableinthe3.0
versionoftheGCDFdatasettoidentifyallloancommitmentsassignedtothe“centralgovernmentdebt,”
“centralgovernment-guaranteeddebt,”and“otherpublicsectordebt"categories).
WhilesimilartoTableA27inMaliketal.(2021),TableA16inthisreportrepresentsanupdated
methodologicalapproach.Maliketal.(2021)defines“sovereigndebt”ascentralgovernmentand
central-governmentguaranteeddebt,anditdefines“hiddendebt”asdebtincurredbystate-owned
entities(includingSPVs)withanylevelofhostgovernmentownership.InTableA16,wealignmoreclosely
withtheDRSdefinitionofPPGdebtbyseparatelyrecordingloanstomajoritystate-ownedentities
(includingSPVs)—withandwithoutrepaymentguaranteesfrompublicsectorinstitutionsotherthanthe
centralgovernment—aswellasprivatesectorlendingthatisguaranteedbymajoritystate-ownedentitiesin
anew(“otherpublicsectordebt”)category.Weclassifyloanstominoritystate-ownedinstitutionswithout
publicsectorrepaymentguaranteesas“potentialpublicsectordebt”sincesuchloansmaybenefitfrom
implicitformsofhostgovernmentliabilityprotection(Maliketal.2021;MalikandParks2021).
TheAidDataandIDSestimatesofPPGandnon-PPGdebtexposuretoChinathatarereportedinTable
A16shouldbeinterpretedwithcautionastheyarebasedoncumulativeloancommitmentsovera22-year
period.Theydonotrepresentamountsoutstandingandassuchdonotaccountfordisbursementsor
repayments.399
399AidDatarecentlylaunchedanewdatacollectioninitiativetotrackdisbursements,repayments,andamountsoutstandingonaloan-by-loanbasis.Inthe3.0versionoftheGCDFdataset,wehavedocumenteddisbursements,repayments,andamountsoutstandinginthe“description”fieldforasubsetofcountries.However,inthefuture,weintendtopublishloan-leveldataondisbursements,repayments,andamountsoutstandinginauser-friendlyformatforamorecompletesetofcountries.398SectionA-5intheAppendixprovidesmoredetailsonhowAidDatausesthe“levelofpublicliability”measuretocategorizelendingtodifferenttypesofborrowers.397InTableA16,cumulativeloancommitmentsfromthe3.0versionofAidData’sGCDFdatasetexcludetheshort-term"rollover"facilitiesdescribedinBox2candSectionA-3.
387
Table
A16:
Country-by-country
comparison
of
AidData
and
IDS
estimates
of
PPG
and
non-PPG
debt
exposure
to
China
in
constant
2021
USD
millions
TableA.16
Countr y-by-countr y
comparison
of
AidData
and
IDS
estimates
of
PPG
and
non-PPG
debt
exposur e
to
China
in
constant
2021
USD
millions
CountryReportstoDRS(1)PPGdebtexposuretoChina(AidData)(2)PPGdebtexposuretoChina(DRS)(3)PPGandpotentialpublicdebtexposuretoChina(AidData)DifferencebetweenAidData(3)andDRS(2)(1b)PPGdebttoChinaasa%ofGDP(AidData)(2b)PPGdebttoChinaasa%ofGDP(DRS)(3b)PPGandpotentialpublicdebttoChinaas%ofGDP(AidData)DifferencebetweenAidData(3b)andDRS(2b)AfghanistanN0NA000.0%NA0.0%0.0%
AlbaniaN100NA1001000.5%NA0.5%0.5%
AlgeriaY11421141120.1%0.0%0.1%0.1%
AmericanSamoaN0NA000.0%NA0.0%0.0%
AngolaY63,02238,59064,44425,85489.4%54.7%91.4%36.7%
AntiguaandBarbudaN310NA31031021.8%NA21.8%21.8%
ArgentinaY32,8708,58132,90924,3286.7%1.8%6.8%5.0%
ArmeniaY282828(0)0.2%0.2%0.2%0.0%
ArubaN0NA000.0%NA0.0%0.0%
AzerbaijanN288NA1,0641,0640.5%NA1.9%1.9%
BahamasN132NA1321321.2%NA1.2%1.2%
BangladeshY17,16811,84517,1685,3234.1%2.9%4.1%1.3%
BarbadosN216NA2162164.5%NA4.5%4.5%
BelarusY10,82212,63310,946(1,687)15.9%18.5%16.0%-2.5%
BelizeN0NA000.0%NA0.0%0.0%
BeninY1,3971,0371,4203837.9%5.9%8.0%2.2%
BhutanN0NA000.0%NA0.0%0.0%
BoliviaY2,3822,9232,382(541)5.9%7.2%5.9%-1.3%
BosniaandHerzegovinaN1,359NA1,3591,3595.8%NA5.8%5.8%
BotswanaY1,6601941,6601,4669.4%1.1%9.4%8.3%
BrazilY39,24023,82439,24015,4162.4%1.5%2.4%1.0%
BritishVirginIslandsN0NA000.0%NA0.0%0.0%
388
BruneiN0NA2,1022,1020.0%NA15.0%15.0%
BulgariaY911509118611.1%0.1%1.1%1.0%
BurkinaFasoY242178242651.2%0.9%1.2%0.3%
BurundiY4819885(113)1.2%5.1%2.2%-2.9%
CaboVerdeY1986319813510.2%3.2%10.2%7.0%
CambodiaY6,1736,7767,45167523.1%25.4%27.9%2.5%
CameroonY7,4446,8997,48858916.4%15.2%16.5%1.3%
CaymanIslandsN0NA000.0%NA0.0%0.0%
CentralAfricanRepublicY2669326617310.6%3.7%10.6%6.9%
ChadY1,0736031,1595556.5%3.7%7.1%3.4%
ChileN312NA1,0311,0310.1%NA0.3%0.3%
ColombiaN296NA3723720.1%NA0.1%0.1%
ComorosY138138138(0)10.3%10.3%10.3%0.0%
Congo(DRC)Y4,0538,1305,270(2,860)7.7%15.4%10.0%-5.4%
Congo(Republic)Y7,1492,8927,1494,25855.7%22.5%55.7%33.2%
CookIslandsN38NA383811.7%NA11.7%11.7%
CostaRicaY7532657534891.2%0.4%1.2%0.8%
Coted'IvoireY7,0145,4047,0141,61010.1%7.7%10.1%2.3%
CubaN4,086NA4,0864,0863.2%NA3.2%3.2%
CuraçaoN0NA000.0%NA0.0%0.0%
DjiboutiY2,1711,5982,17157358.6%43.2%58.6%15.5%
DominicaY7658761813.8%10.5%13.8%3.3%
DominicanRepublicN0NA000.0%NA0.0%0.0%
EcuadorY25,67215,45125,67210,22124.2%14.6%24.2%9.6%
EgyptY14,3606,50314,3607,8573.4%1.5%3.4%1.8%
ElSalvadorN0NA000.0%NA0.0%0.0%
EquatorialGuineaN9,009NA9,0099,00972.5%NA72.5%72.5%
EritreaY1,0532701,17490446.7%12.0%52.1%40.1%
EswatiniN0NA000.0%NA0.0%0.0%
EthiopiaY17,44012,15417,4405,28617.6%12.2%17.6%5.3%
FijiY4443904445410.3%9.1%10.3%1.3%
FrenchPolynesiaN0NA000.0%NA0.0%0.0%
GabonY2,8682,1232,88075715.5%11.5%15.6%4.1%
389
GambiaY303030(0)1.5%1.5%1.5%0.0%
GeorgiaY999(0)0.0%0.0%0.0%0.0%
GhanaY7,8547,9118,3974879.9%10.0%10.6%0.6%
GrenadaY84798467.6%7.0%7.6%0.5%
GuamN0NA000.0%NA0.0%0.0%
GuatemalaN0NA000.0%NA0.0%0.0%
GuineaY3,0162,1813,01683618.8%13.6%18.8%5.2%
Guinea-BissauN27NA27271.7%NA1.7%1.7%
GuyanaY408493408(85)5.1%6.1%5.1%-1.1%
HaitiN0NA000.0%NA0.0%0.0%
HondurasN336NA3363361.2%NA1.2%1.2%
IndiaN968NA9689680.0%NA0.0%0.0%
IndonesiaY22,6204,50325,86921,3661.9%0.4%2.2%1.8%
IranY24,8252,31024,82522,5154.2%0.4%4.2%3.8%
IraqN6,782NA6,7826,7823.3%NA3.3%3.3%
IsraelN0NA000.0%NA0.0%0.0%
JamaicaY1,7041,7091,704(5)11.6%11.7%11.6%0.0%
JordanY367236(36)0.1%0.2%0.1%-0.1%
KazakhstanY27,1134,85654,62149,76514.0%2.5%28.3%25.8%
KenyaY11,79311,47211,79332110.7%10.4%10.7%0.3%
KiribatiN0NA1151150.0%NA50.4%50.4%
KosovoN0NA000.0%NA0.0%0.0%
KyrgyzstanY2,6982,5462,69815230.9%29.1%30.9%1.7%
LaosY10,4409,25316,9667,71354.7%48.5%88.9%40.4%
LebanonN317NA3173170.8%NA0.8%0.8%
LesothoY3172953172213.4%12.4%13.4%0.9%
LiberiaY627262(10)2.5%3.0%2.5%-0.4%
LibyaN451NA4514511.2%NA1.2%1.2%
MadagascarY557564557(7)3.9%3.9%3.9%0.0%
MalawiY6414646411775.3%3.8%5.3%1.4%
MalaysiaN12,584NA12,58412,5843.4%NA3.4%3.4%
MaldivesY1,9161,2531,91666335.5%23.2%35.5%12.3%
MaliY1,2571,2071,257496.6%6.3%6.6%0.3%
390
MarshallIslandsN144NA14414456.0%NA56.0%56.0%
MauritaniaY9631,223963(260)9.6%12.2%9.6%-2.6%
MauritiusY887833887547.7%7.2%7.7%0.5%
MexicoN832NA8328320.1%NA0.1%0.1%
MicronesiaN4NA440.3%NA0.3%0.3%
MoldovaN16NA16160.1%NA0.1%0.1%
MongoliaY5,4781,8065,4783,67236.3%12.0%36.3%24.3%
MontenegroY1,1631,1661,163(2)20.0%20.1%20.0%0.0%
MoroccoY1,5861,5021,586841.1%1.1%1.1%0.1%
MozambiqueY3,1773,7585,1391,38220.1%23.8%32.6%8.8%
MyanmarY10,6403,46910,8837,41418.2%5.9%18.6%12.7%
N.MarianaIslandsN0NA000.0%NA0.0%0.0%
NamibiaN2,020NA2,0202,02016.5%NA16.5%16.5%
NauruN0NA000.0%NA0.0%0.0%
NepalY498418498801.4%1.2%1.4%0.2%
NewCaledoniaN0NA000.0%NA0.0%0.0%
NicaraguaN0NA000.0%NA0.0%0.0%
NigerY1,7191,7463,1161,37011.5%11.7%20.9%9.2%
NigeriaY12,4728,76713,3084,5412.9%2.0%3.1%1.1%
NiueN0NA000.0%NA0.0%0.0%
NorthKoreaN76NA76760.5%NA0.5%0.5%
NorthMacedoniaY1,3198961,3194239.5%6.5%9.5%3.0%
OmanN6,944NA7,1307,1307.9%NA8.1%8.1%
PakistanY67,22145,99967,22121,22219.6%13.4%19.6%6.2%
PalauN0NA000.0%NA0.0%0.0%
PanamaN105NA1051050.2%NA0.2%0.2%
PapuaNewGuineaY1,6501,5534,6393,0866.2%5.8%17.4%11.6%
ParaguayN56NA56560.1%NA0.1%0.1%
PeruN421NA4214210.2%NA0.2%0.2%
PhilippinesY2,5043,1803,6855050.6%0.8%0.9%0.1%
PuertoRicoN0NA000.0%NA0.0%0.0%
RomaniaN292NA2922920.1%NA0.1%0.1%
RussiaN129,191NA129,191129,1917.3%NA7.3%7.3%
391
RwandaY719652719676.5%5.9%6.5%0.6%
Saint-MartinN0NA000.0%NA0.0%0.0%
SamoaY2882722881633.6%31.7%33.6%1.9%
SaoTomeandPrincipeN0NA000.0%NA0.0%0.0%
SenegalY3,6102,9383,61067213.1%10.6%13.1%2.4%
SerbiaY4,7764,4064,7763707.6%7.0%7.6%0.6%
SeychellesN2NA220.2%NA0.2%0.2%
SierraLeoneY8781131,07396020.7%2.7%25.3%22.6%
SintMaartenN0NA000.0%NA0.0%0.0%
SolomonIslandsN0NA000.0%NA0.0%0.0%
SomaliaN0NA000.0%NA0.0%0.0%
SouthAfricaY11,6784,81211,6786,8662.8%1.1%2.8%1.6%
SouthSudanN5,283NA5,2835,283122.8%NA122.8%122.8%
SriLankaY17,35613,09717,7874,69020.3%15.4%20.8%5.5%
St.KittsandNevisN0NA000.0%NA0.0%0.0%
St.LuciaN0NA000.0%NA0.0%0.0%
St.VicentN0NA000.0%NA0.0%0.0%
SudanY18,0015,92618,00112,07450.2%16.5%50.2%33.7%
SurinameN1,514NA1,5141,51447.0%NA47.0%47.0%
SyriaY15077150740.8%0.4%0.8%0.4%
TajikistanY3,5952,4134,2781,86541.1%27.6%48.9%21.3%
TanzaniaY2,8593,0172,859(157)4.1%4.3%4.1%-0.2%
ThailandN3,267NA3,2673,2670.6%NA0.6%0.6%
Timor-LesteN0NA000.0%NA0.0%0.0%
TogoY1,1288901,12823813.8%10.9%13.8%2.9%
TongaY1931761931841.0%37.2%41.0%3.8%
TrinidadandTobagoN646NA6466462.6%NA2.6%2.6%
TunisiaY3141893141250.7%0.4%0.7%0.3%
TürkiyeY17,6643,09318,45315,3602.2%0.4%2.3%1.9%
TurkmenistanY12,21711,72712,21749022.6%21.7%22.6%0.9%
TurksandCaicosN0NA000.0%NA0.0%0.0%
TuvaluN0NA000.0%NA0.0%0.0%
U.S.VirginIslandsN0NA000.0%NA0.0%0.0%
392
UgandaY4,1694,1374,169329.8%9.7%9.8%0.1%
UkraineY3,0071,7613,0071,2461.5%0.9%1.5%0.6%
UruguayN130NA1301300.2%NA0.2%0.2%
UzbekistanY10,2323,92017,44113,52114.8%5.7%25.2%19.5%
VanuatuY3232703235332.9%27.5%32.9%5.5%
VenezuelaN112,782NA112,782112,782100.9%NA100.9%100.9%
VietnamY18,5955,36520,95615,5915.1%1.5%5.7%4.3%
WestBankandGazaStripN0NA000.0%NA0.0%0.0%
YemenY357448357(92)3.6%4.5%3.6%-0.9%
ZambiaY12,3177,14113,0135,87257.8%33.5%61.1%27.6%
ZimbabweY3,8993,8733,9093616.2%16.1%16.2%0.1%Notes:FordetailsonAidData’s“levelofpublicliability”measure,seeSectionA-5.Columns1b,2b,and3bpresentcumulative2000-2021loancommitmentsfromChina(overthe2000-2021period)asapercentageofrecipientcountryGDP.GDPdataaredrawnfromUNstats.un.organdmeasuredinnominal2021USD.
393
Document ID: rivals-and-responders-the-us-china-and-global-crisis-management