untying aid
TRANSCRIPT
Evaluation
of the implementation of the
Paris Declaration
Thematic Study
The Developmental Effectiveness of Untied Aid:
Evaluation of the Implementation of the Paris Declaration
and of the 2001 DAC Recommendation on Untying
ODA to the LDCs
Ownership, Alignment, Harmonisation, Results and Accountability
This report presents the results of a thematic study undertaken within the framework of the Evaluation of the Paris Declaration and in response to the request by the DAC for an evaluation of the effects of untying of aid to LDCs.The report builds on extensive research carried out in a first phase, pub-lished in October 2008 and subsequent case studies in Burkina Faso, Ghana, Laos, South Africa, Vietnam and Zambia.
Untying aid: Is it working?
Evaluation of the Paris Declaration
Untying aid
: Is it wo
rking?D
ecemb
er 2009
Aid untying: is it working?
tHEMAtiC study on tHE dEvElopMEntAl EffECtivEnEss of untiEd Aid:
EvAluAtion of tHE iMplEMEntAtion of tHE pAris dEClArAtion And of tHE 2001 dAC
rECoMMEndAtion on untying odA to tHE ldCs
synthesis report
ISBN: 978-87-7605-352-9e-ISBN: 978-87-7605-356-7
Copyright©2009DanishInstituteforInternationalStudies,DIIS
Suggestedcitation: Clay,EdwardJ.,MatthewGeddesandLuisaNatali: UntyingAid:Isitworking?AnEvaluationoftheImplementation oftheParisDeclarationandofthe2001DACRecommendation ofUntyingODAtotheLDCs,Copenhagen,December2009.
Photocopiesofallpartsofthispublicationmaybemadeprovidingthesourceisacknowl-edged.
GraphicProduction: Turbineforlaget,Filmbyen21,8000AarhusC,DenmarkGraphicDesign: ph7kommunikation,www.ph7.dkPrint: Scanprinta/s,JensJuulsVej2,8260Viby,Denmark
Thereportcanbedownloadeddirectlyfrom:
http://www.oecd.org/dac/evaluationnetworkandmaybeorderedfreeofchargeonlineatwww.evaluation.dk.
ThisThematicStudywasundertakenundertheauspicesoftheSecretariatfortheEvalu-ationoftheParisDeclarationandtheOECD/DAC.Theviewsexpressedaresolelythoseofthestudyteamanddonotreflecttheviewsofthesponsoringorganisations.ThestudymakesextensiveuseofinformationfromtheOECDCreditorReportingSystem(CRS)data-baseasupdatedon29September2009.
Aid untying: is it working?
tHEMAtiC study on tHE dEvElopMEntAl EffECtivEnEss of untiEd Aid:
EvAluAtion of tHE iMplEMEntAtion of tHE pAris dEClArAtion And of tHE 2001 dAC
rECoMMEndAtion on untying odA to tHE ldCs
synthesis report
Edward J. Clay, Matthew Geddes and Luisa Natali
Untying aid: Is it working?iv
ContentsList of Acronyms....................................................................................................................................................................................................... vii
Summary......................................................................................................................................................................................................................... x
1. Introduction.............................................................................................................................................................................................................11.1 Background........................................................................................................................................................................................................ 11.2 Purpose,scopeandtimetable..................................................................................................................................................................... 11.3 Approach,activitiesandoutlineofreport.............................................................................................................................................. 2
2. The international rules and agreements concerning the tying status of ODA........................................................................5
3. A statistical analysis of the untying/tying status of ODA...................................................................................................................83.1 AnincompletepictureofuntyingandtyingstatusofODA............................................................................................................ 83.2 Tyingstatus:trendsandrecentdonorperformance........................................................................................................................113.3 Donorsrespondingtothe2001Recommendation..........................................................................................................................11
4. Donor policies and practices: how does untying work? A survey of five donor experiences of untying............ 184.1 Objectivesandscope...................................................................................................................................................................................184.2 Donorcountryexperiencesofuntying..................................................................................................................................................184.3 Findings.............................................................................................................................................................................................................24
5 .The impacts of tying and untying: a literature review...................................................................................................................... 255.1 Introduction:objectivesandscope........................................................................................................................................................255.2 Theeffectsoftying:conceptualissues..................................................................................................................................................255.3 Theeffectsoftying:evidence...................................................................................................................................................................265.4 Theimpactsofuntyingaid.........................................................................................................................................................................28
6. Uses of untied aid in six partner countries.............................................................................................................................................. 316.1 Objectivesandscope...................................................................................................................................................................................316.2 Profileofsixcasestudycountries............................................................................................................................................................336.3 Aidmodalitiesanduseofcountryprocurementsystems..............................................................................................................336.4 Econometricanalysisoftheimplicationsofbilateralaidfordonorexports............................................................................376.5 Conclusions......................................................................................................................................................................................................39
7. An assessment of the effectiveness of untied ODA: findings of project studies.................................................................. 407.1 Introduction.....................................................................................................................................................................................................407.2 Inputfeatures:contractingarrangementsandtyingpractice......................................................................................................427.3 Consequences:interconnectionsandaideffectiveness.................................................................................................................487.4 Aidefficiency:costeffectivenessandtime..........................................................................................................................................487.5 Aideffectiveness:ownership,alignment,harmonisationandtransparency...........................................................................517.6 Developmentalimpacts..............................................................................................................................................................................527.7 Awkwardissuesaboutuntying:possibletrade-offsandrisks......................................................................................................537.8 Conclusions......................................................................................................................................................................................................53
8. Conclusions............................................................................................................................................................................................................ 55
AnnexesAnnexA:TermsofReference...................................................................................................................................................................................60AnnexB:Statisticaltables........................................................................................................................................................................................64AnnexC:Econometricsanalysisoftheimplicationsofbilateralaidfordonorexports....................................................................68AnnexD:ExecutiveSummariesofCountryStudies.......................................................................................................................................73AnnexE:References..................................................................................................................................................................................................93
Untying aid: Is it working? v
list of tablesTable2.1 DACdefinitionsoftyingstatusofODA,otherofficialflowsandofficiallysupportedcredits...................................... 5Table2.2 ChronologyofOECDagreementsontyingstatusofaidincludingcreditsthatqualifyasODA................................. 6Table3.1 DACdonorcountries:tyingstatusofbilateralODAtoLDCsandnon-LDCsin1999-2001and2005-07 (percentagesbasedonthreeyearaverages)................................................................................................................................. 8Table3.2 DACdonorcountries:tyingstatusofbilateralODAin2007(includingdonoradministrativecosts).....................10Table3.3 DACdonorcountries:untiedaidaspercentofdonorODAin2007....................................................................................12Table3.4 DACdonorcountries:tyingstatusofbilateralODAtoLDCsaccordingtogrant/non-grantstatusin2007.........13Table3.5 DACdonorcountries:tyingstatusofbilateralODAtoLDCsaccordingtotechnicalcooperationcomponentin2007...13Table3.6 DACdonorcountries:bilateralODAbyconsolidatedsectorsin2007:tyingstatusaspercentofsectorODA...14Table3.7 Globalcerealsfoodaidbysourceofsupplyin2001and2007.............................................................................................15Table3.8 DACdonors:geographicaldistributionofcontractsawardedin2007..............................................................................17Table6.1 Profileofsixcasestudycountries.....................................................................................................................................................32Table6.2 TyingstatusofODAinsixcasestudycountries(2005-07)......................................................................................................33Table6.3 BilateralDACdonors(includingEC)–PercentofaidtogovernmentusingCountryProcurement Systems(CPS)in2007...........................................................................................................................................................................34Table6.4 AidandotherinfluencesonDACdonorexports:comparativeeconometricresultsforsixpartner countriesduring2002-07....................................................................................................................................................................38Table7.1 Projectcasestudies:recipientrelationship,implementingentity,headandTCcontracts.........................................43Table7.2 WorldBankcontractsbysupplierdevelopmentcategorysignedin2003-08inthesixcasestudy countries(averagepercentageshareoftotal).............................................................................................................................44Table7.3 Projectcasestudies:procurement,subcontracting,aidmodalityandmajorinputs....................................................46
TableB.1 DACdonorcountries:bilateralODAbyconsolidatedsectorsin2007.TyingstatusaspercentofsectorODA...61TableB.2 Formalfeaturesofthe21projectcasestudies............................................................................................................................62TableB.3 DACdonors:representativenessofthepurposivesampleofsixpartnercountries......................................................67TableD.1 Ghana:bilateralDACdonorsandECdisbursementsbyaidmodality2006-08(US$mn)...........................................75TableD.2 BurkinaFaso:bilateralDACdonorsandEC:disbursementsbymodality2007(US$mn)............................................77TableD.3 SouthAfrica:donoruseofcountrysystems,2005.....................................................................................................................80TableD.4 Zambia:disbursementsofODAbyselecteddonorsin2007/8.............................................................................................84TableD.5 LaoPDR:bilateralODAdisbursementsbysource,2005-06....................................................................................................88TableD.6 Vietnam:aidcommitmentsanddisbursementsbytypeofassistance2006-07.............................................................90
list of figuresFigure3.1 Donorshareoftotalbilateral‘tied/notreported’ODAin2007............................................................................................... 9Figure3.2 Shareoftotalbilateralaidthatisuntied,threeyearaverages(1984-2007):DACandCRSdatacomparison......11Figure3.3 RelationshipbetweenshareofuntiedODA(untiedaidratio)andeffort-sharing(ODA/GNI)in2007...................16Figure4.1 Percentageofbilateralaidreportedasuntiedfrom1996to2007......................................................................................19Figure7.1 Flowdiagramofaid,untyingpracticeandsourcingoutcomes............................................................................................47
list of BoxesBox4.1 ThedeclineofMixedCreditArrangements..................................................................................................................................20Box4.2 Transparency:serviceprocurementbyDanidaHQ...................................................................................................................21Box4.3 UntyingCanadianfoodaid.................................................................................................................................................................22Box4.4 InformationaboutAusAIDopencompetitivetenders.............................................................................................................23Box5.1 Developmentalimpactoflocalandregionalprocurementoffoodaid............................................................................29Box7.1 DFIDSurveyofconstraintsonlocalfirmsbiddingforinternationalcontracts................................................................50Box7.2 Untying,OwnershipandAlignment:countrystudyconclusions.........................................................................................51
1
2
3
4
5
6
7
8
Untying aid: Is it working?vi
Acknowledgements
T heauthorswouldfirstofallliketothankallthoseofficialsinthegovernmentsofthesixpartnercountries,Burkina
Faso,Ghana,LaoPDR,SouthAfrica,VietnamandZambia,fortheircooperationduringthecountrystudiesduringNovem-ber2008-September2009.PreparationofthestudyhasdrawnextensivelyonideasdiscussedandinformationprovidedduringPhaseOneofthestudyontheuntyingofaid,espe-ciallybydevelopmentcooperationagenciesinAustralia,Canada,Denmark,NorwayandSwitzerland,aswellasinPhaseTwowithUKDFIDandNetherlandsMinistryofForeignAffairsandthemembersofthemanagementgroupforthestudy.Thecooperationofdonoragencystaff,especiallythosewhofacilitatedtheworkofthestudyteamsinthesixpartnercountrieshasbeenmuchappreciated.MembersoftheInternationalReferenceGroupfortheParisDeclarationEvaluationalsomadeusefulcomments.TheauthorswouldliketoexpresstheirthankstoNielsDabelsteinoftheSecre-tariatfortheEvaluationoftheParisDeclarationandFrans
LammersenofDevelopmentCooperationDirectorateforthegenerousadviceandassistancetheyhaveprovidedoverthecourseofthisstudy.TheStudyhasbenefittedfromtheadviceofthoseresponsiblefortheCRSdatabaseandtheExportCreditSecretariatintheOECD.Dr.FlorianWienekeatKfWforhelpfullysupportingtheGhanapilotstudy.AtODItheadviceofIsabellaMassaandDirkW.teVeldehasbeenhelpfulandDaniPhilipsandMahaliaRimmerprovidedadministrativesupport.MavisClayandJennyJoneshelpedwithediting.JimBoyce,StephenBiggs,RichardManning,TrevorSibumbaandKenStephensonprovidedvaluableinsightsand/oradviceonanalyticissues.AnespeciallywarmthanksisofferedtothemembersofthesixcountrystudyteamsincludingEmmanuelJ.MensahandPeterQuartey(Ghana),YashRamkolowanandMatthewStern(SouthAfrica),OliverSaasaandChrispinMphuka(Zambia),PietLanser,JudithMattiejsesenandEliPare(BurkinaFaso),andAdamMcCarthyandcolleaguesatMekongEconomics(LaoPDRandVietnam).
Untying aid: Is it working? vii
AAA AccraAgendaforAction(2008)ACP Africa,CaribbeanandPacific(Countrypartyto LoméAgreementwithEU)AFD FrenchDevelopmentAgency(French DevelopmentAgency)ANZCERTA AustraliaNewZealandCloserEconomicRelations TradeAgreementANZGPA AustralianandNewZealandGovernment ProcurementAgreementASEAN AssociationofSouthEastAsianNationsAusAID AustralianAgencyforInternationalDevelopmentB2B BusinesstoBusinessBMZ GermanFederalMinistryforEconomic Co-operationandDevelopmentBOP BalanceofPaymentsCEA CostEffectivenessAnalysisCHF SwissFrancCIDA CanadianInternationalDevelopmentAgencyCPG CommonwealthProcurementGuidelines(Australia)CPS CountryProcurementSystemsCRS CreditorReportingSystem(OECD)DAC DevelopmentAssistanceCommitteeDanida DanishInternationalDevelopmentAssistanceDBI DirectBilateralInstrumentsDFI DirectlyFinancingImportsDFID DepartmentforInternationalDevelopment(UK)DKK DanishKronerDPE DonorPartnerEnvelopeECOWAS EconomicCommunityofWestAfricanStatesEEA EuropeanEconomicAreaEFTA EuropeanFreeTradeAssociationEU EuropeanUnionFAC FoodAidConventionFAO FoodandAgricultureOrganizationFDI ForeignDirectInvestmentGATT GeneralAgreementonTariffsandTradeGBS GeneralBudgetarySupportGDP GrossDomesticProductGNI GrossNationalIncomeGST GoodsandServiceTaxGTZ GermanTechnicalCooperationHIPC HighIndebtedPoorCountryHLM HighLevelMeetingHQ Headquarter(s)ICB InternationalCompetitiveBiddingICRC InternationalCommitteeoftheRedCrossICT InternationalCompetitiveTenderIDA InternationalDevelopmentAssociation(WorldBank)
list of Acronyms
IDRC InternationalDevelopmentResearchCentreJICA JapaneseInternationalCooperationAgencyKfW GermanDevelopmentBank(Germany)LDC LeastDevelopedCountryLIC LowIncomeCountryLRPs LocalandRegionalPurchasesMERX CanadianPublicTenderingSystemMCA MixedCreditsArrangementMCC MillenniumChallengeCorporation(US)MIC MiddleIncomeCountryMoFA MinistryofForeignAffairsMFI MultilateralFinancialInstitutionNAFTA NorthAmericanFreeTradeAgreementNCB NationalCompetitiveBiddingNDFI NotDirectlyFinancingImportsNGO Non-GovernmentalOrganisationNMFA NorwegianMinistryofForeignAffairsNOK NorwegianKronerNorad NorwegianAgencyforDevelopmentCooperationODA OfficialDevelopmentAssistanceOECD OrganisationforEconomicCooperationand DevelopmentOLICs OtherLowIncomeCountriesO&M OperationandMaintenancePD ParisDeclarationPDE ParisDeclarationEvaluationRTE ResourceTransferEfficiencySADC SouthernAfricanDevelopmentCommitteeSBS SectoralBudgetarySupportSDC SwissAgencyforDevelopmentandCooperationSDR SpecialDrawingRights(IMF)SECO StateSecretariatforEconomicAffairs(Switzerland)Sida SwedishInternationalDevelopmentCooperation AgencySSA Sub-SaharanAfricaSWAP Sector-WideApproachTC TechnicalCooperationToR TermsofReferenceUMIC UpperMiddleIncomeCountryUNDB UnitedNationsDevelopmentBusinessUSAID USAgencyforInternationalDevelopmentVNG AssociationofNetherlandsMunicipalitiesWFP UNWorldFoodProgrammeW&SS WaterandSanitationSectorWTO WorldTradeOrganisation
Untying aid: Is it working?viii
summary
Background
T heOECDDAC,inmakingitsRecommendationin2001foruntyingofaidtoleastdevelopedcountries,envis-
agedacomprehensiveevaluationofitsimplementationandimpactby2009.TheParisDeclaration(PD)reaffirmedtheRecommendationfirstlybystatingthatuntyingaidgenerallyincreasesaideffectivenessbyreducingtransactioncostsforpartnercountriesandimprovingcountryownershipandalignment,andconsequentlybymakingfurtherstepstowardsaiduntyingIndicator 8ofprogresstowardsincreasingaideffectiveness.AccordinglyOECDcommissionedthisstudyaspartofthePDevaluation.
Thisstudy,undertakenduring2008-09,hasincludedare-examinationofeffortstowardsuntyingasreportedstatisti-callybyDACdonorstotheOECD,complementedbyareviewofdonorpoliciesandpracticesthroughapurposivesurveyoffivedonorswhohadeitherlargelyuntiedtheiraid,orwereactivelycommittedtountyingafter2001(Australia,Canada,Denmark,NorwayandSwitzerland).Thewidespreadpercep-tionthatuntyingincreasesaidefficiencyandeffectivenesshasbeenre-examinedthroughaliteraturereviewontheeconom-icsoftyinganduntyingpractices,followedbyaninvestigationintountyingpracticesinsixpartnercountries(SouthAfrica,Zambia,BurkinaFaso,Ghana,VietnamandLaoPDR).Twenty-onedevelopmentprojectsinthesecountries,supportedindi-viduallyorjointlybythirteenDACmembersaswellastheEUandtheWorldBank(IDA),andreportedtotheOECDasfullyuntiedorwithuntiedcomponents,wereexaminedindetail.ECadministeredaidisincludedtoensurecoverageofpartial untyingandIDAaidisincludedasabaseforcomparisonwithfully untied funding.MostprojectsareintheWaterandSanita-tionSector(W&SS),whichcombinesbothinfrastructureandsocialdevelopmentobjectives.
Key findings TheoverallpictureisverypositiveintermsofprogressbyDACdonorsintheformal untyingoftheiraid,removinglegalandregulatoryimpedimentstotheprocurementofgoods
andservicesoutsidethedonor’sownmarket.Theproportionoffullyuntiedbilateralaidroseprogressivelyfrom46%in1999-2001to76%in2007,andforLDCsithasincreasedfrom57%to86%.Takingintoaccountmultilateralaid,thepropor-tionofuntiedODAhasrisento83%overall.Thechangesoverthepast10yearsindicatethatthe2001Recommendationonuntyingofaid,togetherwithotherinternationalagreementssuchastheParisDeclaration,havehadanoverallpositiveim-pactonfurtheruntyingaid.Butthereareimportantqualifica-tions.Free-standingtechnicalcooperation(TC)andfoodaidwereexcludedfromtheRecommendationandbothremainsignificant‘greyareas’,with30%andatleast50%respectivelystillreportedastied,andthetyingstatusofsignificantpropor-tionsnotbeingreportedbysomedonors.
Thestudyplacestheseoverallpositivefindings(whicharebasedondatareportedbydonorstotheOECDandasurveyofdonorviewsabouttheirownactions)intothecontextofre-cipientperspectivesandfindingsabouttheactualorde factopracticeofaiddeliveryinpartnercountries.
Country partner stakeholdersstronglyaffirmedthatuntyingisabouttransferringresponsibilityforplanningandmanagingfundsfromdonorstorecipients.Practically,untyingisthenamatterofcontracts,modalities,useofcountrysystemsandofferinglocalbusinessanopportunitytocompetesuccessfullyforcontracts.Theyarelessconcernedaboutremovingbiasesbetweendonorpartnersintheirtrade,exceptwhereexcessiveinefficienciesoftyingsignificantlyreducetheresourcetransfervalueofaid.
Therearesystematicdifferencesintherevealedpreferencesofdonorsaboutaidusesinthecountriessurveyed,wheretheiraidisoverwhelminglyformallyuntied.Genuine(de facto)untyingwastypicallyfoundconvincinglylinkedtoevidenceontheadoptingofprogrammaticandpoolingmodalities,forwhichuntyingisanecessarycondition,combinedwitheffortstostrengthenpartnercapacitiesinfinancialmanagementandprocurement.
Untying aid: Is it working? ix
Project-type aidisstillthedominantmodalityinallsurveyedcountries.Manyformallyuntiedprojectswerefoundtobede factotiedortohaveonlysomeuntiedcomponents.InmostinvestmentprojectstheprimaryorheadcontractsandmostoftheTCcomponentsarestillprocuredfromthedonorcountry,evenifprocurementischannelledthroughrecipientsystems.Thiscallsinquestiontoacertainextentthegenuine-nessofuntyingeffortsandreflectstherealitythat,evenwhereprocurementisbeinghandedovertopartners,mostdonorstrytoinfluenceprojectimplementation,throughlong-termtechnicalassistanceormanagementconsultantsfromtheirhomecountry.Reasonsforthisde factotyinginclude:(i)do-norregulations;(ii)lackoflocalcapacity;(iii)localandregionalcontractorsbeingunabletocompeteinternationally(espe-ciallyiflarge-scalecontractsarebeingprocured,e.g.becauseofweaklocalsecondarymarketsforcreditandprofessionalinsurance);(v)unequalaccesstoinformation;(vi)potentialriskaversionatdonorHQ;and(iv)pressureforspeedyimplemen-tation.
TheWorldBank(IDA)wasmoresuccessfulthanDACdonorsinusingInternationalCompetitiveBidding(ICB)andNationalCompetitiveBiddingtosourcingcontractsbothlocallyandregionallyinotherdevelopingcountries,againpointingtosomecombinationofintendedandunintendedconstraintsonde factountyingofbilateralaid.
Value for money:theneargeneralconclusionineconomictheoryandthelimitedevidence-basedliteratureconcludethatuntiedaidislikelytobecost-effective,andthisisbroadlyconfirmedbythestudy.Inmostcasesprocurementcostswerecompetitivewithlocalandinternationalmarketprices.Occasionallycostswerehigher,reflectingsmall-scaletender-ingprocessesandsocialnetworks.ICBwasnotalwayscost-effective,assomebilateraldonorsstilltend(butnotnecessar-ilyintentionally)toadvantagedonor-basedcompanieswithassociatedhighercosts.
Ownership and alignment:thecountrystudiesfoundthatapositiveapproachtotheuseofuntiedfunds,includingusingprogrammaticmodalitiessuchasgeneralbudgetarysupport,sectoralbasketfundingandjointpoolingbydonors,aswellasadoptingcountrysystemsforprocurement,werepromot-ingownershipandfacilitatingalignmentwithpartnercountrypriorities.
Developmental impacts:thecountrycasestudiesarecautious-lypositive,findingevidenceofdirectemploymentandincomestreamgeneration.Byuntying,thebalanceofsourcinggoodsandservicesisshiftedtowardslocalsuppliers,therebypromotinglocalbusinessdevelopment.Inaddition,capac-itybuildingandknowledgetransferappeartobemorelikelywherelocalandregionalfirmsofmediumcapacityundertakeprojectinvestmentactivities.However,thelinkageeffectsarelimitedduetothelowdevelopmentofmanylocalmarkets.Furtherexpandinguntyingthereforemustgoinlinewithlocalmarketdevelopment.
Overall,successfuluntyingemergesfromthecountrystudiesasacriticallyimportantcontributoryfactortowardsachievinggreateraideffectivenessandmustthereforenotonlybejusti-fiedongroundsofcost-efficiency.
The way forward Donorsandtheirpartnerscandomuchmoretocompletetheprocessofuntyingandtoexploittheopportunitiesofferedbyuntyingtoincreaseaideffectivenessanditsdevelopmentim-pacts.Theseactionsneedtobesetclearlywithinthechangingglobalcontextofaidincluding:
• ‘Theelephantintheroom’–abetterunderstandingisneededaboutthegrowingroleofconcessionalandlargelytieddevelopmentfundingprovidedbyemergingnon-DACdonors;
• NewverticalfundingmechanismsthatcouldbeatoddswiththePDaideffectivenessprincipleswhichprovidetherationaleforuntiedaid;
• Newformsoffinancingseemlikelytoberequiredtoad-dressthechallengesposedbothbytheglobalfinancialcrisisintheshort-termandclimatechangeinthelongerterm.
The2008AccraAgendaforAction(AAA)onaideffectivenessreaffirmedthecommitmenttocontinuetheprocessofunty-ingandcallsuponDACMembersinrespondingtoelaborate individual plans by 2010 to untie their aid to the maximum extent and to improve reporting on the 2001 Recommendation (OECDDAC,2009).Theseproposalsarestronglyendorsedandfurtherspecificmeasuresaresuggestedtosupportgenuineuntyingandreduce,ifnoteliminatede factotyingpractices.
Donorsshould increase the transparency oftheirreportingtotheOECD,forexamplebyincludingthetyingstatusofcatego-riesofaidexemptfromtheRecommendation,includingallTCandfoodaidandexpandingtheroleof ex antereporting.
Donoruntyingplansshouldincludespecificmeasuresformakinggreateruseofcountrysystems,andtospecifythemeanstheyhaveadopted,orplantoadopt,forremovingob-staclesandencouraginglocalandregionalfirmstoparticipatesuccessfullyinprocurement.
Donorsshouldconsiderindividualorjointinitiativestostrengthenlocalandregionalcapacitytocompeteforaidsup-portedcontracts.BroadeningthesourcingofTCandmanage-mentofprojectsshouldbepriorityareas.
Agencieswhoseuntyingpracticeisstillconstrainedbydo-mesticregulationsshouldincludeintheirplansproposalstoremoveorobtaingeneralexemptionsfromsuchrestrictionsonthedisbursementofpublicfunds.
Untying aid: Is it working?x
Reportingonimplementationofplansforcompletingtheun-tyingofaidshouldbesubmittedtobecomepartoftheannualreporttotheDAC,andalsoshouldbemadeaspecificitemforinclusioninfutureDACmemberpeerreviews.
TheDACshouldconsideradding,totheannualprogressre-portsonuntying,acountrypartnerfocusedreviewofprogressinuntying,undertakenwiththefullcooperationofanLDCorHIPCcountryandalsoDACmembers.
DACmembersindividuallyneedtomonitorandrigorouslyevaluateboththeirprogressinuntyingandinmakinguseofthecapacitieswithinpartnercountries.
Untyingisnotanendinitselfbutameanstoincreaseaideffectivenessanditsdevelopmentalimpact.TherehasbeenmuchprogressbymostDACmembers,butsomehavedonelittlemorethannominallyuntyingtheiraidinalegalorregulatorysense.Sothereismuchscopeforfurtheractionstowardsuntyingaidinpracticalwaysandexploitingtheopportunitiesthatthisoffersforstrengtheningaideffective-nessthroughgenuinepartnershipsbetweendonorsandaidrecipients.
Untying aid: Is it working? 1
Chapter 1
1.1 Background
F orseveraldecadesdebatesonaideffectivenesshavefocusedontheissueofthetyingstatusofaid(Seebelow
Chapter2andTable2.1).Ithasbeenclearlydocumentedthattiedaidraisesthecostofgoods,servicesandworksby15%to30%onaverage,andbyasmuchas40%ormoreforfoodaid.1Thisisaconservativeestimateoftherealcostsoftying,sinceitdoesnotincorporatetheindirectcosts.Furthermore,therehavebeenwidelydiscussedconcernsabouttyingreducingeffectiveness:actingasaconstraintondonorcooperationandthebuildingofpartnershipswithdevelopingcountries,byinhibitingtheownershipandresponsibilityofpartnercountriesinaidsupporteddevelopment,aswellashamperingbroadereffortstopromotetheirintegrationintotheglobaleconomy.2
Asaresult,afterextendedanddifficultnegotiations,theOECD/DACadoptedin2001aRecommendationtountiemuchOfficialDevelopmentAssistance(ODA)toLeastDevel-opedCountries(LDCs).TheRecommendation(amendedin2006andin2008)alsoinvitesDevelopmentAssistanceCom-mittee(DAC)memberstoprovideuntiedaidinareasnotcov-eredbytheRecommendation,andtostudythepossibilitiesofextendinguntiedaidinsuchareas.Basically,areasnotcoveredaretechnicalcooperation(TC),foodaidanddonoradministra-tivecosts.3Countrycoveragemightalsobewidened.ProgressachievedintheproportionofODAthatisuntiedistrackedinthecontextoftheMillenniumDevelopmentGoals,astarget
1 The estimates of the average costs of tying were presented in a survey of research and evaluations up to c. 1990 published by the OECD (Jepma, 1991) and those for food aid in a more recent OECD (2006a) study; for further discussion see Chapter 5 below.
2 See, for example, Helleiner (2000) as a statement of such broader concerns about tying practices inhibiting effective partnerships.
3 It might be thought that the tying status of donor administrative costs is irrelevant, as these are virtually by definition tied. However, as donors report more fully on the tying status of their aid, some are found to have begun to source administrative services more widely (see below Table 3.6).
1. introduCtion
number35.The2005ParisDeclarationonAidEffectiveness(PD)reiteratedthe2001DACRecommendationandenvisagedthatprogressintheshareofaidthatisuntiedbemonitoredthroughIndicator8.4
TheWorkingPartyonAidEffectivenessandtheDACNetworkonDevelopmentEvaluationaskedforastudyoftheextenttowhichdevelopmentpartnershaveuntiedtheirassistanceandthekeyfactorspromotingorimpedingprogressonfullyuntyingdevelopmentassistance.Furthermore,theyhaveaskedforthestudytoidentifyexamplesofbenefitsoffullyuntiedaid.The2001DACRecommendationtountieODAtotheLDCshadalsoincludedarequestforacomprehensiveevaluationofitsimpact,whichismandatedforsubmissiontothe2009DACHighLevelMeeting(HLM).5SointhelightoftheclosecommunalitiesbetweenthestudywithintheframeworkoftheevaluationoftheParisDeclaration,andtherequestforanevaluationinthe2001DACRecommendation,thisstudyisbeingundertakentoassesstheeffectivenessofuntiedaid.
1.2 Purpose, scope and timetableThepurposeofthestudyistoprovidetheDACandthe2009HLM,aswellasthewiderdevelopmentcommunity,withacomprehensiveassessmentofcurrentdonorpoliciesandpracticesregardingthetyingstatusofaid,andanassessmentoftheeffectsoftheuntyingstatusonaideffectiveness.Thestudy’sintendedfocusisontheresults of untied aid,andex-aminesif,andtowhatextent,thepresenteffortforuntying aid has contributed to aid effectiveness.Thestudyalsoexplorestheprospectsforincreasingtheshareofuntiedaid,and,where
4 “Untying aid generally increases aid effectiveness by reducing transaction costs for partner countries and improving country ownership and alignment. DAC donors will continue to make progress on untying as encouraged by the 2001 DAC Recommenda-tion on Untying Official Development Assistance to the Least Developed Countries (Indicator 8).” OECD DAC, 2005, Para 31.
5 ‘The comprehensive evaluation will also pay attention to the implementation of this Recommendation with respect to achieving a balance of efforts among DAC Members and promoting and sustaining ODA flows to LDCs.’ OECD DAC, 2001, Para 20.
Untying aid: Is it working?2
Chapter 1
relevant,providespolicyrecommendationsonhowtoachievethisobjective.Thequestionsaddressedbythestudyinclude:
• Towhatextenthasdonorbehaviourchangedasaresultofthe2001DACRecommendationonUntyingAidtoLeastDevelopedCountries,i.e.towhatextenthavedo-norsuntiedtheiraidfurther?
• Whatfactorshaveenhancedorimpededfullyuntyingdevelopmentassistance?
• Whatistheevidencethatuntiedaidhasresultedinanincreaseintheeffectivenessandefficiencyofaid?
Theevaluationstudyhasbeenconductedintwostages.
• Thefirststage,PhaseOne,fromMaytoSeptember2008,providedacomprehensiveheadquartersleveloverviewofthecurrentpoliciesandpracticesofDACmembersandnon-DACdonorsregardingthetyingstatusoftheiraid,andtheeffectsonaideffectiveness.DuringthisfirststageamethodologyforPhaseTwo,includingpartnercountrycasestudies,wasdevelopedforpresentationtoDACmembersinDecember2008.
• PhaseTwoconsideredtheeffectsofthetyingstatusofaidonaideffectivenessonthebasisofarepresentativegroupofsixevidence-basedcasestudiesinpartnercountries(althoughoriginallyonlyfourwereenvisagedintheTermsofReference(ToR)).Thestudieswereorganisedintwostages.TheODIcoreteam,incooperationwithlocalresearchers,didapilotstudyinGhanaduringNovember2008toFebruary20096Fivefurthercasestudieswereundertaken,againincooperationwithlocalresearch-ersbetweenFebruaryandOctober2009.Thisphaseexaminedtheextenttowhichuntiedaidhasresultedinprocurementfromlocal/regionalcompaniesanditseffectonaideffectiveness.7
ThecomprehensivefinalreportcoveringPhasesOneandTwoincludespracticalpolicyrecommendationsforincreasingtheshareofaidthatisuntied.
1.3 Approach, activities and outline of reportItwasenvisagedthatthestudywouldbedonethroughaconsultativeprocesswithdonors,partnercountriesandcivilsociety,andsothisintentionhasshapedthewayPhasesOneandTwohavebeenundertaken.
First,thereisastatisticalanalysisoftyinganduntyingpracticeasreportedtotheOECDdatabases,toprovideanindepend-
6 The Ghana study was made possible by support to ODI under a research a grant from KfW for investigating the trade implications of aid instruments.
7 The Executive Summaries of the country studies are included in Annex D.
entreassessmentofprogresstowardsuntying.Thepreliminaryfindings,reportedinPhaseOne,havebeenupdatedtoincludedatafor2007.ThisanalysiswasundertakenafterdiscussionswithOECDstatisticiansaboutthecoverageandlimitationsofthedataasreportedbymembergovernments.
Second,thereisadesk-basedreviewofpubliclyavailabledocumentation,includingresearchandevaluationstudiesonimpactsoftyinganduntyingpractices.
Third,thereisanexploratorysurvey,basedoninterviewsanddocumentationmadeavailablebyagencies,ofhowfivedonors,Australia,Canada,Denmark,NorwayandSwitzerland,haverespondedtothe2001Recommendation.Thesecoun-triesarealsopartytodifferentregionalandbilateraltradeagreementsthatincludeprovisionsregardingpublicprocure-ment,whichcouldbeaninfluenceonuntyingpractices.
Fourth,therearesixcountrypartnerstudies,whichprovidethroughacombinationofstatisticalinformation,interviewsanddocumentation,furtherevidenceonuntyingpractices,includingchoiceofaidmodalitiesandorganisationalarrange-ments.Thesixpartnercountrieswerepurposivelyselectedtoberepresentativeoftherangeofaidrecipientsandalso,sinceuntyingpotentiallyhastrade,includingregional,implications,aspairswithineconomicandtradingregions–WestAfrica(ECOWAS),SouthernAfrica(SADC)andSEAsia(ASEAN)(seeChapter6,Table6.1formoredetail).8
Fifth,inthecasestudycountries,morein-depthinvestigationswereundertakenfor21developmentprojects,supportedlargelywithbilateraluntiedfunds.Theprojectswerepurpo-sivelyselectedtocovermanyDACmembers,includingthelargestdonors,andtofocusmostlyononesectortoprovidecomparabilitybetweenprojectcomponents.WaterandSani-tationSector(W&SS)waschosenasitincludesbothinfrastruc-tureandsocialdevelopmentaspects,andallprojectswereon-goingin2008-09,thusfacilitatingdatacollectionandtheavailabilityforinterviewofpersonnelfromtheimplementingagencies.Therationaleforfocusingoninvestmentprojectsissetoutbelow.
Project-type aid as an evaluation priority:Discussionswithdonorofficialsandreviewingtheavailabledocumentation,includingevaluationsduringPhaseOne,suggestedfourdiffer-entschematicmodels,reflectingtypesofaidandassociatedmodalities,wherethepotentialimpactofuntyingshouldbeconsideredseparately.ThisschemawasusedtodeterminecountrylevelevaluationprioritiesforPhaseTwo.
Model A: Discrete investmentswhichcanbecomparedwithinvestmentselsewhere,buthavetobetailoredtoaspecificsituation.Theseincludeinfrastructureprojects,suchaspower
8 The ToR originally envisaged at least four country studies. The proposal for Phase Two hoped to include six, including one in Latin America, provisionally Nicaragua, subsequently Honduras. This proved impractical because of time constraints and other unforeseen difficulties.
Untying aid: Is it working? 3
Chapter 1
stations,portsanddams,someindustrialunitse.g.fishingfa-cilitiesandtelecommunicationssystems.Multilaterallyfundedprojectsofferastandardforcomparison,andsoWorldBankprocurementwasrecognisedasprovidingacounterfactualtobilateralprocurement.IndividualprojectshavebeenrelativelylittleassessedfromaperspectiveofuntyingpracticesandweremadethemajorfocusofPhaseTwothroughasurveyof21projectsinthesixcountries.
Model B: Highly divisiblegoodsthatarewidelytradedoncompetitive,butnotnecessarilyperfectlycompetitivemarkets(mostobviouslyinthecaseoffoodaid).Extensivelyre-searchedandevaluated,cost-effectivenessanalysisdemon-stratesefficiencyofunrestrictedsourcinginprocurementbyadonor,itsagentorpartner.Therearealsoeffectivenessissuesintermsofselectionofappropriategoodsanddeliverytimes.AsfoodaidisexemptfromtheRecommendationdiscussionofthiscaseisconfinedtothestatisticaldonorpolicyandliteraturereviews.However,highlydivisibleandwidelytradedgoodssuchascementandpipingwererecommendedforcost-effectivenessanalysesinthecountrystudies.
Model C: Free standing technical cooperationbetweendonorandrecipientpartnercountryorganisations,whereforexamplethedevelopingcountryorganisationisdisadvantaged in terms of information aswellasinaccesstofunding.Thepartnershipsareseenasawayofovercomingtheseobstacles.Twinningar-rangementsbetweenpublicsectororganisationsandBusiness to Business(B2B)involvingprivatesectorenterprisesisbasedonthisthirdmodel.9AsfreestandingTCisoutsidetheRecom-mendationonlyafewTCprojectshavebeenincludedtoensurewiderdonorcoverageintheprojectsurvey.
Model D: Direct bilateral instruments (DBI)wherefundsaremadeavailabletotherecipientpartnercountrygovernment,usuallyasgeneralbudgetary(GBS)orsectoralbudgetary(SBS)orprogrammaticsupporttodisburseaccordingtothepart-ner’sprocedures.Sometimesthesearedescribedas‘basket’arrangementswhereseveraldonorscontributetoacommonprogrammaticfund.Fundsarealsopooledwithinaprojectframework(ModelB).DebtreliefshouldalsobeconsideredaswithinthiscategoryofDBI.Programmaticsupporttoapartnercountrycivilsocietyorganisationaccordingtothepartnerorganisation’sproceduresfollowsthismodelandmaybeimportantinsomecountrieswherethereareproblemsofgovernance.However,theremightnotbegeneralagreementonthislattercase,whichisusuallyoff-budget.TheuntyingstatusofDBIsisrelativelyunambiguousandGBSandSBShavebeensubjecttoinvestigations.Thereforetheircoverageinthisstudyislimitedtostatisticalandliteraturereviewsandcountrylevelsurveystodetermineincidence.
Methods of analysis:Afteracarefulreviewofthestatisticaldataitwasdecidedtoconfinetheanalysistosimpletabularand
9 The OECD study on promoting local and regional procurement identifies informa-tion and dialogue obstacles to local firms’ successful participation in procurement (OECD, 2008a).
graphicalpresentationsofactualvalues,proportions,distri-butionsandtrends,allbroadlysimilartothoseintheannualprogressreportsonuntying.Aneconometricanalysisquanti-fiesincrosssectionforrelationshipsbetweenuntyinganddonorandrecipientcharacteristicsandisreportedmorefullyinPhaseOneChapter3.Amoreambitiousquantitativeanalysisoftrends,asenvisagedintheToR,wasprecludedbydatathatwereincreasingincompleteasonegoesbackwardsintime.Respondingtothewidelyexpressedviewinbothdonorandrecipientcountriesthatinformaltyingpracticesarewidespread,asecondeconometricanalysisisundertakenoftherelationshipbetweenthebilateralaidandexportflowsofindividualdonorsduring2002-07toeachofthecasestudycountries.
Therecipientcountrypartnerstudiesprovidedanopportu-nitytoexplore,throughacombinationofstatisticalanalysisandqualitativeevidencefrominterviewswithstakehold-ers,theextentandinfluenceofuntyingonaidoutcomes.Againtheapproachiseclecticandlimitedbyalackofstrictlycomparablestatisticalreporting,forexampleontheuseofaidmodalitiesandlackofdataonprocurementpractices.Theinvestigationsintoaid-effectivenessanddevelopmentalimpactsarelargelybasedonacombinationofqualitativeevidencefromthe21projectcasestudieswithselectivecost-effectivenessanalyses.10Specialattentionwasgiventoprojectreviewsasawayofprovidingconcrete,case-specificevidence.TheliteraturereviewandinterviewevidencefromPhaseOnehadhighlightedtheproblemoftheabsenceofpreviousevaluativeevidenceontheimpactsofuntyingandthatinvestmentprojectswereprobablythemostworthwhileareafordetailedinvestigation.
Thereportisorganisedasfollows.InChapter2thereisabriefoverviewoftheinternationalrulesandagreementsconcern-ingthetyingstatusofODA.Chapter3providesastatisticalanalysisoftyinganduntyingpracticesofDACmembers,showinghowuntiedaidhasquicklycometopredominatewithinbilateralaid,buthowcontinuingdataproblemspre-cludeapreciseestimateoftheextentofuntying.InChapter4,thefindingsofasurveyofchangingdonorpoliciesandprac-ticesofthefivedonorsaresummarised.Chapter5presentsthefindingsofareviewoftheresearchliterature.Thisreviewconfirmsthatattentioninbothresearchandevaluationhasalmostentirelyfocusedoneffectsoftyingpracticesandverylimitedattentionhasbeengiventountyingpractices:hencetherationaleforthisstudy.11
10 A note of clarification is useful on the uses of qualitative data. When a survey is purposive, as in this case where six countries and 21 projects have been selected on grounds of representativeness, then the appropriate way to use such data is as if it were non-parametric, and so give equal weighting to each observation. We do that and draw attention to what we have done (full transparency), and so it is easy for anyone viewing the data to make an alternative calculation if they so wish. Otherwise the overall picture could be dominated or heavily biased by any country or project value that just happened to be relatively larger.
11 The lack of evidence on the consequences of untying is confirmed by the ‘Evalu-ation of the implementation of the Paris Declaration Phase One Synthesis Report’ (Wood et al., 2008).
Untying aid: Is it working?4
Chapter 1
Chapter6providesafactualreviewbasedoncountryexperi-enceofhowuntyingworksinpracticefromrecipientaswellasdonorperspectivesintermsofmodalitiesadopted,pro-curementpracticeandusers.Inresponsetothewidespreadscepticismaboutthegenuinenessofuntying,aneconometricanalysisexaminestherelationshipbetweendonorsandtheirexportstoeachofthecasestudycountries.InChapter7,projectlevelevidenceispresentedonthewaysuntiedaidisorganised,inputsaresourcedandtheoutcomesintermsofaidefficiency,effectivenesscriteriaanddevelopmentimpacts.Finally,Chapter8setsouttheconclusionsandrecommenda-tionsforincreasingtheshareofaidthatisfullyuntied.
Untying aid: Is it working? 5
Chapter 2
T hischapterprovidesabriefoverviewofthemainagreementsandrulesfordeterminingthetyingstatusof
aidandforregulatingthetyingpracticesofdonorcountries.
Theattemptstoreachinternationalagreementonrulescover-ingthetyingstatusofODAhavealonghistorybecauseoftwodistinctbutultimatelyoverlappingsetsofconcerns.First,therearethedevelopmentcooperationimplicationsoftyinganduntyingpracticesfortheefficiencyandeffectivenessofaid,aswellasburdensharingamongstdonorcountries.Second,therearetheexportcompetitionissues:thepoten-tialtradedistortingimplicationsofconcessionallendingandgrantaidtiedtotheexportofgoodsandservicesfromdonorcountries.AttheinternationallevelthefirstsetofconcernshasbeenafocusofdiscussionintheOECDDAC,whilethesecondhasbeenaddressedprimarilybytheParticipantstotheArrangementonOfficiallySupportedExportCredits.
Definitions of tying status:Theimportantcurrentlyoperativeagreementsarethoseofthe1987DACGuidingPrinciplesforAssociatedFinancingandTiedandPartiallyUntiedOfficialDevelopmentAssistance(OECDDAC,1987)summarisedinTable2.1.
Thecategoriesoftiedanduntiedaidare,atleastformally,rela-tivelyunambiguousfocusingonthekeyissueofwhetherthe
sourcingofgoodsandservicesiseitherrestrictedorfreefromrestrictionandopentosuppliersinotherdonorcountries.Thepartiallyuntiedcategoryismorecomplex.Itwasdevisedtotakeaccountofasituationinwhichthedonorwishestoallowlocalprocurementintherecipientcountryorpossiblyagroupofdevelopingcountries,forexampleinthesameregion.Donorsmayalsoagreetopermitprocurementfromwithineachother’smarketonabasisofreciprocity,aswellastodevelopingcountries.Conventionallysuchtransactionsarestillregardedaspartoftiedaid.However,inthisstudytheyarebothseparatelyidentifiedandalsogroupedwithuntiedaidasameasureofmovementsintyingstatustoallowprocurementwithintherecipientcountryandotherdevelopingcountries.
Reporting on tying status:theGuidingPrinciplesincludeanagreementtoreporttotheOECD’sDACaggregatetablesonthetyingstatusofaid.ReportingonTechnicalCooperation(TC)andadministrativecostswasspecificallymadevoluntary.However,reportingtotheOECDCreditorReportingSystem(CRS)onaprojectleveldoescontaininformationonthetyingstatusofmostODA.Apartfromtechnicalproblemsofensur-ingconsistentreporting,somedonorsinitiallychosenottoreportonthetyingstatusoftheirbilateralaidtotheCRS,andevenmoredidnotreportfullyontheirTCoradministrativecosts.ThecombinationofthesefactorshaslimitedtheabilityoftheOECDDACtomonitorpreciselyprogressonuntyingof
2. tHE intErnAtionAl rulEs And AgrEEMEnts ConCErning tHE tying stAtus of odA
Table 2.1 DAC definitions of tying status of ODA, other official flows and officially supported credits
TYING STATUS CATEGORY DEFINITION AND COVERAGE
UNTIED AID Loans and grants whose proceeds are fully and freely available to finance procurement from all OECD countries and substantially all developing countries.
PARTIALLY UNTIED AID Loans and grants which are tied, contractually or in effect, to procurement of goods and services from a restricted number of countries which must include substantially all developing countries and can include the donor country.
TIED AID All other loans and grants are classified as tied aid, whether they are tied formally or through informal arrangements.
Source: OECD DAC (1987).
Untying aid: Is it working?6
Chapter 2
aid.TheconsequentdataproblemsthathampermonitoringoftyingpracticesaredescribedmorefullyinSection3.1.
In2001DACmembersagreeduponaRecommendationtountiemostcategoriesofaidtoLDCs(OECDDAC,2001).TheRecommendationhastheusualexceptionsoffree-standingTCandfoodaid.ThresholdlevelsontheapplicationoftheRecommendationwerealsoinitiallysetatSDR700,000(SDR130,000inthecaseofinvestmentrelatedTC).Thesethresholdsweresubsequentlyremovedin2006(OECDDAC,2006b).In2008coverageoftheRecommendationwasextendedtoin-cludenon-LDChighlyindebtedcountries(HIPCs)andtherebyprovidedequaltreatmentforthewholeHIPCgroup(OECDDAC,2007).12TheDACalsourgeddonorstopromotegreaterlocalandregionalprocurementindevelopingcountriesandinvitednon-DACdonorstorespecttheprinciplesoftheRec-ommendation.
The2005ParisDeclarationincludesafurtherqualitativecom-mitmentwithoutanyspecifictarget(Indicator8)tocontinuetheprocessofuntying.The2008AccraAgendaforAction(AAA)onaideffectivenessreaffirmedthiscommitment,andcalleduponDACmemberstoelaborateindividualplansby2010tountietheiraidtothemaximumextentandtoimprovereportingonthe2001Recommendation(OECDDAC2009).
Table 2.2 Chronology of OECD agreements on tying statusof aid including credits that qualify as ODA
WHO WHAT WHEN
DAC Official flows split into:- ODA, and- Other Official Flows
1969
Introduction of 25% Grant Element (10% discount rate)
1972
Participants Tied Aid Disciplines- LDCs 50% concessionality- Others 35% concessionalityDifferentiated discount rate (cost of money)
1987
Participants/DAC New Measures in the Field of Tied Aid - No tied aid loans for commercially viable projects
1992
DAC Recommendation to Untie ODA to LDCs (exc. TC and Food Aid)
2001
Somespecificissuesforinvestigationarisefromthestepwiseprocessofcommitmenttowardsuntying.TowhatextenthastheRecommendation,includingcoverage,exemptionsandthresholds,beenaninfluenceondonorpoliciesandpractices?
12 Eight non-LDCs are part of the HIPC Initiative: Bolivia, Cameroon, Cote d’Ivoire, Ghana (case study country), Guyana, Honduras, Nicaragua, and Democratic Republic of Congo.
Incompletereporting,asthestatisticalreviewinChapter3indicates,seriouslyrestrictsanyattempttomonitordonorresponsestotheRecommendation.Promotinglocalandregionalperformanceisakeyissueinassessingtheimpactofuntying,asdiscussedbelowinChapter7.
Issuesoftyingstatusarealsodiscussedinotherinternationalfora.MultilateralagenciessuchastheWorldBankhaveagreedrulesregardingtheacceptabilityandeventuallynon-acceptabil-ityoftiedaidundertrustfundingarrangementsandsoforth.
Thepotentialfortradedistortionwheredonorsusedaidtoseekacompetitiveadvantagefortheirexportershasbeenrec-ognisedandaddressedinanumberofways.Forofficiallysup-portedexportcreditsOECDmembersenteredintoasequenceofagreementsthatsetoutcriteriafordefiningandrestrictingtheuseofcreditsasaformofexportcompetition.Thesecriteriaincludetheconcessionalitylevelandnotificationprocedures.TheprocessculminatedintheHelsinkiPackageofTiedAidDisciplinesin1991thatsetoutprinciplesforofficiallysupportedexportcreditsandtiedandpartiallyuntiedaid,complementedin1992bycorrespondingmeasuresagreedbytheDAC.Theseprohibitedtiedandpartiallyuntiedcreditsforricherdevelopingcountriesaswellas,onacase-by-casebasis,projectsthatarecommerciallyviableandforwhichcommercialfinancingwouldbeavailable.13Thepackagealsoreinforcedmonitoringproce-duresfortiedaidcredits.ThecriteriaagreedforcategorisingtiedcreditsasODAintroduceddifferentialtreatmentforuntiedloansqualifyingwith25%concessionality,tiedcredittoLDCsrequiringaminimumof50%andcreditstootherdevelopingcountriesaminimumof35%.Achronologyofthemainagree-mentswithintheOECDregardingthetyingstatusofaidandofficialcreditsaresummarisedinTable2.2.
Untied aidwouldincludeODAthatdirectly financesimportswherethesourceofprocurementisunrestricted.Inaddition,transactionsthatarenotdirectly financing imports(NDFI)areas-sumedtobewhollyfungibleandthereforefreefromanytradedistortingimplications.Theseformsofaidincludebudgetaryandbalanceofpaymentssupport(theprovisionoffreelyusableforeignexchange),local cost financingand debt relief (includ-ingrefinancingandrescheduling).Inaddition,contributionstoNGOsandofficial funds in support of, or intended for, direct equity investmentarealsoconventionallyregardedasuntiedornon-distorting.Theseareinpractice‘greyareas’wherethereissomeelementofimplicittying(Chapter4).
Internationalconcernsaboutpotentialtradedistortingeffectsoftyingpracticesandtheintroductionofregulatorypracticestocombatthesehavehistoricallyfocusedontwobroadcat-egoriesofaidwithrestrictiveconditions.Theseare:• transactionsthatdirectly finance imports(DFI),which
specifytheprocurementwithaidfundsofspecificgoodsandservices;and
13 See OECD (1991), OECD (2008a) and, for an account of the process that led to the Helsinki Accord, see Ray (1995). The transparency criteria have been extended since 2005 on a voluntarily basis to untied aid credits that qualify as ODA.
Untying aid: Is it working? 7
Chapter 2
• aid-in-kindwheregoodsarepurchasedinthedonorcountryandwhicharereadyforconsumptionoruseinarrivalintherecipientcountry.Conventionally,asubstan-tialpartofbilateralemergencyaid/humanitarianreliefhasbeensourcedinthisway.
ThereisapotentialoverlapofthevoluntaryOECDagreementsandthetreatyobligationsunderWorldTradeOrganisation(WTO)agreementsandproceduresfordisputes.However,thishasbeenineffectresolvedbyexemptingODAfromtheAgreementonGovernmentProcurementinGATT(1979)andalsointheWTOAgreementonSubsidiesandCountervailingMeasures(ASCM),includingtiedaid,whichrecognisestheOECDarrangementasanexemptiontotherules.14
Agriculturehasthroughoutbeentreatedseparately,asreflect-edintheexemptionsforfoodaidfromthe2001Recommen-dationonuntyingofODA,andtheHelsinkidisciplines.Thereareparallelandseparateproceduresforavoiding,oratleastminimising,commercialtradedisplacementthroughthesup-plyoffoodaidthroughtheFAOPrinciplesonSurplusDisposal.FoodaidtyingpracticeshavealsobeenanissueintheexportcompetitionpillaroftheagriculturalnegotiationsfortheDohaDevelopmentRound.TheDACdefinitionsoftyingstatushavebeenadoptedindraftmodalitiesforfoodaid(WTO,2008).
14 See Annex 1, Footnote K of the WTO Agreement (WTO, 1995).
Untying aid: Is it working?8
Chapter 3
3.1 An incomplete picture of untying and tying status of ODA
T hischapterincludes,first,astatisticalanalysisoftrendsinthetyingstatusofODAofDACdonors,largelybasedon
thereportingbydonorstotheOECD.ThisiscomplementedbyamoredetailedreviewofthetyingstatusofODAin2007,themostrecentyearforwhichOECDCRSdataareavailable.15DonorperformanceiscomparedtotheRecommendationandrelatedtargetsagreedbytheDAC.Becauseoftheseverelimitationsoftheavailabledata,especiallypriortotheRecommendation,thisstatisticalexplorationisrestricted:trendscannotbemeasuredwithanyprecisionandcautionisrequiredinmakinginferencesabouttheextenttowhichformalreportingofuntyingisassociatedwithchangesinactualpractices.
15 The statistical information used in this chapter is based on data downloaded from the CRS database on 29 September 2009. A few modifications have been carried out to the raw data (see Annex B for details).
ThedataontyingstatusofODAandotherofficialflows,asreportedtotheOECD,arequiteincomplete,asiswellknownwithintheDAC,butperhapslesswellunderstoodoutsidedo-norcircles(OECD,2008a).Nevertheless,thereportingbyDACmembersonthetyingstatusoftheiraidisimproving,soitispossibletobemoreconfidentindescribingandanalysingthecurrentstatus,asthisconformswithOECDdefinitions(OECDDAC,1987).However,trendsintyingstatushavetobeinferredwithsomecaution,firstbecauseofthesubstantialreductioninnon-reportingasshowninTable3.1(thenon-reportingdeclinedfrom38%during1999-2001toonly7%in2005-07).Secondly,thereareexamplesofdonorsreclassifyingthetyingstatusoftheiraidtobemorefullyconsistentwiththepracticeofotherdonors.16Thirdly,asdiscussedabove,
16 For example, CIDA used to report bilateral aid grants to Canadian based NGOs as ‘tied’ until 2005, when these were reclassified as ‘untied’ to be consistent with the usual practice of other DAC members. There is really no reason why aid to a national NGO should be treated differently from aid to a national enterprise, so should be considered as a ‘grey’ area in examining tying practices.
Table 3.1 DAC donor countries: tying status of bilateral ODA to LDCs and non-LDCs in 1999-2001 and 2005-07 (percentages based on three year averages)1
Period Recipient country grouping Untied (%) Partially tied (%) Tied (%) Not reported (%) Total
1999-2001 LDCs 57 3.5 8 32 100
Non-LDCs 49 2.0 9 40 100
All DCs 51 2.3 9 38 100
All DCs (including EC) 46 7.5 8 39 100
2005-2007 LDCs 83 0.3 13 4 100
Non-LDCs 70 0.8 22 7 100
All DCs 73 0.7 20 6 100
All DCs (including EC) 65 12.1 18 5 100
Source: OECD Creditor Reporting System (CRS) database.Notes: 1. Data are based on commitments in current US$.
3. A stAtistiCAl AnAlysis of tHE untying/tying stAtus of odA
Untying aid: Is it working? 9
Chapter 3
thereareinconsistenciesinthestatusofaidreportedaspartiallyuntied.
Thecurrentextentoftheincompleteinformationandnon-reportingofthetyingstatusofODAisindicatedinFigure3.1forindividualDACdonors.ThemostwidelyavailableannualassessmentoftyingstatushasbeentheannualDevelopmentCooperationReport,whichisbasedonaggregatecommit-mentsreportedtotheDAC,andexcludestechnicalcoopera-tion(TC),onwhichdonorreportingisoptional,anddonorad-ministrativecosts.Around76%ofbilateralODAwasreportedasuntiedfor2007,includingdonoradministrativecosts17.
ThisstudyreviewsdonorperformanceusingdatafromtheOECDCreditorReportingSystem(CRS),towhichdonorsreporttheircommitmentsonanactivitybasis.TheCRShasamorecomprehensivecoverageoftyingstatusthantheDACtables,butstillremainsincomplete.FormostDACmembersthecoverageiscompleteorclosetocomplete.Afewdonors(e.g.FranceandSwitzerland)experiencedifficultiesinreport-ingaidattheactivitylevelprovidedbysomegovernmentdepartmentsandlocalgovernments.EvensothereportingoftyingstatustotheCRSfor2007isover90%ofbilateralODA,incontrasttoaround60%adecadeago.Howeverpreviousunder-reportingmeansthatatthetotalDAClevel,trendsinuntyingcanbeinferredonlyveryapproximatelysince1987,whencurrentdefinitionsoftyingstatuswereadopted(Figure3.2).
Thestatisticalcoverageofthestudyincludesthebilateralaidof22membersoftheDAC,aslistedinTable3.2,andalso
17 Slight differences with figures reported in the 2009 DAC Progress Report are explained by our decision to include donor administrative costs; moreover, a modi-fication, regarding the tying status of non reported Budget Support, was applied to OECD CRS data (See Annex B for further details on data modifications).
insomeinstances,forpurposesofcomparison,EuropeanUnion(EU)aidasorganisedbytheEuropeanCommission.EUaidismultilateral,butitstyingstatusisstillwithreferencetotheSingleEuropeanMarket,andstakeholders,includingofficialsofsomeDACmembercountriesandrecipientpartnercountries,madereferencetotheEUintheirdiscussionoftyingpractices.Assomedonorsreportthatelementsoftheiradmin-istrativecostshavebeenuntied,sothiscomponentofbilateralaidisalsoincludedinthestatisticalanalysis.
Source: OECD Creditor Reporting System (CRS) database.Notes: 1. Data are based on commitments in current US$. 2. Only bilateral donors who report less then 1% (or 0%) tied or not reported ODA are not included.
United States
Germany
Spain
Netherlands
France
Canada
Japan
Austria
Italy
0% 10% 20% 30% 40% 50% 60%
Japan
Germany
Canada
Portugal
Sweden
Italy
Greece
Spain
Denmark
France
Belgium
Finland
0% 10% 20% 30% 40% 50%
(A) Donor share of tied ODA (B) Donor share of not reported ODA
Figure 3.1 Donor share of total bilateral ‘tied/not reported’ ODA in 20071
Untying aid: Is it working?10
Chapter 3
Table 3.2 DAC donor countries: tying status of bilateral ODA in 20071 (including donor administrative costs)
Donor
Tying status as % of donor ODA Tying status as % of bilateral ODA
Bilateral ODA2 (US$
million)
Untied Partially tied
Tied Not Reported
Untied Partially tied
Tied Not Reported
Australia 1,710 96 0 4 0 2.4 0.0 0.4 0.0
Austria 1,382 76 0 24 0 1.5 0.0 2.0 0.0
Belgium 1,587 92 0 4 4 2.1 0.0 0.4 1.1
Canada 3,338 69 1 17 13 3.4 12.8 3.6 8.2
Denmark 1,481 88 0 4 8 1.9 0.0 0.4 2.4
Finland 661 85 0 7 8 0.8 0.0 0.3 1.0
France 8,223 90 0 9 1 10.8 0.1 4.6 1.7
Germany 9,644 73 0 20 6 10.3 0.0 12.1 12.2
Greece 248 13 2 8 77 0.0 2.1 0.1 3.7
Ireland 832 95 5 0 0 1.2 14.7 0.0 0.0
Italy 1,440 52 13 21 13 1.1 68.4 1.9 3.7
Japan 12,503 75 0 4 21 13.8 0.0 3.2 50.2
Luxembourg 253 100 0 0 0 0.4 0.0 0.0 0.0
Netherlands 4,800 78 0 22 0 5.5 0.0 6.6 0.0
New Zealand 279 85 2 13 0 0.3 2.0 0.2 0.0
Norway 2,898 100 0 0 0 4.2 0.0 0.0 0.0
Portugal 605 38 0 8 54 0.3 0.0 0.3 6.4
Spain 3,778 67 0 29 4 3.7 0.0 6.8 3.2
Sweden 2,324 86 0 0 14 2.9 0.0 0.0 6.2
Switzerland 1,524 98 0 2 0 2.2 0.0 0.2 0.0
UK 5,712 100 0 0 0 8.4 0.0 0.0 0.0
USA 24,724 63 0 37 0 22.7 0.0 56.8 0.0
EC 13,373 0 100 0 0 - - - -
All donors (excluding EC)
89,945 76 0 18 6 100.0 100.0 100.0 100.0
All donors (including EC)
103,318 66 13 16 5
Source: OECD CRS database.Notes: 1. Data are based on commitments in current US$ millions. 2. There seems to be a misreporting problem with EC commitments: the tying status should be reported on a commitment basis however the amount of aid from the EC for which the tying status is
reported exceeds the commitments reported by EC.
Untying aid: Is it working? 11
Chapter 3
3.2 Tying status: trends and recent donor performance
3.2.1 The continuing trend towards the untying of ODATheDACdonorshavemovedfromalargelytiedregimeforbilateralODAtoonethatisuntied.ODAreportedasuntiedrosefromaround40%inthemid80stoover70%in2005-07(Figure3.2).Thereiscontinuingprogresswithaidreportedasuntiedrisingto76%in2007.
Acomparisonofthethreeyearsimmediatelypriortothe2001Recommendationandthemostrecentperiodforwhichdataareavailableindicatesthatthereportedincreaseinuntyingispartly‘real’,partlyreflectsachangeindonoractionsandpartlyimprovedreporting(Table3.1).TheportionofODAtoLDCsreportedasuntied,thefocusoftheRecommendation,increasedfrom57%to83%andfrom49%to70%fornon-LDCs.Therewasanapparentincreaseintyingassociatedwiththesharpfallinnon-reportingfrom32%to4%.ThesetrendssuggestthattheRecommendationhasbeenassociatedwithasignificantincreaseinuntiedaidtoLDCsandtodevelopingcountriesmoregenerally.
Thefullextentofthechangeisclearerinfocusinginmoredetailontheoverallandindividualdonorpictureanddifferentformsofaidfor2007.Unfortunately,suchwasthepreviousextentofnon-reporting,itisnotfeltthatmeaningfulde-tailedcomparisonscanbemadeatdonorcommunitylevelabouttypesandformsofaidfromtheOECDdataalone.Such
temporalcomparisonsmaybepossible,drawingonarangeofqualitativeandquantitativeevidenceforindividualdonors,asintheDACPeerReviews,otherrobustbutnotstrictlycom-parabledatasuchasWFPonfoodaid,andintheexploratorysurveyofuntyingpoliciesoffivedonorsinChapter4.
3.2.2 Tying status of ODA in 2007: a cross-sectional analysisTheprogresstowardsuntyingisindicatedinTable3.2:threequartersofallbilateralaidwasreportedasuntied(76%),18%astiedand6%notreported.Eightmembersreported90%ormoreuntiedaid.CanadathathaduntilrecentlyprovidedahighproportionoftiedaidandtheUSAthathadalmostentirelytiedaidhavebothreportedsubstantialprogress.Toconclude,theDACdonorshavewithfewexceptionsmovedtowardsalargelyuntiedframeworkforprovidingbilateralaid.
3.3 Donors responding to the 2001 Recommendation
3.3.1 Targets exceededInordertoimproveeffortsharingamongdonors,theRecom-mendationsetoutreviewmechanismsthatwouldassessmembers’effortsagainsta60%benchmarkforuntyingaidtoLDCs.By2007thishasbeensubstantiallyexceededbyalmostallmembers(exceptItaly)(Table3.3).
1984-1986 1987-1989 1990-1992 1993-1995 1996-1998 1999-2001 2002-2004 2005-2007
100
90
80
70
60
50
40
30
20
CRS data
DAC data
Figure 3.2 Share of total bilateral aid1 that is untied, three year averages (1984-2007): DAC and CRS data comparison2
Source: OECD DAC and CRS databases. Notes: 1. Data are based on commitments in current US$. 2. The two figures are not strictly comparable as DAC data (Table 7b of DAC online dataset) do not include administrative costs and TC. ‘Table DAC 7b is used to report the tying status of bilateral ODA
commitments. Members have agreed that administrative costs and technical co-operation expenditure should be disregarded in assessing the percentages of tied, partially untied and untied aid. These items should therefore not be included in the data reported in this Table.’ (OECD, 2008g)
Untying aid: Is it working?12
Chapter 3
TheextentofuntyingofbilateralaidtoLDCs(83%)issubstan-tiallygreaterthanfornon-LDCs(70%)andoverallwas76%in2007(Table3.2).If(untied)multilateralaidistakenintoaccount,then83%ofallODAwasuntied.ThebroadconclusionisthatformallyDACmembers’ODAtotheLDCsisoverwhelminglyuntied.EvendonorssuchasAustraliaandCanadathathadalargeproportionoftiedaidhavemoveddecisivelytowardsuntying(seeChapter4below).TheUSAisthemajordonorthatcontinuestoformallytie,37%in2007.However,throughdevelopmentssuchastheMillenniumChallengeCorporation
thatmakesavailableuntiedbilateralgrants,therehasbeenasubstantialopeningofaidtointernationalsourcing.18Thereisstillsubstantialscopeforfurtherprogressonformaluntyinginrespondingtothe2001Recommendationandmorebroadly,asindicatedinTable3.3.Thereisstillalackoftransparencyre-gardingsomedonorsnon-reporting;forinstance,asFigure3.1shows,morethan50%ofnon-reportedaidisbyJapan.
18 For example, all Millennium Challenge Corporation compact assistance, which accounted for about 6% of total US ODA in 2006, is untied and open to competitive international bidding.
Table 3.3 DAC donor countries: untied aid as percent of donor ODA in 20071
Donor Untied aid Untied aid ratio2
as % of donor aid to all DCs
as % of donor aid to LDCs
as % of donor aid to non-LDCs
Luxembourg 100.0 100.0 100.0 100.0
UK 100.0 100.0 100.0 100.0
Norway 99.9 100.0 99.9 99.9
Switzerland 98.2 96.5 98.8 98.4
Australia 95.8 97.6 95.1 96.7
Ireland 95.1 100.0 89.2 96.6
Belgium 92.3 99.7 86.6 94.4
France 89.8 87.9 90.3 92.8
Denmark 87.6 94.0 83.5 92.2
Sweden 86.3 97.0 83.1 91.6
Finland 84.9 97.5 75.3 90.7
New Zealand 84.7 81.3 86.3 87.8
Netherlands 77.8 98.9 70.5 85.6
Austria 76.1 90.0 75.4 82.5
Japan 75.2 86.3 72.5 77.5
Germany 73.2 80.1 72.0 81.9
Canada 68.8 84.9 59.5 75.1
Spain 66.5 75.7 65.3 77.6
USA 62.9 74.8 58.5 66.8
Italy 52.2 52.5 52.2 83.5
Portugal 38.2 91.1 23.6 64.2
Greece 12.6 86.7 3.5 56.5
All donors (excluding EC) 76.0 86.0 72.8 82.1
Source: OECD CRS database.Notes: 1. Data are based on commitments in current US$ millions. 2. The untied aid ratio is the sum of: the percent of bilateral aid to LDCs reported as untied (to the CRS) multiplied by bilateral ODA (as reported to the DAC) and DAC multilateral ODA, divided by DAC
total ODA.
Untying aid: Is it working? 13
Chapter 3
3.3.2 Aid instruments and types of aidDifferentialprogresstowardsuntyingmightbeexpectedforspecialaidinstrumentsandtypesofaid.Historically,loansandnon-grantformsofaidwereassociatedwithtyingpractices,involvingelementsofexportpromotion.Theproportionofloansuntiedin2007(97%)wassubstantiallyhigherthatforgrantaid(86%)(Table3.4),andhasincreasedconsiderablysince2006,whenonly48%wasuntied,asnotedinthePhaseOnereport.Thescaleofbilateralloanaidhas,however,alsocontractedsharplysincethemid1990s.Twoexplanatoryfac-torsinthecontractioninloanODAandthemovetountyingareconcernsaboutindebtednessandanarrowingofcommer-ciallyrelatedlendingopportunitiesundertheHelsinkiagree-ment.Mixedcreditarrangementscontinuetobealimitedareaoftyingpracticesforafewdonors,includingtheUSAand
Denmark,asconfirmedbyreportstotheOECDCommitteeonExportCredits(OECDDAC,2008a:Table1).Tiedfoodaidloans,whichareexcludedfromtheOECDreportingonexportcreditsarestillprovided,butonamuchreducedscale(WFP,2008).
TCisexplicitlyexcludedfromthe2001RecommendationandreportingtotheDAContyingstatusisvoluntarythoughrecommended.Evenso,morethan71%ofaidasTCorinclud-ingaTCcomponentwasreportedasuntiedin2006(Table3.5).However,thehighproportionofaidforwhichapossibleTCcomponentisnotreported(61%)putsthisfigureindoubt.Amoredetailedreviewofdonorpractices(SeeChapter4)confirmseffortstountiemanyformsofTC.
Table 3.4 DAC donor countries: tying status of bilateral ODA to LDCs according to grant/non-grant status in 20071
(US$ million) (%)
Bilateral ODA Share of bilateral ODA
Untied Partially tied
Tied Not Reported
Share of untied bilateral ODA
ODA Grants 21,181 96.6 85.7 0.2 11.5 2.6 96.2
ODA Loans 712 3.2 96.9 0.7 2.4 0.0 3.7
Equity Investment 23 0.1 100.0 0.0 0.0 0.0 0.1
Total ODA 21,916 100.0 86.0 0.3 11.2 2.5 100.0
Total ODA (including EC) 25,454 74.1 14.1 9.6 2.2
Source: OECD CRS database.Notes: 1. Data are based on commitments in US$ millions.
Table 3.5 DAC donor countries: tying status of bilateral ODA to LDCs according to tech nical cooperationcomponent in 20071
(US$ million) (%)
Bilateral ODA Share of bilateral ODA
Untied Partially tied
Tied Not Reported
Share of untied bilateral ODA
Without TC component 4,589 20.9 96.6 0.6 2.7 0.0 23.5
With some TC component 183 0.8 91.4 3.0 5.6 0.0 0.9
Wholly TC2 3,754 17.1 70.4 0.1 16.6 12.9 14.0
TC component not reported 13,390 61.1 86.7 0.2 12.6 0.5 61.6
Total bilateral ODA (excluding EC) 21,916 100.0 86.0 0.3 11.2 2.5 100.0
Total bilateral ODA (including EC) 25,454 74.1 14.1 9.6 2.2
Source: OECD CRS database.Notes: 1. Data are based on commitments in US$ millions. 2. Projects where TC is the only component.
Untying aid: Is it working?14
Chapter 3
Asectoral classificationofbilateralODAfor2007indicatesareaswhereuntyingismoreextensive,aswellastypesofaidinwhichformaltyinghasbeenretainedtoagreaterextent(Table3.6andB.1).
Aswouldbeexpected,theextentofuntyingisgreatestfordebtrelief(100%)andgeneralbudgetarysupport(99.3%).Incomparing2006,asreportedinPhaseOne,with2007inTable3.6,therehasbeenadeclinefrom23%ofbilateralODAin2006to13%in2007intheseformsofdirectassistancetogov-ernments.Meanwhiletheoverallproportionofaidreportedasuntiedhasincreasedby2%yearonyear,underliningthemoregeneralmovetowardsuntying.Itisalsointerestingtonotethat20%ofdonoradministrativecostsarereportedin2007asuntied,althoughthiscategoryisexemptedfromtheRecom-mendation.
Thehighlevelofnon-reportinginsectorsinwhichtyingofprojectaidmightbeexpected,infrastructure,transport,in-
dustryandenergy,indicatethatcautionisrequiredinmakinginferencesfromstatisticaldataaboutuntyingpracticesatanaggregateorindividualdonorlevel.
Food aidisexemptfromthe2001Recommendation.ItisalsodifficulttodeterminethefullextentofuntyingfromOECDCRSforaformofaidthatisreportedaspartofatleastthreesectors,includingemergencyassistance(21%tying),develop-mentfoodaid(53%tying)andcommodityaid(tyingstatusalmostentirelyunreported).TheOECD(2006)studyfoundwidespreadmovestountying,withtheexceptionoftheUnitedStates.TheWFPINTERFAIS,whichreportsonfoodaiddeliveriesinphysicaltermsastonnagesofcommodities,pro-videsameasureofprogressonuntying(Table3.7).Totalcere-alsfoodaiddeclinedby46%from9.5milliontonnesin2001to5.1milliontonnesin2007.Commoditiessourcedindonorcountries(directtransactions)alsodeclinedsharplyby65%,from8.5milliontonnesto2.9milliontonnes.Incontrast,theshareofuntiedandpartiallyuntiedcerealsaidsourcedinre-
Table 3.6 DAC donor countries: bilateral ODA by consolidated sectors in 2007: tying status as percent of sector ODA1.
Consolidated Sector (US$ million)
(%)
Bilateral ODA Share of bilateral
ODA
Untied Partially tied
Tied Not Reported
Share of untied bilat-
eral ODA
Budget Support 2,769 3.1 99.3 0.0 0.7 0.0 4.0
Debt 9,010 10.0 100.0 0.0 0.0 0.0 13.2
Economic Infrastructure 11,131 12.4 76.0 0.1 21.2 2.6 12.4
Production Sectors: - Agriculture, Forestry & Fishing- Other
4,2661,337
4.71.5
86.169.0
0.40.0
8.2
16.7
5.3
14.2
5.41.3
Social Infrastructure & Services 26,850 29.9 77.6 0.2 17.5 4.7 30.5
Multisector 6,313 7.0 82.6 1.1 12.5 3.8 7.6
Government & Civil Society 11,287 12.5 71.0 0.3 25.8 2.9 11.7
NGO Support 799 0.9 84.4 3.0 1.5 11.1 1.0
Commodity Aid / Program Assistance 278 0.3 1.2 0.0 0.2 98.6 0.0
Emergency Assistance & Reconstruction 7,201 8.0 78.9 0.2 20.9 0.0 8.3
Food Aid 1,020 1.1 46.9 0.0 53.0 0.0 0.7
Refugees in Donor Countries 1,968 2.2 59.2 0.0 31.9 8.9 1.7
Donor Administrative Costs 4,357 4.8 20.4 0.9 46.3 32.4 1.3
Unspecified 1,360 1.5 44.6 0.5 8.0 47.0 0.9
Total (excluding EC) 89,945 100.0 76.0 0.3 18.0 5.7 100.0
Total (including EC) 103,318 66.2 13.2 15.6 5.0
Source: OECD CRS database.Notes: 1. Data are based on commitments in current US$ millions.
Untying aid: Is it working? 15
Chapter 3
cipientcountries(localpurchases)andinthird,almostentirelydevelopingcountries(triangulartransactions)increasedfrom10%to42%andthequantitydeliveredalsoroseby121%fromunder1milliontonnesto2.1milliontonnes.19Thesedevelopmentsarestriking:inaperiodofincreasinglytightglobalmarketsandrisingprices,boththeshareandreallevelsofuntiedandpartiallyuntiedfoodaidsourcedindevelopingcountriesincreasedstrongly.
3.3.3 Effort-sharingTherearewidedifferencesinaideffort-sharingamongstDACmembers.Thisisreflected,forexample,inthemostwidelycitedindicator,thepercentageshareofODAinGrossNationalIncome(GNI)inrelationtotheUNtargetof0.7%ofGNI.Morespecificallytheimplicationsofadoptingthe2001Recom-mendationcanbeexpressedintermsoftheproportionofaiddirectlyaffected.TheDAChasadoptedtwoindicatorsandrelatedtargetsforeffort-sharing.ThefirstistheshareofuntiedODAintotalaidtoLDCs,withastraightforwardtargetof60%.Asshownabovealmostalldonorshavereachedthistargetwithinsixyears,includingthosethathistoricallyprovidedaidonalargelytiedbasis.ThesecondindicatoristheratioofuntiedbilateralLDCODAandmultilateralLDCODAwithGNI.Atargetequivalentto0.04%ofGNIwasagreedandisreportedtohavebeenexceeded(OECDDAC,2009).However,thisindicatorissomewhatopaque.Donorsmaychoosetodisproportionatelysupportpartsofmultilateralaid,includingregionalfinancialinstitutionsthatproviderelativelymoreorlessassistancetoLDCs.
InconsideringoveralleffortsharingintheuntyingofODA,andnotjusttheaidtoLDCs,theproportionofaidchannelledmultilaterallyisasignificantfactor.Sotoprovideanindicatorofuntyingoverallan‘Untiedaidratio’isestimated,whichistheproportionofuntiedbilateralaidplusthataidchannelled
19 WFP reports that over 90% of local and triangular purchases are in developing countries. The changes between 2001 and 2007 in ‘non-cereals’ food aid deliveries are broadly similar to those for cereals, but the heterogeneity of this category, which includes pulses, vegetable oils, sugar and dairy products, makes interpretation of statistical data more difficult.
multilaterallyexpressedasashareoftotalODA(Table3.3).ThescatterplotinFigure3.3showsthattheuntiedaidratioispositivelycorrelatedwitheffort-sharing,theoverallODA/GNI–thatisthosegivingrelativelymoreaidasapercentageofGNI–appearalsotoproviderelativelymoreuntiedaid.Thisconjecturethatthereisapositiverelationshipbetweenunty-ingandaideffort-sharingwasconfirmedinaneconometricanalysisundertakeninPhaseOneofthisstudy.20However,asdonorswhoseODAisarelativelylowerpercentageofGNI,havealmostfullyuntiedtheiraid,forexampleAustralia,sothisrelationshipisbecominglesspronounced.
20 The analysis for the bilateral aid of 22 donors in 2006 found that approximately two thirds of the inter-donor variations in untying were explained statistically by the following influences in order of significance. First, the share of aid received by LDCs is highly significant, suggesting that the Recommendation is an influence on the ex-tent to which aid is untied. Second, the share of food aid in bilateral ODA is negatively and significantly associated with untying. This confirms our expectation that a high share of bilateral food aid is associated with tying practices. Third, a higher share of TC and fourth a lower GNI per capita are associated with a lower share of untied aid. (Clay et al., 2008, Section 3.5).
Table 3.7 Global cereals food aid by source of supply in 2001 and 2007
Source ‘000 tonnes in 2001(% of total)
‘000 tonnes in 2007(% of total)
Change 2001- 2007 (%)
Developing countries1
(Untied/partially untied)967(10)
2139(42)
+121
Direct transfers2
(Tied)8537
(90)2949
(58) - 65
Total 9499(100)
5088(100)
- 46
Source: WFP INTERFAIS.Notes: 1. Includes local purchases in recipient countries and triangular purchases in third countries. 2. Sourced in the donor country or in the case of the EU within the single European market.
Untying aid: Is it working?16
Chapter 3
3.3.4 Additionality of tied aid?Aconcernoftenvoicedbythosedefendingtyingpracticesorcautioningagainstchangeisthatuntyingcouldresultinalossofdomesticsupportforaidoverall,orhavediversion-aryeffectsawayfromLDCstootherpartnercountries.ThesepossibilitieshavebeenconsideredinthereportingtotheDACorimplementationofthe2001Recommendation.Theannualreviewssuggestmorepositivedevelopmentsandnoevidencetosupporttheseconcerns.AbsolutelevelsofODAtoLDCshadalmostdoubledby2007sincethebaseperiodof1999-2001.TheshareofLDCsintotalODAwasalso5%higher,andevenhigher(13%)whenthesomewhattempo-rarysurgesindebtreliefduring2005-06areexcluded(OECDDAC,2008a).Asthehistoricaldataonuntyingaresoweakbecauseofthehighlevelofnon-reportinginearlieryears,itisunlikelythatanymoresophisticatedstatisticalinves-tigationwillbeattemptedthatmightoverturnthissimplerefutationoftheadditionalityargumentbasedontrendsinaid.Furthermore,theprocessofuntyingiscontinuingbeyondtheRecommendation–intermsofaidtonon-LDCs,TCandfoodaid.
Thechangingpatternofsourcingwithinsteeplydecliningfoodaidlevelscastsfurtherdoubtontheadditionalityargu-mentfortying.Thereallevelsofuntiedfoodaidmorethandoubledbetween2001and2007inaperiodofincreasingly
tightmarketsandhigherglobalprices.Incontrast,levelsoftiedaidsourcedindonorcountriestumbled(Table3.7).Atadonorlevel,thosewhohadalreadyshiftedtodevelop-ingcountrysourcingandremovedresidualtyingpracticessustainedtheirreallevelsoffoodaid,whilstflowsfromthoseprovidingdomesticallysourcedcommoditiessharplycon-tracted(WFP,2008).Iftherewereanyadditionalityoftiedfoodaid,thenthisphenomenonwouldbeunstable,contingentontransientsurpluses(FAO,2006).
Takentogetherthefindingsofthesestatisticalanalysessug-gestthattheRecommendationhashadasignificantimpact,andsoextendingthecoverageoftheRecommendationintermsofcountryandformsofaidcurrentlyexemptwouldleadtofurtheruntyingofaidasenvisagedintheParisDeclara-tion.
3.3.5 Formal untying and de facto tying: saying it does not make it soTheformaluntyingofaiddoesnotnecessarilyhaveadevelop-mentalortradeimpact,unlessitisaccompaniedbyachangeinactualpractice.Herethereisaseriousissue,theactualinformation,whichonlysomedonorschoosetosubmitaboutsomeoftheiraidtotheOECDaboutcontractsthatareopentocompetitivetender,indicatesthat62%ofcontractsand60%byvaluewereawardedwithinthedonorcountry(Table3.8).In
Source: OECD CRS and DAC database.Notes: 1. Data are based on commitments in US$ millions.
GRC
POR
USA
JPN
ITA
NzL
AUSSUI
Uk
FRA
FIN
AUT
IRL
NL
DkSWE
NORLUx
BEL
CAN
ESP
DEU
60
70
80
90
100
.2 .4 .6 .8 1
Untie
d ai
d ra
tio
ODA/GNI
Figure 3.3 Relationship between share of untied ODA (untied aid ratio) and effort-sharing (ODA/GNI) in 2007
Untying aid: Is it working? 17
Chapter 3
theabsenceofcontraryevidence,itwouldseemreasonabletoinferthatanevenhigherproportionofallcontracts,includ-ingthosethatarenotreported,areawardedwithinthedonorcountry.Thereisthusapresumptivegapbetweentheformaldeclaratoryuntyingofbilateralaidandactualpractice,whichimpliesaconsiderableelementofintendedorunintendedde
Table 3.8 DAC donors: geographical distribution of contracts awarded in 2007
Donor Total contracts awarded
Within donor country Other OECD and non-DAC
countries
Developing countries
(excl. LDCs)
LDCs
No. of con-
tracts
US$ mn
No. of con-
tracts
US$ mn
as % of total contracts
awarded (n)
as % total con-tracts awarded
(value)
No. of con-
tracts
US$ mn
No. of con-
tracts
US$ mn
No. of con-
tracts
US$ mn
Australia 12 100.5 10 96.3 83.3 95.8 0 0.0 1 1.4 1 2.8
Austria (2) .. .. .. .. .. .. .. .. .. .. .. ..
Belgium 5 18.2 1 2.8 20.0 15.4 0 0.0 2 2.1 2 13.3
Canada 1 16.1 1 16.1 100.0 100.0 0 0.0 0 0.0 0 0.0
Denmark (3) 6 6.5 5 6.5 83.3 100.0 1 0.0 0 0.0 0 0.0
Finland 3 5.5 3 5.5 100.0 100.0 0 0.0 0 0.0 0 0.0
France 66 350.3 16 57.4 24.2 16.4 7 119.7 32 134.3 11 38.9
Germany 52 171.8 29 75.9 55.8 44.2 3 6.7 4 6.4 16 82.9
Greece (4) 0 0.0 .. .. .. .. .. .. .. .. .. ..
Ireland (2) .. .. .. .. .. .. .. .. .. .. .. ..
Italy (5) 0 0.0 .. .. .. .. .. .. .. .. .. ..
Japan (6) 23 .. 20 .. 87.0 .. 2 .. 0 .. 1 ..
Luxembourg .. .. .. .. .. .. .. .. .. .. .. ..
Netherlands 1 4.0 1 4.0 100.0 100.0 0 0.0 0 0.0 0 0.0
New Zealand 7 10.5 6 9.1 85.7 86.7 1 1.4 0 0.0 0 0.0
Norway 1 6.0 0 0.0 0.0 0.0 0 0.0 1 6.0 0 0.0
Portugal 1 1.4 0 0.0 0.0 0.0 0 0.0 1 1.4 0 0.0
Spain (4) 0 0.0 .. .. .. .. .. .. .. .. .. ..
Sweden (4) 0 0.0 .. .. .. .. .. .. .. .. .. ..
Switzerland (2) .. .. .. .. .. .. .. .. .. .. .. ..
UK 54 293.7 44 258.4 81.5 88.0 4 13.3 4 12.6 2 9.3
USA 95 1917.0 65 1207.4 68.4 63.0 6 20.2 20 650.2 4 39.2
Total DAC 327 2901.3 201 1739.4 61.5 60.0 24 161.2 65 814.3 37 186.4
Source: Table 5 of DCD-DAC(2009)21-REV2-ENG from Contract Awards Questionnaire, 2006. (Provisional data).Notes: 1. This table does not include information on small contracts (below SDR 700,000). 2. No contract award information provided. 3. Two contracts for Denmark did not include project amounts. 4. No contracts from Greece, Spain and Sweden were reported above the SDR 700,000 threshold. 5. No contracts awarded pertaining 2007 ex ante notifications. 6. No contract amounts have been provided.
factotying.Thatapparentgapbetweendeclarationsandout-comesimpliesthattheinvestigationmustseektounderstandhowthecurrentuntiedaidregime,whichdonorshavevolun-tarilyagreed,worksinpracticewithindonoragencies(Chapter4)andatacountrypartnerlevel(Chapter6)beforeattemptingtodetermineitsimpacts.
Untying aid: Is it working?18
Chapter 4
4.1 Objectives and scope
T heToRcallsfora‘comprehensiveoverviewofthecurrentpoliciesandpracticesofDACmembersandnon-DAC
membersregardingthetyingstatusoftheiraidandtheeffectsonaid-effectiveness’(Para10).
Thisreview,undertakeninPhaseOne,drawslargelyontheannualprogressreportstotheDACthatmonitordonorperformanceinrelationtothe2001Recommendation21.Thesereportsprovideapicturebasedlargelyondonors’ownreportingontheformaltyingstatusoftheirODA,andsuggeststeadyprogresstowardsmeetingtheRecommendation.Thereis,however,lessunderstandingofwhatthisformaluntyingmeansintermsoftheorganisationandmanagementofbilat-eralaidinrelationswithpartnercountriesanditsconsequenc-es–inshort,howuntyingactuallyworks.Acomprehensivesurveyofcurrentuntyingandtyingpracticeswasprecludedbytimeconstraintsofreporting.Itwasthereforedecidedtoexplorethisquestion,aswellasexaminetherecentevolutionofdonoraidpoliciesandpractices,forarepresentativesampleoffiveDACmembers,withadditionalexamplesfromotherdonorpractices.Thememberssurveyedare,accordingtotheirownstatements,committedtotheRecommendation,butareatdifferentstagesinuntyingtheiraidoverall:Australia,Canada,Denmark,NorwayandSwitzerland.
ThethreeEuropeandonorswerealreadycommittedtountyingmostoftheiraidandreportedconsiderableprogresspriorto2002.CanadaandAustraliahaveexplicitlymodifiedtheirsourc-ingpoliciesforLDCsinresponsetotheRecommendation,aswellasmovingtountieTCandfoodaid,whicharenotcoveredbytheRecommendation.Thesefivecountriesarealsopartytodifferentregionalandbilateraltradeagreementsthatincludeprovisionsregardingpublicprocurement,whichcouldbeaninfluenceonuntyingpractices:Australia–bilateralagreements;Canada–NAFTA;Denmark–EU;andNorwayandSwitzerland–EEAandEFTA.Canada,DenmarkandNorwayarealsopartof
21 The latest report is OECD DAC (2009)
theinformalNordic+groupingofdonorscommittedtoworkingjointly,includinginsomecasesmanagingeachother’said.
ThisbriefsurveyisbasedonastatisticalanalysisofODAasreportedtotheOECD,plusareviewofofficialdocumentationandinterviewsortelephoneconversationswithaidofficialsatdonorHQlevelundertakenduringJuneandJuly2008.IthasservedtohighlightbothcommonanddistinctivefeaturesofdonorpoliciesandpracticesandtoidentifymorepreciselyissuesappropriateforexplorationatapartnercountrylevelinPhaseTwooftheStudy.Thesurveyresultsarepresentedfirstintheformofabriefaccountoftheevolutionofpolicyontyingandinfluencesforeachdonor,thenacomparisonofcurrentpracticesand,third,identificationofissuesforfurtherexploration–byextendingthesurveyonalimitedrangeofuntyingpracticefeaturestoincludeallmembers,andinvesti-gationatpartnercountrylevel.
4.2 Donor country experiences of untying
4.2.1 Norway Untyingaidbeganinthe1980s,andby2002therewereonlyafewresidualareasoftying(Figure4.1).TheseincludedpartofNorway’sfoodaidcommitmenttoWFP,emergencyaid-in-kind(materielandpersonnel)andsomeTC,includingsupportforprivatebusinessdevelopmentorB2B22,post-secondaryeduca-tionandresearch.
22 These are called ‘Matching Making Programmes’ in Norway (Norwegian Ministry of Foreign Affairs (NMFA), 2004)
4. donor poliCiEs And prACtiCEs: How doEs untying work? A survEy of
fivE donor EXpEriEnCEs of untying
Untying aid: Is it working? 19
Chapter 4
Apartfromthegeneraleconomicdevelopmentargumentsinfavourofuntying–suchasresourcetransferefficiency,provid-inggreaterflexibilitywithinagencyprogramming,aswellaspromotinggreaterresponsibilityandownershipforrecipientpartnercountries–morecountryspecificfactorswereinfluen-tial.First,therewasaconcern,initiallyexpressedinthe1970s,toavoidsectoraldistortionswithinanexportsurplusoil-economy.Second,anegativeexperienceofthe‘commercialisationofaid’–highprofilerepaymentdifficultiesoverconcessionalloanguaranteestofinanceshippingconstructioninNorwaythatledtodebtcancellationin1987–gaveadiscretepushtowardsuntying(Stokke,1989;2005;Abildsnes,2007).23ThelimitedpoolofNorwegianexpertiseisafurtherargumenttojustifybroadersourcingofTC,whichisalsoformallyuntied.However,politicalconcernsaboutsustainingpublicsupportrecentlyledtogivingspecialemphasistoareaswhereNorwayisseentohavecom-parativestrength–petroleumexploitationandenvironmentmanagement–whichmakesitmorelikelythatNorwegiantechnicalexpertisewillbefunded(Norad,2007a;2007b).
Bilateralaidwasreportedaswhollyuntiedin2007(Table3.2).Thelasttiedfoodaidwasphasedoutoncost-effectivenessgroundsin2006(NMFA,2007).Thereare,however,specialprogrammesfortheprovisionofNorwegiansourcedin-kindemergencyaidandpersonnel;theseinvolveassuredsupplyandstockpilesofalimitedrangeofmaterialswhicharede factotied.Arecentevaluationdrewattentiontothecomplexissuesinvolved.Therearehighcostsofaid-in-kindwhichhavetobebalancedagainstthedifficultiesofassuringthetimelyprovisionofsuppliesofhighlyspecificgoodstomeetunpre-dictabledisasterneedsintheabsenceofawell-establishedmarket(Norad,2008).Problemsalsoariseinensuringthat
23 There are parallels with the Pergau Dam affair that led to a political commitment to complete untying of UK bilateral aid first by abolishing the Aid and Trade Provision in 1997 and full final untying in 2000.
productinnovationandchangingbestpracticearetakenintoaccountwherethereistyingtospecificsuppliers.24Thisisseenasagreyareainwhichbalancingthebenefitsandcostsofty-ingpracticesarebothlessclearandsignificantlycasespecific.
Civilsociety,includingNGOs,researchinstitutionsanduni-versitiesandtheprivatesector,hasamajorroleinNorwegianODA.In2006,almost25%ofbilateralODAwaschannelledthroughNorwegianNGOs.Theseagenciesarefreetosourcegoodsandservicesanywhere.However,therewillbeaproportionofdonorcountryexpenditureinthisassistance.Fromarecipientcountryperspectivesuchpracticesmayalsoappearagreyareainthesourcingofaid.ApossibleissueforfutureinvestigationistheeffectthatchannellingofaidthroughNGOshasonsourcingofgoodsandservices.
TheshareofHQlevelcontractingisquiterestricted,amount-ingtounder5%ofODA.TheEEAandEFTAtradeframeworksmeanthatNorwayandalsoSwitzerland(seebelow)followEUguidelinesquitecloselyonthresholdsforcompetitivetender-ing.25ThispracticeexplainswhytherearesevereandEurope-widelimitsondonortyingpossibilities.
24 A similar problem of lack of incentive to innovate was found in an evaluation of US food aid support for nutritional improvement. Products developed in the 1960s were still being supplied from dedicated processing facilities three decades later, without taking into account advances in nutrition and food fortification (Marchione, 2002).
25 The current thresholds (according to Norwegian Government regulations) require national tendering over NOK 500,000. Above NOK 1.2 mn an open international tendering is required. Contracts for tender are posted in English as well as Norwegian on the Norad web-site and also forwarded to Brussels to post on the EU’s procure-ment web-site. Norad does not post contracts on the UN business web-site. Direct contracting is allowed below NOK 100,000. Between NOK 1,000,000 and NOK 500,000 Norad is expected to invite at least three proposals on the basis of a simpli-fied invitation.
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
100
90
80
70
60
50
40
30
Source: OECD CRS database.
Australia Canada Denmark Norway Switzerland
Figure 4.1 Five DAC Donors: bilateral aid reported as untied from 1996 to 2007 (percent)
Untying aid: Is it working?20
Chapter 4
MixedCreditArrangements(MCAs)areanexampleofhowuntyingmayleadtothedeclineinuseofaidinstruments.Norwayandothersthathavesofarretainedthemcouldphasethemout(Box4.1).
Theresidualareasofde factotyingandcontractingpracticeunderanuntyingregimeraisetheinterestingandimportantissue– when is the market the most appropriate regulator?
4.2.2 Denmark DevelopmentisaccordedhighpriorityinDenmark;thisisreflectedinanetODA/GNIratioof0.8evenafteradecadeoflittlegrowthinrealODA.OvertwothirdsofallDanishbilateralaidhadbeenuntiedpriorto2002(Figure4.1).Thisuntyingprocesshascontinued,includingbothTCandfoodaid,andin2007over88%ofbilateralaidwasreportedasuntied(Table3.2).Tyinghasbeenretainedformixedcredits.Supportforprivatesectordevelopment(B2B)isanambiguous‘grey’areainbilateralprogrammes,becauseassistanceisusuallydirectedtobuildinglinksbetweenbusinessintherecipientanddonorcountries.ThesetwoareasamountedtorespectivelyUS$32mnandUS$20mnin2006.
Historically,ahighproportionofDanishaidwastied,withstronglinks,forexample,toagricultureandagribusinessaswellashightechareas.TheEUregulationsonpublicprocure-
ment,introducedfollowingthecreationoftheEuropeanSingleMarketin1992,providedastrongimpetusfortheuntyingofallDanishbilateralaid,includingTCandcontributionstomultilateralfoodaid.TherequirementsforopencompetitivetenderingwithintheEUfor‘large’publiccontractsoverthethreshold,whichisnow€125,000,weakenedthelinksbetweenaidanddomesticexporters.Theargumentofcost-effectivenesswasusedtojustify,forexample,switchingDenmark’sshareoftheEU’sFACcerealsaidcontributionfromHQprocurementtogivingcashtoWFP(ColdingandPinstrup-Andersen,1998).Thetiedsupplyofdairyproductsasaidwasalsoendedin2003.
Denmarkhasdeliberatelyretainedsometyinginsupportofprivatesectordevelopmentandtofundinfrastructure,accountingforaround2-3%ofODA.26TheMixedCreditProgramme,largelydirectedtonon-LDCs,isseenassustainingdomesticbusinesssupportforaproportionatelyhighlevelofaid.Tenderingrulesareincreasinglyflexible:contractorsareal-lowedtosourcegoodsandservicesoutsideofDenmark,and,ifthereisinsufficientdomesticinterest,non-Danishregisteredcompaniescancompete.27
B2Bisseentobelargelydemanddriven.Developingcountrybusinessesareinvitedtoseekassistancefromdonorcountryofficesinfindingpartners.Suchdirectedassistanceisseenasrespondingtoinformationproblemsthathamperbusinessdevelopmentinpartnercountries.BoththeMixedCreditProgrammeandB2Bhavebeensubjecttorecentevaluations,leadingtoadaptationsinpracticetoimproveperformance(Danida,2002;2004).
Thehighproportionofuntyingisseenasassociatedwithdecentralisationofbilateralaidprogrammingandmanage-menttocountrylevel.Enhancedprocedureshavebeenestablishedformonitoringcountryofficeperformance.Wherethereisstronggovernanceandcapacityinplace,policyistochannelfundsthroughpartnerinstitutionsad-ministeredaccordingtotheirownsystems,asexemplifiedinSouthAfrica(PhaseOne,Box4.2).Denmarkisalsosupport-ingthestrengtheningofrecipientprocurementcapacity,forexampleinBangladeshandTanzania.Asecondrouteisjointfundingwherecontractingbecomestheresponsibilityofanotherleaddonor.ThethirdpossibilityiscontractingaccordingtoDanishprocedures.EUdirectivesonpublicprocurementgoverncontracting.Thehightransactioncosts
26 The government elected in 2000 was more questioning than its predecessors about the value of aid, cutting the aid budget by around a quarter; and that was a possible reason for retaining instruments that command support in the business community.
27 The Danida Board, which advises the Minister for Development Cooperation on matters related to Danish development cooperation, is a distinctive institutional fea-ture which is indicative of links between domestic stakeholders and aid strategy. The Board reviews and comments on new projects and programmes (or new phases of these) and discuss major strategy issues. It is comprised of nine members appointed by the Minister. Members come from different organisations (interest groups) in the Danish society such as NGOs working on development issues, the Danish Industry, universities and workers organisations.
Box 4.1 The Decline of Mixed Credit Arrangements
MCAs were widespread and have been contracting sharply since the adoption in 1991 of the Helsinki Ac-cord Rules on export and aid credits. This instrument was designed to operate in a tied aid environment, funding infrastructure and power projects, for example. Currently the Norwegian MCA provide around NOK 100 mn annually, although this amount is variable. Prior to untying the level was around NOK 300 mn a year. Projects of over NOK 5 mn are considered. All contracts are internationally tendered and this has proved prob-lematic. Tied MCA could offer a predictable package in-cluding timing, guarantees, private loans and suppliers. Untied, the MCA is not a full financing package – as the agency is unable to assure more than the interest sub-sidy on part of the funding with the remainder coming from a commercial source that in turn needs guarantees in place. Thus a project for Lao PDR in 2006, as part of a World Bank assisted project, was very slow to organise under international competitive tendering rules. China ended its Norwegian MCA when it became untied and China could no longer specify the exact source of supply for the equipment. Switzerland has phased out MCAs, but Denmark continues to provide them mainly for infrastructure to non-LDCs. Canada provides only grant aid and Australia has ceased to make new com-mitments as loans.
Untying aid: Is it working? 21
Chapter 4
andtimerequiredmightexplainthedeclininguseofcon-tractedprojectsthatrequiredopencompetitivetenderinginfavourofothermodalities.28Danidaisexemplaryinbeingtransparentaboutitscontracting:ICTcontinuestoawardpredominantlytoDanishcompaniesandtheirpartners(Box4.2).ThisoutcomeisconsistentwithotherdonorpracticeasreportedtotheDACandissuggestiveofapersistent,butinthiscasedeclining,advantageheldbydomesticorganisa-tions.
DenmarkhasalsoengagedineffortstountieTC,whichby2006wasfreeofsourcingrestrictions.Researchandhigher
28 In 2006 procurement through HQs had fallen to around 5% of bilateral ODA (OECD DAC, 2008b). Contracts of less than DKK 500,000 can be contracted directly. Then under DKK 975,000 (€ 125,000) there is a letter of notification; and evidence is required of three offers, but not a formal tender procedure. Over DKK 975,000 Danida is required to make an international openly competitive tender (equivalent to € 125,000 under EU directives); and this typically needs seven months for full procedure, a committee established involving partners, embassy and technical advisory services.
educationremainareaswherefundingisprimarilydirectedtoDanishinstitutions.29
4.2.3 SwitzerlandSwissODAhasincreasedby47%inthepastdecade.However,theshareofODAinGNIis0.4%andSwitzerlandhasmadelittleprogresssince2001towardsmeetingtheUN0.7%ofGNItargetforODA.98%ofSwissbilateralODAcommitmentsin2007werereportedasuntied(Table3.2).
Switzerland’saidwasoverwhelminguntiedpriorto2002(Figure4.1).30Thekeymomentinmovingonuntyingwas1994,whenthecurrentframeworkwithinwhichalmostallSwissODAisfor-mallyuntiedwasestablishedinresponsetothreedevelopments:theHelsinkiAccordin1991,theintroductionoftheEUsinglemarketin1992andthe1994WTOUruguaySingleUndertak-ing(i.e.agreementontrade)(Switzerland1994;1995;2007a).Switzerland,asanEFTAcountry,broadlyadoptedEUrulesonpublicprocurementofaid,similartotheEU,enablingcompaniesregisteredwithintheEU(15)in1995andtheEECtocompeteonanequalfooting.31Tyingpracticesweregraduallyphasedout.TheStateSecretariatforEconomicAffairs(SECO),whichisresponsibleforSwitzerland’sODAundereconomicdevelopmentcooperation,phasedoutmixedcreditsfromthelate1990s.InmeetingFACcommitmentstheSwissAgencyforDevelopmentandCooperation(SDC)begantoreplacecerealsfoodaidin-kindwithcashfrom1995(FerreiraDuarteandMetz,1996).
Remainingareasofde factotyingincludedairyaidandemer-gencyrelief.32In-kinddairyaid,costingUS$15mnin2006,wasdirectedsolelyforuseinnutritionalprogrammes,therebyminimisingrisksoftradedistortion,andjustifiedassustainingsupportforaid.In-kindemergencyaidincludesthehighprofiledirectassistanceprovidedbytheSwissHumanitarianAidUnit.
Untyingisnon-controversial,withSwissbusinessperceivingitselfascompetitiveinitsareasofspecialisation.Nevertheless,since
29 Grants are primarily awarded to research in Danish programme countries and within the yearly prioritised research themes. Priority is given to larger strategic research projects/programmes with partners in developing countries, and collabora-tion between several Danish institutions. Capacity building of these partners has for many years been a high priority for Denmark and so the majority of the budget is provided to them. Apart from this type of support, international research institutions are supported e.g. the Consultative Group for International Agricultural Research (CGIAR).
30 Swiss ODA is determined within a four year ‘credit framework’ under which annual appropriations are made. A large share of the bilateral portfolio is taken up with follow-on actions, suggesting that international commitments such as the MDG and the 2001 Recommendation have not been significant influences on the real level of ODA.
31 While a member of EFTA, Switzerland is not party to the European Economic Area Agreement. Regulations governing trade with the surrounding EEA countries tend to meet or exceed the requirements of EU Directives. http://www.euro.cauce.org/en/countries/c_ch.html
32 Switzerland did not report to the CRS the tying status of TC and administrative costs commitments in 2006 explaining the high level of non-reporting shown in Figure 4.3.1.
Box 4.2 Transparency: service procurement by Danida HQ
In 2006 Danida entered into 20 service contracts with a total value of DKK 254 mn. Danish companies account-ed for 15 contracts with a total value of DKK 199 mn. About one third of the contracts with Danish companies were in joint venture with foreign companies. Danida received 118 applications for the 20 tenders, with 41 being from foreign companies.
In 2007 Danida entered into 14 service contracts with a total value of DKK 163 mn. Danish companies account-ed for 12 contracts with a total value of DKK 157 mn. About half of the contracts with Danish companies were in joint venture with foreign companies. Danida re-ceived 68 applications for the 14 tenders with 29 being from foreign companies, which suggests an increasing interest from foreign companies.
Note: * The contracts are tendered according to the rules and procedures of the EU-directive on procurement. Contracts under the threshold of the EU-directive are awarded according to Dan-ish procurement rules. The contracts are completely untied. In 2008 Danida entered 520 such contracts to a total contract value of DKK 190 mn.
Service contracts tendered by Danida* 2006 2007
Total number of contracts 20 14
Number of contracts with Danish companies 15 12
Total value of contracts in DKK mn 254 163
Total value of contracts Danish companies in DKK mn 199 157
Untying aid: Is it working?22
Chapter 4
1994SDChascommissionedfouryearlyinvestigationsintotheimpactofODAontheSwisseconomy(GDP)andemployment.Thesestudieshaveindicatedaratioofbetween1.40-1.60ofout-putinassociationwitheveryCHFinODAwhichhavenotbeenaffectedbyprogressiveuntyingofaid(Zarin-Nejadanet al.,2008).
CivilsocietyhasamajorroleinSwissODA,includingNGOs,researchinstitutionsanduniversities,theprivatesectorandtheGenevabasedICRC,withwhichtherearestronghistoricties.In2006,almost21%ofSwissODA(24%ofSDCand8%ofSECOfunding)waschannelledthroughSwissbasedorgani-sations.Sourcingofgoodsandserviceswaslargelyuntied,exceptingdairyfoodaid(Switzerland,2007b).
Therelativelyhighproportionofmultilateralaid,NGOchan-nelling(together40%ofODA)andtheuseofdirectbilateralassistanceinstrumentslimitedSDCHQ’sprocurementofgoodsandservicestounder11%ofSDCbilateralaidin2006.Around85%ofthisfundingwasthendirectlycontracted,largelybecauseofthepracticeofexemptingon-goingprojects(74%)fromcompetitivetenderingrequirements.33Internationalcompetitivetenderingincludedonly2%ofcontracts.Non-Swissorganisationsaccountedfor9.6%ofcon-tractedfunding.Consequently,theeffectsofuntyingpracticesarelargelyintermediatedthroughthechannellingofODA,in-cludingformsofdirectbilateralassistancesuchassectoralandbudgetarysupportaswellasunrestrictedsupporttoNGOsandtheICRC.Untyingisseenasassociatedwithastrongpolicycommitmenttojointfundinganddecentralisationofmanagementtoacountrylevel.Thereisconsiderableinterestinabetterunderstandingofhowuntyinghasfunctionedanditsconsequencesatacountrylevel.
4.2.4 Canada LevelsofCanada’stotalandbilateralODAhaverisensubstan-tiallyduringtheprocessofuntying,especiallysince2004.Onlyfoodaid,acategoryofODAexemptfromtheDACagreementsonuntying,andpriorto2006virtuallyalltiedaid,declinedinrealterms.Untyingandtyingpolicieswouldappeartobeanissueaboutthecompositionofaidandnotaboutoveralllevelsofaid.
Historically,Canadawasadonorthatgaveitsbilateralaidlargelytied.Since2002,significantstepshavebeentakentountieODA,bothinacceptingtheRecommendationandgoingbeyondit;aidtonon-LDCs,TCandfoodaidhavebeenuntiedinresponsetoevidenceoneffectivenessandefficiency.Priorto2002,around50%ofbilateralODAwasuntiedand6%par-tiallyuntied(Figure4.1).By2007,atleast69%ofbilateralaidwasuntied,17%tied,1%partiallytiedand13%notreported(Table3.2).TheremainingareasoftyingareoutsidetheRec-ommendation:TCandemergencyordisasterrelief.
33 The SDC categories are that procurement under CHF 50,000 can be directly contracted. Contracts up to approximately CHF 250,000 (limited to the EU direc-tive € 125,000) are subject to an invitation procedure. Above that level a public call to tender is required. However, ‘ongoing’ (Folgephasen) projects are exempt (Switzerland, 2007a). In 2006, 9.5% of contracts were under CHF 50,000, 74.2% were ongoing, 14% subject to invitation and 2.2% open to competitive tender.
TheRecommendationhadanimportantinfluenceonuntyingpolicy.Priorto2002,thegreaterpartofCIDA’sODAwastiedtoprocurementinCanada,witharequirementthatgoodsandserviceshadamajorityofCanadiancontentcoveringabout8%oftotalODA.Theprevious1987policy,whichcoincidedwiththeDACdeterminationoftyingstatus,calledforpartialuntying(50%inthecaseofLDCsandSSA).Procurementintheindustrialisedcountrieswasrestrictedto10%andpermis-sibleonlywhenlocalorCanadiansourcingwasimpractical.TheRecommendationwasincorporatedin2002intoCIDAoperationalguidance,andeligibilitytobidonCIDAcontractsextendedtoLDCsandSSA.InternationalcompetitivetenderoutsidetheRecommendationcanbesanctionedonacase-by-casebasis.Foodaidwasexempt,butuntyingofTCpermitted.However,untyinghasbeenseparatelyextendedtofoodaid,initiallyto50%offundingonapartialuntyingbasisin2005,andthenthefulluntyingofallfoodaidin2008.Laterin2008,CIDAcommittedtofullyuntyingallitsbilateralaidby2012/13.
Partial untying toallowdevelopingcountrysourcinghasbeenanimportantsteptowardsfulluntying.TheabsenceofreciprocityonthepartofCanada’sneighbourandmajortrad-ingpartner,theUSA,wasamajorfactorinexemptingODAfromNAFTArulesthatwouldhaveallowedUSorganisationstocompete.Sourcingrestrictionswereinitiallymodifiedtoallowdevelopingcountrysourcing.Fulluntyingwasseenaspoliti-callydifficultasillustratedbythecaseoffoodaid(Box4.3).
PresentlythemainareasoftyingareTC,76%untiedin2007(asreportedin2009DACProgressReport),andsectorallybotheducationandsupporttogovernanceandcivilsocietywhichhaveahighTCcomponent,aswellasemergencyassistanceandrecon-
Box 4.3 Untying Canadian food aid
Canada has been a food aid donor since the 1950s, exporting cereals, oil seeds and dairy products, both as direct distribution and as budgetary/BOP support. Then, in 1998, a Performance Review highlighted resource transfer inefficiencies of Canadian programme food aid and the similar experience of other donors. At that stage untying was not considered a practical op-tion, so the focus shifted to making targeted nutritional improvement the primary goal and programme aid for sale was phased out. There was very limited partial untying to allow local procurement. In 2005, in the con-text of WTO agricultural negotiations and citing further OECD evidence on resource transfer efficiency, a policy decision committed CIDA to partial untying of 50% of funding. In 2008, food aid was fully untied. These steps were made following NGOs meeting with domestic agricultural interests to emphasise the benefits of local and regional sourcing for developing country small farmers and the likely negligible effects on Canadian agricultural exports.
Untying aid: Is it working? 23
Chapter 4
struction.ThefullextentofchannellingofaidthroughCanadianNGOsisunclearbut,incontrasttopracticeonthepartofEuro-peandonors,untilrecentlyCIDAreportedsuchaidastied,evenwheretheywereinfactfreetosourceindevelopingcountries.34
Thedecisiontountie,inaccordancewiththeRecommenda-tion,requiredtheformulationofnewoperationalguidanceandanundertakingwiththeTreasuryBoardonexemptionsfromnormalpublicprocurementguidelines:aconditionofthesechangeswasacommitmenttomonitoringandevaluatingtheconsequencesofuntying,sotheprocessiswelldocumented.Inpreparationforareviewtheevaluatorsformulatedalogicframeworktomakemoreexplicittheobjectivesofuntyingfromadonorperspective.ThishasbeenofassistancetoPhaseTwooftheOECDthematicstudy(seebelowChapter5).
IncontrasttoEuropeandonors,forwhomregionaltradearrangementshavebeenanimportantinfluenceonaidasaformofpublicprocurement,aidprogrammingliesoutsideoftheNAFTAobligationsbecauseofahistoriclackofreciprocity.Otherwise,CanadiancompanieswouldhavehadaccesstobidonUSaidcontracts–buttheydonot.35
4.2.5 AustraliaPriorto2001about55%ofAustralianbilateralaidwasuntied.Historically,onlyasmallshareofODAwasmultilateral(15%)andtherewasageographicalfocusonneighbouringcoun-trieswitharoundhalfintheformofTC.AustraliathenmovedtountiesomecomponentsofitsaidtoLDCsinconsistencywiththe2001Recommendation,andin2004Australiare-ported89%ofitsaidasuntied(Figure4.1).
In2006,thepolicyofuntyingwasextendedtoincludevirtuallyallODAinrecognitionofitsroleinimprovingeffectivenessandefficiency.Untyingisseenasincreasingpartnershipsbetweendonorsandpartnergovernments,strengtheningownershipandachievinggreatervalueformoneythroughcompetition(AusAID,2006).Inaseparatedecision,foodaidhasbeenwhollyuntied.Inconsequence,almostallbilateralODAhasbeenuntiedwithafewspecificexceptionssuchasTCinpost-sec-ondaryeducationandtheAustralia-IndonesiaPartnershipforReconstructionandDevelopmentthatcontinuesupto2010.36In200798%ofcommitmentstoLDCsand96%ofallbilateralaidwerereportedasuntied,indicatinghowquicklythepolicytountieaidhasbeenimplemented(Table3.3).
34 As of the beginning of 2008 all Canadian partner agreements are free of conditions on grant use and so are from the donor perspective untied. Past differences in donor reporting are another indication of the difficulty of inferring trends in tying practices.
35 However if funding is from CIDA’s Operation and Maintenance (O&M) budget, then there is an obligation to allow for competitive bidding by firms from NAFTA partner countries – as this represents the procurement of goods and services for CIDA’s own account and not that of developing countries. There is a threshold of CND 76,500, however, for NAFTA (including GST – Goods and Service Tax) for which CIDA have to go through a competitive process through the Canadian Public Tendering System (MERX) that allows for NAFTA competition.
36 Under this agreement, which predates the untying policy, Australia also provided A$ 500 mn worth of concessional loans.
Throughouttheprocessofuntying,netaidhasrisenbroadlyinlinewitheconomicgrowthratessotheODA/GNIratiohasremainedstableataround0.3%.Thispatternreflectsdomesticpublicexpenditureprioritieswhichconstrainedthelevelsofaid.Atthesametime,anemphasisonvalueformoneyinuseofthere-sourcesunderpinnedthecommitmenttountying(AusAID,2006).
Priorto2001,bilateraltradearrangementswithNewZealandgavefavourednationstatus.37Theseagreementsinfluencedpubliccontractingunder(national)CommonwealthProcure-mentGuidelines(CPG)(Australia,2004).Since2005,ODAhasbeenspecificallyexemptfromDivision2oftheCPGs(theMan-datoryProcurementProcedures),whichestablishanA$80,000opentenderthreshold(AusAID,n.d.).Duetothisexemption,AusAIDhasestablisheditsownthreshold(A$500,000)belowwhichdirectcontractingispermissible.Again,toprovideflex-ibility,procurementcanbeundertakenbelowthethresholdbycountryoffices.Follow-onexemptionsfromcompetitivetender-ingareallowed,butonlyasexceptionstotherule.Effortsaremadetomaketenderinggenuinelyinternational(Box4.4).
TheimmediatebenefitsofuntyingareseenfromtheAusAIDperspectiveasbothgreateroperationalflexibility,allowingnotonlybudgetaryandsectoralsupport,butalsocontracting,asinVietnam,accordingtopartnergovernmentprocedures.Untyingwidensthepoolof internationalexpertise andhence
37 These are the Australia New Zealand Closer Economic Relations Trade Agreement (ANZCERTA) and Australian and New Zealand Government Procurement Agreement (ANZGPA). There are post 2001 arrangements with USA (2005), Thailand (2005), Singapore (2003) and with Chile that comes into force in 2009: all are contributing to a context favouring non-discrimination as a key principle underpinning procurement.
Box 4.4 Information about AusAID open competitive tenders
Recognising that the problem of information can restrict competition, invitations to tender for large contracts are posted not only on the government’s own tender website (AusTender) but more widely. The fol-lowing locations are typically used:· DAC Untied Aid Website;· dgMarket (http://www.dgmarket.com/ – the tenders
and consulting site attached to the US Development Gateway);
· UN Development Business (UNDB) website (recently signed up, but will now be standard);
· Newspaper advertising – usually in the Weekend Australian, and may include other Australian or overseas papers as appropriate for the tender (e.g. The Economist, The Sydney Morning Herald, The Canberra Times, The Age, Australian Financial Re-view, etc. as well as local papers around Asia and the Pacific.); and on occasion
· Other websites (such as the Australian Development Gateway), email newsletters and industry journals.
Untying aid: Is it working?24
Chapter 4
reapsbettervalueformoneyfrom aiddeliveryteams.Untyingisalsoexplicitlyassociatedwithstrengtheningofjointfundingwithotherdonors.Forexample,theeffectivenessofTC,his-toricallythemajorareaofaidtying,hasbeenjointlyevaluatedbyAusAIDwithDanidaandBMZ,providingfurtherguidanceonimplementingthesepolicies(Land,2007).
4.3 Findings Performance on untyingAllfivedonorshavesuccessfullyrespondedtotheRecom-mendation,infourcasesvirtuallyuntyingalloftheirbilateralaidinthe‘eligiblecategories’andthefifth,Canada,iscom-mittedtodosoby2012/13.Allhaveextendeduntyingbe-yondthecategoriescoveredbytheRecommendation,bothtoincludenon-LDCsanduntyingofTCandfoodaid.Formaltying,namelysourcinginthedonorcountry,haslargelybeenphasedout.
Sofarthewidespreadmovestowarduntyinghavenotbeensubjecttosystematicmonitoringorevaluation.Thereareafewprojectorsectoralevaluationsthatmayofferinsights.Ratherthanprovidinganswers,discussionswithofficialsatdonorHQleadtothequestion–howdoesuntyingactuallywork?Whatistheexperienceofotherdonors?Soitwouldbeprematuretodomorethanspeculateonthelikelyconsequencesoftheextensiveuntyingofaidinthepastdecadethathasgatheredpacesincethe2001Recommen-dation.
Additionality of tied aid?Firstofall,inconsideringoverallODA,thereisnoevidencethatuntyinghasbeenassociatedwithanyreductioninbilat-eralaidorinoverallODA,andwithoneexception,Denmark,ODAhasrisensignificantlyinrealtermssincetheRecommen-dation.InDenmark’scase,domesticpoliticalconsiderationsratherthanuntyingofaidperhapshadanegativeimpactonODAatthebeginningofthedecade.
The evolution of policy on untyingTheexperienceofthefivedonorcountriesillustrateshowtheRecommendationhashaddifferentimplications,especiallysignificantforthoseDACmembersthathadpreviouslytakenverylimitedstepstowardsuntying.Canadahadtointroducelegalandinternalregulatoryinstrumentsandthendevelopproceduresfortheadministrationofuntiedaid.Incontrast,forAustraliauntyingrequiredaministerialdecisiontomovetheaidprogrammeintolinewithrecentgovernmentguidelines,whichhavenon-discriminationasakeyprincipleunderpin-ningprocurement.
Untyingiswidelyseenascloselylinkedwithdecentralisationofresponsibilityforaidprogrammingtoacountryofficelevel.Somanyofthequestionsabouthowitworksandimpactsarefelttobeanswerableonlyatapartnercountrylevel.ThisissueistakenupinChapters6and7.
Thesurveysuggeststhatuntyinghasbeenassociatedwithashifttoinstrumentsthatpassresponsibilityfordisburse-menttocountrypartners,otherjointdonorpartners,andcivilsocietyorganisations,andsoreducesthedonor’sroleincontractingorganisationstoimplementprojects(Denmark,Norway,Switzerland).However,inthemuchreducedlevelofHQcontractingtheproportionofawardstoforeignorganisa-tionsisverylow.Twofactors,policychoicesintheselectionofprioritysub-sectorsandthewaycontractingisorganised,especiallyinprovidingTC,appeartoinvolveasubstantialele-mentofde factotying.
Thereareseveralremainingrelativelysmallgrey areasoftyingpractice.Thesetendtobejustifiedonconventionalgroundsofsustainingdomesticsupportforaidthroughthesupportofinterestgroups,andofprovidingvisibleevidenceofadirectlinkbetweenbilateralaidandimpactsinbeneficiarycountries.Thisisafactorbehindthesubstantiallevelsofemergencyandhumanitarianaidprovidedbothin-kindandthroughperson-nel.ApossibleissueforfutureinvestigationistheeffectthatchannellingofaidthroughNGOs,presumedtobeuntied,hasonsourcingofgoodsandservices.AlthoughoutsidetheRecommendationallfivedonorshavelargelyremovedformaltyingrequirementsforfoodaid,whereasTC,supporttogovernance,post-secondaryeducationandresearchareareasofcontinuedtying.However,thereareotherargumentsthatrequirecarefulconsideration.ThejustificationofBusinesstoBusiness(B2B)citesproblemsofaccesstoinformationaboutglobalmarketsthatdisadvantageprivatebusinessdevelop-mentinmanydevelopingcountries.Arethereinfactareasinwhichmarketsalonearenotnecessarilyabletodeliver(i.e.emergencyequipment)orarenotthebestregulators?
Untying aid: Is it working? 25
Chapter 5
5.1 Introduction: objectives and scope
T hischapterreviewstheliteratureontheimpactoftyinganduntyingpracticesonthedonorandtherecipient
country.Theearlyliteraturehasbeenreviewed(Jepma,1991)andsotheintentionistoconsiderwhethermorerecentstudiesconfirmearlierfindings,providebetterexplanationsofwhatarecomplexrelationshipsoraregoinginfreshdirec-tions.Whenmostbilateralaidwastiedtheeffectivenessandtheimpactofuntiedaidwasatbestasubjectforspeculationandhypothesising.Sotheobjectivehereistoprovideapreliminaryaccountofastilllittleinvestigatedfield.
Thereviewcoversacademicliterature,contributionsfromcivilsociety,informationgatheredthroughvisitstodonoragenciesandinternaldonorreports.Itfocusesfirstontheconceptualframeworkunderpinningtheliterature,thenconsidersthecontributionofmorerecent(since1991)quantitativeinves-tigationsintoeffectsoftyingontherecipientandthedonorcountry,andfinallyconsiderstheextremelylimitedliteratureontheimpactsofuntying.
5.2 The effects of tying: conceptual issues Themostsignificantcontributionsintheliteraturefromananalyticperspective(Chilchiniski,1983;Bhagwatiet al.,1983;KempandKojima,1985;Schweinberger,1990;Hatzipanayo-touandMichael,1995;LahiriandRaimondos,1995;Brak-manandvanMarrewijk,1995)havesetouttheanticipatedconsequencesoftyingandestablishedtheconceptualframeworkforquantitativeestimatesoftheeffectsoftyingpractices.Thebroadconclusionofallthesestudiesisthattyingpracticesarelikelytoresultinwelfarelossesforrecipi-enteconomiescomparedwiththealternativeofunrestrictedaidtransfersmadeintheformofforeignexchange.Regard-ingthedonoreconomyaspartofathree-economyworldinwhichtherecipient,donorand‘othercountries’trade,thewelfareeffectsareunclear.Theglobalconsequenceislikelytobeareductioninwelfarewhencomparedwithunrestrict-edtrade.Thelimitationofsuchmodellingistheinherent
difficultyofinferringwhatwillbeabetterpolicyinasecondbestworld.38
Adegreeofcautionisnecessarywhenexaminingtheas-sumptionsandapproachesemployed,astheymayreducetherelevanceofthefindingsandrestricttheirpracticalapplicabil-ity.Theuseofthesemodelsbecomesproblematicwhencon-ventionalassumptionsdonothold.Perfectcompetitionandinformationareoftenassumed;findingsmaydifferdepend-ingonwhetherthetransferistiedtomanufacturedgoodsorprimarycommodities.Findingsmayalsovaryaccordingtowhetherthegoodsareprivateorpublic,consumptionorinvestment,tradableornon-tradable,onwhetherflowsareconsideredtobetemporaryorpermanent,ontheextenttowhichdifferentcharacteristicsoftheeconomiesaretakenintoaccount,orontheexistenceofquantitativetraderestrictions,andsoforth.Finally,themethodologicalapproachusedintheanalysis(comparativestaticversusacomparativedynamicmodel)mayinfluencefindings,underminingtheirrelevanceinpracticalanalysis.
Aconsequenceofthegulfbetweentheoryandrealityisamultiplicityofpossiblespecialcases.Forexample,thoseargu-inginfavouroftyingpracticesmayclaimthatinaparticularcasetheefficiencygainsofuntyingarenegligibleandshouldbesetagainstthepracticalrealitythattyingsustainshigherlevelsofaid.Afurtherissueisthat,ascircumstanceschange,theattempttoempiricallyexploretheconsequencesofuntyingmaychange.Forexample,thepolicyliteratureasitrelatestotying,pointstotheneedtounpackthethird‘other’economy.Itmaynolongerbeotherdonorindustrialisedeconomies,theimplicitassumptioninagreementstocontaincompetitivetyingsuchastheFAORulesonSurplusDisposalforfoodaid(1955)andtheHelsinkiRulesonexportcredits(1991).Insteadinsomecontextsthe‘other’economymaybeneighbouringdevelopingcountriesoremergingeconomies,includingnonDACdonors.
38 For a more detailed review of studies on foreign aid and trade policies see Suwa-Eisenmann and Verdier, 2006.
5. tHE iMpACts of tying And untying: A litErAturE rEviEw
Untying aid: Is it working?26
Chapter 5
5.3 The effects of tying: evidenceMostinvestigationsthatattemptquantificationhavefocusedeitherontheimpactoftyingontherecipientcountriesintermsofitsefficiencylosses,oronitsimpactonthedonors,mainlyintermsoftheassessmentofpotentialeconomicbenefits.Soitisappropriatetoconsiderthesestrandsoftheliteratureseparately.
5.3.1 Effects of tying on recipient countriesThefirststrandoftheliteratureontiedaidhasfocusedontheimpactoftiedaidasadistortionimposedbyaidpolicyontherecipient.Mostoftheempiricalstudieshavetriedtodeterminetheextenttowhichtyingpracticesrepresentacostonrecipientsthroughoverpricing,asaid-financedgoodsandservicessourcedinthedonorcountrymayinfactbeconsist-entlymoreexpensivethanthoseavailabletotherecipientfromalternativesources.Thus,tyingreducesthevalueofaidifnotnecessarilywhollyoffsettingthebenefitsofconcessional-ity.Beyondassessingdirectcostsofrestrictedsourcing,thereissomeexplorationoftheindirectcostsandotherfactorsthatcouldreducethepotentialnetbenefitsofaid.
Theattemptstoquantifythecostsoftyinghavebeenham-peredbyseveralpracticallimitations.Firstofall,dataavail-abilityconstrainstheestimates;secondly,todeterminewithprecisionthelikelyimpactoftiedaidontherecipientrequiresaknowledgeabouttheexistenceofinformaltyingpractices(JepmaandDeHaan,1984)andthedegreeoffungibility,ortheextenttowhichaid-financedimportsmightsubstituteforcommercialimportswhichwouldhaveoccurredanyway(Bhagwati,1985).Therefore,estimatesofthecostsoftyingtotherecipientsshouldberegardedasroughapproximations(Jepma,1991).
Resource transfer efficiency:Jepma(1991)inthemostwidelycitedreviewoftheearlyliterature(1960-90)concludesthattheexcesscosttorecipientsisintherangeof15-30%.Thisestimatewasmadenotwithstandingtheexistenceofstud-ieswhichreportsubstantiallyhighercosts,implyingthatthemostwidelycitednumbersareaconservative,lowerlimitestimateofthecostsoftying.
After1991,therearefewerstudiesontheimpactoftiedaidonrecipients.Somestudiesontheresourcetransferefficiencymodel(RTE)estimatethedirectexcesscostoftying(Yeats,1990;Yassin,1991;Osei,2004;Osei,2003;OECD,2006a).Thisapproachquantifiesthecostsoftyingbycomparingpro-curementpricesundertiedaidwiththecostofalternativecommercialtransactionsthatcouldhavebeenprocuredintheabsenceoftyingrestrictions.ThefindingslargelyreconfirmJepma’sconclusion:procurementpricesoftiedaidaretypi-callyhigherandthemark-upinpricesoffundedgoodsandservicesisoftenover20%.Asimilarconclusion(overpricingintherangeof5-25%),althoughusingadifferentapproach,isfoundbyAryeeteyet al.(2003)whointerviewprojectmanagerstoprovideestimatesoftheadditionalcostoftyingwithrespecttoopencompetitivebidding.Furtherqualitative
evidenceonthedirectexcesscostoftyingisalsoprovidedbycountry-basedexamples(Chinnock,1998;ChinnockandCol-linson,1999;ActionAid,2003;ActionAidInternational,2005;ActionAidInternational,2006;ODA,1996).
Themark-upsmaydiffersignificantlyamongfundedgoodsandservices;inparticular,thepricedifferentialisespeciallyhighforthoseproductsprovidedonaone-offbasis(Osei,2003),orcharacterisedbyahighertechnologycomponent(Osei,2004).Moreover,theexcesscostsoftyingaidmaybesystematicallyhigherthanJepma’s15-30%bandforparticu-lartypesofaid,suchasTC(Williamset al.,2003;RiddellandStevens,1997)orfoodaid(BarrettandMaxwell,2005;OECD,2006a;USGAO,2005).
Somestudieshavealsoexaminedtheimpactofexcesscostsontheconcessionalityofaidusingtheshadowgrantelementapproach,whichinvolvescomputingthegrantelementofaid,takingintoaccountthattheworthofatiedloanisactuallylessthanitsnominalvaluebecauseoftheeffectoftheexcesscostoftying(Yassin,1991;Osei,2003;MorrisseyandWhite,1996).Theyfindthattherealworthofaidisseverelyreducedbytheattachedtyingrestrictions(Yassin,1991;Osei,2003),althoughnotallthebenefitsofconcessionalityseemtobewipedoutbydirectexcesscosts(MorrisseyandWhite,1996).Giventhehardertermsonwhichmixedcreditsareofferedandtherelatedhighexcesscostsoftying,associatedfinanceisfoundtoreducetheconcessionalityofaidmorethanothertiedaidcredits(MorrisseyandWhite,1996).
Manyinvestigationshighlightadditionalwaysinwhichtyingindirectlyreducesthenetvalueoftheaidtransfer.Thereareindirectcostssuchasadditionalrecurrentcostsandshippingexpenses.Othernon-pricefactorscanentailactualreductionsinthestreamofpotentialdevelopmentalbenefitsofthetrans-fer,suchasdelaysindelivery,thequalityoffundedgoodsandservices,theinefficientallocationofresourcesfollowingfromhighlyinappropriateorlowerprioritypurchasesorselectionofprojects;environmentalimpactofsomesectoralaidflows,e.g.infrastructure;factorsrelatedtothetradeimpactsoftying,whereitbecomesabarriertoaccessingmarketsandpromot-inginter-regionaltrade,orimpactsonthelocalmarket(Jepma,1991;Jepma,1994;Bhagwati,1985;Ensign,1992;Hendra,1987;Belfrage,2007;Barrett,1999;ActionAid,2003;Chinnock,1998;ChinnockandCollinson,1999;ActionAidInternational,2005;ActionAidInternational,2006).Tiedaidcreditsmayalsoimpactonarecipientcountry’sexternaldebt(Larrú,2003;Hendra,1987);someevidencesuggeststhatbeingarecipientoftiedaidincreasesacountry’sexternaldebt,atleastinthemediumterm(Alonso,1999;LarrúandGonzales,2004;Abildsnes,2007).
Asemergingdonorsincreasetheiraidtodevelopingcoun-tries,theirpoliciesandpracticesbecomemorerelevant.Theappearanceofnew(nonOECD)donorswhomaynotfeelobligedtocomplywithvoluntaryDACrules,bringsforLDCstheriskofincurringonceagainthecostsofunsustainable,un-productivecapitalprojects(Manning,2006;Pehnelt,2007;Gill
Untying aid: Is it working? 27
Chapter 5
andReilly,2007;McCormick,2008).Itshouldbehighlighted,however,thatemergingdonors’tiedaid(especiallyChineseandIndian),withitspackagedapproach,iswhollyunevaluat-edfromeitheranefficiencyoreffectivenessperspective(PauloandReisen,2009;Collier,2008).Thustheargumentaboutwhetherthistiedaidisinfactcost-effectiveascomparedwithpreviousexperienceisinconclusive.
Finally,tyingpracticesmayalsohaveanimpactonrecipients’behaviourorattitudetowardsaid,oftencharacterizedasa‘lackofownership’,whichisoutsidetheformaleconomiccalculations.Recipientslearntorequestaidfinancingfromdonorsinthosespecificsectorswhichdonorsarewillingtofinance(Hendra,1987)andtowastefully‘cherrypick’tiedaidpackages.Suchpracticesreducethevalueofaidfromthedonorperspective.However,theyalsocontradicttheviewthattyingoftenrendersrecipientsaspassiveactorsinthedevelopmentprocesswiththestandardexampleofmachin-eryinappropriatelyprovidedbydonorslyingbrokenwhilerecipientsareleftwaitingforsparepartsortechnicalassist-ance(examplesinHendra,1987;ActionAid,2003;Chinnock,1998;ChinnockandCollinson,1999;ActionAidInternational,2005;ActionAidInternational,2006).
Tosumup,therecentliteratureontheimpactoftyingontherecipientcountrybroadlyreconfirmsearlierfindingsthatprocurementcostsoftiedaidaretypicallyover15%-30%higherthanactualandhypotheticaluntiedalternatives.Pricedifferentialsmayvarysignificantlyamongfundedgoodsandmaysystematicallybehigherforsomeformsofaid.Thereissupporttooforearlierfindings,usingmorerefinedevidence-basedapproaches,includingonthereductionintherealconcessionalityofassociatedfunding.Theimportanceofindirectcostsandnon-pricefactorsinassessingtheimpactonrecipientcountriesisconfirmed.
Overall,thereisbroadagreementintheliteraturethattiedaidinvolvesasubstantiallossofvaluetorecipients,butlittlesys-tematicinvestigationintohowdifferentformsoftiedaidmayimpactonrecipientsandonparticularsectors.Thisgreaterspecificityoffocusandtheimpactofemergingdonors'aidpracticesonrecipientswouldseemtobeprioritiesforfurtherinvestigation.
5.3.2 Effects of tying on donor countriesTheprimaryintentionintyinganaidtransactionistofavoursuppliersinthedonoreconomyrelativetotherestoftheworld.Thiseffectwouldeitherdisplacecommercialexportsthatwouldhaveoccurredwithouttheaidtransactionor,assumingahigherlevelofimports,entailadisproportionateincreaseinthedonor’sexportscomparedwithcompetitors.
Soiftyingiseffective,donorswillderivebenefitsfromin-creased tradethataremorethanproportionaltotheirshareofworldexportmarkets.Morerecentstudiesindicatethattherealimpactoftyingondonors’exportsisquitelimited(Tajoli,1999;Johanssonet al.,2006;Lloydet al.,2000;Lloydet al.,
2001;Oseiet al.,2004);oftenwhenaidflowsarefoundtohaveasignificantpositiveimpactonexports,thisisevenmoresoastheproportionofuntiedaidincreaseswithtime39(SchönherrandVogler-Ludwig,2002;Zarin-Nejadanet al.,2008;Martínez-Zarzosoetal.,2008).Sectoralstudies,andinparticularthoseonfoodaid,offersupportforthisthesis;foodaidisusuallyfoundtobenotfullyadditional.40Somestudies,however,findasubstantialpositiverelationshipbetweenaidflowsandexportsfromdonorcountries(Nilsson,1997;Wagner,2003;Larrú,2003;LarrúandGonzález,2004;Nowak-Lehmannet al.,2008).Sectoralstudies,againforfoodaid,offersomeempiri-calevidenceofthispositiverelationship.Acrediblesequenceofimpactsisthatcommercialfoodimportsarereducedintheshorttermandthereisalaggedincreaseinexportsinthelongterm(Barrett,1998;1999).Theoverallimpressionisthat‘italldepends’onthetemporalspecificsofgoodsandmarkets.
Similarlycertainformsofaid(debtrelief,budgetsupport,technicalassistanceandremainingmostlyprojecttypeaid)areexpectedtohavedifferentimpactsondonors’exports;duetothelackofdataavailabilityonlyafewstudies(Johanssonet al.,2006;JohanssonandPettersson,2009)haveinvestigatedthishypothesisanditisthereforenotpossibletodrawwell-foundedconclusions.
Economicliteraturehasbeenlargelysilentontradedistortingimpactsofdifferentaidinstruments.AnODIstudyhowevershowsthatgrantsandloans,andthereforethelevelofconces-sionalityper se,donotseemtohavedifferentialtradedistort-ingimpactsbeyondtyingpractices(ClayandTurner,2007;MassaandTeVelde,2009;Clayet al.,2009).
Aid,andinparticulartiedaid,isalsolikelytohavedifferentimpactsdependingonthesector(economicinfrastructure,socialsectorandsoforth).Sectoraldifferences,althoughac-knowledged,havenotbeenthoroughlyanalyzed.JohanssonandPettersson(2009),inordertotakeintoaccountthataidflowsimpactmightbeheterogeneous,disaggregateexportsbycategory.
Ingeneral,disaggregatedanalysishasbeenlimited,con-strainedbylackofdata.Therearefurtherimportantqualifica-tionstobemadeontherobustnessofthesefindings.First,studiesrarelydisentangletheimpactofaidonthedonorcountryintermsofexportsdirectlyfinancedbytiedaidandthosethatarenot(Jepma,1994).Secondly,fungibilityisusu-allynottakenintoaccount;yetthiscouldconsistentlyaffectfindings.Thereareindicationsthatthesubstitutioneffectisfairlylarge(Jepma,1991;KhiljiandZampelli,1991;PackandPack,1993;Feyziogluet al.,1998;Devarajanet al.,1999).However,theremaybecasesinwhichtherecipientcountry
39 This is known as the goodwill effect; as a result of untied aid flows, the recipient might feel more inclined to buy goods and services directly from the donor.
40 When food aid is provided for sale on the recipient market the ratio of aid to displaced imports has been found to be around 0.6 tonnes for each tonne of aid (FAO, 2006). The displacement is lower, around 0.3 of directly distributed food.
Untying aid: Is it working?28
Chapter 5
isnotableorwillingtoexploitsuchfungibilitypossibilities(Bhagwati,1985).Finally,thedisplacementeffect,theextenttowhichthedomesticorexportdemandfordomesticinputsisdisplacedbythedemandtomeettiedaidexports,shouldnotbeunderestimated,asitcouldentailareductioninthenetimpactoftyingonthedonor(Morrissey,1989;Jepma,1994).
Moreover,empiricalinvestigationshaveusuallyfocusedontheimpactofaidondonorsexportflows;however,aidcouldalsoinfluencerecipientexports(Johanssonet al.,2006;Jo-hanssonandPettersson,2009).JohanssonandPetterssonfindthataidflowshaveapositiveeffectnotonlyondonor-recipi-entbutalsoonrecipient-donorexportsandconcludethatthedonor-recipientaidrelationshipreduces‘theeffectivecostofgeographicdistance’,impactingpositivelyontradeflowsandnotonlyonexportflows.
Afewstudiesattempttoassesswhetherdonors’tyingprac-ticesleadtoanetincreaseinemploymentintheircountries(Jepma,1991;Maleket al.,1990;Morrissey,1990;MorrisseyandRudaheranwa,1998;focusonmixedcredits:MacQuaideandToye,1986;LoveandDunlop,1990;Morrissey,1991).Theconclusionsarethattyingofaidisnotsuccessfulingenerat-ingsubstantialemploymentindonorcountriesbutfinancialbenefitsseemtoaccruetoparticularfirmsandgroupsinthedonorcountry.Tyingisawayofmakinginternaltransfersfunded,usuallythroughgeneraltaxation,tothesebeneficiaryinterestgroups.
Astiedaidhasbeenjustifiedforthepositiveimpactexpectedondonors’balanceofpayments(BOP),someearlystudiesmakeanattempttoempiricallytestthispropositionandfindanegligibleimpact(Hopkin,1970;Levitt,1970).Associatedfinancingseemstohaveaproportionallymorepositiveimpactonthedonor’sBOP(Jepma,1991;ToyeandClark,1986).
Finally,itshouldberecognizedthattyingaid,andspecificcasesofaidmalpractice,canhaveanimportantpoliticalratherthananarrowlyeconomicimpactonthedonorcountry,asillustratedbythefollowingexamples.ThePergaudamscandalintheUKhadlongertermrepercussionsasitaddedmomen-tumtotheexistingdevelopmentalconcernswithinthegov-ernment,pushingtowardsapolicyshiftandthedecisiontountieBritishaid(ChinnockandCollinson,1999).InNorwaytheShipExportCampaignhadimportantrepercussions,leadingNorwaytoofferdebtforgivenesstothedevelopingcountriesinvolvedandrecognizingirresponsiblelending(Abildsnes,2007,andseeaboveChapter4).
Literatureontiedaidisstillverythinfromalegalperspective;apartfromafewarticlestiedaidhasrarelybeenreferredtoinlegalliteraturewhereathoroughanalysisoftyingpracticesandtheconnectionstothecurrentWTOlegalframeworkisstillmissing(LaChimia,2009).
Tosumup,themorerecentliteratureontheimpactoftyingondonoreconomiesbroadlyreconfirmsearlierfindings;the
macroeconomicimpactisfoundtobefairlylimitedwhereasthe‘limited’commercialbenefitsmaybeimportanttoparticu-lardomesticinterestgroups.
5.4 The impacts of untying aidTherehavebeenveryfewempiricalinvestigationsintotherelativeeffectivenessofuntiedandtiedaidonacomparablebasis,oroftheimpactsoftheactualuntyingofaid.Theexist-ingliteratureincludesstudiesdifferinginscopeandmeth-odologyandsofindingsarerarelygeneralisableorstrictlycomparable.Furtherevidence-basedstudiesareneededontheconsequencesoftheextensiveuntyingthathasoccurredsincethemid1990s.
Therearepurelyanalyticortheoreticalexplorationsthatcomparethehypotheticalwelfareandterms-of-tradeeffectsoftiedanduntiedaid(e.g.ChatterjeeandTurnovsky,2004;MichaelandvanMarewijk,1998;Chatterjeeet al.,2003;ChatterjeeandTurnovsky,2007).Theseanalysestypicallyconcludethatunder‘plausibleconditions’,untiedaidben-efitstherecipientandtheworldeconomymorethantiedaid.Somestudiesareempiricalinthesensethattheymodel,usuallyexante,thepotentialconsequencesofuntying,usingactualvaluesforkeyvariables.Untiedaidispresumedtoben-efittherecipientandsostudiesusuallyhaveadonorfocus.Theprocessofuntyingdoesnotseemtoentailawelfarelossforthedonorcountry(MorrisseyandRudaheranwa,1998).Onthecontrary,somestudiesdemonstratethattheprovisionofuntiedaidmightevenincreasedonorexportstotherecipi-entcountryasaconsequenceofgoodwillgenerated(ArvinandChoudhry,1997;Arvinet al.,2000;ArvinandBaum,1997;Arvinet al.,1996;Arvinet al.,2003).Jepma(1988)examinedtheimpactofahypotheticalcoordinatedshifttowardspartialuntyingofaidonanEUscaleondonors’tradeflowspriortotheintroductionofthesinglemarketthatwouldopennationalprocurementtoothermemberstates.Heconcludedthatsuchashiftinpolicywaslikelytobebeneficialfromadonor’sperspective.
Withregardtotherelativeeffectivenessoftiedversusuntiedaid,Miquel-Florensa(2007)concludesthatuntiedaidismoreeffectivethantiedaidincountrieswithpolicyenvironmentsmorefavourabletoeconomicgrowth.Similarly,Amegashieet al.(2007)highlighttheinefficiencyoftiedaid,butnoteitisbetterabletocontrolmoralhazardwheretheremaybeproblemsofcorruptionanddirectionoffundsawayfromtheenvisagedaidpriorities.
Thedebateovertheeffectivenessoftiedversusuntiedaidhasalsogivenrisetoanormativeliteraturebasedoncaseexamples.Quartey(2005)seesbudgetarysupportasawaytoovercometheproblemsoftiedaid;Hendra(1987)con-trastsCanada’sbilateraldevelopmentassistance(highlytied)withScandinavianpractices(moreuntied)andhighlightsthesuperiorityofthelatter.Morecomprehensivestudiesof
Untying aid: Is it working? 29
Chapter 5
untyingoffersomeusefulinsightsonthepotentialbenefitsofuntiedaid(e.g.Puttersideet al.,2004),oftenusingasastartingpointthenegativeimpactsoftiedaidincountry-casestudies(ActionAid,2003;Chinnock,1998).
Theexistingliteratureinadequatelycoverstheimpactsofuntiedaid;thefewstudiescarriedoutseemtohavefocusedontheimpactofuntyingfromthedonor’sperspective.Incontrast,evidence-basedstudiesarelackingoftheexpost,economy-wideimpactsofactualuntyingforeitherindividualrecipienteconomies,groupingssuchasLDCsorregionalcom-munities.Theliteratureislargelysilentontheeffectofuntyingonrecipientcountries(impactonlocalmarkets,procurementand/orpromotionofregionaltrade)andontheefficiencyofuntiedaid.Thisisapowerfulargumentinfavourofcountryfocusedstudiesofuntyingpractices.Animportantexceptionistherapidlygrowingliteratureontheimpactsoflocalandregionalpurchasesmadepossiblebytheuntyingorpartialuntyingoffoodaid(Box5.1).
Mostimportantly,whatseemstobemissingisafullydevel-opedanalyticframeworkdesignedtoprovidethebasisforexploringtheimpactofuntying.Intheanalysisoftheimpactoftiedaid,thestartingpointwasrepresentedbyanidealiseduntiedsituationinwhichtyingwasintroducedasadistortion.Intheanalysisoftheimpactofuntyingwehavethereversecondition:inaworldcharacterizedbytiedaid,onetriesto
predictwhatwouldhappenifthesourcingrestrictionwereremovedandaiduntied.However,thereisnoreasontobelievethatwhatisobservedinthetiedaidframeworkissimplyreversible.Thisisnotareturntoanidealisedworldinwhichgoodsandservicescanbesimplysourcedincompeti-tive,undistortedmarkets.Furthermore,aidisanadminis-teredprocesswithmanyformalandinformalrestrictionsthatarenotnecessarilyalteredwhentheformalrestrictiononsourcingisremoved.Thewholediscussionislocatedwithinahighlyconstrainedsecond-bestframework.Theinvestigationintotheconsequencesofamodificationinanaidregimehastofocusonlocaloptimality:efficiencygains(bettervalueformoney)withinaceteris paribusframework,andsimilarlywitheffectiveness.Forthisreasonagain,ananalyticframeworkthatsetsoutthesepossibilitiesinasystematicandlogicalwaywouldbeparticularlyhelpfultothesecondphaseofthisevaluationandthewiderassessmentofthemovestountying.
Ashighlighted,studiesontheconsequencesofuntyingarelimited.However,thereisaparallelliteraturethatfocusesontheuseofaidmodalitiesforwhichuntyingisanecessary(pre)condition.Thesemodalitiesincludewhataresometimescalled‘directbilateralinstruments’,financialaidforGBSandnewmodalitiesofaidsuchaspooledfunding.41Severalstudiesdiscussthepotentialdevelopmentalbenefitsandimpactsofbudgetsupport:reductionintransactioncosts,facilitationofdonors’co-ordination,improvementinthepredictabilityofaidflowsandhigherallocativeefficiencyofpublicexpenditureaswellasstrengtheningofpartners’ownershipandinstitutionaldevelopment(forexampleOECDDAC,2006b;NascholdandBooth,2002;Nilsson,2004).Anin-depthevaluationoftheefficiencyandeffectivenessofGBSbasedonsevencountrycasestudiesovertheperiod1994-2004(IDDandAssociates,2006)concludes,whilstemphasis-ingsomeintrinsiclimitations42oftheinvestigation,thatGBSisoften‘anefficient,effectiveandsustainablewayofsup-portingnationalpovertyreductionstrategies’.TheevaluationalsoconcludesthatGBShasbeensuccessful:instrength-eningownershipandaccountability,aswellasimproving‘operationalandallocative’efficiencyofpublicspending;inimprovingharmonisationandalignmentofdonors.ThusGBSbenefitsseemtooutweighsomenegativeunplannedeffects(suchasfeedingcorruptionandunpredictability),
41 Strictly we are talking about financial aid as budgetary support. In the past pro-gramme commodity aid that the recipient government sells on the local market to generate ‘counterpart funds’ has been used to provide budgetary support. The most important case has been US program food aid. The record on impact was positive for the Marshall Plan, very negative especially in the Vietnam war. Such programme aid as was found to be poor value for money as development aid in 1980s and 1990s and drastically reduced by the US and phased out by other donors (Barrett and Maxwell, 2005; OECD, 2006a).
42 The evaluation is limited to seven countries and it is difficult to draw robust conclusions as GBS experiences are relatively recent. Differently from project aid, it can be difficult to isolate GBS and distinguish its effect from the aggregate ones as there are several interactions and forces at play. For a fuller analysis of the caveats see page 117 of the evaluation.
Box 5.1: Developmental impact of local and regional procurement of food aid
There has been a substantial increase in the scale and relative importance of local and regional purchases (LRPs) of food aid, notably in East and Southern Africa (WFP, 2006; see also Table 3.7). Evidence-based research has demonstrated efficiency gains from LRPs, and that these result directly from actual domestic and import parity prices of commodities sourced locally (40%) and within the region being typically 30% below delivered costs of foodstuffs procured on donor markets (OECD, 2006a). These efficiency differentials have been main-tained even during the current period of escalating food prices (Tschirley, 2008). So far studies have been positive about greater net benefits for rural and urban populations of local (untied) rather than imported (tied) aid (NRI, 2005; Coulter et al., 2007; WFP, 2006). The qual-ity of commodities offered for purchase by local traders rises. Local processing is encouraged and distortions of local markets largely avoided. However, assuring sustained benefits for local agriculture and trade may be difficult where intervention is episodic (Tschirley and del Castillo, 2006). Findings are, of course, context spe-cific and so generalisation will require an accumulation of case specific evidence for an under-researched area.
Untying aid: Is it working?30
Chapter 5
whilstrecognisingthattherearepoliticalandotherriskstoitssustainedusethatshouldbecarefullyconsidered.
Inconclusion,thereviewoftheliteratureonuntiedaidbroadlyshowsthattherehasbeenlittleformalinvestigationoftheeffectivenessoftiedversusuntiedaidorontheimpactofuntying.Studiesdifferinscope,coverageandmethodol-ogyandfindingscannotbegeneralized.Thediscussionaboutuntyinghasalsofailedtotakeintoaccounttheevidenceaboutaidmodalitiesforwhichuntyingisanecessarycondi-tion.Furtherempiricalstudiesareneededinordertodrawmorerobustconclusions,especiallywithrespecttotheimpactofuntyingontherecipientcountries.Mostimportantly,whatseemstobemissingisaconceptualframeworktobeusedforsystematicempiricalinvestigationsintotheconsequencesofuntying.
Untying aid: Is it working? 31
Chapter 6
6.1 Objectives and scope
O nlyasmallminorityofdonorsreporttyingforafractionoftheirbilateralODA.Formallytiedaidis,apartfrom
foodaidandfreestandingTC,exceptional.Howthendoesthisregimeofformallyuntiedaidworkinpractice?ThesurveyoffivedonorsattheHQlevel,andwiderinformalsoundingswiththedonorsfoundthatagencypersonnel,drawingontheirownexperience,seeuntyingasassociatedwithotherchangesinaidmanagement.Theseincludeashifttodecentraliseresponsibilityatarecipientcountrylevel,jointprogramming,usesofrecipientcountryproceduresandareducedshareoffundingthroughdirectcontractingandopentenderingbydonoragencies.Butintheabsenceofanysystematicinvesti-gationtheseviewswererecognisedasspeculative.Theimplicationisthatthefirsttaskinthecountrylevelinvestiga-tionswastoundertake a factual review of how untying works at this level.Theproductofthisinitialpartoftheinvestigationisanaccountoftheuntyingregime,itsconsistencywiththeOECDRecommendationandothervoluntaryunderstandingssuchastheguidelinesonaidprocurement.
ThedonorcountryHQperspectiveasreflectedinreportingtotheOECD(Chapter3)isthatmostaidisnowformallyuntied.Incontrast,partnercountrygovernmentofficials,academicsandothersincivilsocietypresentedininitialcontactsaquitedifferentview.Theyarguethattheprocessofformaluntyingwasunsatisfactoryinthreeways:itfailedfirst,togiveowner-shipandsecondly,toachievealignmentofmuchoftheaidofmany,includingmajor,donors;whilstthirdlytherewasstillextensivede factotying.Thislatterconcernwasinformallyreiteratedbysomeagencypersonnelwithregardtocertainaspectsoftheirownprogrammesandespeciallytotheprac-ticesofotherdonors.
Thecountrystudies,inprovidingafactualreviewoftheextentofformaluntying,firstsoughttoconfirmwhatwasreportedfromHQ,bycomparingcountrylevelsourceswithCRSdata.Second,thereviewwouldbeusedtoestablishhowtheunty-ingregimeisworkingintermsofmodalitiesandusersofaid
atacountrylevel.Forexample,isuntyingactuallyassociatedwiththeadoptionofmodalitiesthatareconditionalonunty-ingsuchasdirectbilateralinstrumentstoprovidebudgetarysupport,poolingofprojectorprogrammaticfundsandtheuseofcountrypartnerssystems?Inpractice,monitoringmaynotdistinguishmodalitiesinthisway.Therefore,asappropri-ate,theteamsweretocomplementavailableinformationwithapurposivesurveyofmodalitiesanduseofdifferentprocure-mentsystems.AfterprovidingcomparativeprofilesofthesixcountriesinSection6.2,theresultsoftheinvestigationsintomodalitiesandusesarereportedinSection6.3.
Thewidespreadscepticismencounteredregardingthegenuinenessofsomesupposedlyuntiedaidnecessitatedaninvestigativeresponse.Thisresponsetakestwoforms.First,itwasdecidedtoundertakeacountry-leveleconometricinvestigation,reportedinSection6.4,intotherelationshipbetweenbilateralaidanddonorexportstotestforevidenceofcontinuedtradedistortingpractices.Second,thestudywouldconsiderforarepresentativeselectionofdonorprojectsthegeographicalsourcingof‘head’contractors,whomanageprojects,sub-contractingandsourcinggoodsandserviceswithinprojects.ThisisreportedinChapter7.
Smalltwoorthreepersonteamsundertookthecountryinves-tigationsovertwomonths,involvingcollectingofinformationandinterviewsoveraperiodofthreetofourweeks.Theyhadtobehighlyselective,andheavilydependentonthedataanddocumentationthatwasreadilyaccessiblethroughtheco-operationofcountrypartnersanddonors.Theseinvestiga-tionsshouldthereforeberegardedasexploratory.ForamorecomprehensiveassessmentasimilarlevelofresourceswouldbeneededforconsideringtheconsequencesofuntyingbyasingleDACdonorinanyonerecipientpartnercountry,andperhapsthroughrepeatedassessmentonceabenchmarkwereestablished.
6. usEs of untiEd Aid in siX pArtnEr CountriEs
Untying aid: Is it working?32
Chapter 6
Regi
onCo
untr
yDe
velo
pmen
t st
atus
1
DAC D
onor
sTo
tal O
DA
(bila
tera
l an
d m
ulti-
late
ral)
US$
mn
Mul
tilat
-er
al O
DA
(%)
Bila
t-er
al
ODA
(%)
Bila
tera
l OD
AGD
P
per
capi
taTo
tal
ODA
ODA
pe
r ca
pita
Mul
ti-la
tera
l OD
A (%
)
Aid
as
% o
f GN
IBi
lat-
eral
OD
A (%
)
Aid
as %
of
Pub
lic
Expe
ndi-
ture
2
Wat
er &
Sani
tatio
n
DAC
Dono
rsTo
tal
ODA
US$
mn
Mul
tilat
-er
al O
DA (%
)
Bila
tera
l OD
A (%
)
LDCs
OLIC
sLM
ICs
UMIC
sHI
PCs
> U
S$ 2
mn
Wes
t Af
rica
Burk
ina
Faso
X
X2,3
0233
.866
.216
416
5312
.8-
92.2
50.4
49.6
Gha
na
X
X
4,441
26.0
74.0
1657
166
11.7
26.0
228.6
34.3
65.7
Sout
hern
Af
rica
Zam
biaX
X
4,636
9.690
.418
808
129
17.6
39.9
153.5
30.4
69.6
Sout
h Af
rica
X
2,8
544.6
95.4
225,5
1520
0.40.9
227.4
0.010
0.0
Sout
h Ea
st As
ia La
o PDR
X
85
028
.471
.615
592
478.4
89.7
39.7
1.298
.8
Viet
nam
X
8,4
9734
.865
.221
718
344.8
-85
8.954
.046
.0
Sour
ce:
OECD
Cred
itor R
epor
ting
Syst
em (C
RS) d
atab
ase;
Wor
ld D
evel
opm
ent I
ndica
tors
(WDI
).No
tes:
1.
The d
evel
opm
ent s
tatu
s of e
ach
coun
try i
s defi
ned
acco
rdin
g to
the D
AC li
st o
f ODA
recip
ient
s (w
ww.
oecd
.org
/dac
/sta
ts/d
aclis
t).
2.
Aid
as %
of P
ublic
Expe
nditu
re re
fers
to 2
007
and
not t
o 20
05-0
7.
3. D
ata a
re in
curre
nt U
S$ m
illio
ns; O
DA d
ata a
re b
ased
on
com
mitm
ents
.
Table 6.1 Profile of six case study countries
Untying aid: Is it working? 33
Chapter 6
6.2 Profile of six case study countriesThesixcountrieswereselectedtobebroadlyrepresentativeoftherangeofdevelopingcountrysituations.Theselectionalsotakesintoaccountregionalandeconomicgroupings,bearinginmindtheimportanceaccordedintherationaleforuntyingassetoutintheRecommendationtopromotinglocalandregionalmarketsandenterprises.Thesurveyincludesfivelow-incomecountries(LICs)andoneuppermiddleincomecountry(UMIC),SouthAfrica,withinthreeregionalpairinginWestAfrica,SouthernAfricaandSEAsia(Table6.1).
ThethreeAfricanLICsarealsofullACPmembers.BurkinaFaso,LaoPDRandZambiaasLDCs,arecoveredbythe2001Recommendation,asisGhanaasanon-LDCHIPCsince2008.Thesefourcountriescanalsobecategorisedasrelativelymoreaiddependentintermsofaid/GNIandaid/governmentexpenditureratios(Table6.1).ODAislessim-portantforVietnam:disbursementswere1.2%ofGNIdur-ing2005-07.AidisofmarginalsignificanceforSouthAfrica(under1%ofcentralgovernmentexpenditure).TherelativeimportanceofloanandgrantODAreflectsvoluntarycom-mitmenttoprioritisegrantaidtoLDCs(BurkinaFaso,LaoPDRandZambia).ThehighproportionofloanODAforVietnamisaccountedforbythethreelargestdonorsbe-ingJapan,FranceandGermany,andthattheycontinuetoprovideacomparativelyhighshareofloanODA.ThestudywastofocusspeciallyontheWaterandSanitationSector(W&SS),andsothereisasubstantialamountofODAinallcases.Finally,thepurposivelyselectedcountriesasagroupreflectbroadlytherangeofreporteduntyingofbilateralaidforindividualcountries(Table6.2)andtogethertheshareofuntied,tiedaidandnon-reportingoftyingstatus(TableB.3inAnnexB)issimilartothosereportedforallODA(Table3.2).
Morebackgroundinformationonthesixcountriesisprovidedintheexecutivesummariesofthecountrystudies(AnnexD).Fivecountries,thatisexceptingSouthAfrica,werealsoincludedinthepre-Accra2008ParisDeclarationEvaluation(PDE)surveysonaideffectiveness.
6.3 Aid modalities and use of country procurement systemsAreviewofaidmodalitiesandusesofCPSonacasebycasebasisprovidesevidencefirstlyonwhetherthereareinfactdifferencesbetweencountriesand,second,onwhethertherearesystematicdifferencesamongstindividualdonorsintheiruntiedaidpractice.ThisreviewbeginswithGhana,thepilotcasestudy,andthenconsidersthethreeLDCscoveredbytheRecommendation–Zambia,BurkinaFaso,andLaoPDR–andnextthetwonon-LDCs,VietnamandfinallySouthAfrica.ThestatisticalevidenceonusesofmodalitiesforeachofthesixcountriesisincludedintheexecutivesummariesasTablesD.1-D.6andusesofCPSsummarisedinTable6.3.
Ghanahasrelativelygoodmonitoringofaid.ThegovernmentwithWorldBanksupportcloselymonitorsaiddisbursementsintheformofthedonorpartnerenvelope(DPE)andreportstotheConsultativeGroup(CG).Thereisahighlevelofreportedfullunty-ing,77%in2005-07,with8%partiallyuntied(almostentirelyEU),11%tiedandonly3%withnon-reportedtyingstatus(Table6.2).Thefindingsofasurveyonbilateraltyingpracticesin2006werealmostidenticaltotheuntyingreportedtotheCRS,confirmingthatHQdataareusedtodetermineuntyingstatus(MAPConsult,2007).Afterthissurveyinterestinmonitoringuntyinglapsed.
Agroupofeightdonors,includingtheEC,providesubstan-tialgeneralbudgetsupportandmakeextensiveuseofother
Table 6.2 Tying status of ODA in six case study countries (2005-07)
Region Country Bilateral ODA
Loans as % of total bilateral ODA
Untied ODA as % total bilateral ODA
Partially Tied ODA as % total bilateral
ODA
Tied ODA as % total bilateral ODA
Not Reported as % total bilateral ODA
West Africa Burkina Faso 2.7 80.5 5.5 7.6 6.4
Ghana 8.1 77.1 8.7 11.0 3.2
Southern Africa Zambia 3.5 85.6 9.0 3.6 1.8
South Africa 14.2 64.8 20.7 10.7 3.7
South East Asia Lao PDR 5.7 70.4 3.0 4.4 22.2
Vietnam 57.9 64.9 6.9 19.4 8.7
Source: OECD Creditor Reporting System (CRS) database.Notes: 1. Data are based on commitments in current US$.
Untying aid: Is it working?34
Chapter 6
modalitiesassociatedwithuntyingsuchasbasketfunding.Collectively,55%oftheiraidin2006-08wasprovidedthroughthesemodalities(TableD.1).Thesedonorsalsoreportedsubstantialuseofcountryprocurementsystems(CPS)intheiraidtogovernment(Table6.3).Twoofthefivecasestudydonors,CanadaandSwitzerland,alsoconfirmedthishighlevelofuntiedmodalitiesanduseofCPS.Incontrast,twomajordonors,Japan,almostentirely,andtheUSA,wholly,continuetouseaprojectapproachandmadelittleuseofCPS(13%and0%respectively).Threeotherdonors,Belgium,ItalyandSpain,providedonlyprojectaid,anddidnotreportdisburse-mentsusingtheCPS.Ghanaianinformants,whilstwishingtocommentinformally,arguethattheonlygenuinelyuntiedaidintermsofownershipandalignmentwereunrestrictedfundsmadeavailabletogovernmentanddisbursedusingtheCPS.
Zambiahasthehighestreportedlevelofuntyingamongstthesixcasestudycountries,86%ofcommitments,withlessthan5%reportedastiedornon-reportedduring2005-07
(Table6.2).GovernmentanddonorswithintheConsultativeGroupagaincooperateinmonitoringaidperformance.Intheabsenceofdonorspecificinformationonuseofmodalitiestendonors,includingtheECandtheWorldBank,wereaskedtoassistinprovidinginformationinresponsetoasurveyaboutmodalitiesandprocurement,andalldidso(TableD.4).ThesurveycomplementedbyinformationcollectedforthePDEconfirmsthattherearewidevariationsindonorpractices.Eightdonors,includingtheECandtheWorldBank,providedsubstantialsupportthroughprogrammaticandpooledfundingmodalities.Inadditiontoprogrammaticsupport,somealsousedtheZambianCPSforpooledprojectaid.Twomajordonors,JapanandtheUSA,continuetouseaprojectapproach,withprocurementoutsidetheCPS.Japanusestheagency’sownprocurementrules.TheUSAfundsaredisbursedthroughnon-stateproceduresofimplementingagencies,contractedunderinternationalcompetitivetender;buttheseprocedureshavetoconformtoanyapplicableUSlegislationonsourcinggoodsandservices.Thesedifferencesindonor
Table 6.3 Bilateral DAC donors (including EC) – Percent of aid to government using Country Procurement Systems(CPS) in 20071
Donor zambia Lao PDR Ghana Vietnam Burkina Faso2 South Africa1 Six countries (average3)
Ireland 100 100 100 100
Norway 96 100 100 99
Netherlands 94 100 97 79 100 94
Spain -- 83 83
Canada 100 82 100 10 100 78
Denmark 87 56 44 99 75 72
UK 93 99 44 29 66
Switzerland 0 91 48 96 88 65
Belgium 39 76 75 63
New Zealand 0 80 100 60
France 80 100 19 86 7 58
Sweden 76 72 0 79 42 54
Germany 90 69 14 55 33 52
Finland 84 0 30 85 50
EC 74 0 32 58 77 42 47
Japan 0 27 13 92 0 26
Austria 24 -- 24
Italy 69 0 0 23
Luxembourg 0 -- 41 21
Australia 0 40 20 20
USA 22 54 0 -- 0 4 16
Source: OECD (2006b); OECD (2008d); OECD (2008e); OECD (2008f); OECD (2008g); OECD (2008h).Notes: 1. Percent of aid using CPSs in 2005 (South Africa did not participate to PD Survey 2008 and therefore no data was collected for 2007). 2. Computed as ‘aid disbursed using CPSs’/’aid for government sector’. However, in some cases aid disbursed using CPSs was higher than aid for government sector and therefore % of aid using `CPSs
was computed as ‘aid for government sector’/’aid disbursed using CPSs’ (See OECD, 2008d for further details). 3. Computed as an unweighted average.
Untying aid: Is it working? 35
Chapter 6
practiceareconfirmedinthePDEsurvey,whichindicatesahighlevelofuseofCPS,74%to100%,exceptingJapan(0%)andtheUSA(22%)(Table6.3).AgainZambianinformantsstressedtheimportanceofunrestrictedprogrammeaidanduseofCPS,implyingownershipandalignment,asindicatorsofgenuinelyuntiedaid.
Burkina Faso:againtheevidenceindicateswidedifferencesindonorpracticesassociatedwithahighlevelofreportedformaluntying,81%(Table6.2).GBSasaproportionofallODAdisbursements,under18%inthelatenineties,hasvariedbeen27%and32%during2001-07.GBShasbeenprovidedsince2001byDACbilaterals,includingin2007Canada,Denmark,France,Germany,Netherlands,SwedenandSwitzerland,aswellastheEC(TableD.2).Theproportionofbilateralcommit-mentsreportedtotheCRSasGBSisevenmorevariable.In2007GBSwas29%ofODAandbasketfunding(Fondscom-muns)werereportedseparatelyforthefirsttimeasafurther7%ofODA.Meanwhileaidfollowingaconventionalprojectapproach,andincludingadecliningproportionofTC,stillaccountedfor64%ofODA(57%forthebilateralsgivingGBSandbasketfunding).Twomajordonors,JapanandtheUSA,aswellasAustria,Belgium,ItalyandLuxembourgcontinuetouseonlyaprojectapproach.TheuseofCPSamongstdonorshasalsobeenvariablewithtwoDACdonors,ItalyandtheUSAmakingnouseofnationalsystemsandothersrangingbetween55%and96%ofsupporttogovernment(Table6.3).Whereinvestmentfeasibilitystudieswereundertakenonasectoralbasis,thiswasmorelikelytoleadtosectoralbasketsupport,orpoolingbydonorscommittedtosuchanap-proach.
Lao PDR:thepatternofbilateralaidanduntyingappearstobequitedifferentfromthoseofthethreelowincomecountriesinAfrica.ThereisasmallproportionofDACloanODAfundingsimilartotheotherLDCs.However,theextentofformalunty-ingislower(70%)withahighlevelofnon-reporting(TableB.3inAnnexB).Thenon-reportingismostlyforJapan(56%),thelargestbilateral,whichprovidesalsoahighproportionofTCrelatedassistance(CRS)andwhich,accordingtolocalinform-ants,isde factotied.BilateralaidisoverwhelmingprojecttypewithlittlecommitmentofGBSinrecentyears.France,SwedenandtoamorelimitedextentJapanuseCPS.ThelowlevelsofGBSandCPSuseareexplainedbydonorsasrelatedtolimitedcountrypartnerfinancialmanagementcapacity.
Governmentreportingonbilateralaidincludesnon-DACAsiandonorsprovidinganuncommonsnapshotoftheiractivities(TableD.5).Asiandonorsprovidedalmost60%ofbilateralaidin2005-06:Japan,27%;Vietnam,11.1%;China,9.5%;Thailand,8.4%;andKorea,2.6%.Informantsindicatedthatvirtuallyallofthenon-DACaid(30%ofbilateralaid)wasde factotiedandalsoJapaneseassistance,apartfromlocalprocurement.43Thisestimatewouldimplythatatleasttwo
43 A different perspective is provided by JICA that “it is not appropriate to treat Ja-pan, a DAC member which has untying obligations, as other Asian non-DAC countries that do not have reporting obligations in the context of untying extent of ODA.”
thirdsofallbilateralaidisde factotied.Furthermore,takingintoaccountapparentlysubstantial,butunquantifiedofficialexportcreditsfromnon-DACsources,theimplicationisthatinpracticeexternalfundingfordevelopmenttoLaoPDR,anLDC,isoverwhelminglytied.
VietnamasarapidlydevelopingLICreceivesahighproportionofDACmemberODAasconventionalprojectaid,mostlyfo-cusedoninfrastructureandeconomicsectors.Thisaidincludesahighproportionofloans(58%ofbilateralDACODAin2005-07)andalsosubstantialtying(19%)andnon-reporting(9%)(Table6.2).ThehighlevelofconcessionalloanODAisexplainedbythethreemajorbilateraldonors,Japan,France(AFD)andGermany(KfW),beingthosethathaveretainedasubstantialloanelementintheirportfolio.Programmaticsupportisarelativelylow8.5%ofODAduring2006-07(TableD.6)andDACbilateralODAin2007includingonly2%ofbilateralGBS.
TheuseofCPSbydonorsisextremelyvariable(Table6.3).Dif-ferentviewswereencounteredtooaboutcapacity,especiallybelowthelevelofcentralgovernment,andcorruptionprob-lems.IncontrasttotheothercasestudycountriestherewasahighusageofCPSunderJapanesesupportedprogrammesin2007underabilateralagreementthatexperienceddifficultiesduring2008.
Vietnamalsoreceivessubstantialbutseparatelyandpartlyreportedconcessionallendingfromnon-DACdonors,includingespeciallyChinaaswellasIndia,Korea,Kuwait,SingaporeandThailand.ForexampleKoreacommittedoverUS$100mnofconcessionalexportcreditsayearduring2007-09whichexceedsthelevelofODAbyallexceptthethreelargestDACdonors.Thelevelsofconcessionalcreditsimplythattheproportionoftiedfundingfordevelopmentis,asinLaoPDR,verymuchhigherthanthereportedlevels.Theextentofuntyinganddomesticorregionalsourcingverymuchdependonthewaysinwhichprojectcontractingandprocurementwithinprojectisorganised.ThegovernmentiswidelyconsideredtoadoptastrongnegotiatingstanceonissuessuchascontractingandsourcingofODAfundedgoodsandservices.
South Africa:bilateralODAisrelativelyunimportantforanup-permiddleincomecountry.Incontrasttothefivelowincomecountries,governmentinsistsontakingdirectresponsibil-ityforcoordinationnegotiatingandorganisingonadirectbilateralbasis.Thereisunusualnichefunding,suchassupportforsomeofthepreparationsforthe2010WorldCup(Clayet al.,2008,Box4.2).AnotherconsequenceislackofstatisticalreportingonaidmodalitiesandSouthAfricachosenottoparticipateinthe2008PDEsurveyofaideffectiveness.
BilateralODAisequivalenttolessthan1%ofpublicexpendi-ture.AlowproportionofODAisreportedastied(10%)withlittlenon-reporting(3%).Bilateralaid,asreportedfor2005,alsoshowsvaryingdonorpracticesinusingpublicfinancemanagementsystemsandCPSprocurement(TableD.3).Three
Untying aid: Is it working?36
Chapter 6
bilateraldonorsrespondedtothestudysurveyindicatingamixofmodalitiesandprocurementarrangementsthroughtheirown,governmentandnon-governmentpartnerproce-dures.Indiscussionthesechoices,andsoassociatedvarianceinCPSuse,werefoundtoberelatedtothewayindividualorasmallnumberofbilateralprojectshadbeendesigned.Soitisdifficulttodrawmoregeneralconclusions.Nevertheless,somedonorswereagainfoundtomakelittleuseofgovernmentsystems,preferringtheirownproceduresorthoseofcontrac-torsthathadtoconformwiththefunder’sownregulationsandlegislationonsourcingandreporting.
TheSouthAfricangovernmenttakesastrongposition,activelypromotinglocalcontractingandquestioningdonorchoiceswheretheseinvolveotherformsofsourcing.Finally,theviewofgovernmentofficialsontyingpractices,andtak-ingEUproceduresonlyasanexampleofawiderissue,werenoteworthyinexpressinginaforthrightwayaperspectivecommontomostcountrypartners:
“FormanySouthAfricanGovernmentofficials,theprinci-pleoftyingoftenreferstotheproceduresandprocessesusedinaidprogrammes.Thefocusisthereforeonthealignmentandownershipofaidwithcountrysystemsoverandabovetheprocurementandsourcingofgoodsandservices.AnexampleofthisistheEUaidmodel,citedbysomeGovernmentofficialsasparticularlytied.ThisperceptionrelateslargelytotheECrequirementthatunlesscertaineligibilitycriteriaaremetandabudgetsupportaidmodalityisused,EUaidis,forthemostpart,governedbytheEU’sPracticalGuidetoContract.44How-ever,inSouthAfrica,thefactthattheEUimposesitsownrulesandconditionsonhowODAisspentisinitselfseenasproblematic.Insuchcases,theuseofdonorcountryprocurementsystemsissometimessynonymouswithtying.” (Ramkolowan and Stern, 2009).
Overall Findings:Therearecertainiterativepracticesthatemergefromareviewofthesixcountries.First,theextentofformalunty-ingandchoiceofaidinstrumentsisbroadlysensitivetocountrydevelopmentstatusandcurrentfinancialcircumstances:thehighestlevelofuntyinganduseofgrantaidisfoundinthefourHIPCs,BurkinaFaso,Ghana,LaoPDRandZambia.Second,theex-tenttowhichdonorshaveadopteddirectbilateralfundingmo-dalitieswithuntiedfundsisverycountrycontextspecific,whichthejointevaluationofGBSconfirms(IDDandAssociates,2006).Third,therearehoweverwidedifferencesindonorpracticeonlypartlyreflectedinthefivedonorcasestudies.TheNordic+groupincludingCanadawithFrance,Germany,LuxembourgandSwit-zerland,andmostrecentlyAustraliahavefundedsucharrange-ments,alongwiththeMFIs,especiallytheWorldBank.
Fourth,thereisapaucityofreliablequantitativeinformationontheextentofpooledfundingarrangements,whichdonot
44 Procedures for EC external actions (commonly referred to as PRAG). Under these guidelines EU aid to South Africa is only partially tied to the EU states; the state that is a beneficiary of assistance and developing countries as specified by OECD DAC.
relatetoconventionalclassificationsofODAsuchassectors,instruments,TCcomponentsortyingstatus.Ontheevi-denceprovidedthesemodalitiesappeartoinvolvesubsetsofdonorswithinthewidergroupthatarealsowillingtoprovidedirectbudgetarysupport.The‘experiments’withpooledfundinggoingbeyondconventionalparalleljointfundingarealsoinwhatdonorsperceivetobefavourablecountry,sectoralandprojectspecificcircumstances.Thesedevelopmentscannotbesaidtobetheresultofuntying,whichisformallynowalmostgeneral.Rathertheseinitiativesarenowmorelikelywhereasectoralapproachisadoptedforinvestmentplanning.Thosemorelikelytoparticipatehaveregulatoryflexibilityinthewaysfundscanbeused,includingcriticallytransferringthesetoanotheragencyorthecountrypartnertodisburse.Somedonors,includingthetwolargest,theUSAandJapanhavenotparticipatedsubstantivelyinthesearrangementsorGBS.NorhavesomeEuropeandonorssuchasAustria,Italy,Spain,Portugal,theCzechRepublicandGreece,eventhoughpoolingwouldseemtoofferrelativelythesmallerdonorbothanopportunitytopartici-pateinmajorinitiativesandtoachieveefficiencysavingsinaidadministration.Thereissofaralsoalackofsubstantiveevaluativeevidenceontheaideffectiveness,letalonethedevelopmentalimpactofsuchinitiatives.
ThelinksbetweenchoiceofmodalitiesandtheuseofCPSareapparentbutcannotberigorouslydemonstrated,becausetheavailabledatadonotfocusonthesepotentialconnections.ThePDEcountrysurveysprovidesomeusefulcross-sectionalevidenceonuseofcountrysystems,withwhichthefindingsofthecasestudiesarebroadlyconsist-ent.Againtherearewidedifferencesindonorpracticethatarebothprogrammaticandtosomedegreereflectcountrycircumstances.
Thestudysoughttoobtainmoredetailedinformationaboutpoolingandbasketsandprocurementarrangementsatanagencylevelwithincountries,butwithverylimitedsuccess.TheexplanationappearedtobesomecombinationofdonorsurveyfatigueafterworkpreparatorytotheSeptember2008Accrameeting,theextentofin-countrydonorcoordinationoninformation,andsensitivityinsomecasesoverprovid-ingevidenceaboutprocurementissues.Someprojectswerethereforeselected(seebelowChapter7)formorein-depthreviewasexamplesofthewaypoolinganduntyingmaybelinked.
Thewidespreadscepticismexpressedinallsixcountries,butofteninformallyandofftherecord,aboutthegenuinenessofuntyingonthepartofsomeifnotmanydonorspromptedthestudyteamtoconsiderthisissueintwoways.First,ifthereisextensivede factotying,thentheeffectmaybeobservableasanaidassociatedbiasinfavourofdonorexportsortradedis-tortion(Section6.4).Second,asprojecttypeaidstillpredomi-nates,actualpracticesshouldbemoretransparentataprojectlevel,asconsideredinChapter7.
Untying aid: Is it working? 37
Chapter 6
6.4 Econometric analysis of the implications of bilateral aid for donor exports
6.4.1 Introduction to econometric analysisIfoneconsiderstheexportsfromdevelopedindustrialdonorcountriesthatconstitutetheDACtoaidrecipientdevelopingcountries,thenonewouldexpecttheflowofexportsfromadonorcountrytoarecipientcountrytobeexplainedbyfac-torssuchasthesizeoftheexportingeconomy,theproximityordistance,andthesharingofacommonlanguage.Eventhetotallevelofaidmightbeaninfluenceontheexportsofalldonorcountries.However,inamulti-donorworld,wherenosingledonordominatesastheUSdidduringtheMarshallPlanera,thereisnoreasonwhythelevelofaidfromadonortoarecipientwouldinfluencethatdonor’sexports,unlessbilateralaidwerehavingatradedistortingeffectthroughintendedorunintendedtying45.Tyingiswidelyconsideredtoadvantageexportersinthedonorcountry.Untyingisthereforeexpectedtoreducethedistortingbiasofaidinfavourofthedonor’sexports.Similarly,particularaidinstruments,notablybilateralconcessionalloans,werethoughttobeoftenassociatedwithdonorexports,forexamplethroughamixedcreditarrange-ment.Sophasingoutofsucharrangementswithadeclineinloansagainoughttoenddistortions.
Thissectionthereforeinvestigateswhetherrecentbilateralaidflowsandassociatedpracticeshavetradedistortingeffects.TheaimistounderstandwhetherbilateralODAisaninfluenceonadonor’sexports,evenwithcurrenthighlevelsofunty-ing,whilsttakingintoaccountothervariablesthatmightbeinfluencingthisrelationship(suchasFDI,exportcredits,ECdisbursements).Conventionally,tyingstatusandthedifferentformsinwhichaidisprovided,forexampleloansandgrants,havebeenhypothesisedtoinfluenceaggregatedonorexportflowsatthecountrylevel.
Inordertotakeintoaccountcountryspecifics,asimilareconometricinvestigationhasbeingundertakenforeachofthesixcasecountrystudiesandthecomparativeresultsoftheanalysisforthesixcountriesarereportedinthissection.Thesixcountriesarepurposivelypairedaccordingtotradingregiontoisolatepotentialregionaleffects:GhanaandBurkinaFasoinWestAfrica,LaoPDRandVietnaminSouthEastAsiaregionandASEAN,andZambiaandSouthAfricainSADCregion.Moreover,thesixselectedcountriesalsodifferfortheirDACclassification:BurkinaFaso,LaoPDR,ZambiabeingLDCs,SouthAfricaasUMIC,andVietnamandGhanaasOLICs;BurkinaFaso,GhanaandZambiaarealsobeingclassifiedasHIPCs,aswellasbeingfullACPmembers.Thescopeofthein-vestigationanditstheoreticalframeworkaresetoutinAnnexC,withonlysummaryfindingspresentedinthissection.
6.4.2 DataEmpiricalestimationincludesaidfromupto22DACdonorstorecipientcountries,withobservationscoveringtheyears
45 This conjecture is explored further through the project level analysis in Chapter 7.
2002-0746becauseofdataqualityissuesandavailability47.ThebilateralODAdataaretakenfromtheOECDCRSdatabase.Theanalysisusesaiddisbursementsandnotcommitments,becausetradeeffectsareexpectedtofollowactualdisburse-mentsandnotcommitments.Theavailableexportseriesincludeonlygoodsandsounder-estimateeconomicflowsfromthedonortorecipient,whichinvolvealsoTCandotherservicesin-kindfinancedatthedonorend.ThesedataissuesarediscussedfurtherinAnnexC.
6.4.3 FindingsThestatisticalresultsattheindividualcountrylevelsuggestprobabilisticandnotnecessarilyrobusteconomicrelation-ships.Therelationshipswhicharefoundmayinturnbeonlyaproxyformorecomplexunderlyingeconomicrelationshipsbetweendonorandrecipientcountry.However,acomparisonofresultsfromallsixcountries,aspresentedhere,providesafullerandpossiblymorerobustanalysis,highlightingsimilarpatternsandsystematicdifferences.Theseresultsaresumma-risedinTable6.4anddiscussedmorefullyinAnnexC.
GDP of donor countries,whichproxiesfortheireconomicsize,is,asexpected,animportantdeterminantofdonorbilateralexports.Thelargerthedonorcountry’sGDPthehigherthelevelofexports,ascountriessuchasJapanorUnitedStatesclearlyexportmorethanLuxembourgorSwitzerland.
46 2002-06 for South Africa.
47 Further details can be found in Annex C and the background paper (Natali et al., 2009).
Untying aid: Is it working?38
Chapter 6
Astrikingdifferenceinresultsamongcountriesrelatestothecommonlanguage.Language,aswellasmoregeneralcom-monculturalfactors,canhelptofacilitatetradeflowsbetweencountries.SharingthesameofficiallanguageishoweveronlystronglyandpositivelylinkedwithbilateralexportflowsforFrancophonedonorstoBurkinaFasoanddoesnotholdforthethreeAnglophonecountries48,wherenoobservableadvantageisconferred.ApartfromthesharedlanguagetherearethelinkswithFranceastheformercolonialpower;itisthereforehypothesizedthatasimilarrelationshipmightbefoundbetweensomeSpanish-speakingLatinAmericancountriesandSpain.
Otherexpectedinfluencesarehoweverfoundtobemostlyinsignificant:the level of development of donor countries,meas-uredbytheirincomepercapita,hasnoclearinfluenceonbilateralexportflowsandonlyinVietnamthereisevidence,inaccordancewiththetheory,thatanincreaseinthedistanceisassociatedwithalowerlevelofexportsfromdonorcountries.
Turningtotheinfluenceofaidflows,itisimportanttodis-tinguishbetweenadonor’sODAandODAthattherecipient
48 In Vietnam and Lao PDR the common language variable is automatically dropped as no donors share a common official language with these countries.
countryreceivesfromallotherbilateraldonors.Thebasicideaistoconsiderwhetherthereareanytradedistortingeffectsofaidatanaggregatedlevelbyexaminingwhetherbilateralaidfromaparticulardonortoarecipientcountryaffectstradebetweenthatdonorandrecipientpairdifferentlyfromaidfromotherdonors.
Resultsshowthataggregatedaidflowsfromadonorhaveasignificantimpactonthatdonor’sexports,whereasODAfromtherestofbilateraldonorsisnotasignificantinfluenceonthatdonor’sexports.Thissuggeststhatanincreaseindonoraggregateaidflowsisassociatedwithariseinexportsevenwhereaidisoverwhelminglyuntied.AlsoincountriessuchasSouthAfricaandZambia,wherethisaid-traderelationshipdoesnotseemtoholdatanaggregatedlevel,adisaggrega-tionshowsthatgrantsfromthedonorcountryarepositivelyassociatedwithitsexports,whereasgrantsfromtherestofbilateraldonorsarenotsignificant.Suchresultsareagainsuggestiveofsomeformoftradedistortion:itisdifficulttoexplainwhybilateralgrantsfromaparticulardonortoarecipientcountryshouldaffecttradebetweenthatdonorandrecipientpairdifferentlyfromgrantsfromotherdonors.Thisresultissuggestiveofinformalorde factotyingofgrants.Amongotherpossibleexplanationstherecouldbethegood-
Table 6.4 Aid and other influences on DAC donor exports: comparative econometric results for six partner countriesduring 2002-07
West Africa Southern Africa – SADC South East Asia – ASEAN
Dependent variable: Exports Ghana Burkina Faso South Africa zambia Lao PDR Viet Nam
GDP (+)*** (+)*** (+)*** (+)*** (+)*** (+)***
GDP per capita (-)***
Distance (-)***
Common language (+)***
FDI (+)***
Import residuals (+)*** (+)*** (+)***
Tying status % (+)** (+)** (+)**
Grants i (+)*** (+)** (+)*** (+)** (+)*** (+)***
Grants All-i
Loans i (+)** (-)*
Loans All-i (+)**
Export credits
EC disbursements (+)** (+)*** (+)** (-)***
N observations 117 102 75 95 103 126
Period under analysis 2002-07 2002-07 2002-06* 2002-07 2002-07 2002-07
Notes: 1. All variables are in natural logarithms (except Common language, Import residuals and Tying status). All regressions include a time trend not reported.Legend: * significance at 10 percent **significance at 5 percent *** significance at 1 percent.
Untying aid: Is it working? 39
Chapter 6
willontherecipient’sparttowardsthedonor;still,theaidrelationshipmightsimplyfacilitatetradebetweenthedonorandtherecipient,increasingrecipient’sproclivitytoprocuregoodsfromthedonor(reinforcingcommercialties)49.Loans,whichareanunimportantformofaidexceptinVietnam,arenotasignificantinfluenceondonorexports.Thisresultmightbeexplainedinpartbytherelativelevelsofgrantsandloansinthecountriesanalyzed,andsoadisaggregationofODAintoloansandgrantsisnotnecessarilyverymeaningful.
Tosumup,theaidvariablesarefoundtobeanimportantin-fluenceondonorexports,suggestingthataid,despiteformaluntying,isneverthelessasignificantpartofthecomplexcom-mercialrelationshipbetweendonorandrecipient.Althoughthereisevidenceofapositiveimpactofformal tied aid ondonors’exports(especiallyinSub-SaharanAfricancountries),ratheritisaidoverall,evenifformallyuntied,whichappearstohaveatradedistortingimpactonbilateralexports.
Theanalysisalsofoundthatdisbursements from the ECposi-tivelyassociatedwithexportsinfullACPmembercountries(BurkinaFaso,GhanaandZambia)supportingtheideathattheACP-EUrelationshipisapositiveinfluenceonexportsfromEUmemberstatesinACPrecipientcountries.
Foreign direct investments(FDI)areanotherpotentiallyimpor-tantinfluenceontradebetweencountries.DuetolackofdataforindividualDACmembersthisrelationshipisstudiedonlyforSouthAfrica,whereFDIarefoundtohaveapositivebear-ingonexportflows.ThereareseveralwaysbeyondtheinitialinvestmentinwhichFDIandexportscouldbestronglylinked.Noevidenceisfoundthatexport creditshaveanimpactonbilateralexportflowsinthosecountrieswheredatawereavail-able,aresultperhapsrelatedtothepoorqualityofthedata50.
Thestatisticalanalysiscanonlyhighlightinfluencesthatarecapturedwiththeavailabledata.Thereareotheraspectstotheeconomicrelationshipbetweendonor-exporterandaidrecipientimporter,oftendifficulteithertoidentifyormeasure,soanattemptismadetocontrolforsuchunderlying relation-shipsbytakingintoaccountimportsbythedonorfromtheaidrecipient51.InSouthAfrica,LaoPDRandVietnam(middleincomeand/orSEAsiancountries),theresultsaresuggestiveofacomplexreciprocalcommercialrelationshipcomparedwithaperhapsmoreaiddependentoneforthethreeAfricanACPcountries.Ifitistruethatasaresultofthedonor-recipientrelationship‘theeffectivecostofgeographicdistance’isre-duced(JohanssonandPetterson,2009),thenfurtherresearchisneededoninter-countrytradeflowsbetweendonorandrecipientsandnotjustondonorexports
Theresultsofthiseconometricanalysissuggestthatbilateralaidisstilladistortinginfluenceontrade,evenafteruntying
49 See theoretical framework in Annex C for further details.
50 See Annex C for further discussion.
51 Through the inclusion of an import residual variable (Annex C).
ofmostbilateralODA.Thatfindingjustifiesacloser,disaggre-gatedexaminationoftheusesofaidandsourcingofgoodsandserviceswithaidfunds.Theresultsespeciallyfordisag-gregatedaidflowssuggesttheexistenceofsomecontinuingtradedistortingeffectsofaidinthepost-tyingworld,atleastforgrants.Loans,whichareanunimportantformofaidinmostofthecountriesanalyzed,arenotasignificantinfluenceondonorexports.
6.5 ConclusionsTosumup,thecountrylevelanalysishighlightshowtheper-ceptionsofdonorsandpartnercountriesontyinganduntyingdiffer.Whereasfromadonors’perspectiveuntyingisdirectlyrelatedtoremovingsourcingrestrictionsonprocurementofgoodsandservices,recipientcountriesperceiveuntyingasabroaderconceptrelatedtoalignmentandownershipofaidandlocalcompaniesbeingabletocompetesuccessfullyforcontracts.
ThereviewofaidmodalitiesanduseofCPSdrawsattentiontosystematicdifferencesamongstbilateraldonors.Donors,suchastheNordic+Group,France,GermanyandSwitzerland,whomakelargeuseofprogrammaticapproachesandCPScontrastwiththetwomajordonors,theUSAandJapan,whoarefoundtohavemainlyaproject-typeapproachandmaderelativelylittleuseofCPSinthecountriesunderreview.Otherdonorstooarefoundtofavouraproject-typeapproachbutarenotactiveinallcountries.
Thereviewalsoindicateduntyingpracticesarecountry-specif-ic.Therearedifferencesbetweenpartnercountriesaccordingtotheirdevelopmentstatus(thehighestlevelofuntyinganduseofgrantaidisfoundinthefourHIPCs);totheirnegotiationpower(asSouthAfricaandVietnamcasestudiesshow,theex-tentofuntyingisrelatedtopartnercountries’ownershipandstrengthofnationalsystems);andtotheirin-countrycapaci-ties(theextentofuntyingcanbelimitedbylackofcountrycapacity).InLaoPDRlittlecommitmenttotheuseofGBSandCPSismainlylinkedtolimitedcountrypartnerfinancialmanagementcapacityandpossiblytotherelativeimportanceofnon-DACdonors.
TheanalysisofaidmodalitiesanduseofCPSalsohighlightthepaucityofexistingdataonaidmodalitiesandinparticularonpooledfunding.Somecountrystudies,suchasBurkinaFaso,LaoPDRandVietnamalsoindicatediscrepanciesbetweenthepatternofaidsuggestedbytheCRSandlocaldatatowhichgovernmentsanddonorsmakereferenceatacountrylevel.
Finally,doubtsonthegenuinenessofaiduntyingseemtobesupportedbyaneconometricanalysis:aidisfoundtohavesometradedistortingimpactinaposttyingworld.
Untying aid: Is it working?40
Chapter 7
7.1 Introduction
Objectives and scope
T hischapterpresentstheevidencefromthe21projectcasestudiesundertakenaspartofthesixcountry
studies52.Anumberofkeyfeaturesofuntyingpracticehavealreadybeenconsideredandthischaptertakesasitsstartingpointthetheoretical,econometric,donorandrecipientcountryanalysesofthepreviouschapters.Themainaimistocomplementtheseperspectiveswith‘reallifeexamples’fromcurrentpracticeattheprojectlevel.ThepilotstudyforGhanashowedhowmanyofthevariationsinpracticecanonlybeproperlyunderstoodintheirindividualcontext.
Theprojectlevelreviewisorganisedintwoparts.Section7.2considersthecasestudiesintermsofinputfeatures;whofundedwhat;andthetermsandcontractingarrangementsemployed.Thentheconsequencesareconsideredofthewayprojectsareorganisedandinputssourcedforaidefficiency,effectivenessanddevelopmentinSections7.3to7.7.
Case study approachThecasestudymethodwasdevelopedduringtheGhanapilotcountrystudyontheassumptionthatmanyofthefeaturesofuntyingcanonlybeunderstoodintheirprojectcontext.Foreachcountrytheprojectswereselectedjointlyinconsultationwithcountrypartneranddonoragenciesusingtheavailableinforma-tionfromtheCRSandcountrydatabases.Selectedprojectsweretobepredominantlyuntied,intheWater&SanitationSector(W&SS),involveprocurementofgoodsand/orservicesandbecurrentlyactive(2008).Countryteamsthencarriedoutaseriesofstructuredinterviewswithrecipientanddonorpersonnel,aswellasendusers,contractors,NGOsandothergovernmententities,includingthepublicprocurementagencies.Fromtheseinter-viewsandsupplementaryinformationprovided,aprofileofeachprojectwaspreparedcoveringdisbursementoffundsateverystage,thedecision-makingprocessandfactorslikelytoinfluence
52 The EXSUMS are in Annex D and full versions of the reports will be available on the PDE website in due course.
outcomes.Forsomeprojects‘Ground-Truthing’visitstoprojectswerecarriedouttoverify‘capitalcitylevel’informationandviewsaboutwhatwashappening.
Practically,eachcountryteamwasabletoinvestigatethreeorfourprojectsindepthduringthetwomonthperiodallocated.Thepilotconcludedthattoensurefocusandcomparabilitybetweenprojectsthestudiesshouldconcentrateprimarilyononesector,watersupplyandsanitation.Thissectorwasse-lected,becauseitisbothaninfrastructureandgrowthsectorandhasastrongsocialorientationwithbothruralandurbancoverage.Itinvolvestheprocurementofsubstantialamountsofbothphysicalgoodsandcivilengineeringservices,aswellashavingasignificantTCcomponentalongsideavarietyofaiddeliveryapproaches.ODAusuallyplaysakeyroleandmanybilateraldonorsareinvolved.Theprojectselectionproc-essaimedtoprovidewidestcoverageofdonorsacrossthesixcountries,andsosomeprojectsareinothersectors.Thecasestudiesshouldberegardedasprovidingexamplesofwhatagenciesconsidergoodpractice.
Tying and untying in practice53
Tyingisultimatelyaboutoutcomes,thegeographyofsourcing,andnotjustaboutlegalorallowablepossibilities.Whenthereisformaltying,becausecontractsarespecificallyrestrictedtoorganisationsregisteredinthedonorcountry,ortheuseoffundsislimitedtobuyinggoodsandservicesfromthedonormarket,theoutcomeisunambiguous.Aidcanbeconsideredtobeeffectivelyuntiedwherethecontractisawardedtoanorganisationregisteredintherecipientcountryortosomethirdparty.54Thecontractmaybetendered
53 Unlike at the theoretical level, it is also possible to consider the effects of partial untying on a case by case basis because there is an answer to the question of which countries are included. Depending on which set of countries are unrestricted, partial untying can be closer to either tied aid or untied aid in terms of its effects.
54 There may be disagreement as to the genuineness of such a registration. Some informants specifically questioned whether ‘Chinese’ companies registered in Ghana were genuinely local, having regard to the possible implications for employment and sourcing of goods.
7. An AssEssMEnt of tHE EffECtivEnEss of untiEd odA: findings of proJECt studiEs
Untying aid: Is it working? 41
Chapter 7
internationallyoronlynationallybythefunderorthroughthecountryprocurementsystem(CPS),orevendirectlyawarded.
De factotyingariseswherefundsareformallyuntiedorpartiallyuntiedbutthecontractisinfactawardedtoadonorcountryorganisation,orgoodsandservicesdirectlyprocuredfromthedonormarket.Thisoutcomeisprima facieevidenceofde factotying,butnotconclusive.Theoutcomemaybetheintendedorunintendedconsequenceofthewayinwhichthecontractisorganised.Someinformantsarguethattheveryactivitiesselectedforfundingmayleadtoanunintendedbiasinfavourofsourcinginthedonorcountry,andreflecteitherthedonor’sownortherecipient’spercep-tionofthefunder’scomparativeadvantage:forexample,NorwegianTCforoilexploration,ChineseroadconstructionorJapanesecarsormotorcycles.
Thecontractingprocesshastobeexaminedtodeterminewhetherthereareobstaclestoorganisationsoutsidethedonorcountryobtainingacontract,suchasprequalificationrequirementsorlackofinformation.Theseconsiderationsleadtocategorisingcontractingorprocurementoutcomesas‘untied’ifsourcedotherthanfromthedonorcountry.Butifacontractgoestoadonorcountryentity,thatisatleastanambiguousoutcome,potentiallysuggestiveofde factotying.
Todemonstratede factotyingisdifficult.Acase-by-caseinves-tigationcanindicatewhetherthereisabiasinfavourofdonorcountryorganisations.Practicesareagencyspecific,boundbyorganisationalandnationalregulatoryframeworks,asillustrat-edinChapter4,andsoacomparisonofapurposivesampleofprojectswillhighlightdifferencesandtheirimplications.Whatisdifficulttoestablishisacounterfactualexample,tocomparewithselecteddonorprojects.However,theIDA(WorldBank)reportsonthegeographicalsourcingofcontracts,usuallybyICB,NCB,qualitybasedselectionand‘shopping’andthisinfor-mationprovidesasetofoutcomesagainstwhichtocomparesurveyedbilateralprojects55.
Finally,cost-effectivenessanalysiscanprovideanexposttestoftheefficiencyofchoices.Ifoutcomesofrestrictivesourcingdecisionsareinefficientcomparedwithactualorlocalmarketpricesorcostingofhypotheticalalternativetransactionsthatissuggestiveofabias.
Aprojectcaseapproachispotentiallyverytimeandresourceintensive,andsoisonlypossibletoundertakeonahighlyselectivebasis.Thisapproachalsorequirescooperationoffunders,partnersandimplementingagentsinmakingavailablewhatmayberegardedasprivilegedorcommer-ciallysensitiveinformation.Thepilotindicatedthatprojectanalysiswasonlyfeasibleforcurrentprojectswithpersonnelanddocumentationavailablein-country.Sothatlimitedthe
55 Another reference point to compare with IDA sourcing is the reporting of ex ante contract awards to the DAC. This outcome measure of untying could provide a valu-able form of verification to compare with the formal commitments (input measure) also reported by donors.
choiceofprojectsanddonors.Countryteamsalsofoundthatcooperationdependedonanorganisationhavingacultureoftransparency,andthatsomeinvestigationswereconstrainedbyslowandincompleteresponses.
Theinvestigationsrecognisetheimportanceoflevelsofcontractingandsofocusonthreelevelsofprocurement.First,thereisthehigherlevelimplementingentity.Insomecasesthisisthecountrypartnerministryortheagencytowhichithasdelegatedresponsibility.However,inmanycasesthereisaheadcontractforthemanagementoftheprojectorthepro-videroftechnicalcooperationtothelocalmanagingagency.Thiscontractcommonlyinvolvessupervising,disbursementandmonitoringofexpenditureonbehalfofthefunder.Thebilateralfunderconventionallyawardstheheadcontract,andthereportedtyingstatusofprojectsisbasedontheformalcontractingrulesadopted,andwhetherornottheseatleastintheoryallowunrestrictedsourcing.Second,sub-contractingoccurredwithinalmostallinvestmentprojectsorganisedbytheheadcontractororlocalimplementingorganisations,andsomaybesubjecttodifferentcontractingrules.Third,thereisactualprocurementorpurchaseofgoodsandservicesbyheadorsubcontractors.Anassessmentofthetyingstatusintermsofoutcomesofaprojectshouldtakeintoaccountallthreelevelsofprocurement.
ThelimitedscaleoftiedODAhasresultedinfocusinglargelyonformallyuntiedODA.Thisfocusiscontinuedattheprojectlevel:thoseidentifiedforreviewwereformallyfullyuntied,exceptinfourcases,whichwereexplicitlymixed,orhybrid,in-cludingbothtiedanduntiedcomponents.Despitethisrestric-tion,thecasestudiesencompassawiderangeofvarietyoftypesofaidanddonorpractice,demonstratingtheincreasedrangeofdonorpracticewithuntiedaid.
ItwastheintentiontotryandincludeatleasttwoprojectsforallofthemajordonorsandatleastoneprojectforallotherDACmembers.Limitingfactorswerethenumberofprojectcasestudiesthateachcountryteamcouldcarryout,thedonorsthatwereactiveineachoftherecipientcountriesintheW&SSandtherequirementthattheprojectswereactivein2007/8toensurethatthedocumentationandatleastsomeoftheimplementingstaffwerestillavailablein-country.ThestudywasabletoincludeprojectssupportedbyAustralia(AusAID),Canada(CIDA),Denmark(Danida),theEU(EC),France(AFD),Germany(KfW),Ireland(IrishAid),Japan(JICA,MoFA),Luxembourg(LuxDev),Netherlands(VNG),Sweden(Sida),UK(DfID)andUSA(USAID).AWorldBank(IDA)projectwasalsoincludedasacounterfactualcasestudy.Switzerland(SECO,SDC)providedusefulinformationregardingtheirbudgetarysupportandusefuldiscussionswereheldwithNo-rad,GTZandtheMCC.NoprojectsfundeddirectlybyAustria,Belgium,Italy,NewZealand,Norway,Spain,orSwitzerland,wereincluded.Severalnon-DACdonorswerealsoincludedaspartneragenciesinspecificprojects,aswereseveralothermultilateralssuchastheregionaldevelopmentbanks,UNagenciesandtheGlobalFund.WaterAidandseveralprivate
Untying aid: Is it working?42
Chapter 7
foundationsarealsoincludedaspartfunders.TheprojectcommitmentsrangedfromapproximatelyUS$500thousandtoUS$1.5billion.
Thegeographicalcoverageofthe21projectsisshowninTableB.2inAnnexB,andtheprojectsarealsonumberedforeasyidentificationwhentheyarecited.TableB.2alsoliststheformalcharacteristicsoftheprojectsasreportedtotheOECD,donorandcountrydatabases.
InSouthernAfrica,theprojectsselectedinSouthAfricaincludetwointhehealthsectorandonlyoneintheW&SS,wheretherewerefewbilateraldonorsupportedprojects.Allthreemostlyfundedtechnicalcooperation.InZambiathreeprojectswereselectedintheruralW&SS,mainlycoveringinvestmentactivities.AUSAIDprojectprovidesTCforBusi-nessDevelopment,andaWorldBanksupportedprojectwasincludedforcomparison.
IntheASEANregiontherearefourVietnameseinvestmentprojectsintheW&SS,oneamulti-donorloosepoolingar-rangement.ProjectsinLaoPDRincludedroads,ruraldevelop-ment,energyandoneintheW&SS.TheUS$1.5billionenergyprojectisjointlyfundedandledbytheWorldBankandhasmanyprivatesectorpartners.
InWestAfricatheBurkinaFasostudyincludestwohealthsectorandtwoW&SSprojects,allexceptoneinvolvingmulti-donorfunding.TheGhanapilotstudyconsideredtworuralW&SSprojects,includingonepooledarrangement.
7.2 Input features: contracting arrangements and tying practice
7.2.1 Organisational features and tying statusTheprojectssurveyedareallconsideredintermsoftheiror-ganisationalandcontractingarrangementsthatexplaintheirtyingstatus.Thestudyteamslookedcloselyintoorganisation-alandespeciallycontractingarrangementssoastoestablishsourcingoutcomes.ThesefindingsaresummarisedinTables7.1.and7.3.
Formal tying statusThestudypurposivelyselectedprojectsthatarereportedasformallyuntied.ThirteenbilateralprojectsandanIDAsup-portedprojectwerereportedasfullyuntied.AlltheECsup-porttoprojectsisreportedaspartiallyuntied.Fiveprojectswerereportedashavingbothtiedanduntiedcomponents.AsacomparisontheCRSdatabaseindicated79%ofbilat-eralprojectsbyvalueinthesamesixcasestudycountriesareuntied(Table3.3).Amajorfactordeterminingformaltyingstatuswastheprojectstartupdate.Olderprojectsoriginatinginanearlierdonorpolicycycleweremorelikelytobereportedashavingtiedcomponents.Forexample,theCIDAsupportedW&SSProject21inGhanaincludestiedelements,whereasthemorerecentiterationofthisCIDA
projectsupportisfullyuntied.Similarly,theKfWProject5inZambiaandthejointfundedProject18inBurkinaFasobothbeganbeforethe2001Recommendation.
Untying in practiceGovernmentandotherinformantsinallpartnercountriesex-pressedsurpriseattheextentofformaluntyingasindicatedbytheCRS.Thereis,asdiscussedinChapter6,adifferentconceptionofuntyinglinkedtoownershipandalignment.Additionally,bothlocalandinternationalprivatesectorcon-tractorsonlysawaidasuntiedwhentherewasareasonableprospectthattheycouldbidandwincontracts.So,fromtheperspectiveoflocalfirms,anICBthatwouldbeawardedtoaninternationalcompanyandbeyondtheirreachappearedtohaveatying-likerestriction.Thekeytothisgulfbetweendonordeclarationsandcountrypartnerperceptionsislinkedtode factotying.Makingthesimpledistinctionthatifacon-tractissourcedinthedonorcountrythisissuggestive(prima facieevidence)oftying,theprojectcasestudiesareusedtoestablishfirst,whatistheextentofuntying,andsecond,whatinfluencesexplainthedifferentoutcomes.
I. The extent of untying amongst case study projectsBearinginmindthatthestudyspecificallytargetedformallyuntiedprojects,13ofthesampled21bilateralprojects(62%)plustheIDAprojectwereformallydescribedasbeingfullyuntied(Table7.1).Oftheremainingprojects,thetwoECprojects(2and16)weredescribedas100%partiallyuntied,andfiveprojects,(around25%)numbers3,6,14,18,and21,weredescribedashavingbothtiedanduntiedcomponents.Theclassificationoftyingstatusisdeterminedbywhethertheheadcontractisformallyopentounrestrictedinternationalcompetition.
Incontrasttoreporteduntyingstatus,lookingatthegeo-graphicalsourcingofheadcontracts,generalTCandprojectcomponents,onlyeightoutof21(Projects2,5,7,9,10,15,17,19)wereunambiguouslyfullyuntied.Themajorityofprojectswerehybridormixed(Table7.1).
II. Components of projects that are more likely to be formally tied or de facto tiedThemostcommonlytiedcomponentsorthoselikelytobesourcedinthedonorcountryweretheheadcontracttomanagetheprojectanditsfunds,or,whereaprojectislocallymanaged,theprovisionofeitherservices-in-kindand/orcon-sultancycontractsforTC.Oftentheseaspectswerebundledintoasingleheadcontract.
Ten(around50%)ofthebilateralheadcontractsdirectlysignedbythedonorwereawardedtodonorcountryorganisa-tions.Whereheadcontractswereprocuredthroughframe-workagreements(bythedonorHQ),100%ofthesecamefromthedonorcountry.
GeneralTC/consultancycontractstellasimilarstory.In13(62%)ofthe21projectstheTC/consultancywasofdonor
Untying aid: Is it working? 43
Chapter 7
nationality,infour(19%)projectstheTC/consultancywasofDACmembernationality’leavingonlyfour(19%)projectswheretheTC/consultancywasofrecipientnationalityandnoprojectswheretheTC/consultancywassourcedfromanon-DACthirdparty.
Table 7.1 Project case studies: recipient relationship, implementing entity, head and TC contracts
Project Number
Recipient involvement
Implementing entity
Head contractprocurement
Head contract nationality
TC/Consultant procurement
TC/Consultant nationality
Untied Aid share
(top level)
SA1 + Private Framework agreement
Donor Donor, Framework Agreement
Donor 82%
SA2 ++ Recipient Competitive Recipient Recipient Recipient 100%*
SA3 + Recipient Tied Donor Donor Donor 31%
ZM4 - Private Competitive Donor Unknown Donor 71%
ZM5 + Recipient Competitive Recipient Donor Recipient 100%
ZM6 - Recipient Competitive Donor Donor Donor 15%
ZM7 + Recipient Unknown Unknown Recipient Recipient 100%
VM8 ++ Recipient Competitive Recipient Private Donor 92%
VM9 ++ Recipient Competitive Recipient Recipient Recipient 100%
VM10 ++ Recipient Competitive Recipient Recipient DAC member 100%
VM11 ++ Recipient Competitive Recipient Donor DAC member 98%
LA12 ++ Recipient Competitive Recipient Donor Donor 90%
LA13 - Private Competitive Donor Donor Donor 95%
LA14 + Recipient, Private Tied Donor Donor, Tied Donor 0%
LA15 + Donors, Recipient Non-competitive Donor Donor Donor 100%
BF16 ? Recipient, Donor Not applicable N/A Donor Donor 57%
BF17 ? Recipient, Donor Competitive Donor Donor DAC member 100%
BF18 ++ Recipient Tied Donor Competitive Donor 92%*
BF19 + Recipient Not applicable N/A Donor DAC member 100%
GH20 + Recipient Framework agreement
Donor Donor, Framework agreement
Donor 100%**
GH21 - Private Tied Donor Donor, Tied Donor 74%
Notes: *Partially Untied, ** Excluding services-in-kind
key to headings:Project Number: Reference for project details in Table B.2Recipient involvement: The amount of involvement by the recipient partner (dominant = ++, significant = +, mixed = ?, little = -).Implementing entity: The dominant implementing entity.Head contract procurement: The procurement systems used for the head contract, the primary contract signed with the donor.Head contract nationality: The nationality of the head contractor.TC/Consultant procurement: The procurement system used for the technical cooperation or consultancy services.TC/Consultant nationality: The nationality of the TC or consultant(s)Untied aid share (top level): The share of the project at the top level of contracting and contracts signed by the donor that are both formally untied and untied in practice according to whether there is a reasonable
opportunity for firms other than from the donor nationality to be awarded contact.
Provisionofservices-in-kind(Projects8,12and20)involvedcon-sultantsfromdonorcountries.Althoughcompetitivelytenderedoftenusingframeworkagreements,services-in-kindallowednoopportunitytocountrypartnerstomanage,orlocalfirmstobidforthesecomponents.
Untying aid: Is it working?44
Chapter 7
III. Components that are likely to be untiedThoseprojectcomponentsusingCPSand,significantly,thoseinvolvingsub-contractedprocurementweremorelikelytobeuntied.AllprojectcomponentsusingCPSwerefoundtobeuntied,thatiswithoutawardstodonorcountrybasedentities.
Sub-contractingisamajorsecondarylevelsourceofunty-ing,whetherornottheheadcontractorwasadonorcountrypartnerorganisation.Inmostcasescontractorswereallowedtosub-contractwithouttyingrestrictions.So,forexample,75%offundingwassub-contractedlocallyorregionallyintheKfWfundedProject6inZambia,withaGermansourcedcontractor.
World Bank contracting: a counterfactual.AnIDAfundedprojectinZambiawaspurposivelyincludedtoprovideacoun-terfactualtobilateralpractice.AllTCandothersub-contractsonthisprojectinZambiawereawardedtoregionalorlocalcompaniesaftertheIDAthresholdrequiringICBwasraisedtoencouragelocaltenders.ManyinformantsseetheWorldBankICBtenderingproceduresasamodel,forexampleforCPS.ItisthereforeinterestingtocomparethegeographicaldistributionofWorldBankcontractswiththebilateralsourcinginthesixcasestudyrecipientcountries(Table7.2).Whereasnobilateralcontractwasawardedoutsidethepartnercountrytocompa-niesregisteredinregionalLDC/HIPCsorothernon-DACcoun-tries,theseaccountedfor44%ofBankcontractsintheW&SSand26%ofallBankcontracts.Thesesamecountrieswerealsoespeciallyimportantincivilworks(e.g.VietnamandChinesecontractorsinLaoPDR).ThehighproportionofbilateralTC/consultancycontractsawardedtocompaniesfromthedonoreconomy(71%–12outof17whollybilateralprojectsinTable7.1),comparedwithonly23%forallDACcountriesforBankcontractssuggeststhatformalorde factotyingisaninfluenceonoutcomes.Thissuggeststhatbilateralagenciescoulddosignificantlymoretocreatealevelplayingfieldonbothanindividualandjointfundingbasis.
IV. Factors preventing local or regional firms from winning tendersAkeyissueappearsnottobeoneoflocalandregionallybasedfirmsbeingout-competedbutratherthattheywereoftennotbiddingforaidcontracts.
Onefactorisunequalaccesstoinformation.Thisisinprac-ticelinkedtothelevelofprocurementandresultantspreadofadvertising,andeventheinformalspreadofinforma-tionaboutcontracts.NCB(whichdoesallowinternationalcompaniestotender)attractedfarmorelocalinterestthanICB.Organisationsthatcontributetoprojectdesignweremorelikelytobedonorcountrybased.Projectsmaywellrepresentanexplicitorinformalcontinuationofapreviousprojectandwithahistoryoftiedprojectimplementation.Thosesameentitiesareineffectadvantagedinsidersintheirknowledgeofthedonor’sprocedures,thefocusoftheprojectandinearlyinformalaccesstoinformationaboutthecontract.
Thelocalresponsetotendersisalsorelatedtotheskillsandcapacityofthelocalmarket.LDCsarenotabletosupplyallofthegoodsandservicesrequiredforprojects,sothisdependsonwhatisrequiredandtheprojectsector.
Thesizeoflotsputoutfortenderwasasignificantinflu-enceontheamountoflocalinvolvement.Project7raisedtheICBthresholdtoincreaselocalparticipation.Localfirmssometimesdidnothavethecapacitytobidandwereoftenex-cludedatpre-qualificationstages.Largelotsarelikelytohavehigherbarrierstoentryintheformoflargerdeposits,largerstocksofequipmentandmanpowerandhigherrisk.Forthemostpartthisfavouredinternationalandlargeregionalfirmsfrommoreadvanceddevelopingcountries.
CompaniesbasedintheLDCswerefacedwithotherfinancialbarrierssuchastheweakcreditmarketandunavailabilitylo-
Table 7.2 World Bank Contracts by supplier development category signed in 2003-08 in the six case study countries(average percentage share of total)
Source Country Water & Sanitation Sector All sectors
Civil works Consultancy Goods Total Total
DAC Members 12% 23% 55% 20% 38%
Local (aid recipient) 54% 53% 21% 36% 35%
Other (non-DAC) 18% 15% 1% 13% 15%
Regional LDC/HIPC 0% 4% 0% 1% 0%
Regional Other 16% 4% 23% 31% 11%
Total 100% 100% 100% 100% 100%
Source: World Bank Contract Awards datasetNotes on source country: Local is sourcing from within the recipient partner country; Regional Other refers to non LDC/HIPC countries within the region (e.g. Thailand in SE Asia); Other refers to non-DAC, non-regional third parties. Where there is overlap of categories, then DAC members are recorded as DAC (e.g. Japan in SE Asia).
Untying aid: Is it working? 45
Chapter 7
callyofprofessionalindemnityinsurance,givinginternationalfirmsfurtheradvantages(Box7.1).
V. How the procurement system can affect the level of untyingNineofthe11projectswhichusedtheCPS,alsoinvolvedmajorlocalprocurementcomparedtothreeofthenineprojectswhichdidn’t,suggestingthattheprocurementsys-temaffectsthelevelofuntying.Procurementoptionsseeninthesampleprojectsweredonorprocurement(inallcasesviaHQ,therewaslittleevidenceofdecentraliseddonorprocure-ment),countryprocurementsystemsandprocurementbyaprivateentityfollowingtheirappointmentbythedonor.Sevenprojects(33%)solelyusedtheCPS.Onlyoneproject,whichusedtheCPSforanycomponent,waslessthan90%untied.Thegovernmentwas(orwaspartof )theimplement-ingentityfor16(76%)ofthe21projects.Fourprojectswereimplementedsolelybyprivateentitiesandnoneweresolelyimplementedbythedonor.
VI. Recipient country partner influence and the aid environmentIneverycountrywiththepossibleexceptionofVietnam,therewasawidevarietyofdifferentcontracting,fundingandprocurementproceduresinvolvingdifferentrelationshipswithdifferentdonorsforspecificprojects.However,severalfeaturesstoodout.
Strongnationalsystemsandgovernmentownershipassistsuntying.BothSouthAfricaandVietnam,thetwoleastaiddependentandrichestrecipientcountries,exercisesignificantownershipbychannellingaidfundsthroughtheirnationalfinancialmanagementsystems.
InSouthAfrica(apartfromTCandframeworkheadcontracts)allprocurementwaslocal.Thegovernmentvigorouslychal-lengedtheuseofaframeworkcontract(Project1)acceptingitonlyongroundsthatitwouldexpeditedisbursement.Project3,involvingtwinningofDutchandSouthAfricanlocalauthori-ties,isapartnershiponfloodprotection,wheretheNether-landsisperceivedtohavespecialexpertise.
InVietnam,duetouseoftheCPSandstronggovernmentinvolvementtherewasaveryhighshareofuntiedaidandlocallyledprojects.ThisisexemplifiedinProject10where,despiteafunderpreferenceforcontinuingwiththedonorcountrybasedcompanythathadbeenresponsiblefordesign,themainconsultancyserviceswereputtoICTandwonbyacompanyinanotherDACmembercountry.
IncountriesandsectorswhereSWAPswereinoperationthenpooledfunds(whichtendtobeuntied)weremostlikelytobeorganised,asinGhanaandZambia.
7.2.2 Project characteristicsTheformalfeaturesoftheprojectsaresummarisedinTableB.2inAnnexB.Thisprovidesaconventionalprojectdescriptionintermsoffullprojecttitle,donors,valueofdonorcontribu-tionsectoraluseandformaltyingclassificationasreported
totheOECD.Onthebasisofthedistinctionmadeabovebetweenlevelsofcontracting,projectsarethenprofiledinTable7.1accordingtotheways‘top’levelheadcontractsandTCareprocured,geographicalsourcingandtheirformaltyingstatus.Anassessmentismadeofthelevelofrecipientpartnerinvolvement,whichisqualitativelyindicatedasdominant(++),significant(+),unclear(?)andlittle(-).Thusassuggestedabove,ahighlevelofinvolvementisfoundtobestronglyas-sociatedwiththechoiceofacompetitiveformofcontractingandmorelikelytoleadtoalocallybasedorganisationbeingawardedtheheadcontractand/ortobeproviderofTCandconsultancyservices.Formaltyingstatusprovidesnoindicatoroflikelygeographicalsourcing.
SecondarylevelcontractingandfinalsourcingoutcomesareconsideredinTable7.3.Theimportanceofsub-contractingarrangements,includingespeciallylotsize,ishighlightedasadeterminantofthelikelyextentofcontractawardswhicharelocalandregional.Smallerlotsweremostlyawardedlocally,andthisoutcomeisirrespectiveofwhetherornottheheadcontractorisdonorcountryorlocallybased.Furthermore,thefinaloutcome,takingintoaccountsecondarylevelcontract-ing,indicatesaveryhighlevelofuntying.Thislinkissointhecaseofpartialtyingrestrictions(Project2inSouthAfricaandProject18inBurkinaFaso)orwheretherewasprovisionofTCservicesin-kind(Project20inGhana).ApparentexceptionsarealocalauthoritytwinningprojectinSouthAfrica(Project2)andalargelyTCactivityinZambia(Project4)whichbothreflectthelowproportionofsub-contractingandsourcingofgoodsandservicesintheoverallprojectdesign.Project14inLaoPDR,theonlyoneforwhichinformationonsourcingwasnotavailable,wasseeminglylargelyde factotied.Thekeyun-tyingissuewouldseemtobewhetherinbothprojectdesignandimplementationthereisastrongcommitmenttopromot-inglocalandregionalsourcingofgoodsandservices.
Untying aid: Is it working?46
Chapter 7
7.2.3 Untying and sourcing outcomes: a process approachTheprocessofdeliveringformallyuntiedaidiscomplex,andsoattemptingtotracethelikelysourcingconsequencesoftheexplicitandimplicit(default)choicesmadebydonorandrecipientcountrypartnersistheonlywaytodeterminewhetheraidisinpracticetiedoruntied.Thisalsohastobethefirststeptowardsexploringtheoutcomesintermsofaidand
Table 7.3 Project case studies: procurement, subcontracting, aid modality and major inputs
Project Number
Main procure-ment system(s)
Lot size Subcontracts and their sourcing
Modalities used Project type / major inputs
Local pro-curement
Untied aid share
(final)
SA1 Private Small +, Local Project TC, services + 82%
SA2 CPS Small +, Local Programme/SBS TC, services + 100%*
SA3 Donor Small -, Local Project TC, services - 31%
ZM4 Donor Small +, Local Project TC, services + 80%
ZM5 CPS Small -, Local Programme Civil Works + 100%
ZM6 Donor Large -, Regional Project Civil Works, TC + 75%
ZM7 CPS, IDA Large -, Intl. Project Civil Works - 100%
VM8 CPS Small +, Local Pool, Programme TC, Civil Works + 92%
VM9 CPS Small -, Local Project Civil Works + 100%
VM10 CPS, Donor Small -, Local Project Civil Works, + 100%
VM11 CPS, Donor Small -, Local Project Civil Works + 98%
LA12 CPS Small +, Local Project Civil Works + 0%
LA13 Private Small -, Local Project TC - 95%
LA14 Private Large Unknown Project Civil Works . Unknown
LA15 Private Large -, Intl. Project Civil Works + 100%
BF16 CPS Small +, Local Project Civil Works, TC . 57%
BF17 Donor Small +, Local Project Works, TC . 100%
BF18 IDA rules Large -, Local, Intl. Project Civil works, TC . 92%*
BF19 CPS / IDA Large & Small +, Local Pool, Programme TC, Goods . 100%
GH20 CPS Small +, Local Project Civil Works, TC + 100%**
GH21 Private Large & Small -, Regional, Intl. Project Civil Works, TC + 74%
Notes: *Partially Untied, ** Excluding services-in-kind
key to headings:Project Number: Reference for project details in Table B.2 in Annex BMain procurement system(s): The main procurement system for the project.Lot size: The size of the majority of contracts tendered for implementation Subcontracts and sourcing: The number of subcontracts and their dominant sourcing location (Many = +, few = -, unknown = ).Modalities used: The aid modality.Project type / major inputs: The dominant type of assistance and corresponding major inputsLocal procurement: The level of local procurement (major = +, minor = -).Untied aid share (final): The share of the project at the level of final disbursement for implementation that are both formally untied and untied in practice according to whether there is a reasonable
opportunity for firms other than from the donor nationality to be awarded contact.
developmentaleffectiveness.SomethingofthiscomplexityispresentedinastylisedwayinFigure7.1.Thisflowdiagrampresentsinqualitative,probabilistictermswhathasbeenlearntintheprojectstudiesanddiscussionswithcountrypartnerandagencyofficialsandotherswithexpertiseaboutthemoreorlesslikelyoutcomesofchoicesmadeandtheinfluencesontheprocurementprocess.
Untying aid: Is it working? 47
Chapter 7
Figure 7.1 Flow diagram of aid, untying practice and sourcing outcomes56
56 By their nature flow diagrams are qualitative assessments so additionally, for example, it is important to note that some donors are able to untie some administrative costs and also source them locally which is not shown on the diagram.
Bila
tera
l ODA
Mod
ality
Budg
et Su
ppor
tFu
ll Po
olLo
ose P
ool
Proc
urem
ent E
ntity
Coun
try P
rocu
rem
ent S
yste
mDo
nor P
rocu
rem
ent
Proc
urem
ent D
etai
l
Priv
ate P
rocu
rem
ent
Wai
ver
Tied
Aid
Ratio
of P
roje
ct In
puts
Tend
erin
g
Lot S
ize
Loca
l Sup
ply C
apac
ity
Sour
cing
Outc
omes
Loca
l (Un
tied)
Regi
onal
/ In
tern
atio
nal (
Untie
d)
High
NCB
Larg
e sha
re o
f phy
sical
wor
ks
ICB
Larg
e sha
re o
f TC
Low
Smal
l Lot
Size
Larg
e Lot
Size
Dono
r (De
fact
o Tie
d)Do
nor (
Form
aly T
ied)
Recip
ient
Invo
lvem
ent
<<
<<
Rec
ipie
nt In
volv
emen
t >>
>>
<<
<<
Loca
l Sup
ply C
apac
ity >
>>
>
<<
<<
Tend
erin
g >
>>
>
<<
<<
Inpu
t Rat
io >
>>
>
High
Low
Dono
r Fun
ds
Subc
ontra
ctin
gFr
amew
ork A
gree
men
tsSe
rvice
s-in
-kin
dDo
nor A
dmin
. Cos
ts
Untying aid: Is it working?48
Chapter 7
Keyinfluencesonthesourcingoutcomeare:• Ownership–theextentofrecipientinvolvement;• Choiceofmodality;• Procurementsystemused(CPS,privatenon-stateagency,
donor)anditsdetails(extentofsub-contracting,frame-work,agreements,etc.);
• ProjectactivityintermsoftherelativeproportionofTCandphysicalinvestment(constructionandplant);
• Contractingprocedure(NCB,ICB,directcontracting,qual-itybasedselection,shopping);
• Sizeoftenderedlots;• Extentoflocal(andregional)capacity.
Theoutcomesarereflectedintermsofthelikelyproportionoflocal,regional,otherDACandnon-DACdonorcountrycontractingandsourcingofgoodsandservices.Evenwhereaidisformallytiedatheadcontractleveltheoutcomemaybemixed,ifcontractorsarefreefromrestrictionsinsub-contractingandsourcinggoodsandservices.Similarly,formallyuntiedaidmaybede factotiedtodonorcountrysourcing.
Thechoicestobemadeatdifferentlevelsintheprocessraiseanimportantissue,thegenuinenessofthecommitmenttountying.Istyingperceivedconstructivelyorpositivelywithinadonororganisationorbyitsagents?Alternativelyisunty-ingperceivednegativelyasamodificationtoregulationsoncontractingandprocurementnecessarytoconformtotherequirementofuntyingstatusasintheOECDdefinitionsandthe2001Recommendation?Apositiveapproachwouldimplyencouragingproject-typepreferencesandprocesschoicesthatflowtotheleft-sideuntyingoutcomes.Anegativeorpurelyconformistapproachwouldresultinapreponder-anceofde factotyingoutcomesontherighthandsideofFigure7.1.Ontheevidenceofthe21studies,first,therearestillwidedifferencesindonorpractice,someintendedandsomeunintended,andsecond,theoutcomesarealsostronglyinfluencedbycountrypartnerinvolvement,capacityandthebargainingpowerthattheyfeelvis à visfunders.
7.3 Consequences: interconnections and aid effectivenessTheinterconnectednessoftheprocessesthroughwhichtheorganisationofprojectsiscontracted,inputsprocuredandthesourcingoutcomesinvolvealargenumberofpossiblepathways,asillustratedinFigure7.1.Theseinterconnectionsmakeitdifficulttoisolateanysingleaspectoftheprocessandtoinferoutcomesonapartialbasis.Theoutcomesmaybemisleadinglyattributedtoasingleinfluence.Itisalsoimpor-tanttoreiteratethatformaluntyingisnotonlytherelaxationofaconstraintthatmayconsiderablyincreaseflexibilityinthesourcingpossibilities,butthatitalsoopensupthechoiceofprocedurethatcanbeused.Iftheuntyingisgenuine,thentheuseofthedonor’sfundsiscompatiblewithallprocure-mentsystems,formsofcontrol,aidmodalitiesandfinancialmanagementsystems.
Onecanlookattheconsequencesandeffectivenessofunty-ingasbothapolicyandaprocess.TheToRforthestudyrecog-nisesthreeformsofeffectiveness.Firstly,thereisefficiencyasmainlyreflectedintheliteratureontying(Chapter5)intermsofcosteffectivenessandtimeefficiency,whichisconsidered,inSection7.4.NextthecontributionofuntyingtotheParisDeclarationaideffectivenessobjectivesofownership,align-mentandharmonisationareconsideredinSection7.5.ThentheissueofuntyingandtransparencyisconsideredinSection7.6.Finally,thereisthecontributionofuntyingtodevelop-mentaleffectivenessintherecipientpartnercountryasanoutcome,includingespeciallystrengtheningoflocalmarketsandregionalintegration(Section7.7).Theformsofevidencethatcouldbeassembledmeanthatthefocusofinvestigationismainlyondirecteffectsandonlysuggestiveaboutultimateaideffectivenessanddevelopmentoutcomes.
7.4 Aid efficiency: cost-effectiveness and timeCost-effectivenesshasbeentheconventionalrationaleforuntyingaid.Theoreticalanalysisoftyingimpliesthatitisinefficientandtheverylimitedempiricalliteraturesug-geststhattiedaidisbetween15%and35%morecostlythanunrestrictedmarketprices.Withthisfindingasbasisforcomparison,eachcountryteamundertookatleastonecost-effectivenessanalysis(CEA)attheprojectlevel,lookingattheuseofuntiedfundsforsourcingofgoodsandservicesincomparisontolocalcommercialmarketorinternationalmarketprices.
TheSouthAfricanCEAlookedatthecosteffectivenessoflocalprocurementfortwoprojects.Thecivilworkswereappar-entlycosteffective,butthelowlevelsofcompetitionfromacombinationoflocalonlyadvertising,limitedlocalcapacityandpreferentialprocurementwasseenasapotentialthreattopriceefficiency.Project1,organisedbyadonorbasedconsult-antunderaframeworkagreement,comprisedwhollylocallycontractedTCtothehealthsectorandtheteamfoundthattheconsultancyratespaidunderSouthAfricanprocurementguidelineswerehigherthantheywouldhavebeenunderdonorguidelines.TheECsupportedProject2demonstratedtheadvantageofflexibilityassociatedwithevenpartialunty-ing,asitwaspossibletoshifttousingtheCPSduringprojectimplementation.Project3involvedTCprovidedbyadonorpartnermunicipalityintheNetherlandsinfloodprotectionaswellaslocalprocurementforminorcivilworks.Thecost-effec-tivenessofTCcouldnotbeassessedbecausetheestimatedcostsofTCwerenotavailable,andalsothereisnotransparentmarketforsuchresources,butbothsidesperceivethisasanareaofdonorcomparativestrength.
InZambia,thestudyconsideredarepresentativesampleofgoodsprocuredunderIDAfundedProject7,whichwereallprivatelyimportedfollowingIDAguidelinesbythecontrac-torfromSouthAfrica,theregionallydominanteconomy.TheCEAshowedthecostofthegoodsprocuredusinguntiedaid
Untying aid: Is it working? 49
Chapter 7
waslowerbyatleast6%,relativetotheimportparitypriceforcomparablegoods.
TheVietnamstudy’sCEAsoncomponentsofthreeprojects(8,10and11)concernedlocalprocurementthroughtheCPS.TheCEAindicatedmixedresults,withlargevariationinpricesandgoodsprocuredaroundmarketreferenceprices.Thestudyconcludedthatequipmentandmaterialspurchasedfortheprojectswere,inthemajorityofcases,actuallymoreexpen-sivecomparedtothemarket,duetoproblemsofcorruptionorstrongsocialnetworks,leadingtocostinefficiencies.
InLaoPDR,CEAsonProjects12and13,involvingCPSandprivatelocalsourcing,alsoshowedsignificantvariationinprocurementcosts.Butonbalancecostsavingsweremadeonthepurchaseofequipmentitemsandconstructionmaterials,relativetothemarketreferenceprice.Thisimpliedcost-effec-tiveuseofaidfundsasaresultofuntying.
TheBurkinaFasostudywaslaunchedlaterthantheothersandatthetimeofdraftinghadnocompleteCEAavailable.However,provisionalfindingsindicatequalityissuesconcern-ingtheimportingofgoodsbylocalfirms.Interestingly,Project19involvedacompetitivere-contracting,whichresultedinareplacementbidfromtwolocalcompanies15%belowthemixedcreditarrangementthatitreplaced.
TheGhanastudyconsideredCEAwithinProject20,whichcombinedTCservicesandunrestrictedpooledfunds.ADanishcompanyprovidedTCforprojectmanagementasservices-in-kindfundedbyDanidaHQunderanEU-widetenderedframeworkcontract(formallypartiallytiedandde factotied).ThecostsofTCtotheprojectwereapparentlynotseparablefromthoseofawiderprogrammeanditwasnotpossibletoestimatecost-effectiveness.
TheDFIDpooledcontributionwaspaiddirectlytoDanidafortheprojectteamtodisburse,fundingimplementationofallinvestmentsunderCPS,whichwerethenconsideredtobecontractsratherthansub-contracts.Procurementofcontrac-torsbynationalcompetitivebiddingwascarriedoutthroughsub-contractsatdistrictlevel,withassistancebytheregionalgovernmentwaterdepartment.Sixofeightcontractorswerewhollylocalandtwocontractswereawardedtolocallyregis-teredChineseentities.Contractorswerethenunrestrictedintheirprocurement.Thegoodsandcivilworksprocurementwasbroadlycosteffective,matchingreferencepricesfromthelocalmarket.
TheGhanastudyalsoreportsanexampleofdrainageworkswhereICBledtosubstantiallycheapersourcingthaneithersinglesourcingornon-competitiveselection.ThreestretchesofdrainageweretenderedbyICB,selectivebiddingandsinglesourcing.Thesectioncontractedbyselectivebiddingwaspriced47%higherthanthesinglesourcingsection,whichwasinturn90%higherthantheICBcontract.
Therearebroadlyconsistentfindingsfromthesecountrycas-es.Untiedaidandlocalcompetitivesourcingwerefoundtobecosteffectivefortheprocurementgoodscomparedtomarketprices.Thecostadvantagesincomparisonwiththepreviousevidenceontiedaidweremostapparentwhereuntyingledtoincreasedandtransparentcompetition.Itprovedimpossibletotestthecost-effectivenessofTC,anareaofpoortranspar-encyfromacountrypartnerperspective.Afurthercautionisnecessary.Thecountrystudiesareonlyabletoprovideillustra-tiveexamplesingeneralrelativelyfavourablecircumstancesofgooddonor-recipientrelationships,bettergovernanceandeconomicstability.
Asthestudyspecificallylooksatuntiedprojects,thedirectcosteffectofcompetitiveprocurementisnotsurprising.Pertinentquestionsarethereforehowdoesuntyinginteractwithcurrentsourcingpractices,andwhatarethekeyfeaturesassociatedwithcosteffectiveness?
Costsavingsfromuntyingwerefoundtobefarfromauto-matic.Untyinggivesmoreoptionsbutispronetosomeofthesamepitfallsthatbedevilpubliccontracting,perhapsuniversally.
Costeffectiveprocurementrequirescapacityandskills.Ifunty-ingtriggersamovetousingcountrysystemstheninordertogainthebenefits,CPSneedadequatecapacity.UntyinghighlightedthecapacitychallengesforCPSineachcasestudy.However,untyingalsoprovidedtheopportunity(funding,trainingandamajorincentive)forimprovement.CPSwereidentifiedaslackingcapacityespeciallyattheprovincial(dis-trict)level.CPSwerenoteffectiveforprocuringinternationalTC,buthadbeengivenlittleopportunitytodothis.Thestudyfoundmixedresults.ManyprojectssucceededusingtheCPS,wherethoseresponsibleforotherprojectsusingotherap-proacheswereclaimingthatitwasnotfoundfitforpurpose.
Realisingthepotentialbenefitsofuntyingalsorequiresthesupplycapacitytomakepossiblegenuinecompetition.Sever-alstudiesreportedlownumbersofbiddersbothinprocuringatlocallevelswheretherewerelowskillsets,andalsointerna-tionally,duetothelargesizeofsomecontracts(Box7.1).
Theimportanceofeffectivecompetitionappliedateitherorboththelocalandinternationallevel,dependingonwhatwasbeingprocured.Forwidelytradedcommoditiessuchasaggre-gatesandcementtherequirementsforefficientprocurementwereeffectivelocalcompetitionandalsowideningthepoolofexpertiseavailabletotenderbids.Narrowlydefinedconstruc-tioncontracts,suchasfordrainsorwells,alsoappearasareasinwhichthesamecriteriaforefficiencyapply.
Untying aid: Is it working?50
Chapter 7
Frameworkcontractscanprecludethecompetitiongainsfromuntyingtolocalandregionalsuppliers.Undoubtedlydonorsfindthesecontractsconvenientinachievingadminis-trativeefficiencysavingsandtheymaybecosteffectivefromaperspectiveofmanagingamulti-countryprojectportfolio.Theymayalsobecompetitivelytenderedamongstdevelopedcountrysuppliers.However,attheindividualprojectlevelandfromtherecipientperspectivetheyappeartobeaformofde factotyingthatprecludeslocalsourcing.Undertheprojectswithframeworkagreements,contractshadbeenawardedtodonornationalityfirms.Frameworkcontractsareoftenregardedas‘captured’byasmallgroupofnationalfirms(oftenworkingtogether).Anoptionforachievingcompetitiongainscouldbetoprioritiseincreasingtheabilityoflocalfirmstopartnertheframeworkcontractwinner.
Time efficiencyTheimplicationsofuntyingfortimeefficiencyappeartodependontheprocurementsystem.Untyingallowsachoiceonthespeedofprocurement,balancingthisagainstcostandotherconsiderations.
ICBwasfoundinbothdonorHQandcountrylevelinvestiga-tionstobewidelyregardedastimeconsuming,aswellascostlyandadministrativelyinconvenient.Accordingly,itisonlyadoptedwhereprocurementgoesbeyondminimumthresh-olds.Inpractice,agencieswerefoundtousevariousexpedien-ciestoavoidICBunlessitwasfeltunavoidableorpreferableoncompetitiongrounds.
TheoverwhelmingviewofinformantswasthatCPSweretypi-callyfasterforlocaltendering,whereasmostdonorsystemswerelikelytobequickerforinternationaltendering.Therewereeveninconsistentviewsoverusingdonorprocedures.GovernmentofficialsmanagingProject2commentedun-favourablyontheneedfortheapprovaloftheEC’sofficeinSouthAfrica’sforsomeitems,buttheyalsogreatlyappreci-atedthesingletenderprocurementprocedureunderEUrulesonotheritems.
Anotheralternativefavouredbysomedonorsisusinganexternalagent(asinallprojectswithprivateprocurement)whicheffectivelyunburdenstheagentfrombothdonorandgovernmentprocedures.Privateprocurementwastypicallyseenasveryfast,asonewouldexpectbeinglessencumberedbyagencyrules.However,privateprocurement,usingrulessetbythedonor,wasfoundtobeslow,asinthecaseoftheMCAinGhana,whichprecludedinclusionofoneofthatagency’sprojectsinthisstudy.
Takingawiderviewthestudiesmostlyfoundevidencethatsupportsthecombinationofdonorsprovidingfundingthatisgenuinelyuntiedandsoflexibleandtheuseofcountrypro-curementsystems.TheareaaboutwhichthereisawidespreadlackoftransparencyisTCcostsandtheefficiencyofthewayinwhichTCiscontractedevenformanyoftheprojectsonwhichdonorswerewillingtocooperatewiththestudyteams.
Box 7.1 DFID Survey of constraints on local firms bidding for international contracts
Finding a lack of interest in bidding from companies in partner countries, DFID (UK) conducted a survey and held consultations with local businesses in India, Bang-ladesh, Nigeria and South America. The purpose was to jointly review constraints on local companies bidding for internationally tendered contracts for bilateral aid in their region, and to identify ways of breaking down the barriers to entry. The survey found that even where they were aware of the opportunities, developing country consultancies usually did not bid, perceiving the cost of preparing bids was greater than the reward, taking into account the perceived risk of not winning. Developing country consultancies face problems in submitting bids that were competitive with interna-tional consultancies because of the following:• financial constraints of comparatively smaller firm
based in a developing country with lower turnover (harder to meet prequalification), facing exchange rate risks, managing cash flow of costs and pay-ments, and bid security;
• lack of capacity to understand ToRs and bidding documents, which are often vague and overly legal, in unfamiliar language and bidding practice, with terms not suited to local conditions;
• difficulty of sustaining partners which enable bid-ding on larger tenders or to compete with interna-tional consultancies for these partners; and
• lack of reputation outside their own country, even when experienced.
The problems were greater for LDC consultancies bid-ding for work in other LDCs, with experience but with-out local expertise and links. Suggestions for increasing the number of bids and their success rate included:• raising the threshold for ICBs, and otherwise using
NCB;• circulating a list of local experts to level the informa-
tion playing field;• reducing prequalification requirements;• making advance payments, as do development
banks;• increasing transparency of why the winner was
chosen;• modifying contract terms to suit local conditions;
and • assistance with understanding ToRs, bidding docu-
ments and current practices.
Source: DFID Procurement Group, Value for Money Department
Untying aid: Is it working? 51
Chapter 7
7.5 Aid effectiveness: ownership, alignment, harmonisation, and transparencyUntyingcontributestotheAidEffectivenessagendabecauseuntiedaidisanecessaryconditionformanyofthestepsdonorstaketoimplementtheParisDeclaration.Soitisneces-sarytoconsiderwhetheritisuntyingthatiscontributingtoaideffectiveness,ortheinternalpressureswithinagenciestountiecomefromseekingtheflexibilitytoforwardtheParisDeclarationAgenda.
ThecasestudyprojectsprovidedexamplesofpotentiallyimprovedpracticeinPDeffectivenesstermsbasedonthewaysinwhichuntyingisassistingdonorsinmovingawayfromtheconventionalstand-aloneprojectbasedapproachtoapproachesthatarecooperative.Thismovementfacilitatesalignmentandharmonisationincludingbudgetsupport,pooling,delegatedcooperation,useoftheCPSanddecen-tralisation.Thesesteps,supportedbymanybutnotalldonors(SeeChapter6),werefoundtobestronglysupportiveofPDgoalsofownershipandalignment(Box7.2),whichisoneoftheultimateobjectivesofuntyingaccordingtotheRecom-mendation.
Thestudiesprioritisedprojectfocusedinvestigationsmakingitmoredifficulttoconsiderharmonisationamongstdonors.Nevertheless,thestrongassociationbetweencooperativeuseofprogrammaticmodalitieswithuntiedfunds,towhichmanygovernmentandagencyinformantsdrewattention,confirmstheviewthatgenuineuntyingcontributestoharmonisation.
InLaoPDRwhere,ofthesixcountries,conventionalproject-typeassistancewasmostdominant,theteamconcludedthat“toincreaseaideffectivenessingeneral,itisrecommendedonthebasisofthisstudythatfurtherstepsaretakentoencour-agetheuseof‘newer’aidmodalitiesamongdonors(e.g.budgetsupportandpooledfunding)andincreaseduseofcountryprocurementsystems”(AnnexD,McCartyandJulian,2009).
Transparency in donor procurement isoneofthesixPDEobjec-tivesandakeyaspectofanyuntyingaudit.Thefollowingar-easofdonorcontractingpracticeswerefoundtobelessthanwhollytransparentfromarecipientcountryperspective:• DonorcontractingatHQ• Frameworkcontracts• B2B• Twinning/partnershiparrangements• Privatemanagementcontracts
Alloftheaboveareasarefullycompatiblewithuntiedaid.However,alloftheseofferscopeforde factotyingpractices.Transparencyisneededtoseewhethertheoutcomesfollownotjusttheformalrulesbutresultingenuineuntying.Iftrans-parencyisessentialanditisgoingtobehardtomaketheseareasofcontractingmorefullytransparent,thenthealterna-tivefordevelopmentpartnersistomakelessuseofthem.
Thesewerealsoareaswhereinmanycasesstudyteamsstrug-gledtoobtaininformation,whichwas‘unavailable‘atacoun-
Box 7.2 Untying, ownership and alignment: country study conclusions
Ghana: “Untying is creating policy space for recipient country ownership, use of recipient country systems and align-ment with recipient country priorities. The existence of strong country leadership, such as through the creation of SWAPs, was also seen as a driver of untying.”
Burkina Faso: “Le déliement est souvent accompagné par le transfert de gestion au pays bénéficiaires. Effectivement, l’alignement sur les procédures nationales, que veut la Déclaration de Paris, est difficilement envisageable sans délie-ment de l’aide.”
South Africa: “In an untied environment, donors are more likely to make use of recipient country procedures. The case studies show that the use of recipient country procurement systems enables government to pursue domestic policy pri-orities and improve alignment.” The EC supported Project 2 demonstrated the advantage of flexibility associated with even partial untying, as it was possible to shift to using the CPS during project implementation.
Zambia: “Untying has facilitated the increased use of GBS and SWAPs. This improves the transparency and alignment of aid.”
Lao PDR: “Untying has been associated with a shift away from donor priorities and towards the government’s needs and objectives (alignment), and has encouraged a decentralisation of responsibility and increased recipient ownership, sup-ported by the fact that more government agencies are now responsible for project implementation and management.”
Vietnam: “From the cases reviewed, a significant efforthas been made to ensure that ownership over ODA lies in the hands of government, through the use of country systems, national technical advisers and the greater involvement of government executing agencies.”
Untying aid: Is it working?52
Chapter 7
trylevel.Theextentoftransparencywasvariable.Asthecost-effectivenessanalysesdemonstrate,somelocalagenciesandsomedonorswerewillinginallcountriestobeopenaboutcontractingandprocurement.However,someimplementingentities,especiallyprivatecontractors,wereinpracticeunen-thusiasticaboutwhattheyregardedasprivilegedinformation.Someweresoslowtorespondthatthisprecludedundertak-ingaCEAorfullerexplorationofthecontractingprocess.
Ontransparency,themostopaqueaspectsofdonorpracticeconcernedTCandassociatedconsultancies,theareasofas-sistancewhereformalandde factotyingwerefoundtobestillmostpervasive.Asonecountrystudyconcludes:“thehighleveloftechnicalassistanceandco-operationagreementsinSouthAfricamakesODAlesstransparentandincreasesthepotentialfortying”.
7.6 Developmental impactsTheToRforthethematicstudydrawattentiontothemanyim-pactsthatareaspectsofdevelopmentaleffectiveness:“effectsonthemarketforlocalandregionalproducers,onpromotionofthelinkageofthelocaleconomywithregionalnetworks,globalnetworks,onemployment,onpro-poorgrowthandpovertyalleviation,andongenderandenvironment,impactofuntiedaidonCorporateSocialResponsibilityandonprojectlifetimecostsavings.”
Therearealsomultipleinfluencesatwork,whichwilldeter-minethedevelopmentalimpactofprojects,donorpro-grammesorthewholedevelopmentcooperationeffort.Itis,asindicatedalready,difficulttotracetheconsequencesoftherelaxationofasingleconstraintforalltheseaspectsofeffectiveness.Soinsteadcountrystudieswereaskedtoconsiderwhetheruntying,andspecificallyataprojectlevel,wassupportiveofratherthanlikelytoimpedeeffectivenessintermsofpovertyreduction,advancinggenderequalityandsoforth.Andindeedthecountrystudiesfoundthatmostoftheprojectsunderreviewwerelikelytohavepositivebroaderde-velopmentaloutcomes,butthatthesecouldnotbeattributedtountyingper se.
Theinvestigationsfocusedmoreonareaswhereamoredirectconnectionmightbeobservedbetweenuntyingandoutcomes,specificallyahigherproportionofthetransferflow-ingasemployment,incomesandprofittothelocaleconomy,developmentoflocalbusinessthroughgreateruseoflocalandregionalcontractorsandsourcingfromlocalsuppliersthroughlocalmarketsandusesoflocalsystems.
Country studies are cautiously positive, finding evidence of direct employment and income stream generation.However,theseef-fectsarehighlysectorandactivityspecificandsocouldnotbeattributedtountyingper se.InVietnam,“Locallabourandlocalsourcingofgoodsincreasedtheincomesandemploymentofsomeof(the)poorestprovinces,wherethesedevelopmentinitiativestakeplace.Yetproblemsofseasonalortemporary
employmentquestionthesustainablelong-termdevelop-mentimpact.Difficultiesalsolieindeterminingwhethertheeffectonlocalemploymentisadirecteffectoftheuntiednatureoftheprojects,i.e.thereislittleevidencetoassumethatiftheseprojectsweretiedthattheywouldsourcelabourfromforeignmarkets.”Similarly,theLaoPDRstudyconcluded(butcautiously)thatlocalsourcingofcontracts,consultants,workersandmaterialsasaresultofuntyingcanhavepositivedevelopmentaleffects,byprovidingincreasedemploymentopportunitiesandrevenuesforlocalmanufacturersandsup-pliers.However,intheprojectsstudiedthesebenefitswereseverelyconstrainedbylackoflocalcapacity.TheZambiastudyfound:“theevidenceprovidedbythesefourprojectssuggeststhat,ingeneral,theextensiveuseoflocalsub-con-tractors,whichismadeeasierinanuntiedenvironment,hascontributedtothecreationofemploymentopportunitiesandatransferofskills.”
The balance of sourcing of goods and services is shifted by unty-ing towards local suppliers, as all studies indicate. Thispromoteslocalbusinessdevelopment,butthelinkageeffectsarelimitedbytheunderdevelopedstructureoftheeconomy.ThedirecteffectonthelocaleconomyisdampenedbecausemostLICsdonotproducemanyoftherequiredgoodsdomestically.Forexample,theZambiastudyfoundinadetailedprojectinvesti-gationthatamixoflocalandforeign(butZambianregistered)companieswoncontractsthroughacombinationofICBandNCB,butvirtuallyallthegoodswereimported57.TheBurkinaFasostudyreportedasimilarsituationwhentheylookedindepthataprojectlevel.
The promotion of regional networkingappearstobelimitedandlikelytobenefitthestrongereconomiessuchasSouthAfricaandVietnam.NoprojectsreportedcontractsawardedtoorganisationsfromneighbouringLDCsorHIPCs.TheZambianexampleagainhighlightedtheabsenceofsourcingofgoodsandservicesfromregionalLDCs.Infact,noneoftheprojectstudiesreportedsourcingfromotherregionalLDCs.Againthisfindingmayreflectthehighlysectorspecificbasketofgoodsandservicesunderstudy.Incontrast,foodaidstudieshavereportedregionalprocurementwithinSADCashighlycontingent,reflectingthethencurrentlocationofsurplusesanddeficits.58
UntyingwascloselyassociatedinmostcaseswithuseofCPS,exceptwhereprivatecontractorsweremanagingprojects.AgainstudiesareallcautiouslypositiveaboutthewaythisuseofCPSstrengthenslocalcapacitybybuildingupexperi-ence.ThelackofcapacitywasfelttolimituseofCPSorashifttoprogrammaticmodalitiesinLaoPDR.TheSouthAfricaand
57 Untying in both the OECD definition and the literature refers to the first round effect i.e. no account is made of the origin of the good or service. Some donors do make use of national content requirements when tying their aid in order to capture the second round effects as well as the first round effects.
58 At different times Malawi, Zambia, Zimbabwe and Tanzania as well as South Africa have exported aid funded food to other SADC countries.
Untying aid: Is it working? 53
Chapter 7
Vietnamstudiesemphasisethatspecificeffortsareneededtocombineefficiencyandcapacitydevelopment.
Thecontentofaprojectratherthannominaltyingstatuswasfoundtobecriticallyimportantindeterminingoutcomes.TwoprojectsinZambiaandSouthAfricaprovidedfundingforlocalgovernmenttodisbursedirectly,withapparentsuccess.Incontrast,inanotherSouthAfricanproject,asubstantialproportionofthegainsintermsofexperience,management,trackrecord,reputationandprofitwerecapturedbythedonorcountrycontractorunderaframeworkagreement.Capacitybuildingandknowledgetransferappeartobemorelikelywherelocalandregionalfirmsofmediumcapacityundertakeprojectinvestmentactivities.
7.7 Awkward issues about untying: possible trade-offs and risksAre there efficiency-development effectiveness trade-offs as-sociated with untying? Theconsequenceofuntyinghavingmultipleobjectivesofincreasingefficiencyandaideffective-nessultimatelyisthattheremaybeareasinwhichdonorsandrecipientsfaceimmediatetrade-offs.AnumberoftheprojectsweretradingoffalignmentandharmonisationagainstcostandtimeofprocurementbypreferringtousedonorandprivateprocurementsystemsinsteadoftheCPS.Theadmin-istrativeburdenandinefficienciesofusinglocalsystemswerecommonlycitedbydonorinformants.However,theadminis-trativeburdenimposedbydonorsystemswasalsoregardedasasubstantialissuebylocalinformants.ThecostandtimerequirementsofICBpredisposedsomedonorstofavourframeworkagreementsanddirectcontracting,whilstmanylocalinformantspreferredNCB.
SeveralcountriesincludingSouthAfricaandGhanahavepref-erentialprocurementpolicies.Thesemaydirectlylowertheassessedcostoflocalprocurementrelativetoothersourcesinordertoincreasetheshareoflocalsourcing.TheywereusedextensivelyinSouthAfricaandveryrarelyinGhana.Asimilareffectwasthatlowlocalcapacitymeantthatwhenprocuringlocally,projectscouldbepayingapremiumforcertainskillsduetothelackofcompetition.
Riskaversionappearedtobepartoftheexplanationforwidespreadde factotying.Theprojectsreviewedillustrateawiderangeofdonorresponses,notonlytotheopportunitiesbutalsototheperceivedrisksincontractingandprocure-mentassociatedwithuntying.MostbilateraldonorsstillshowarevealedpreferenceforcontractingmanagementandTCfromwithinknownsocialnetworksongroundsofriskaver-sion.Thisoftenhintedatbutnotexplicitlystatedpreferenceisreflectedinthehighproportionofheadcontractsawardedtoorganisationsregisteredinthedonorcountry.Contractrequirements,includingpreregistration,completionbondsandsoforth,makesuchoutcomesmorelikely.Somedonorsfocusonsuchrisksaspartjustificationforadoptinga‘businessasusual’approach.Forexample,thiswasarguedbyadonor
directlytenderingfromHQlevelforaprivatecontractorastheimplementingentity.
Inmostcases,evenwhereheadcontractswereformallyorde factotied,therewaswillingnesstouseCPSforsub-contractingprojectcomponentsandtheprocurementofspecificgoodsandservices.ThesepositivechoicesinuseofuntiedaidwereconsideredtoreflectcountryandsectorallevelassessmentsofrisksaboutwhenemployingCPSandalsousingdirectbudgetaryinstrumentswereappropriate.Thepossibilityofpooroutcomesduetoinefficiencyincontracting(e.g.delaysorrestricteddisseminationofinformationabouttenders)andcorruptionwererecognisedbyinformantsandreportedincountrystudies.Mostdonorsconsideredtheseriskstobemanageablethroughfinancialreportingandmonitoring.
Becauseofconcernsabouttransparencyandaccountability,untyinghaddonelittletoreducefinancialreportingandmonitoringrequirementsbyrecipientsorwithindonoragen-cies,evenforpooledfundingarrangements.Sofromadonorperspectiveoneoftheanticipatedbenefitsofuntyinghasyettomaterialise.TheSouthAfricaandVietnamstudiessuggesttheappropriateresponsetorisksofinefficiencyormisuseoffundsincontractingisanawarenessofproblemsandcoop-erationwithcountrypartnerstoaddresstheimplications.Inmostcountrieseffortswerenotedtosupportthestrengthen-ingoflocalfinancialmanagementcapacity.
7.8 ConclusionA priori analysisofwhetheraprojectisuntiedisimpractical.Thetyingstatusisformallydeterminedbywhethertheheadcontractisopentounrestrictedinternationalcompetition.Howevercountrylevelanalysisindicatesthatthedetermina-tionshouldtakeintoaccountallthreelevelsofprocurementincludingsubcontracts.Thetyingstatusalsoneedstotakeaccountoftheextentofde factotying.Thischapterthereforeshedslightontheimportance of the genuineness of the commit-ment to untying.Istyingperceivedconstructivelyorpositivelywithinadonororganisationorbyitsagents?AlternativelyisuntyingperceivednegativelyasamodificationtoregulationsoncontractingandprocurementnecessarytoconformtotherequirementofuntyingstatusasintheOECDdefinitionsandthe2001Recommendation?
Apositiveapproachwouldimplytheuseofdirectbilateralinstrumentsandpoolingarrangementswhichgiveownershipandalignment,wherethisisregardedaspracticalandalsobytryingtostrengthenlocalfinancialmanagementandpro-curementcapacity.Aconsiderabledifferenceintherevealedpreferencesofdonorswasfound(Chapter6)withthetwolargestbilaterals,theUSandJapan,makingleastuseofsuchmodalities.
Apositiveapproachalsoimpliesencouragingproject-typepreferencesandprocesschoicessuchasusingreducingtendersizesandraisingICBthresholds,i.e.designingcontract
Untying aid: Is it working?54
Chapter 7
packagesthataremorelikelyresultinuntiedoutcomes.Anegativeorpurelyconformistapproachdoesresultinapre-ponderanceofde factotyingoutcomes.
Theseformsofpositiveuntyingareverymicroinnatureandbasedinaprojecttypeapproach.Buttheyshowthatthespecificusesoftheprocurementsystemareimportantindeterminingthelevelofuntyingaswellasobviouslyuntiedmodalitiessuchasbudgetsupportandfullpooling.
Intermsof value for money or aid efficiencythefollowingcanbeconcluded:localcompetitivesourcingusingNCBorICBwasfoundtobecosteffectiveforgoodsprocurementcomparedtomarketpricesandtheefficiencygainsweremostclearwhereuntyingledtoincreasedcompetitionincompari-sonwithtiedaid.Thecostsavingsfromformallyuntyingaidarefarfromautomatic.
Internationalcompetitivebidding,althoughoftendescribedastimeconsuming,costlyandadministrativelyinconvenient,resultedinincreasedcompetitionthroughamuchwidergeographicalsourcingofcontractorsincludingregionally.However,evenwhereICBwasadopted,theprojectplanningandcontractingprocessesofseveralDACmembersappearedtointentionallyorunintentionallyadvantagedonor-basedcompanies.Someindicationofthisisgivenbytheeconomet-ricresultsinChapter6andalsobythecomparisonwithWorldBankcontractingandtheassociateddifferentdistributionofsourcing.
Issuessuchasunequalaccesstoinformationandfinancialbarriersprecludelocal(andregionalLIC)firmsfromcompet-ingfortenders.Intheprojectcasestudiesthiswasparticu-larlytrueforframeworkcontracts.Anoptionforachievingcompetitiongainscouldbetoprioritiseincreasingtheabilityoflocalfirmstopartnertheframeworkcontract.Otheroptionsincludepositivepracticessuchasreducingtendersizeswhereappropriate.
ProcurementusingCPSminimisedmanyofthesebarriers.However,thebestresultswereachievedwhereCPShadca-pacityinplacetoachievetransparentandeffectivecompeti-tion.
Realisingthepotentialbenefitsofuntyingalsorequiresthesupplycapacitytomakepossiblegenuinecompetition.Thiswassometimesconstrainedbylownumbersofbiddersbothinprocuringatlocallevelswheretherewerelowskillsets,andalsointernationally,duetothelargesizeorspecificnatureofsomecontracts.
Thefollowingareasofdonorcontractingpracticeswerefoundtobelessthanwhollytransparentfromarecipientcountryperspective:donorcontractingatHQ;frameworkcontracts;B2B,twinning/partnershiparrangements;andprivateman-agementcontracts.Alloftheseareasofferscopeforde factotying.Themostopaqueaspectsofdonorpracticeconcerned
TCandassociatedconsultancies,theareasofassistancewhereformalandde factotyingwerefoundtobestillmostpervasive.
Thecountrystudiesfoundmanyexamplesofpotentiallyim-provedpracticeinPDeffectivenesstermsinthewaysinwhichuntyingisassistingdonorsinmovingawayfromtheconven-tionalstand-aloneprojectbasedapproachtocooperativeap-proaches.Thismovementfacilitatesbudgetsupport,pooling,delegatedcooperation,useoftheCPSanddecentralisation.Thesesteps,supportedbymanybutnotalldonors,werefoundtobestronglysupportiveofPDgoalsofownershipandalignmentandharmonisation.Untyingandownershipisatwo-wayrelationship;untyingsupportsownershipandstrongnationalsystemsandownershipassistuntying.
In terms of development effectiveness,countrystudiesarecau-tiouslypositive,findingevidenceofdirectemploymentandincomestreamgeneration.However,theseeffectsarehighlysectorandactivityspecificandsocouldnotbeattributedtountyingper se.
Thebalanceofsourcingofgoodsandservicesisshiftedbyuntyingtowardslocalsuppliers,asallstudiesindicate.Thispromoteslocalbusinessdevelopment,butthelinkageeffectsarelimitedbytheunderdevelopedstructureoftheeconomy.Thepromotionofregionalnetworkingwaslimitedandmoretobenefitthestrongereconomies.
Thecontentofaprojectratherthannominaltyingstatuswasfoundtobecriticallyimportantindeterminingoutcomes.Capacitybuildingandknowledgetransferappeartobemorelikelywherelocalandregionalfirmsofmediumcapacityundertakeprojectinvestmentactivities.
Untying aid: Is it working? 55
Chapter 8
T hisfinalchapterdrawstogetherinsummaryformtheconclusionsfromthewholestudy,PhaseOne,adeskand
donorfocusedreview,andPhaseTwo,partnercountrystudies,whicharesetoutmorefullyonachapterbychapterbasis.Theconclusionsareorganisedhereintermsofthethreequestionsproposedforevaluation:
• Towhatextenthasdonorbehaviourchangedasaresultofthe2001DACRecommendationonUntyingAidtoLeastDevelopedCountries,i.e.towhatextenthavedo-norsuntiedtheiraidfurther?
• Whatfactorshaveenhancedorimpededfullyuntyingdevelopmentassistance?
• Isthereevidenceoftheuntyingofaidhavingresultedinanincreaseoftheeffectivenessandefficiencyofaid?
Theseconclusionsarefollowedbyrecommendationsforwaysofensuringthefulfilmentofcommitmentstofullyuntieaidinanefficientandeffectiveway.
The 2001 Recommendation and untying of aidTheannualreportsonprogressinuntying,basedonstatisti-calinformationsubmittedtotheCRSandtheDAC,indicatecontinuedprogressonuntying,andthatallDACmembersareatleasttosomedegreeintheprocessofuntying.Thepropor-tionofuntiedaidrosefrom46%in1999-2001to76%in2007andforLDCsfrom57%to83%.If(untied)multilateralaidistakenintoaccountthen83%ofallODAin2007wasuntied.Formally,theODAofDACmembers,andespeciallytoLDCs,isoverwhelminglyuntied.
Intermsofeffort-sharingtheDACindicatedtwotargets.Thefirstisastraightforwardtargetof60%shareofuntiedODAintotalaidtoLDCs,whichalmostalldonorshadreachedwithinsixyears,includingthosethathistoricallyprovidedaidonalargelytiedbasis.ThesecondindicatoristheratioofuntiedbilateralLDCODAandmultilateralLDCODAwithinGNI.Atar-getequivalentto0.04%ofGNIwasagreedandisreportedto
havebeenexceeded.ThescopeoftheRecommendationwasextendedin2006and2008andreinforcedbytheParisDecla-ration.IthasbeenasignificantinfluenceonthebehaviourofDACmemberstatessince2001inencouragingandsustainingtheprocessofformaluntyingofbilateralODA.AtthisoverallleveltheresponsetotheRecommendationisanaidsuccessstory.
Thisstudyhasre-examinedthedataontyingstatusasreportedtotheOECD,andcomplementedthisfirstbyase-lectivereviewofdonorpoliciesandpracticesinrespondingtotheRecommendation,andthenthroughseekingviewsandevidencefrompartnercountries.Thereviewbroadlyconfirmstheannualreportsonprogressinuntyingasaformalprocess.
Thereviewoffivedonorcountriespoliciesprovidedmorespecificevidenceonhowtheuntyingprocesshasbeencharacterisedbytheremovalofformaldomesticregulatoryrestrictionsonthesourcingofgoodsandservicesbydonors,andinagreementswithpartnercountriesandcontractswithaidimplementingentitiesthatwereintendedtofavourdomesticorganisationsinthedonorcountries.GovernmentssuchasthoseofAustraliaandCanadathatpriorto2001hadlargelytiedaidmadenotjustsubstantialchangesinagencypolicybutagreedexemptionsfromdomesticrestrictionsonuseofpublicfunds.Thedonorpeerreviewsandthe2008studyonpromotinglocalandregionalprocurement(OECDDAC,2008b)confirmthatallDACmembershaveparticipatedtovaryingdegreesintheprocess.
TherehasbeenfurtherprogressyearonyearsincethePhaseOnereportreviewedtheresponseinthefiveyearsupto2006.Budgetarysupportanddebtreduction,thetwoalmostentirelyuntiedcategoriesofdirectassistancetogovern-ments,declinedfrom23%ofbilateralODAin2006to13%in2007.Nevertheless,theoverallproportionofuntiedaidin-creasedby2%,underliningthemoregeneralmovetowardsuntying.
8. ConClusions
Untying aid: Is it working?56
Chapter 8
Additionality of tied aid?Somewhodefendtyingpracticesorcautionagainstchangearguethatuntyingcouldresultinalossofdomesticsupportforaidoverall,orhavediversion-aryeffectsawayfromLDCstootherpartnercountries.ButabsolutelevelsofODAtoLDCshadalmostdoubledby2007since1999-2001.TheshareofLDCsintotalODAwasalso5%higher,andevenhigher(13%)whenthesomewhattemporarysurgesindebtreliefduring2005-06areexcluded.AgainthedonorreviewinthisstudyconfirmedthatuntyinghadhadnonegativeeffectoneithertotalODAoraidtoLDCs.
Acloserscrutinyindicatesvariouswaysinwhichthenarrativeofuntyingofaidasasuccessstoryshouldbequalified.TheuntyingofformsofaidexemptfromtheRecommendationhasbeenlessthanthoseincluded.Some30%oftechnicalcooperation(TC)andaidwithaTCcomponentwasreportedastied,oritstyingstatuswasnotreported.Itisdifficulttodeterminethefullextentoftyingoffoodaid,reportedaspartofassistanceinthreesectors:emergencyaid(21%tying),developmentfoodaid(53%tying),andcommodityaid(tyingstatusvirtuallyunreported).Butitisalsointerestingtonotethat20%ofdonoradministrativecostsarereportedin2007asuntied,althoughthiscategoryisexemptedfromtheRecom-mendation.
TherearestilllargevariationsinDACmembers’untyingandreportingontyingstatusoftheiraid.SevenDACmembersreportedover90%untyingin2007.TheUSisthemajordonorthatcontinuestotieextensively(37%ofitsbilateralODAin2007and57%ofallreportedtying).Halfofallnon-reportingwasbyJapan.Therewasahighlevelofnon-reportingforTCandsectorsinwhichtyingofprojectaidmightbeexpected:infrastructure,transportandenergy.Thereisalackoftranspar-encythatindicatestheneedforcautioninmakingstatisticalinferences.Improvedreportingisnecessaryasagreedinthe2008AccraAgendaforAction(seebelow).
Beforeconsideringthequestionoffactorsthatenhanceorimpedefullyuntying,thereisthepriorissueraisedbymanyinformantsinthecourseofthisstudy,thegenuinenessofthedeclaredformaluntyingofaid.Changesinpolicyandregula-tionsmayallowthesourcingofgoodsandservicesoutsideofthedonorcountrybut,inpractice,aretheresuchobstaclesastomakethissounlikelythataidremainsde factotied?
Twoformsofstatisticalevidenceindicateapresumptivegapbetweentheformaldeclaratoryuntyingofbilateralaidandactualpractice,whichimpliesaconsiderableelementofintendedorunintendedde factotying.First,thehighreportedproportionofcontractsawardedtocompaniesregisteredinthedonorcountry(Chapter3)findsfurthersupportinthedonorreview(Chapter4)andfrompartnercountries(Chapter7).Second,thereis,afterextensiveformaluntying,acontinu-ingassociationbetweentheamountofbilateralaidfromtheindividualdonorandthevalueofitsexportstothesurveyedrecipientcountries,exceptingonlySouthAfrica,forwhichaidisofmarginaleconomicsignificance(Chapter6).
Factors enhancing and impeding fully untying development assistanceThereviewoffivedonorsandfurtherdiscussionwithbothcountrypartnersandagencypersonnelhighlightedfactorsencouragingthemovestountyingofaid.First,theRecom-mendationandtheextentofitscoveragecontinuetobeapositiveinfluence.Second,movestofulfiltheParisDeclarationobjectivesofgivinggreaterownershipandalignmentwithpartnercountrystrategiesandharmonisationofdonoractionsisassociatedwiththeadoptionofaidmodalities,generalandsectoralbudgetarysupport,basketfundingofprogrammesandpoolingoffunding,forwhichuntyingisanecessarycon-dition.Thirdly,untyingincreasesprogrammaticflexibility,andvirtuallyallcountypartnerandaidagencypersonnelwereinfavourofthat.
Impedimentstoformaluntyingincludetheeffortneededtochangeregulatoryandlegislativeframeworks.Thereisalsothecarryovereffectofcontinuingprojectsandprogrammes.IntheimportantcaseoftheUSA,legislativerestrictionsonuseofpublicfundsremaininplace(whilstallowingforwaivers)andforfoodaidalmostentirely(OECD,2008a),inpartbecauseofeffectiveoppositionfromdomesticinterestgroups.
Thefivedonorsreviewalsoidentifiedseveral‘greyareas’oftyingpractice.ThesearealloutsidetheRecommendationandoftenjustifiedongroundsofsustainingdomesticsupportforaidofinterestgroupsandofprovidingvisibleevidencetothepublicofadirectlinkbetweenbilateralaidandimpactsinbeneficiarycountries,aswellasfosteringaidefficiencywheremarketsareunabletodeliver.Theseareasincludeemergencyandhumanitarianaidprovidedbothin-kindandthroughpersonnel,channellingofaidthroughNGOs,presumedtobeuntied,foodaid(nowexplicitlytiedonlybytheUSA),TC,sup-porttogovernance,post-secondaryeducation,researchandBusinesstoBusiness(B2B).
Thecountrystudyevidencesuggeststhattheobstaclestoun-tyingarenowformostDACmembersnotsomuchtocomple-tionoftheformalprocessofuntying,buttoreducingorevenremovingintendedandunintendedtyingpractices.
Aid effectiveness and developmental impacts of untyingAideffectivenessisunderstoodintermsofaidefficiencyinprovidingvalueformoneyandtheParisDeclarationaimsofownership,alignment,harmonisationmanagingforresultsandaccountability.Thedonorpolicyreview(Chapter4)foundthatagenciesassociateduntyingwithashifttoinstrumentsthateithergivegreaterresponsibilityforfinancialmanage-menttocountrypartners(directprogrammaticsupportanduseofcountryprocurementsystems(CPS)),ortojointdonorpartnersthroughpoolingoffunds,ortocivilsocietyorganisa-tionsemployingtheirownprocedures.Therewouldalsobedecentralisationofmanagementatcountrylevel.TheroleofHQprocurementwasbeingmuchreduced.
Untying aid: Is it working? 57
Chapter 8
Bothgovernmentsandotherinformantsinpartnercountriesstronglyaffirmedthatuntyingimpliestransferringresponsibil-ityforplanningandhandlingfundsfromdonorstorecipients(Chapter6).Thisismoreamatterofcontracts,modalities,usesofCPSandofferinglocalbusinessanopportunitytocompetesuccessfullyforcontracts.Recipientcountrystakeholdersarelessconcernedaboutremovingbiasesamongstdonorpartnersintheirtrade,exceptwhereexcessiveinefficienciesoftyingsignificantlyreducetheresourcetransfervalueofaid.
Theliteraturereviewcoveringtheefficiencyanddevelopmen-talimpactsoftyinganduntyingfoundanextensivebodyofeconomicanalysisandaquitelimitednumberofevidence-basedstudies,withfewrecentcontributions,focusedontheinefficiencyoftyingandpotentialtradedistortingeffects.Untyingwastypicallyfoundconvincinglylinkedtoevidenceontheadoptingofprogrammaticandpoolingmodalities,forwhichuntyingisanecessarycondition.SofarDACmembershavealsodonelittletosystematicallyevaluatetheactualconsequencesoftying.Thereisthereforenowellestablishedconceptualframeworkforempiricalinvestigationintothecon-sequencesofuntying.ThesixcountrystudiesrequiredapilotstudyinGhanatodevelopafeasibleanalyticapproachandshouldberegardedasanexploratoryinvestigation,providingprovisionalfindings.
Thekeyissueraisedbythecountryinvestigationsisthegenu-inenessofthecommitmenttountying.Istheroleofuntyingperceivedconstructivelyorpositivelywithinadonororgani-sationorbyitsagents?AlternativelyisuntyingperceivednegativelyasamodificationtoregulationsoncontractingandprocurementnecessarytoconformtotherequirementofuntyingstatusasintheOECDdefinitionsandthe2001Recom-mendation?Apositiveapproachwouldfirstimplytheuseofdirectbilateralinstrumentsandpoolingarrangements,whichgiveownershipandalignment,wherethisisregardedaspractical.Second,whetherprovidingprogrammaticorproject-typeassistanceitimpliestryingtostrengthenlocalfinancialmanagementandprocurementcapacityratherthanseeinglimitedcapacityasjustaconstraint.ApositiveapproachalsoimpliesmakingchoicesinprojectssuchasusingCPSandde-signingcontractpackagesthataremorelikelyresultinuntiedoutcomes.Anegativeorpurelyconformistapproachislikelytoresultinapreponderanceofde factotyingoutcomes.
Untying aid modalitiesThecriticalchoicefordonorpartnersisthatofaidmodalities.Thereisalackofcomparativeandconsistentdataontheusesofdifferentmodalities(particularlyjointpooling)andalsotheuseofpartnercountryfinancialmanagementandprocure-mentsystems.Nevertheless,takinganeclecticapproachtotheevidence,thecountrystudiesshowthattherearesystem-aticdifferencesamongstdonorsintherevealedpreferencesofdonorswasfoundincountrieswheretheiraidisoverwhelm-inglyuntied(Chapter6).AgroupofDACmembersprovideasubstantialpartoftheiraidtosupportjointlyprogrammaticapproachesandusepoolingarrangements.Thecountry
surveyforthePDEalsoconfirmsthatbroadlythesameDACmembers,theso-calledNordic+group,aswellasFrance,GermanyandSwitzerlandmakeextensiveuseofCPSinaidtogovernment.Others,includingthetwolargestdonors,USAandJapan,restrictthemselvesalmostentirelytoproject-typesupportandmakelittleuseofCPS.Project-typesupportisstillthedominantaidmodalityinallsurveyedcountries.
The project approach and aid effectivenessProgrammaticmodalitieshavebeenevaluatedelsewherewithbroadlypositivebuthighlycountry-specificfindings(e.g.IDD,2006).Thestudythereforefocussedontheuseofformallyuntiedfundingforapurposivelyselectedsampleofprojecttypestoillustratetherangeofcurrentpractices.Thefindingsfor21projectsfundedsinglyorjointlybydonorsinsixcoun-trieswerebroadlyconsistentandsopermitaprovisionalsetofconclusionsonefficiencyandaideffectiveness(Chapter7).
Projects,evenwhenformallyuntied,arerarelyfullyuntied,butincludeelementsofuntyingandde factotying.Inpar-ticular,itishighlylikelythatwherethereareheadcontractsformanagingTCthesewillbesourcedinthedonorcountry.Recipients’powerinnegotiationisakeyinfluenceonwhetherimplementingentitiesarelocalandthetop-levelcontractsaresubjecttoICB.WheredonorsallowahighproportionoffundstobedisbursedthroughCPSusingICBorNCBforsubcontract-ingcontractsaremorelikelytobeawardedtolocalorregionalfirms.
Value for money – aid efficiency:LocalcompetitivesourcingusingNCBorICBwasfoundtobecosteffectiveinmostcasesforgoodsprocurementcomparedtointernalcommercialorinternationalmarketprices,andtheefficiencygainsweremostclearwhereuntyingledtoincreasedcompetition.However,directcontractingandNCBwasinsomecasesmorecostlywheretherearelocalsupplyconstraintsandpowerfulsocialnetworks.Similarlyadonoremphasison‘quality’incontractspecificationratherthanvalueformoneycaneasilybecomeawayofnarrowingtherangeofsupplyoptions.
An unlevel playing fieldInternationalcompetitivebiddingresultedinamuchwidergeographicalsourcingofcontractorsincludingregionally.ThisismostevidentincontrastingthesourcingofbilateralprojectswiththosecontractedbytheWorldBankusingICBandNCBinthesixpartnercountries.However,evenwhereICBwasadoptedtheprojectplanningandcontractingprocessesofseveralDACmembersappearedtointentionallyoruninten-tionallyadvantagedonorbasedcompanies.
ManyfactorspreventlocalandregionalfirmsfromwinningcontractswhentheseareopentoICB,especiallywherethisisundertakenbythedonorundertheirownrulesandatHQlevel.Thesefactorsincludelackofaccesstoinformation,lackofskillsandcapacity,thesizeoflotsputtotender,aswellasotherfinancialbarrierssuchasweakcreditmarketsandnon-
Untying aid: Is it working?58
Chapter 8
availabilitylocallyofprofessionalindemnityinsurance.Wherefirmsarelikelytotenderonlyforcontractsintheircountryorregion,thentheyrunintoaproblemofactuarialriskassoci-atedwithunsuccessfulbids–incontrastwithdonor-basedormultinationalfirmsandnon-profitorganisationstheyareunabletospreadpreparationcostsforproposalsacrossmanyapplications.
Frameworkcontracts(aformofservices-in-kindfromtherecipientandprojectperspective)canprecludecompetitiongainsfromuntying.Anoptionforachievingcompetitiongainscouldbetoprioritiseincreasingtheabilityoflocalfirmstopartnerintheframeworkcontract.
Untying and aid effectiveness: Thecountrystudiesfoundmanyexamplesofpotentiallyim-provedpracticeinPDeffectivenesstermsinthewaysinwhichuntyinghasalloweddonorstomoveawayfromtheconven-tionalstand-aloneprojectbasedapproachtocooperativeap-proaches.Thesemoves,supportedbymanybutnotalldonors,includingbudgetsupport,pooling,delegatedcooperation,useoftheCPSanddecentralisation,increaseownership,facili-tatealignmentandachieveharmonisation.
Thefollowingareasofdonorcontractingpracticeswerefoundtobelessthanwhollytransparentfromarecipientcountryperspective:donorcontractingatHQ;frameworkcontracts;B2B,twinning/partnershiparrangements;andprivatemanagementcontracts.ThemostopaqueaspectsofdonorpracticeconcernedTCandassociatedconsultancies,theareasofassistancewhereformalandde factotyingwerefoundtobestillmostpervasive.
Developmental impacts of untyingCountrystudiesarecautiouslypositive,findingevidenceofdirectemploymentandincomestreamgeneration.However,theseeffectsarehighlysectorandactivityspecificandsocouldnotbeattributedtountyingper se.
Thebalanceofsourcingofgoodsandservicesisshiftedbyuntyingtowardslocalsuppliers,asallstudiesindicate.Thispromoteslocalbusinessdevelopment,butthelinkageeffectsarelimitedbytheunderdevelopedstructureoftheeconomy.
Thecontentofaprojectratherthannominaltyingstatuswasfoundtobecriticallyimportantindeterminingoutcomes.Capacitybuildingandknowledgetransferappeartobemorelikelywherelocalandregionalfirmsofmediumcapacityundertakeprojectinvestmentactivities.
Overall,anddespitereservationsontransparency,genuineun-tyingemergesfromthecountrystudiesasacriticallyimportantcontributoryfactortowardsachievinggreateraideffectivenessandnotonetobejustifiedonlyintermsofefficiencygainsorpromotinglocalbusinessdevelopment.Thepreciseeffectsalsodependverymuchoncountryspecificcircumstances,asisap-parentfromcomparinganyofthethreecountrypairings.
Therelativelynewprogrammaticmodalities:financialaid,asgeneralandsectoralbudgetarysupportandbasketfunding,isde factountiedanddemonstratesgenuinelyuntiedpractices.Fundsdirectedtogovernmentanditsagenciesaremostlikelytobedisbursedusingcountryprocurementsystems.
Incontrast,theformaltyingstatusofaprojectprovideslittlecluetotheactualextentofuntyinganditsconsequences.Projecttypeaidisnowmostlyreportedasuntied.Thatdecla-rationistakentomeanthatatleasttheheadcontractformallysatisfiestheOECDdefinitionforuntiedaidevenifinpracticeitisde factotied.
Inpracticethetypicalproject,asfundedbymostDACdonorsindividuallyorjointlyislikelytobeahybridoramixofaidcomponentsthatareformallyorde factotiedandalsoformallyorde factountied.Thisrealityimpliesaviewoftyingthatisquitedifferentfromthesimplecategoricalnotionthataidistied,partiallytiedoruntied.Fromthisperspectivetheuntiedcontentofaprojectcanbeincreasedthroughchangesincon-tractingatdifferentlevelsanddifferentcomponentactivities.
Technicalcooperationcomponentsofmostprojectsconsid-eredwerefoundtobede facto,ifnotformally,tied.Thisfind-ingimpliesfocusingonfurtherpracticalstepstobroadenthesourcingofTCandthemanagementaspectsofprojects.
Beyond rhetoric: the future of untying in a rapidly evolving aid context Beforeproceedingtomakespecificrecommendationsonhowtomovetowardsmorefullyuntiedaid,thatnarrowlyfocuseddiscussionshouldbesituatedinitswidercontext,avoidingrhetoricandrecognisingpoliticaleconomicrealities.
Fullyuntyingaidobligesdonorstogiveuptwoconvenientpractices,justifyingtheiraidthroughtyingandrelyingontheirnetworksofdomesticallybasedorganisations.Sofarsomedonorshavebeenreluctanttodothisresultinginbothopenandde factotying.
Similarly,untyinginpracticeexposesproblemsofrecipientgovernmentcapacityinfinancialmanagementandgovern-ance.TheuseofCPSraisesissuesofcapacityandalsorisksofcorruption,ascountrystudieshavenoted.Realisingmanyofthepotentialbenefitsofuntyingdependsondevelopinglocalcapacitytosupplybothgoodsandservices.Thisissueextendsalsotosecondlevelservicessuchasefficientlyfunctioningmarketsforcreditandpublicliabilityinsurance.
The changing global aid contextraisesstrategicpolicyissues.Thegrowingroleofconcessionalfundingfordevelopmentprovidedbyemergingnon-DACcountries,suchasChinaandIndia,mostobviouslyinSEAsiabutalsoinAfrica,wasaneverpresentbackgroundissueinundertakingthecountrystud-ies.However,thescaleandtermsofsuchconcessionalflowscouldnotbesatisfactorilydocumented,letalonequantified.
Untying aid: Is it working? 59
Chapter 8
Thisfundingisoftenverylargescaleandmostlydirectedtobasicinfrastructure,facilitatingcommodityexportsandsomeprestigeprojects.Mostlyprovidedasconcessionalloans,itappearedtobelargelytiedbutwithlittleconditionality,andisthereforeveryattractivetomanyaidrecipients.ThesecountriesareurgedtoacceptDACprinciples.Butinanycasethescaleofsuchnon-DACflowsraisesthequestionofwhatisthefuturecomparativestrengthofbilateralODAofDACmembers?Whatcombinationwouldbeappropriateofwhollyuntieddirectbilateralbudgetaryassistanceandproject-typeinvestmentandTCwhichareformallyuntiedbutinpracticearealmostallmixedpackagesofuntiedandde factotiedele-ments.
Newverticalfundingmechanisms,forexamplePEPFAR(President’sEmergencyPlanforAIDSRelief )andtheGFATM(GlobalFundtoFightAIDS,TuberculosisandMalaria),couldbeatoddswiththePDaideffectivenessprinciplesthatprovidetherationaleforuntiedaid.Duetotheirverticalnaturethesemechanismsmaynotbealignednorharmonised,preferringparallelimplementationstructurestoCPS.Theirverticalnaturemayalsoworkagainstindividuallytailoredlocalsourcingar-rangements.
Finally,newformsoffinancingseemlikelytoberequiredtoaddressthechallengesposedbothbytheglobalfinancialcrisisintheshorttermandclimatechangeinthelongerterm.
The way forward: recommendations The2005ParisDeclarationincludesafurtherqualitativecom-mitmentwithoutanyspecifictarget(Indicator8)tocontinuetheprocessofuntying.The2008AccraAgendaforActiononaideffectivenessreaffirmedthiscommitment,andcalleduponDACmembersinrespondingtoelaborateindividual plans by 2010 to untie their aid to themaximumextentand to improve reportingonthe2001Recommendation(OECDDAC,2009).Theseproposalsarestronglyendorsedandfurtherspecificmeasuresaresuggestedtosupportgenuineuntyingandreduce,ifnoteliminatede factotyingpractices.
Donorsshouldincreasethe transparencyoftheirreportingtotheOECD,forexamplebyincludingthetyingstatusofcatego-riesofaidexemptfromtheRecommendation,includingallTCandfoodaid.
TheuntyingplansshouldincludespecificmeasuresformakinggreateruseofCPSandspecifythemeanstheyhaveadoptedorplantoadoptforremovingobstaclesandencour-aginglocalandregionalfirmstoparticipatesuccessfullyinprocurement.
Donorsshouldconsiderindividualorjointinitiativestostrengthenlocalandregionalcapacitytocompeteforaidsup-portedcontracts.BroadeningthesourcingofTCandmanage-mentofprojectsshouldbepriorityareas.
Agencieswhoseuntyingpracticeisstillconstrainedbydo-mesticregulationsshouldincludeintheirplansproposalstoremoveorobtaingeneralexemptionsfromsuchrestrictionsonthedisbursementofpublicfunds,followingthepositiveexampleofAustraliaandCanadasince2001.
Reportingonimplementationofplansforcompletingtheuntyingofaidshouldbesubmittedtobecomepartofthean-nualreporttotheDACandalsotobemadeaspecificitemforinclusioninfutureDACmemberpeerreviews.
TheDACshouldconsideraddingtotheannualprogressre-portsonuntyingacountrypartnerfocusedreviewofprogressinuntyingtobeundertakenwiththefullcooperationofanLDCorHIPCcountryincludedintheRecommendationandalsoDACmembers.Suchcountrystudieswouldbothbuildonthisthematicevaluationandgivemoresubstancetotheannualreports.Thefirstofsuchannualreviewsmightbeun-dertakenincooperationwithacountryintheLatinAmericanandCaribbeanregion,whichisnotincludedinthisstudy.
Thisstudyhasonlybeenabletoprovideaprovisionalassess-mentoftheextentofuntyingandtheimplicationsforaideffectiveness.Theassessmentisonbalancepositivebuttherearemanyunknownsandtoolittletransparency.Therearetheinevitabledataproblems.ButtheseissuesoffullandgenuineuntyingconcernallDACmembers.Untyingandtyingprac-ticesrarelyappearontheevaluationagendaandsoopportu-nitiestoknowmorearesystematicallymissed.DACmembersindividuallyneedtomonitorandrigorouslyevaluateboththeirprogressinuntyingandinmakinguseofthecapacitieswithinpartnercountries.Thisstudycouldprovideabench-markforfutureassessmentswhenDACmembersaresharingtheirfindingsandlessonslearnt.
Untying aid: Is it working?60
Annex A
THEMATIC STUDY – THE DEVELOPMENTAL EFFECTIVE-NESS OF UNTIED AID: EVALUATION OF THE IMPLEMENTA-TION OF THE PARIS DECLARATION AND OF THE 2001 DAC RECOMMENDATION ON UNTYING ODA TO THE LDCSTERMS OF REFERENCE 22 FEBRUARY 2008 (Revised14May2008)
1 Background1. Fordecades,debatesonaideffectivenesshavefocussedontheissueofthetyingstatusofaid.Ithasbeenclearlydocumentedthattiedaidraisesthecostofgoods,serv-icesandworksby15%to30%onaverage,andbyasmuchas40%ormoreforfoodaid.Thisisaconservativeestimateoftherealcostsoftying,sinceitdoesnotincorporatetheindirectcosts.
2. Asaresult,afterextendedanddifficultnegotiations,theOECD/DACadoptedin2001aRecommendationtountiemuchODAtoLeastDevelopedCountries(LDCs).TheRecom-mendation(amendedin2006)alsoinvitesDACmemberstoprovideuntiedaidinareasnotcoveredbytheRecommenda-tionandtostudythepossibilitiesofextendinguntiedaidinsuchareas.ProgressachievedintheproportionofODAthatisuntiedistrackedinthecontextoftheMillenniumDevelop-mentGoals,e.g.targetnumber35.The2005ParisDeclarationonAidEffectivenessreiteratedthe2001DACrecommendationandprogressintheshareofaidthatisuntiedismonitoredthroughIndicator8.
3. Internationalconferencesondevelopment,suchastheMonterreyInternationalConferenceonFinancingforDevel-opment,haveconsistentlyhighlightedthatincreasingtheshareofaidthatisuntiedisasignificantmeanstoimproveaideffectiveness.ThisviewhasbeenreiteratedatsuccessiveG-8Summits,mostrecentlytheJune2007HeiligendammG-8Summitwhichnotedinparagraph21theneedtocontinueenhancingeffortstountieaid.
4. InviewoftheseargumentsandthecallsforfurtheruntyinganumberofDACdonorshaveeitherfullyuntiedtheiraid(e.g.Ireland,Norway,Luxembourg,andtheUnitedKingdom)oralmostcompletelyuntiedtheirbilateralaidpro-grammes(e.g.Australia,Belgium,Denmark,Finland,SwedenandSwitzerland).Inaddition,theEuropeanCommunityhasadoptedtworegulationsgoverningaccesstoECexternalassistanceunderwhichallaidtoLDCsisprovideduntied.Moreover,allexpertise(e.g.technicalco-operation)andfoodaidwillbeuntiedandaidfromtheEUbudget(i.e. exclud-ingtheEuropeanDevelopmentFund)willbeopentootherdonorsonthebasisofreciprocity.Finally,theUSMillenniumChallengeCorporation(MCC)alsoprovidesitsaidinuntiedform.
5. InpreparationoftheThirdHighLevelForumonAidEffectivenesstobeheldinAccra,Ghanaon2-4September2008,MembersoftheWorkingPartyonAidEffectivenessandtheDACNetworkonDevelopmentEvaluationhaveaskedforaThematicStudyoftheextenttowhichdevelopmentpartnershaveuntiedtheirassistanceandthekeyfactorspromotingorimpedingprogressonfullyuntyingdevelopmentassistance.Furthermore,theyhaveaskedtoidentifyexamplesofbenefitsoffullyuntiedaid.
6. Inthecontextofthe2001DACRecommendationtoUntieODAtotheLDCsacomprehensiveevaluation, inter alia,ofitsimpactismandatedforsubmissiontothe2009DACHighLevelMeeting.(Thecomprehensiveevaluationwillalsopayat-tentiontotheimplementationofthisRecommendationwithrespecttoachievingabalanceofeffortsamongDACMembersandpromotingandsustainingODAflowstoLDCs.)
7. InlightoftheclosecommunalitiesbetweentheThematicStudywithintheframeworkoftheevaluationoftheParisDeclarationonAidEffectivenessandtheDACrequestforacomprehensiveevaluationof2001DACRecommenda-tiononUntyingODAtotheLDCs,onethematicstudywillbeundertakentoassesstheeffectivenessofuntiedaid.The
AnnEX A: tErMs of rEfErEnCE
Untying aid: Is it working? 61
Annex A
evaluationstudywillincluderesultsassessmentsbasedoncasestudiesinanumberofpartnercountries.
2 Purpose 8. Thepurposeofthestudyistoprovidethe3rdHLFinAccrainSeptember2008,theDACandthe2009HLM,aswellasthewiderdevelopmentcommunity,withacompre-hensiveassessmentofcurrentdonorpoliciesandpracticesregardingthetyingstatusofaidandanassessmentoftheeffectsoftheuntyingstatusonaideffectiveness.TheThematicStudyshouldfocusontheresultsofuntiedaid,andexamineif,andtowhatextentthepresenteffortforuntyingaidhascontributedtoaideffectiveness.TheStudyshouldalsoexploretheprospectsforincreasingtheshareofuntiedaid,andwhererelevantprovidepolicyrecom-mendationonpromisingapproachesonhowtoachievethisobjective.
9. ThequestionstobeaddressedbytheEvaluationstudyare:
• Towhatextenthasdonorbehaviourchangedasaresultofthe2001DACRecommendationonUntyingAidtoLeastDevelopedCountries,i.e.towhatextenthasdonorsuntiedtheiraidfurther?
• Whatfactorshaveenhancedorimpededfullyuntyingdevelopmentassistance?
• Isthereevidenceofuntyingofaidhavingresultedinanincreaseoftheeffectivenessandefficiencyofaid?
3 Output10. TheThematicStudywillbeundertakenthroughaconsultativeprocesswithdonors,partnercountriesandcivilsociety.Theevaluationstudywillbeconductedintwostages.
• Theoutputofthefirststagewillbeareportprovid-ingacomprehensiveoverviewofthecurrentpoliciesandpracticesofDACmembersandnon-DACdonorsregardingthetyingstatusoftheiraidandtheeffectsonaideffectiveness.DuringthisfirststagetobebasedondocumentaryresearchandtheDACdatabase,ameth-odologyforthepartnercountrycasestudieswillalsobedeveloped.
• Theoutputofthesecondstagewillbearepresentativenumberof,evidence-basedcasestudiesinpartnercoun-triesconcerningtheeffectsofthe(un)-tyingstatusofaidonaideffectiveness.Thecasestudieswillbeundertakenincooperationwithlocalresearchcentres,suchasforinstancetheAfricanEconomicResearchConsortium.Thisstageshouldalsoexaminetheextenttowhichuntiedaidhasresultedinprocurementfromlocal/regionalcompa-niesanditseffectonaideffectiveness.
11.ThecomprehensivefinalreportcoveringStagesIandIIshouldoutlinepracticalpolicyrecommendationsforincreas-ingtheshareofaidthatisuntied59.
4 Activities12. Theoverallassessment(e.g.StagesIandII)willbepre-sentedinasuccinct,butcomprehensivereportthatcontains:
• Anexecutivesummary.
• BriefoverviewoftheinternationalrulesandagreementconcerningthetyingstatusofODA.
• AstatisticalanalysisoftrendsinthetyingstatusofODAincludingtying/untyingpracticesofemergingnonDACdonors.[Thereportshouldalsoaddressissuesofburden-sharingandtransparencyofprocessesandprocedures].
• AsurveyofthepoliticalfactorsandnationalrulesandregulationsbearingonthetyingstatusofODA.
• AnassessmentofthedevelopmentaleffectivenessoftiedversusuntiedODAbasedonaliteraturesurveyandtheoutcomesinarepresentativenumberofcasestudiesinpartnercountries.
• PolicyfocussedconclusionsandrecommendationonpromisingapproachestoincreasetheshareofODAthatisuntiedanddevelopmentallyeffective.
13. ThechapterconcerningtheinternationalrulesandagreementsregardingthetyingstatusofODAwillprovideasuccinctandcomprehensiveoverviewofOECDarrangementsandrecommendationsontiedanduntiedaidaswellasWTOarrangementsandagreements.Thechaptershouldassessthecurrentpositiononbothsidesofthedebateastothede factocredibility,balanceandenforceabilityofuntying.
14. Thechapterconcerningthestatisticalanalysesbydo-norcountryoftrendsinthetyingstatusofODAwillcovertheperiod1985-2005.Itwillaimatidentifyingrobuststatisticalcorrelationsbetweentheextentoftied/untiedODAand;
• financialinstruments,suchasloans,grantsormixedcred-its,andinnovativefinancinginstruments,suchasGlobalFunds(HIV/AIDS)
• typesofaid,suchasfinancialaid,technicalassistance,foodaid,programmeaid,budgetsupportanddebtrelief;
• bilateralandmultilateralflows;
59 PD para 31: “Untying aid generally increases aid effectiveness by reducing transaction costs for partner countries and improving country ownership and align-ment. DAC Donors will continue to make progress on untying as encouraged by the 2001 DAC Recommendation on Untying Official Development Assistance to the Least Developed Countries (Indicator 8).”
Untying aid: Is it working?62
Annex A
• incomecategories,and
• ODA/GNIratios(asaproxyforpublicsupportfortheaideffort).
15. ThechapterprovidinganoverviewoftrendsinDACMembersandnon-DACdonors’policiesandpracticeswillbebasedontheexistingdataanddocuments,especiallythean-nualProgressReportontheimplementationof2001UntyingRecommendationandtheresultsoftheDACstudyconcern-inglocalandregionalprocurementwhichiscurrentlybeingundertaken.
16.ThechapteraddressingthepoliticalfactorsinfluencingthetyingstatusofODAwillprovideananalyticalassessmentoftheargumentsmadeindifferentcountriesinsupportoftiedaidsuchas, inter alia,capturingorpreservingexportmarketsfordonors’companies,generatingdomesticsupportfortheaideffortandstrengtheningtheinteractionbetweencitizensfromdonorandpartnercountries.ToassessthestrengthofargumentsinfavourofuntyingODA,thechapterwilladdressissuesrelatedtotradedistortion,nationalaswellasinternationalpolicycoherence,costeffectiveness,andaideffectiveness.
17. Thesecond stageoftheevaluationstudy(e.g.Chaptervi)willassessthedevelopmentaleffects(oreffectiveness)oftiedversusuntiedaidprojectsandprogrammesthrougharepresentativenumberofcasestudiesinpartnercountries.ThisassessmentwillbebasedontheDACevaluationcriteria,e.g.• Relevance(e.g.appropriatenessoftheprocuredgoods,
worksorservices,andwhetherthetyingstatusofaidisconsistentwiththecounty’sdevelopmentobjectiveandneeds.
• Efficiency(e.g.costandnon-costfactorssuchas,qualityofgoodsandservices,deliveryperiod,guaranteeconditions,maintenanceconditions,costofrepairsandspareparts,safetyandhealthaspects)
• Impact(e.g.effectsonthemarketforlocalandregionalproducers,onpromotionofthelinkageofthelocaleconomywithregionalnetworks,globalnetworks,onemployment,onpro-poorgrowthandpovertyalleviation,andongenderandenvironment,impactofuntiedaidonCorporateSocialResponsibilityandonprojectlifetimecostsavings.
18.Theassessmentslistedabove(e.g.Chaptersii–vi)shouldbediscussedwithrepresentativesfromdonors,partnercountries,civilsocietyandtheprivatesectortoarriveatasmallsetofkeypolicyfocussedconclusionstoincreasetheeffectiveness,quality,transparencyandshareofaidthatisun-tied.60TheserecommendationsshouldmeettheDACcriteria
60 The modalities for this need to be established.
forreviewingpossiblewaystoincreasetheshareofaidthatisuntied.First,thereshouldbebroad-basedsupportfortheproposalandsecond,themagnitudeofthepotentialbenefitsshouldbereasonable.
5 Timing19. TheinitialreportonStageIshouldbereadyintimefortheThirdHighLevelForumonAidEffectiveness,whichwillbeheldinAccra,Ghanaon2-4September2008.ThespecifictimingofstageII,e.g.partnercountrycasestudies,willbedecidedwhenspecificToRhavebeendevelopedandrelevantcasestudycountriesidentifiedandagreedupon.Tentativeschedule:
1August2008:PhaseOne:DraftReport5September2008:DraftPaperonMethodologyforPhaseTwoMidSept-midOctober2008:DAC/EvaluatorsAgreeonDe-tailedPlansforPhaseTwoMidOctober-MidNovember2008:Recruiting/contractingforCountryStudiesNovember-December,2008:LaunchCountryStudiesMarch2009:DraftCountryStudiesApril2009:DraftReportonPhaseTwoMay2009:PeerReviewofDraftReport;AgencyreviewJune2009:PresentationofFinalDraftReport
6 Experts profile 20. Theresearchteamshouldcombinetrainingandexperi-enceindevelopmentpolicy,officialdevelopmentassistancefinancing,evaluationandprocurement,andhaveasoundexperiencerelatedtotheissuesofresultsassessments.Theyshouldalsohaveexperienceregardingtheissueofthetyingstatusofaidandbefamiliarwithrelatedpublicationsandworks.TheteamshouldcombineexpertsfromdifferentDACMemberStatesandfrompartnercountries.
7 Reporting21. Atthestartoftheassignmenttheconsultantswillprovideaninceptionreport.Duringtheassignmentthecon-sultantswillprovideprogressreportsonacontinuousbasis.ProgressinworkandprovisionalresultswillalsobepresentedanddiscussedinregularmeetingswithaManagementGroupcomprisingrepresentativesoftheDACSecretariatandtheRef-erenceGroupfortheEvaluationoftheParisDeclaration.Theconsultantsshouldacceptsuggestionsandguidelinesagreeduponatthesemeetings.
22.AttheendofStageIoftheThematicStudyaninterimreportwillbeproducedinEnglishnolaterthan1August2008.Thisreportwilldescribetheresultsofthedescriptiveanalysis,database,conclusionsandreferencesaccordingtosection4above.AttheendofStageIIafterthecompletionofthecasestudies,acomprehensivereportwillbeproducedinEnglishnolaterthanJune2009
Untying aid: Is it working? 63
Annex A
23. ThereportwilltheneventuallyberevisedfollowingcommentsfromtheManagementGroup.Thefinalreport,whichshouldlinkaclearcommunicationsstylewitharigor-ousanalyticalapproach,willbepublishedandsharedwiththeOECDmembersandpartnercountries.
8 BudgetAmaximumof€310.000isavailable.Distributedapproxi-matelyasfollows:
PhaseI: EUR52,000PhaseII:CoordinationandSynthesis: EUR90,000FourCountryCaseStudies: EUR168,000
Untying aid: Is it working?64
Annex B
AnnEX B: AdditionAl stAtistiCAl MAtEriAlTable B.1 DAC donor countries: bilateral ODA by consolidated sectors in 20071. Tying status as % of sector ODA.
Consolidated Sector (US$ million) (%)
Bilateral ODA Share of bilateral ODA
Untied Partially tied
Tied Not Re-ported
Untied share of bilateral ODA
I.1.a. Education, Level Unspecified 1,822 2.0 85.0 0.3 7.8 6.9 2.3
I.1.b. Basic Education 1,800 2.0 83.5 0.3 13.8 2.3 2.2
I.1.c. Secondary Education 764 0.8 68.7 0.0 10.3 21.0 0.8
I.1.d. Post-Secondary Education 3,781 4.2 48.9 0.0 39.5 11.6 2.7
I.2.a. Health, General 1,486 1.7 86.7 1.0 6.3 6.0 1.9
I.2.b. Basic Health 2,911 3.2 85.3 0.2 12.2 2.3 3.6
I.3. Population Pol./Progr. & Reproductive Health 6,482 7.2 90.9 0.1 8.3 0.7 8.6
I.4. Water Supply & Sanitation 4,402 4.9 89.0 0.4 8.1 2.5 5.7
I.5.a. Government & Civil Society-general 8,783 9.8 71.9 0.3 25.3 2.5 9.2
I.5.b. Conflict, Peace & Security 2,504 2.8 67.8 0.2 27.6 4.4 2.5
I.6. Other Social Infrastructure & Services 3,401 3.8 53.6 0.1 40.7 5.6 2.7
II.1. Transport & Storage 3,863 4.3 72.2 0.0 24.6 3.2 4.1
II.2. Communications 276 0.3 60.2 0.0 26.5 13.3 0.2
II.3. Energy 3,881 4.3 82.2 0.2 16.8 0.9 4.7
II.4. Banking & Financial Services 1,655 1.8 92.4 0.2 6.1 1.3 2.2
II.5. Business & Other Services 1,456 1.6 54.2 0.0 40.3 5.4 1.2
III.1.a. Agriculture 3,508 3.9 86.6 0.4 8.6 4.5 4.4
III.1.b. Forestry 565 0.6 92.2 0.0 1.5 6.3 0.8
III.1.c. Fishing 194 0.2 59.8 1.1 21.6 17.6 0.2
III.2.a. Industry 703 0.8 71.1 0.1 15.0 13.9 0.7
III.2.b. Mineral Resources & Mining 42 0.0 42.8 0.0 14.9 42.3 0.0
III.2.c. Construction 85 0.1 75.4 0.0 20.2 4.4 0.1
III.3.a. Trade Policies & Regulations 462 0.5 69.5 0.0 17.5 12.9 0.5
III.3.b. Tourism 46 0.1 43.9 0.0 31.2 24.8 0.0
IV.1. General Environment Protection 2,237 2.5 78.9 2.2 14.7 4.2 2.6
IV.2. Other Multisector 4,075 4.5 84.7 0.5 11.2 3.6 5.0
VI.1. General Budget Support 2,769 3.1 99.3 0.0 0.7 0.0 4.0
VI.2. Dev. Food Aid/Food Security Ass. 1,020 1.1 46.9 0.0 53.0 0.0 0.7
VI.3. Other Commodity Ass. 278 0.3 1.2 0.0 0.2 98.6 0.0
VII. Action Relating to Debt 9,010 10.0 100.0 0.0 0.0 0.0 13.2
VIII.1. Emergency Response 6,270 7.0 76.8 0.3 22.9 0.0 7.0
VIII.2. Reconstruction Relief & Rehabilitation 799 0.9 92.5 0.0 7.5 0.0 1.1
VIII.3. Disaster Prevention & Preparedness 132 0.1 95.4 0.0 4.0 0.6 0.2
IX. Administrative Costs of Donors 4,357 4.8 20.4 0.9 46.3 32.4 1.3
X. Support to NGO’s 799 0.9 84.4 3.0 1.5 11.1 1.0
XI. Refugees in Donor Countries 1,968 2.2 59.2 0.0 31.9 8.9 1.7
XII. Unallocated / Unspecified 1,360 1.5 44.6 0.5 8.0 47.0 0.9
All sectors 89,945 100.0 76.0 0.3 18.0 5.7 100.0
All sectors (including EC ODA) 103,318 66.2 13.2 15.6 5.0
Source: OECD CRS database.Notes: 1. Data are based on commitments in current US$ millions.
Untying aid: Is it working? 65
Annex B
Table B.2 Formal features of the 21 project case studies61 62
Project Number
Project Name Country Partner
Donor(s) Value of Donor contribution (millions)
Formal untied aid share61
Sector Major Input Type
1 RRHF (Rapid Response Health Fund) South Africa DFID UK£ 15 100% Health TC
2 EPDPHCP (Extended Programme for Delivery of Primary Health Care)
South Africa EC € 70 100% partially untied
Health TC
3 Buffalo/Leiden Storm Water & Drainage
South Africa VNG62 € 0.3 31% untied Water & Sanitation
TC
4 PROFIT (Production, Finance and Technology)
Zambia USAID US$ 17.5 100% Business TC
5 WSSNP (Water Sector Support In Northern Province)
Zambia Irish Aid € 2 100% Water & Sanitation
Investment
6 SWSP (Southern Province Water Supply and Sanitation Project)
Zambia KfW DM 27 Tied and untied aid (shares unknown)
Water & Sanitation
Investment
7 WSPIP (Water Sector Performance Improvement Project)
Zambia World Bank (IDA) US$ 33 100% Water & Sanitation
Invest-ment
8 NTP RWSS II (National Target Programme for Rural Water Supply & Sanitation
Vietnam AusAID, Danida, Netherlands
US$ 125 100% plus services-in-kind
Water & Sanitation
Investment
9 Saigon River Lowlands Flood Prevention Project
Vietnam AFD US$ 11.8 100% Water & Sanitation
Investment
10 Phan Ri-Phan Thiet Irrigation Project
Vietnam JICA US$ 51.9 100% Water & Sanitation
Investment
11 Nam Tuan Irrigation Project Vietnam LuxDev € 3.3 100% Water & Sanitation
Investment
12 Lao-Swedish Road Sector Project Phase III
Lao PDR Sida US$ 20 100% plus services-in-kind
Roads Investment
13 ICBRDP (Integrated Community Based Rural Development, Namor & Xay Districts, Oudomxay Province)
Lao PDR EC € 1.1 100% partially untied
Rural Development
Investment
14 Vientiane Water Supply Develop-ment Project
Lao PDR JICA, MOFA US$ 30.2 Tied and untied aid (shares un-known)
Water & Sanitation
Investment
15 Nam Theun II Lao PDR AFD, WB, Thai-Exim, ADB, EIB, Private Banks and Shareholders
US$ 1,450 100% untied Energy Investment
61 It proved impossible to consistently identify projects in the CRS database therefore this column is populated with the study team’s best estimate of the formal reporting of the untying status for each project
62 Association of Netherlands Municipalities.
Untying aid: Is it working?66
Annex B
Table B.2 (continuation) Formal features of the 21 project case studies6364
Project Number
Project Name Recipient Foreign Funder(s) Value of Donor contribution (millions)
Formal untied aid share64
Sector Major Input Type
16 PADSEA (Programme d’Appui au Développement du Secteur Eau et Assainissement)
Burkina Faso
Danida, Sida CFAF 30.7 100% untied Water & Sanitation
Investment
17 CNTS (Le Projet d’Appui au Centre National de Transfusion Sanguine)
Burkina Faso
LuxDev CFAF 1.3 100% untied Health Investment
18 ZIGA (Le Projet d’Approvisionnement en Eau potable de la ville de Ouaga-dougou à partir du barrage de Ziga)
Burkina Faso
WB, AFD, EIB, EDF, IDB, BADEA, KFAED, KfW, Belgium, OPEC, BERT
€ 228 92% untied and partially untied
Water & Sanitation
Investment
19 PADS (Le Programme d’Appui au Développement Sanitaire)
Burkina Faso
WB, Netherlands, Sida, AFD, ENFPA, PMNCH, GAVI, Gates Foundation, GFATM
CFAF 38 100% untied Health Investment
20 Damanko-Kpassa Ghana Danida/DFID US$ 2.5 65 100% plus services-in-kind
Water & Sanitation
Investment
21 NORWASP/NORST Ghana CIDA CDN$ 16.4 74% untied Water & Sanitation
Investment
63 It proved impossible to consistently identify projects in the CRS database therefore this column is populated with the study team’s best estimate of the formal reporting of the untying status for each project.
64 DfID contribution only as the value of the Danida supplied TC is not known.
Untying aid: Is it working? 67
Annex B
Table B.3 DAC donors: representativeness of the purposive sample of six partner countries
Donor All Developing Countries Countries under review
Commitments 2005-07 Untied aid share of donor ODA (%)
Commitments 2005-07 Untied aid share of donor ODA (%)
Australia 5391 83.5 138 78.2
Austria 3725 77.5 67 43.6
Belgium 4710 94.9 282 73.6
Canada 8454 62.1 301 67.3
Denmark 4394 90.4 755 93.5
Finland 1941 83.3 212 87.8
France 26080 78.5 2174 72.7
Germany 28404 74.3 1527 80.7
Greece 652 28.0 4 55.3
Ireland 1948 97.9 196 100.0
Italy 6667 65.7 117 51.1
Japan 41882 78.2 3880 70.4
Luxembourg 670 100.0 105 100.0
Netherlands 18595 90.8 1266 88.1
New Zealand 875 64.4 35 65.2
Norway 7500 99.8 370 99.1
Portugal 1066 56.5 1 99.0
Spain 8390 49.7 120 48.3
Sweden 8078 92.0 440 98.8
Switzerland 4182 68.0 184 76.1
United Kingdom 26033 100.0 1566 100.0
United States 76758 45.5 2494 73.2
Total DAC donors 286393 71.6 16235 79.3
Source: OECD Creditor Reporting System (CRS) database.Notes: 1. Data are based on commitments in current US$ mn.
Untying aid: Is it working?68
Annex C
1 Introduction – objectives and scope of investigationIfoneconsiderstheexportsfromdevelopedindustrialdonorcountriesthatconstitutetheDACtoaidrecipientdevelopingcountriesthenonewouldexpecttheflowofexportsfromthedonorcountrytotherecipientcountrytobeexplainedbyfactorssuchasthesizeoftheexportingeconomy,proximityordistance,andsharingacommonlanguage.Eventhetotallevelofaidmightbeaninfluenceontheexportsofalldonorcoun-tries.However,inamulti-donorworldwherenosingledonordominates,astheUSdidduringtheMarshallPlanera,thereisnoreasonwhythelevelofaidfromadonortoarecipientwouldinfluencethatdonor’sexport,unlessbilateralaidwerehavingatradedistortingeffectthroughexplicitorimplicittying.Tyingiswidelyconsideredtoadvantageexportersinthedonorcountry.Untyingisthereforeexpectedtoreducethedistortingbiasofaidinfavourofthedonor’sexports.Similarlybilateralconcessionalloanswerethoughttobeoftenassoci-atedwithdonorexports,forexamplethroughamixedcreditarrangement.
Thisannexthereforeinvestigateswhethercurrentbilateralaidflowsandassociatedpracticesarehavingtradedistortingeffects.TheaimistounderstandwhetherbilateralODAisaninfluenceondonorexport,evenwithcurrenthighlevelsofuntying,whilsttakingintoaccountothervariablesthatmightbeinfluencingthisrelationship(suchasFDI,exportcredits,ECdisbursements).Conventionally,tyingstatusandthedifferentformsinwhichaidisprovided,forexampleloansandgrants,havebeenhypothesisedtoinfluenceaggregatedonorexportflowsatthecountrylevel.
Inordertotakeintoaccountcountryspecifics,asimilareconometricinvestigationisbeingundertakenforeachofthesixcasecountrystudies65andanoverallanalysisforthesixcountriesisprovidedinthissection.Thesixcountriesarepur-posivelypairedaccordingtotradingregiontoisolatepotential
65 The econometric analysis for each country can be found in Country Reports which will be placed on the PDE web-site in due course.
regionaleffects:GhanaandBurkinaFasoinWestAfrica,LaoPDRandVietnaminSouthEastAsiaregionandASEAN,andZambiaandSouthAfricainSADCregion.Moreover,thesixse-lectedcountriesalsodifferfortheirDACclassification:BurkinaFaso,LaoPDR,ZambiabeingLDCs,SouthAfricaasUMIC,andVietnamandGhanaasOLICs;BurkinaFaso,GhanaandZambiaarealsobeingclassifiedasHIPCs.
TheAnnexisorganizedasfollows.Thefirstsectionbrieflysetsoutthetheoreticalframework.Sections3and4respectivelybrieflydescribethemethodologyandthedataused.Section5discussesandcomparestheresultsoftheempiricalanalysisoftherelationshipbetweenaidandtradeflowsforthesixcountriesunderreview.
2 Theoretical framework Thebasicideaistoconsiderwhetherthereisanytradedistort-ingeffectsofaidatanaggregatedlevelbyexaminingwhetherbilateralaidfromaparticulardonortoarecipientcountryaffectstradebetweenthatdonorandrecipientpairdifferentlyfromaidfromotherdonors.
TherecouldbeseveralreasonswhyOfficialDevelopmentAssistancecouldleadtoanincreaseinthedonor’sexports(positive impact).Themostobviousexplanationisthataid is formally tiedtoex-portsfromthedonorcountry(direct effect);bydefinitiontiedaidresultsintradedependency66.However,thereareseveralwaysforadonortoeffectivelytieaidwithoutaformaltyingagreement.Therefore,alsode factotyingbycovertrestrictionsonsourcingcangenerateincreasedtradewiththerecipient.TherearethenotherindirecteffectsthatmightleadtoapositivecorrelationbetweenODAanddonor’sexports.Aidmighthaveamacroeconomic impact intherecipientcountries;higherratesofeconomicgrowthandstructuraleconomic
66 Tied aid may simply finance donor exports that would have been procured from the donor country anyway; however, we would expect tied aid to usually increase donor’s exports.
AnnEX C: EConoMEtriC AnAlysis of tHE iMpliCAtions of BilAtErAl
Aid for donor EXports
Untying aid: Is it working? 69
Annex C
reformssuchastradeliberalizationmightindirectlyimpactondonor’sexports.Moreover,aidflowscouldresultinanincreaseintradewiththedonor,reflectingthegoodwilloftherecipienttowardsthedonor.Finally,theaidrelationshipmightfacilitatetradebetweenthedonorandtherecipientincreasingrecipi-ent’sproclivitytoprocuregoodsfromthedonor(reinforcingcommercialties).Intheory,bilateralaidcouldalsodecreasedonor’sexportstotherecipientcountry(negative impact);ifaidweretiedtheremightbeastraightforwardsubstitutioneffect(Oseiet al.,2004;Wagner,2003;Martínez-Zarzosoet al.,2008;Zarin-Nejadanet al.,2008;).Alsoiftherewereapartialtyingrestriction,forex-ampletolocal,regionalordevelopingcountrysourcing,thatcouldhaveanegativeeffect.Ingeneral,earlierstudies(Nilsson,1997;Wagner,2003;Martín-ez-Zarzosoet al.,2008;Zarin-Nejadanet al.,2008;Nowak-Leh-mannet al.,2008;PettersonandJohansson,2009)havefoundapositivecorrelationbetweendonoraidanddonorexports;theseresultshaveusuallybeenseentocorroboratethehy-pothesisthataidistied(formallyorinformally)toexportsfromthedonorcountry(PettersonandJohansson,2009).
3 Methodology – Empirical specificationThegravityapproachhasbeenoftenpraisedforitsremarkableexplanatoryvariableanditissolidlybasedintheempiricallit-erature(asopposedtoTajolimethodologywhichisonlyami-norapproach).Grangercausalityitselfhasseverallimitations67andwouldbedifficulttoimplementgiventhelimitednumberofobservationsavailable.Forthisreasonitwasdecidedtouseagravitymodelandadaptittothecountrylevel.
Inthissection,theempiricalspecificationusedtotestempiri-callythehypotheses68issetout.Thisfollowscloselythemeth-odologyadoptedbyMassaandTeVelde(2009).First,thefollowingregressioniscomputedinordertoun-derstandwhetheraidflowshaveanimpactontotaldonors’exportflowstotherecipientcountryatanaggregatelevel.
67 See ‘Problems in estimating the determinants of exports – the aid and trade relationship’ section of the background paper for a full discussion on the direction of causality as well as Granger causality drawbacks.
68 See ‘problems in estimating the determinants of exports related to the nature of the aid and trade relationship’ section of the background paper (Natali et al., 2009). The variable for the recipient having been in a colonial relationship, is also dropped; at the country level this variable loses its interest and moreover it is likely to be highly correlated with the common language variable.
ThedependentvariableExpirepresentstheexportflows69fromcountryitotherecipientcountry;YiandYpcimeasurerespectivelytheGDPandGDPpercapitaofdonorcountryi;DististandsforthedistancebetweentheexportingandtherecipientcountrywhereasComLiisadummyindicatingwhetherthedonorcountryandrecipientshareacommonofficiallanguage70;FDIistandsforforeigndirectinvestmentstotherecipientcountry;ExpCistandsforexportcredits;ImpRistandsforimportresiduals;ODAiisthecountryi’sOfficialDevelopmentAssistancetotherecipientcountrywhereasODAAll-irepresentsODAtherecipi-entcountryreceivesfromallthebilateraldonorsotherthancountry i.εiistheerrorterm,normallydistributed,withmean0andvarianceσε
2.
Theremightbeanumberofunmeasuredinfluencesthataffectbothaidandexportsinthemodel.Theomissionofvariablesrepresentingsuchinfluences(eitherdifficulttomeasureandunquantifiablerelationshipssuchaspoliticaltiesbetweenthedonorandtherecipient,orsimplyunderlyingrelationshipsnotidentifiedaspotentialdeterminants)couldleadtoattributingtoogreatanimportancetotheexplana-torypowerofforeignaid.Aidcouldbeproxyingforacomplexrelationshipincludingseveraloftheseomittedvariables.InordertoovercomethisproblemanOLSregressionofimportstothedonorfromtherecipientisestimatedandtheresidualsfromthisregressionareincludedintheoriginalequations.Thisprocedureiscarriedoutontheassumptionthatunderly-ingrelationships,andmoreingeneralunmeasuredvariables,betweenthedonorandtherecipient(tradingpartnersinthiscase)affectimportsandexportsinthesameway(Wagner,2003).
69 Exports include all traded goods but not services. Ideally, exports of goods and services would have been used as a dependent variable; this exercise would have proved particularly interesting considering aid is often tied to consultancy services. However, exports of services had to be excluded due to the unavailability of disag-gregated data (see data issues in the background paper for further details).
70 In order to adapt the original gravity model to the country level, the recipient’s GDP and GDP per capita (which only vary across years and not across donors) are not included in the regressions. The dummy ‘colony’, indicating whether the donor and the recipient have been in a colonial relationship, is also dropped; at the country level this variable loses its interest and moreover it is likely to be highly correlated with the common language variable.
(ODA aggregated)
Untying aid: Is it working?70
Annex C
Second,theaboveregressionisre-estimatedincludingapotentiallyimportantextravariableTS(tyingstatus).TSistandsfortheshareofdonorODAreportedastied.Thereisadirectlinkbetweenformallytiedaidandtrade.Whenadonorprovidestiedaid,therecipientisconstrainedtothepurchaseofdonorcountryexports.
Third,theregressionaboveisre-estimateddisaggregatingODAintoloansandgrants71totestwhethercertainaidinstru-mentsimpactdifferentlyontradeflows.
Aidinstrumentsaredistinguishedbasedontheoriginoftheflows.Grantsiisthecountryi’sgrantstotherecipientcountrywhereasGrantsAll-i representsthegrantstherecipientcountryreceivesfromallthebilateraldonorsotherthancountryi.Similarly,LoansiandLoansAll-i arerespectivelytheloanstherecipientcountryreceivesfromexportingcountryi,andfromallthecountriesotherthancountryi72.
However,theexclusionofcertainpotentiallyimportantfactorsmayconfoundtheinfluenceoftheexplanatoryvariablesonexportflows.Therefore,theaboveregressionisre-estimatedincludinganECvariable;somedonorpersonnelinGhanasuggestedthattheACP-EUrelationshipmightbeaninfluenceonexportsfromEUcountries.InordertoexplorefurtherthisrelationshipavariableforECdisbursementsisincluded.
4 DataTheeconometricanalysis includes annualflowsofex-ports and ofaidandotherdevelopmentfinancefrom upto22DACdonors tothesix recipientcountrieswithobservationscovering2002-07 becauseofdata availability and qualityis-suesforearlieryears. ThebilateraltradedataarefromtheIMFDirectionofTradeStatisticsdatabase,ODAdataaretakenfromtheOECDCRSdatabase,FDIcomefromtheOECDInterna-tionalDirectInvestmentStatisticsandofficialexportcreditsareasreportedtotheOECD.ThebilateraltradedataarefromtheIMFDirectionofTradeStatisticsdatabase,ODAdataaretakenfromtheOECDCRSdatabase,FDIcomefromtheOECDInternationalDirectInvestmentStatisticsandofficialexportcreditsareasreportedtotheOECD.
71 Aggregated ODA includes loans, grants, equity investment and grant-like (see definitions section). However, with the disaggregated analysis information about equity investment and grant-like is lost.
72 One cannot compute the log(loans) or log(grants) when the variable is equal to zero (the log is undefined). To handle this type of problem we add 1 to the data before logging it – log(0+1). We assume that such adjustment is immaterial in the dataset as all positive values in the data are large numbers.
Theanalysisusesaiddisbursementsandnotcommitments,becausetradeeffectsareexpectedtofollowactualdisburse-mentsandnotcommitments.Dataonthetyingstatusofaid,however,isnotavailableindisbursementform73.Therefore,aproxyforthetyingstatusiscomputedbyapplyingtheshareoftiedcommitmentstototalcommitments(averageoftheprevioustwoyears)andthenassumingthetiedshareindis-bursedaidtobethesame.
DataforGDPandGDPpercapitaaretakenfromtheWorldDevelopmentIndicators(WDI)Onlinedatabase.Allnominalvariables(exports,exportcredits,GDPandGDPpercapita,FDI)aredeflatedinto2000constantUSdollarsusingDACdeflators.DatafordistanceandcommonlanguagedummyaretakenfromtheCEPIIDistanceDatabase.
5 Comparative econometric results Limitationsontheavailabilityandqualityofthedata,thesmallsampleandpotentialproblemswiththeeconometricmodel,warrantcautionininterpretingtheresultsatthecoun-trylevel74.Over-interpretationofresultsfromsmalldatasetsshouldbeavoidedespeciallygiventheincompletenatureofthetradedata:theanalysisisrestrictedtotradeingoodsandthereforethepotentialtradedistortingimpactoftechnicalas-sistance/cooperation,whichisbothasignificantshareofODAandanevenlargershareoftiedODA,isnotpickedup.
Moreingeneral,findingsatthecountrylevelsuggestproba-bilisticandnotnecessarilyrobusteconomicrelationships.However,thissectionwillprovideafullerandpossiblymorerobustanalysislookingatthecomparativeresultsforthesixcases,highlightingsimilarpatternsandsystematicexplicabledifferences.
73 In the OECD Creditor Reporting System, tying status is reported against new commitments only.
74 See data section in background paper for further details.
(Tying status)
(ODA disaggregated)
Untying aid: Is it working? 71
Annex C
Itisinterestingtocompareresultsforsixcountriesdeliber-atelypairedaccordingtotradingregion:GhanaandBurkinaFasoinWestAfrica,LaoPDRandVietnaminSouthEastAsiaregionandASEAN,andZambiaandSouthAfricainSADCregion.Moreover,thesixselectedcountriesalsodifferfortheirDACclassification:BurkinaFaso,LaoPDR,ZambiabeingLDCs,SouthAfricaasUMIC,andVietnamandGhanaasOLICs;BurkinaFaso,GhanaandZambiaarealsobeingclassifiedasHIPCs.Econometricresultsbycountryarereportedineachcountryreportswhereasasynthesistable(Table6.4)withtheeconometricresultsforthesixcountriesisreportedinChapter6.
Inlinewithexpectationsandpreviouspublishedfindings,thesizeofdonoreconomyseemstobeoverallanimportantde-terminantofbilateralexports;anincreaseinGDPisassociatedwithariseinexports.However,thelevelofdevelopmentofdonorcountries,measuredbytheirincome per capita level,hasnoimpactonbilateralexportflows75.
Distanceisnotfoundtobesignificantanddoesnotseemtoaffectthelevelofexportflows.OnlyinVietnam,inaccordancewiththetheory,distanceisfoundtobehighlysignificantandnegative;anincreaseinthedistanceleadstoadecreaseinexportsfromdonorcountries.
Onestrikingdifferenceinresultsamongcountriesconcernsthecommonlanguagevariable.ThelanguagevariableprovestobestronglysignificantonlyforFrancophoneBurkinaFaso;thisdoesnotholdfortherestofthecountries76:withEnglishbeingthelinguafrancaofinternationalcommerce,noobservableadvantageisconferredonAnglophonecountries.Sharingthesameofficiallanguageisstronglyandpositivelylinkedwithbi-lateralexportflowsfromFrancophonedonorsandBurkinaFaso;suchrelationshiphoweverisnotsimplydrivenbyalanguageissuebutbythelinkswithformercolonistFranceastheestima-tionofanalternativespecification,wherethelanguagevariableissubstitutedwithacolonyvariable,shows.
Thedirectionoftherelationshipbetweenformallytiedaidandexportsisasexpectedbutnotatallrobust:theestimatedtying statusvariableispositivebutnotalwayssignificant.ThepositiverelationshipimpliesthatthehighertheshareoftiedaidoverdonorODAthehighertheexportflowsfromthedonorcountrytotherecipient(Ghana,SouthAfricaandZambia).Thisresultislesstrivialthanmightappear.Itisnotclearthattiedaidalwaysleadstoanincreaseindonorexportsastiedaidmaysimplyfinancedonorexportsthatwouldhavebeenprocuredfromthedonorcountryanyway(PettersonandJohansson,2009).
75 In Burkina Faso, however, the direction of the relationship between exports and the level of development of donor countries, measured by their income per capita level, is unexpected although apparently robust.
76 In Vietnam and Lao PDR, the common language variable is automatically dropped: no donors share the common official language with either Vietnam or Lao PDR.
Whenaidisdisaggregatedintermsofloansandgrants77,grantsfromthedonorcountryarefoundtobepositivelyassociatedwithdonors’exportswhereasgrantsfromtherestofbilateraldo-norsarenotsignificant.Suchresultsaresuggestiveofsomeformoftradedistortion:itisdifficulttoexplainwhybilateralgrantsfromaparticulardonortoarecipientcountryshouldaffecttradebetweenthatdonorandrecipientpairdifferentlyfromgrantsfromotherdonors.Grantsresultinanincreaseintradewiththedonorcouldbeseenasmostlikelyreflectinginformalorde factotyingofgrants;amongotherpossibleexplanationstherecouldhoweverbethegoodwillontherecipientparttowardsthedonor;still,theaidrelationshipmightsimplyfacilitatetradebetweenthedonorandtherecipient,increasingrecipient’spro-clivitytoprocuregoodsfromthedonor(reinforcingcommercialties)78.Incontrast,loans,whichareanunimportantformofaidinmostofthecountriesanalyzed,arenotasignificantinfluenceondonorexports.Thismightinpartbeexplainedbythegrant/loanlevelinthecountriesanalyzed:adisaggregationofODAintoloansandgrantsisnotnecessarilyverymeaningfulsincetheloancomponenthasbeenverysmallandthereforeanunimpor-tantformofaidinmostcountriesapartfromVietnam.
Tosumup,theaidvariablesareprovedtobeanimportantinfluencesuggestingthataiddespitetheformaluntyingisneverthelessasignificantpartofthecomplexcommercialrelationship.Results,asreportedinthebackgroundpaperwhereanaggregateODAvariablehasbeenincluded,high-lightthattheoverallexplanationdoesnotincreasemuchindisaggregatingfromthetotalaidvariable.
Otherflowsseemhowevertohaveabearonbilateralexportflows.DuetodatarestrictionsthisrelationshipisstudiedonlyintheSouthAfrican79casewhereFDIarefoundtobeaposi-tiveinfluenceonbilateralexportstowardthehostcountry.Thiscouldbeexplainedbypatternsofintra-firmtradewithinmulti-nationalenterprises(MNE),thatisexportsbetweenhomecountryparentsandtheirforeignaffiliates(HejaziandSafarian,2001).Forinstance,newfacilitiessetupbyaMNEsubsidiarycouldbeequippedusingimportedcapitalgoodsfromthehomecountry;therearehoweverseveralotherwayswellbeyondtheinitialinvestmentinwhichFDIandexportscouldbestronglylinkedinSouthAfrica.Otherpotentialinfluencesarefoundtobewhollyinsignificant:export creditshavenoimpactonbilateralexportflowsinthosecountrieswheresuchdatawasavailable.Thismighthoweverberelatedtothepoorqualityofthedata80.
77 Econometric results for an aid aggregate variable are reported only in the background paper.
78 See theoretical framework.
79 An attempt was made to include FDI as an explanatory variable in the empirical specification for the remaining countries; however, reviewing that data it was found this would have led to a substantial loss of observations as 2007 data are still not available from the OECD International Direct Investment Statistics, and there are also many missing values in the data for some specific donors in particular years.
80 Very few observations and missing values have been treated as zeros (see data section in background paper).
Untying aid: Is it working?72
Annex C
Thestatisticalrelationshipsfoundhighlightinfluencesreflect-ingtheavailabledata;intherealitythereareotherrelation-shipsoftendifficulteithertoidentifyormeasure.The import residualsvariable,whichisusedtocontrolforsuchunderlyingrelationshipsbetweenthedonorandtherecipientcountries,ishighlysignificantandpositiveforSouthAfrica,LaoPDRandVietnam(nonACPcountries),implyingthatspecialrelation-shipsbetweentradingpartnersdoexistandimpactpositivelyontheexportslevel.Thisresultissuggestiveofacomplexreciprocalcommercialrelationshipforthesecountries;bilat-eralaidcouldincreaserecipientexportstodonorcountriesinSouthAfrica,LaoPDRandVietnam.Ifitistruethatasaresultofthedonor-recipientrelationship‘theeffectivecostofgeographicdistance’reduces(PettersonandJohansson,2009),furtherresearchisneededtoanalyserecipienttradeflowsandnotjustexportflows.
Finally,overallresultssuggestthatdisbursements fromthe EC impactpositivelyonbilateraldonorsexportsinfullACPmem-bercountries(BurkinaFaso,GhanaandZambia)supportingtheideathattheACP-EUrelationshipisapositiveinfluenceonexportsfromEUmemberstatesinACPrecipientcountries.
6 ConclusionOveralltheseresultssuggestthataidisstilladistortinginfluenceontrade,evenafteruntyingofmostbilateralODA,justifyingacloser,disaggregatedexaminationoftheusesofaidandsourcingofgoodsandserviceswithaidfunds.Inthepost-tyingworld,alsodisaggregateaidflowsestimatessup-porttheexistenceofsometradedistortingeffectsofaidatleastforgrants;loans,whichareanunimportantformofaidinmostofthecountriesanalyzed,arenotasignificantinfluenceondonorexports.
Untying aid: Is it working? 73
Annex D
d.1 ghana Country study81
IntroductionThisstudyisacontributiontoPhaseTwoofthewiderOECDDAC–PDEThematicStudyonUntiedAid82.TheThematicStudyonUntiedAidisdesignedtoprovidebothacomprehensiveassessmentofcurrentdonorpoliciesandpracticesregardingthetyingstatusofaidandanassessmentoftheeffectsoftheuntyingstatusonaideffectiveness.TheGhanaCountryStudyisfundedpartlyunderagrantfromKfWtotheOverseasDevelop-mentInstitute(ODI),London,andpartlybytheSecretariatfortheParisDeclarationEvaluationandtheDCDoftheOECDinsupportingtheThematicStudyonUntiedAid.ThePhaseOnereport83focusedontheHQlevelandtheex-tentoftheformaluntyingthathasbeenreportedbydonors.PhaseOneconcludedthatODAtoLDCs,andtodevelopingcountriesmoregenerally,isnowoverwhelminglyformallyuntied,exceptingexemptedcategoriesunderthe2001Rec-ommendation.However,therehasbeenverylittlesystematicinvestigationintotheeffectivenessofuntiedaid,orevenintounderstandinghowuntyingworksatthecountrylevel.Thiscountrystudyispartofawiderattempttoaddressthisgap,andwasdesignedtoserveasthepilotforthesubsequentseriesoffivefurtherrecipientcountrystudies.InselectingGhanaforthepilotcountrystudy,severalfactorsweretakenintoaccount:
• Africanstatesmakeupalargeshareofaidrecipientcoun-tries.
• Ghanaiswidelyregardedashavingrelativelygoodgov-ernance,andhasnotsufferedanyrecentlarge-scalehu-
81 The study was undertaken by Matthew Geddes, Emmanuel J. Mensah, Luisa Natali and Peter Quartey ([email protected]) (Geddes et al., 2009).
82 The full title is: The Developmental Effectiveness of Untied Aid: Evaluation of the Implementation of the Paris Declaration and of the 2001 DAC Recommendation on Untying ODA to the LDCs.
83 http://www.oecd.org/dataoecd/5/22/41537529.pdf
manitariancrisisorothershockthatwouldhaveresultedinahighproportionofnon-developmentaid.
• Donorsemployawiderangeofinstrumentsandmodali-ties.
• Ghanahasarecentlyreformedprocurementlawand,inaddition,arangeofprocurementoptionsshouldbeavail-able.
• TheGovernmentisstronglycommittedtothePDprocessforaideffectiveness,asreflectedinithostingtheAccrasummitin2008,thereforeensuringtherequisiteco-op-erationforastudyfocussedontheeffectivenessofdonorpractices.
Ghanahasapopulationofapproximately23millionandinrecentyearshasenjoyedsteadyandincreasingeconomicgrowth,aswellasbeingthefirstSub-SaharanAfricancountrytoachievetheMillenniumDevelopmentGoal(MDG)onpov-erty.GNIpercapitawasUS$450in2005,andODAwas10.4%ofGDPin2005.ODAtoGhanaamountedtoUS$1,236mnin2007,equivalentto10%ofGNIandoverUS$50percapita.Ghanaisregardedasa‘star’partnerofthedevelopmentco-operationcommunity.Supportisbroadbased:theWorldBankprovided20.4%ofODAdisbursementsin2007;Netherlands,12.8%;UK,11.9%;USA,9.4%;EC,6.7%;AfDB,5.4%;andDen-mark5%,withfiveotherdonors–Canada,France,Germany,JapanandSpain–eachcontributingbetween2.5%and4.7%.BilateraldonorsincreasinglydominateaidsupportinGhana.Asthepre-Accraaidperformancereviewrecognised,thismul-tiplicityofdonorsexemplifiestheneedforactiononimple-mentationoftheParisDeclarationonaideffectiveness.
MethodologyUntyinginpracticeinvolvestherelaxationofjustonecon-straintontheprovisionofaid–onthesourcingofgoodsandservices.Theapproachadoptedistoconsiderthequestionofeffectivenessintermsoffivesequentialissues.Firstly,it
AnnEX d: EXECutivE suMMAriEs of Country studiEs
The full country reports can be downloaded from http://www.oecd.org/dac/evaluationnetwork
Untying aid: Is it working?74
Annex D
examinestheassociationbetweenformaluntyingandaidmo-dalitiesandusesinstatisticalterms.Secondly,itlooksattheconnectionsbetweennewerandpotentiallybetterpracticesinthePDaideffectivenesssenseforwhichuntyingisaneces-sarycondition.Thirdly,itinvestigatestheeffectsofuntyingonprocurement,intermsofproceduresused,thesourcingofgoodsandservices,andtheimplicationsforthelocalandwiderregionaleconomy.Thisleadstoaprovisionalexplora-tionofboththecostefficiencyandfinallydevelopmentaleffectivenessimplicationsofuntying.Itishopedthatthisstudywillprovideabenchmarkagainstwhichfurthermovestowardsuntyingcouldbeassessed.Inaninitialreview,includingdocumentationanddiscus-sionswithstakeholders,itbecameclearthattoensurefocusandcomparabilitybetweenprojectstudiesthisstudyshouldconcentrateononesector.WaterandSanitationwasselected,becauseitisaninfrastructureandgrowthsectorwithastrongsocialorientation,whichinvolvesprocurementofsubstantialamountsofbothphysicalgoodsandservicesaswellasasig-nificantTCcomponentandavarietyofapproaches.ItisalsoasectorinwhichODAplaysakeyrole,manybilateraldonorsarestillinvolved,andithasbothruralandurbancoverage.
Asanexploratoryinvestigation,thestudyadoptedaneclecticapproach,combiningdifferentformsofanalysistodocumentandthenexaminetheconsequencesofcurrentaidpractice.Thisevidence-basedinvestigationinitiallyinvolvedadeskreviewoftheavailablestatisticalinformation,publishedandgreydocuments.Thestudyteamundertookaseriesofstructuredinterviews,betweenNovember2008andFebruary2009,withmajordonorsandthosesignificantinwaterandsanitationandgovernmentanditsagencies,aboutcurrentpolicyandpractice.Atwo-stepprojectanalysisincludedaninitialreviewofsixdonoractivitiesinthewatersectorand,duetodataaccessproblems,onlytwoin-depthstudiesofcurrentdonor-supportedprojects,toinvestigateprocurementpracticesandtheirimplicationsforcosteffectiveness.Inonecase,theteamundertook‘ground-truthing’fieldvisits.
Aid modalities and tying practicesAreviewbasedonOECDCRSdatafoundthat,formally,DACbilateralODAtoGhanawas75%untied,11%partiallyuntied,11%tiedand3%notreportedduring2005-07.ThispatternisbroadlysimilartoaveragereportedtyingstatusforbilateralODAforLDCsandHIPCsoverall.Thereis,however,significantvariationamongdonors,bothintermsoftyingandalsoofnon-reporting.ThereisanexistingmonitoringsystemforaidinGhana(theDevelopmentPartnerEnvelope),reportingtotheGovernmentanddonorsintheConsultativeGroup(CG),focusedonboththeMDGandPDobjectivesthatfacilitatedthestudy.A2006surveytotheCGbroadlyconfirmedthetyingstatusofaid,asreportedtotheCRS.Thedisbursementbyaidmodalityshowsover40%ofbilateralODAchannelledthroughprogrammaticmodalities,forwhichuntyingisnecessaryandalsolikelytoinvolvecountryprocurementsystemsincludinggeneralandsectoralbudgetarysupportandsomeformsofbasketfund-
ing(TableD.1).Theremainderofaidisdisbursedthroughaprojectapproach.Pooledfundingofprojects,anewermodal-itywhichislikelytobeassociatedwithuntying,wasidentifiedasgrowinginimportancebutisnotyetseparatelymonitored.TherearewidedifferencesindonorpracticeswithCanada,Denmark,France,Germany,Netherlands,Switzerland,theUK,aswellastheECandWorldBank,stronglysupportingtheprogrammaticarrangements.Otherdonorsprefertocontinuelargelywithestablishedproject-relatedarrangements(Japanandsomeminordonors),oreventoestablishnewparallelproject-typearrangements(USA),whichisseeminglyoutofstepwiththePDgoalofharmonisation.Thereisextensiveuseofcountryprocurementsystems(CPS),includingover50%ofaidtotheGovernmentsector,imply-ingthatatleast20%ofbilateralaidinaprojectapproachalsousesCPS.Anattempteddonorsurveybytheteam,toquantifybothuseofpoolingarrangementsanddifferentprocurementpractices,wasunsuccessfulduetopoorresponse.Thiswaspartlybecausedonors’ownfinancialandstatisticalmonitoringappearnottobeorganisedtoreportdirectlyoneithertyingpracticesoruseofdifferentprocurementarrangements,andpartlyduetoassessmentfatiguefollowingthesurveyunder-takeninpreparationfortheAccrameetingonthePD.
Thereremainconsiderableinconsistenciesbetweenthehighlevelofreportedformaluntyingandtheviewsofmanystakeholders(includingmostGhanaians)that,inpractice,tyingisstillwidespread,evenpervasive,especiallyonthepartofsomedonors.Ghanaiansuniformlyseeuntyingnotsomuchasavoidingaid-transferinefficienciesandtradedistor-tions,butratherintermsofownershipandalignmentwithnationalprioritiesfordevelopment,whichisonlyassuredwithunrestrictedprogrammaticaidandtheuseoftheGhanaianprocurementsystem.Muchtechnicalcooperation(TC)iscloselyassociatedwithtiedaid:between2005and2007only35%ofstand-aloneTCaidwasuntied,comparedwith80%ofnon-TCaid.Interviewevidencesuggeststhatthisassociationisaconsequence,notnecessarilyintended,ofanumberoffactors.Donorschooseareaswheretheyhavecomparativeprofessionalstrength,e.g.hydraulicsandwaterengineering.Frameworkcontractsfavournationalconsultantsandtheprovisionofservicesinkindisexplicittying.Therearenoevidentformallinksbetweentheuseofdifferentaidinstruments,grantsorconcessionalloansandtyingprac-ticesamongstDACdonors.However,thisisapparentlynotsofornewdonors.Thescaleandmodalitiesofconcessionalfundingbynewdonors,suchasChina,remainsoutsideoftheCGandDACmonitoringprocesses,butappearstobelarge,growingandintheformoftiedconcessionalexportcredits.Tyingis,bydefinition,apracticethatconfersacompetitiveadvantageondonorexportersofgoodsandservices.Ifde factotyingwerewidespread,despiteformaluntying,thenthiswouldbereflectedindonorexportscomparedwiththeircompetitors.Aneconometricanalysiswas,therefore,under-takentodeterminewhetherODA,formaltyingstatusandthedifferentformsinwhichaidisprovided(loansandgrants)
Untying aid: Is it working? 75
Annex D
hadadiscernibleimpactonaggregatedonorexportflowsatacountrylevel.TheoverallresultssuggestthattheODAingeneral,andgrantsinparticular,haveatradedistortingeffect,andwerefelttojustifycomparableinvestigationsforthefiveothercasestudycountries.
Project analysisUntyingishighlyinfluencedbythemodalityofaid.Budgetsupportandpoolingarrangementswherefunding,oncecom-mitted,isnotdistinguishableascomingfromanyonedonor,arebothhighlyassociatedwithuntiedaid.Incontrast,theextentofde factotyingoruntyingwithinprojectsisunclear.Aninitialreviewofthewatersectorrelatedprojectssupportedbysevendonors,Canada,Denmark,theEC,Germany(KfW),Netherlands,theUK(DFID)andtheUSAconfirmedthatmostprojectsare,whatevertheirformalstatus,typicallyhybrid,
comprisingamixofaidelementswithdifferenttyingstatus,procedures,implementingagenciesandrulesets.Morein-depthinvestigationintoprocurementpracticesandoutcomesisresourceintensive,andonlypracticalforanon-goingprojectwheredocumentationislocallyaccessibleandpersonnelinplace.Thestudy,therefore,restricteditselftotwocasestudiesofwaterimprovementprojects:aDanida-DFID(UK)pooledfundedprojectandaCIDA(Canada)supportedproject.Thecasestudiesfocusedonprojectinputsandout-putsandoutcomes,suchastheextentoflocalsourcingandcost-effectiveness.The‘headcontract’thatdeterminestyingstatus,asreportedtotheCRS,wasfoundinpracticetobetiedinbothcasestudyprojects.Inbothcases,the‘headcontract’accountedforthemajorityoftheTCresources.However,subcontractingpracticesandcontractornationalitieswithintheprojectswere
Table D.1 Ghana: bilateral DAC donors and EC – disbursements by aid modality 2006-08 (US$ mn)
Donor Project Approach
Basket-Funding
Sector Budget Support
earmarked
Sector Budget Support
unearmarked
General Budget Support
IMF2 Total disbursements
Project Approach as % of donor
disbursements
Belgium 2.0 0.0 0.0 0.0 0.0 - 2.0 100.0
Canada 81.4 0.0 55.8 0.0 46.2 - 183.4 44.4
Czech Republic 0.5 0.0 0.0 0.0 0.0 - 0.5 100.0
Denmark 1.7 161.9 12.3 0.0 24.4 - 200.3 0.9
EC 230.8 0.0 0.0 2.9 79.5 - 313.1 73.7
France 71.6 0.0 0.0 0.8 57.7 - 130.0 55.0
Germany 76.8 24.8 0.0 0.0 39.3 - 140.9 54.5
Italy 52.8 0.0 0.0 0.0 0.0 - 52.8 100.0
Japan 116.5 0.0 0.0 0.0 3.1 - 119.6 97.4
Netherlands 274.2 55.8 0.0 38.4 91.1 - 459.5 59.7
Spain 84.2 0.0 0.0 0.0 0.0 - 84.2 100.0
Switzerland 8.8 4.4 0.0 0.0 22.7 - 35.9 24.5
UK 94.0 24.7 12.7 88.6 227.0 - 447.0 21.0
USA 369.3 0.0 0.0 0.0 0.0 - 369.3 100.0
IMF/UN** 1083.5 40.2 0.0 45.0 433.9 116.6 1719.2 63.0
Total (excluding IMF/UN)
1464.5 271.6 80.9 130.7 591.0 0.0 2538.6 57.7
Total (including IMF/UN)
2548.0 311.8 80.9 175.7 1024.8 116.6 4257.7 59.8
Source: WB DPE, 2008Notes: 1. 2006-07 data is based on actual disbursements whereas 2008 data is based on projected disbursements. 2. Support to Balance of Payments (only IMF) 3. It includes: AfDB, EIB, FAO, Global Fund, IFAD, ILO, IMF, IOM, NDF, UNAIDS, UNDP, UNESCO, UNFPA, UNHCR, UNICEF, UNIDO, WB, WFP, WHO.
Untying aid: Is it working?76
Annex D
moreimportantindetermining,forexample,theextentoflocalsourcingofservicesandgoodsandlikelytradeeffects.Useofcountrysystemsemergedasbothanoutcomeofandaninputtotheuntyingprocess.Projectsandpartsofprojectswhichprocuredgoodsandservicesthroughthecountrypro-curementsystem,usingNationalCompetitiveBidding,wereuntiedandpredominantlycontractedlocallyandregionally.ThisappearedtobemainlyaresultofsmallerlotsizesandlocaladvertisingunderCPS.AnexploratorycosteffectivenessanalysisofprocurementwithintheDanida-DFIDsupportedprojectusingCPSindi-catedthat,forthemostpart,goodsandcivilworkscontractswereboughtatcompetitivepricesanddonesolocally.Onceruleshadbeenagreed,thepoolingarrangementbetweenDanidaandDFIDalsoledtosavingsfromsharingauditingandmonitoring,aswellasmatchingresponsibilitiestosectorexpertise.
ConclusionsUntyinginGhanaiscontributingsignificantlytotheaideffectivenessagenda.Untyingiscreatingpolicyspaceforrecipientcountryownership,useofrecipientcountrysystemsandalignmentwithrecipientcountrypriorities.Theexistenceofstrongcountryleadership,suchasthroughthecreationofSWAPs,wasalsoseenasadriverofuntying.Knowingwhetheraidsupportisformallyuntiedprovidesaquitelimitedindicationastothebalanceofactualtyinganduntying,andthedevelopmentalimplicationsofthesepractices.Tounderstandtheserelationshipsrequireslookingatthechoiceofmodalityand,forproject-typeaid,theactualarrangementsforcontracting,fromtheheadcontracttothelowestlevelofsub-contractingandtheprocurementofgoodsandservices.Asthereisadearthofempiricalstudiesonhowformallyuntiedaidisworkinganditsimplications,thisconclu-sionwasfelttojustifyaccordingahighprioritytoprojectanalysisinthecountrystudies,foracrosssectionofdonors.Atamacro-level,thedonorchoiceofmodalitiesandanex posteconometricinvestigationintotherelationshipsbetweenaidandtradeflowsprovidesfurtherevidenceonuntiedaidpracticeanditsconsequences.Ghanahasrelativelygoodco-operativemonitoringofaidbyGovernmentandthedonorcommunity.Nevertheless,whatthesereportingarrangementstellusaboutmodalitiesandusesofnationalandotherformsofprocurementisquitelimited.ItalsoprovedalmostimpossibletorelateindetailaidasreportedfromdonorHQtotheOECDandaidasbeingmonitoredatacountrylevel.Theconclusionfromthispilotstudyisthatinvestigatingtheextentofgenuineuntying,continuingde factotyingandthedevelopmentalimplicationsofthesepracticesisgoingtobefarfromeasy,evenwithfullco-operationfrommajorstakeholders.
Untying aid: Is it working? 77
Annex D
d.2 Etude de pays Burkina faso84
IntroductionCerapportprésentelesrésultatsdel’étudedepaysduBurkinaFaso.L’étudemiseàmettrecesgrandesquestionsenavantàtraversunebrèverevuestatistiqueetéconométriquedesdon-néesdisponiblesetparl’évaluationdequatreétudesdecas.L’accentdesétudesdecasétaitsurlespratiquesdesdonneursenmatièred’approvisionnementetdesarrangementslocauxquipourraientavoiruneffetsignificatifsurlafaçond’acheterlesbiensetservices.Lesétudesdecasconcernaientdeuxpro-jetsausecteurdel’eauetdel’assainissementetdeuxausec-teurdelasanté.Ils’agitdedeuxprojetsbilatéraux,unprojetfinancéparuncofinancementconjointparungrandnombredebailleursetunprojetfinancéparunfondscommun.
La revue statistique et économétriqueLeBurkinaFasoreçoitdesmontantsimportantsenaidepubliqueaudéveloppement(APD).Depuis1998l’APDapratiquementdoubledeEUR433millionàEUR862millionen2007.Ainsi,leBurkinaFasodépendlargementdel’aide
84 The study is undertaken by Piet Lanser ([email protected]), Judith Mat-tiejsesen ([email protected]) and Elie Pare (Lanser et al., 2009).
extérieurequireprésentait15%duproduitintérieurbrut(PIB)et47%dedépensesdel’Etaten2007.Environtroisquartsdel’aideestfourniesousformededons.Parmilesdixpremiersbailleursdefondsontrouvecinqbailleursbilatéraux,ycom-prislaCommissionEuropéenne,troisinstitutionsfinancièresinternationales(IFI),leSystèmedesNationsUnies(SNU)etuneorganisationnongouvernementale(ONG).LaplupartdesIFIoctroientprincipalementdesprêts,maiscertaines,tellequelaBanquemondiale,fournissentégalementdesdons.En2007,l’aidedébourséeétaitrepartieentreappuibudgétairegénéral(29%),aideprojet,ycomprisl’assistancetechnique(63%),fondscommuns(7%)etaidealimentaire(1%).
Lesdonnéesstatistiquesdusystèmedenotificationdespayscréditeurs(SNPC)del’OCDE/CADetdesrapportsdelaDirectionGénéraledelaCoopération(DGCOOP)montrentdegrandesdifférences.UngrandnombredepartenairesduBurkinaFasonesontpasmembresoupartenairedel’OCDEetparconséquentneparticipentpasauSNPC.QuantauxpaysdontlabasededonnéesduSNPCcomprendlesdonnéesonconstatequeleschiffresfournisenmatièrededéboursements
Table D.2 Burkina Faso: bilateral DAC donors and EC: disbursements by aid modality 2007 (US $ mn)
Donor Project Approach
Basket-Funding
Sector Budget Support
earmarked
Sector Budget Support
unearmarked
General Budget Support
Total disbursements
Project Approach as % of Total
disbursements
Austria 4.4 0.0 0.0 0.0 0.0 4.4 100%
Belgium 5.4 0.0 0.0 0.0 0.0 5.4 100%
Canada 7.2 11.3 0.0 0.0 0.0 18.5 39%
Denmark 30.6 4.7 0.0 0.0 8.7 44.0 70%
EC 129.5 0.0 0.0 0.0 74 203.5 64%
France 44.4 6.6 0.0 0.0 9.5 60.5 73%
Germany 19.7 0.0 0.0 0.0 6.7 26.4 75%
Italy 2.9 0.0 0.0 0.0 0.0 2.9 100%
Japan 11.5 0.0 0.0 0.0 0.0 11.5 100%
Luxemburg 6.6 0.0 0.0 0.0 0.0 6.6 100%
Netherlands 14.7 26.4 0.0 0.0 19.9 61.0 24%
Spain 0 0.0 0.0 0.0 0.0 0.0 0%
Sweden 7.9 4.1 0.0 0.0 9.3 21.3 37%
Switzerland 3.1 2.6 0.0 0.0 7 12.7 24%
UK 0 0.0 0.0 0.0 0.0 0.0 0%
USA 12.9 0.0 0.0 0.0 0.0 12.9 100%
Total 300.8 55.7 0.0 0.0 135.1 491.6 61%
Source: DGCOOP, 2008
Untying aid: Is it working?78
Annex D
auSNPCsontsouventsupérieursàceuxcomprisauxrapportsdelaDGCOOP.Apartdesdifférencesduesauxerreursetomissions,ilparaîtqu’unepartiedesl’APDdestinéeauBurkinaFason’yestpasenregistrée.LaDGCOOPprésenteunchiffredel’aidenonliéede92%.Cependant,ilestapparuquecechiffreestunesimpleestima-tion.Lesquestionnairespériodiquesnecontiennentpasunequestionenlamatièreetl’indicateur8dusystèmedesuividelaDéclarationdeParisn’estpasrenseigné.L’OCDEprésenteunchiffredel’aidebilatéralenonliéede80,5%.Enincluantl’aidedesinstitutionsmultilatéralescechiffremonteà87,1%àraisonducaractèrenonliédesprêts.Enréalitéletauxdel’aidenonliéeseraitmoindrecommequelquespaysdéclarentqueleuraideestliécontrairementauxdonnéesdel’OCDEetunepartiedesinstitutionsfinancièresnonpartenairesdel’OCDEdonnentdesprêtsquisontpartiellementliés.Lebutdel’analyseéconométriqueestdedéterminersil’APD,lasituationduliementetlesdifférentsmodesdel’aide(donsetprêts)ontunimpactperceptiblesurlesfluxcumulésdesexportationsdesdonneursauniveaudupays,danscecasleBurkinaFaso.Lesrésultatssuggèrentquel’APDengénéraletlesdonsenparticulierpourraienteffectivementavoircertainseffetssurlecommerceparlesrelationsdonneur-bénéficiaire.
Types de l’aide et liementSelonlesstatistiquesdelaDGCOOPletypedel’aidepré-dominantestl’aideprojet.Cependant,cetypecomprendégalementl’aideprogrammequin’estpasfournisousformed’aidebudgétaireoucommefondscommun.L’aidebaséesuruneapprocheprogrammeauraitété57%en2007selonlerapport2008surlesuivideladéclarationdeParis.Danslapanopliedesinstruments,l’appuibudgétairegénéralprenduneplaceimportantavec29%desdéboursementsen2007etestappliquépar11bailleurs.Cetyped’aideestdépenséenutilisantlesprocéduresnationalespardéfinition.Lesbailleursayantacceptél’appuibudgétairegénéralsouventutilisentégalementaumoinsunepartiedesprocéduresnationalespourd’autrestypesd’aide.Lepourcentagedel’aideutilisantlesprocéduresnationalesdelapassationdesmarchésétait54%selonlerapport2008del’OCDEenreculde6%parrapportaurapport2006.Cechiffrecontrasteavecceluiconcernantl’aidebilatéralenonliéede92%selonlemêmerapport.Ceciimpliquequ’unepartiedel’aidenonliéeestutiliséeselonlesprocéduresdudonneur.NotonsquelesprêtsdesIFIsontsouventnonliés,maisquel’utilisationdesprocéduresdesbanquesestsouventobliga-toire.
Analyse des projetsQuatreprojetsontétéétudiés,quionttouspasséunnombreimportantdemarchéspourl’acquisitiondesbiensetdeser-vicesoudel’exécutiondetravaux.Troisbénéficientdel’aidenonliée,lequatrièmeconnaissantunecombinaisondel’aideliée,partiellementetnonliéeetnonliée.
1. LeProgrammed’AppuiauDéveloppementduSecteurEauetAssainissement(PADSEA)acommeobjectifd’assurer
l’accèsdurableàl’eaupotablepourenviron300.000person-nesenmilieururaletpourenviron100.000personnesenmilieusemi-urbaindanstroisrégionsduBurkinaFaso.LeprojetestuncofinancementduDanemarketdelaSuède.Lapartiedestravaux,ycomprisl’achatdesbiensetservicesyafférents,estgéréeparlesinstitutionsburkinabèselonlesprocéduresnationales.Lescomposantesconcernantdel’assistancetechniquesontgéréesauniveaudel’AmbassadeRoyaleduDanemarkselonlesprocéduresdelaDanidapourlapartiedanoiseetauniveaudel’ASDIpourlapartiesuédoise.Lesaidesconsistentendonsnonliés.Touslesmarchéspassésparlesservicestechniquesontétéadjugésauxentreprisesburkinabè.Enmatièredesmatériaux,lagrandepartieestimportée,maislamissionn’apasd’informationquantàlaprovenanceexacte.Entoutcas,lesdossierd’appeld’offresneposentpasdeconditionsàceniveauetlaissentlechoixauxsoumissionnaires.Larestrictionlaplusimportantestauniveaudespompesmanuellesquidoiventêtred’unedesmarquesutilisésdanslesrégions.Lesmarchéssontdetaillerestreinte,cequilimitelessoumissionséventuellesdel’extérieureàcausedefraisgénérauxélevésqu’engendraientlesentreprisesétrangères.
2. Leprojetd’ApprovisionnementenEauPotabledelavilledeOuagadougouàpartirdubarragedeZigaestungrandprogrammedeplusdeEUR230millionspourlapremièretranchedelapremièrephaseconsistantenlaconstructiond’unbarrage,d’unestationdepompageetdetraitementdel’eau,desadductionsjusqu’àOuagadougouetleréseaudedistributionetdesbranchementsàOuagadougou.Ycom-prisl’Etatetl’OfficeNationaldel’Eauetdel’Assainissement(ONEA)lenombretotaldesfinanciersétaitde15.Parmelesfinancements,unétaitlié,septétaientpartiellementliésetseptétaientnonliés.Pourdesraisonsd’harmonisationetdecoordinations,lesinstitutionsfinancièresarabesontfaitunedérogationdelaconditiond’éligibilité.Malgrél’étendudescontrats,lesfournisseurssesontlimitéàdesentreprisesétabliesenFrance,enAllemagneetauBurkinaFaso.Lesfraisd’approcheetdemobilisationdeschantiersontgonflélesoffresdesentreprisessansbasesolideauBurkinaFaso.Unedeuxièmeraisonpourraitavoirétél’origineallemandeetfran-cophonedesbureauxd’études,financéssurdesressourcesliées,ayantrésultédansdesdossierstechniquesfacilitantlasoumissiondesentreprisesdecespays.
3. LeProgrammed’AppuiaudéveloppementSanitaire(PADS)apourobjectifl’améliorationdel’étatdesantédespopulationsàtraverslacréationetlagestiond’unfondscommun.LePADSreçoitdesfinancementsd’unnombrededonneursbilatérauxetmultilatérauxetdesfondationsinternationales.TouslesfinancementssontdesdonsnonliéssauflefinancementdelaBanquemondiale(IDA)quiestenpartieunprêt.LesfondssontdépensésselonlesprocéduresnationalessauflesfondsIDAquisuiventlesprocéduresdelaBanque.En2008lePADSalancé39appelsd’offresrésult-anten60marchés.Surles39appelsd’offrestroisseulementontfaitl’objetd’unappeld’offresinternationalrésultantdans
Untying aid: Is it working? 79
Annex D
quatrecontratdontlesattributaireviennentdelaFrance,del’Allemagne,duSwazilandetduBurkinaFaso.Ils’agitdesmarchésfinancésparIDA.Lesprocéduresnationalesfixentdesseuilsaudessuslesmarchésdoiventpassentparappeld’offresouvert,maispasderèglesobligeantàdesappelsd’offresinternationaux.Laplupartdesfournisseurssontdesétablissementsburkinabè,dontcertainssontdesconcession-nairesdesentreprisesmultinationales.Quatremarchésd’unmontantimportantconcernantlalivraisondesmédicaments,vaccinsetcontraceptifs,passentdegréàgréavecl’UNICEFoulaCentraled’AchatdeMédicamentsEssentielsetGénériques.
4. LeprojetdAppuiauCentrenationaldeTransfusionSanguine(CNTS)estunprojetbilatéraldelacoopérationLuxembourgeoise.Sonobjectifestd’assurerunedisponibil-itécontinuedesproduitssanguinsdequalitéetenquantitésuffisantepourl’ensembleduterritoire.L’appuiestundonnonlié,quiestgéréensembleavecLuxDevelopmentselonlesprocéduresdecetteagence.Depuisjuillet2007onaenregistré35marchéspourunmontanttotaldepresqueEUR1million.Onzemarchéssontpassésparappeld’offresouvertreprésentant91%dumontant.Lagrandepartiedesfournis-seursviennentduBurkinaFasooudelaFrance,quoiquequelquesétablissementsburkinabèsoientdesconcession-nairesdesentreprisesmultinationales.Lesautresfournisseursviennentdel’AllemagneetduLuxembourg.Lagrandepartiedel’aideavaituncaractèrenonliéparori-gineougrâceàunedérogation.LebénéficiaireprincipaldudéliementparaîtleBurkinaFasodontlesfournisseurssesontaccaparésd’unegrandepartiedesmarchés.Cephénomèneseraitliéàunnombredecauses.Lepaysétantenclavélesfraisd’approcheetdemobilisationdechantierssontélevés.Enplus,lemarchénationalestrestreintcequilimitel’intérêtdumonded’affairesinternational.Lalanguepourraitégalementconstituerunobstacleàunmilieudeprédominanceanglo-phone.Lemorcellementdesacquisitionsendesmarchésdevolumelimitédiminuel’intérêtéventueldesfournisseursinternationauxquin’arriventpasàrécupérerleursfraisgéné-rauxsurdesmarchésrestreintsàunprixacceptable.Enfin,lapublicationdesappelsd’offresdansdesjournauxnationauxourégionauxauniveaudelacommunautéouest-africainefavoriseralesétablissementsayantdéjàunœilsurcemarché.
Efficacité en termes de coûts et de développementL’informationgénéraleindiqueunediminutiondescoûtsgrâceaudéliement.LeprojetZigaafourniunexempleconcretd’unmarchéquiestdevenumoinscherdûauchange-mentducaractèredeliementaucoursduprocessus.Cepen-dant,legainestannuléenpartieparlesprocéduresnationalesgénéralementconsidéréespluslongues.Aceniveauilfautnoterqueladécisiondudéliementestparfoiscoupléeàuneconditiond’unavisdenonobjectionducôtédubailleur,cequicontribueégalementàlaprolongationdeladuréedesprocédures.Uneautrerestrictionestquelemoindreprixestparfoisaudétrimentdelaqualité.Ledéliementpermetaupaysdemieuxdéfinirlesproduitsquiluisontutiles.Telsestlecasdepompesàeauàmotricité
humaineoupourlesbranchementsdel’eauenmilieuurbainoùunecertainestandardisationdeséquipementss’estavérénécessaireafind’enassurerunniveauminimumdemécani-ciensforméspourletyped’appareilsprécisetl’approvisionne-mentcontinudepiècesdétachées.L’attributiondesmarchésàdesentreprisesnationalesacertainementuneffetpositifsurledéveloppementécono-mique.Ceteffetseproduirasurtoutauniveaudel’assemblagedesproduitsimportés,delameilleureutilisationdescom-pétencesnationales,delagestionetdel’intermédiation.Ilyauraégalementunepartiedeprofitsquiparviennentaupays,quoiqu’unepartiedeceux–cidisparaissentdanslecasdesfilialesdescompagniesinternationales.Ledéliementestsouventaccompagnéparletransfertdegestionaupaysbénéficiaires.Danslaperceptiondesinterlo-cuteursnationaux,leliementconcerneraitplutôtlemodedegestionquel’originedeprojets.Effectivement,l’alignementsurlesprocéduresnationales,queveutlaDéclarationdeParis,estdifficilementenvisageablesansdéliementdel’aide.AuBurkinaFasoilyaunetendanceversl’appuibudgétaireetdel’approcheprogrammeavecdefinancementsousfondscommun.Cetyped’aideestincompatibleavecl’aideliée.Lapassationdeplusdebailleursavecuneplusgrandepartiedeleursfondsverscestypesd’aidefavoriseraégalementl’avan-céedansledéliementdel’APD.
Untying aid: Is it working?80
Annex D
d.3 south Africa Country study85
IntroductionThisstudyaimstohighlightkeyissuesthroughashortstatisti-calandeconometricreviewoftheavailabledataandthroughtheassessmentofthreeprojectcasestudies.Thefocusofthecasestudieswasontheprocurementpracticesofthedonors,governmentalexecutingagenciesandimplementingagents,thatmayhaveasignificantimpactonhowgoodsandservicesarepurchased,withimplicationsforaideffectiveness.Thethreeprojectsselectedascasestudieswere:• ahealthsectorprojectfundedbyDFID• ahealthsectorprojectfundedbytheEU• awatersectorprojectfundedbytheAssociationofNeth-
erlandsMunicipalities(VNG)International.
Statistical and econometric analysisODAinSouthAfricahasincreasedprogressivelyanddiversi-fiedsince1994-95,whenofficialaidwasfirstusedtosupportnationalpolicydevelopment.YettotalODAcommitmentsasapercentageofGDPremainverysmall,atlessthan0.4%in2007.ODAcommitmentshaveprimarilybeenintheformofgrants,andmostaidhasbeenfrombilateraldonors.TheUnitedStatesandtheEUareSouthAfrica’slargestdonorpartners,respectivelycontributing29.5%and20.6%oftotalbilateralODAtoSouthAfricabetween2005and2007.Asanon-LDC,SouthAfricawasnotoriginallycoveredbythe2001DACRecommendation,butODAcommitmentsreportedasuntiedhaveshownageneralupwardtrendsince2001(thoughthereremaininconsistenciesbetweendonorreport-ingofuntiedaidandstudiescommissionedbytheSouthAfricanGovernment).AlargeproportionofdonorODAtoSouthAfricaisintheformoftechnicalco-operation(TC),andthetyingstatusofmuchofthiskindofODAisnotreportedtotheDAC.Projectswheretherewasnoreportingonprojecttypeaccountedfor39%oftotalODAbetween2005and2007.InSouthAfricaalmostallGovernmentofficialsexpressedcon-cernabouttheuseofTCasameanstotieaidtohomecountryconsultants.TheaimoftheeconometricanalysisistodeterminewhetherODA,thetyingstatusandthedifferentformsinwhichaidisprovided(loansandgrants)haveanydiscernibleimpactonaggregatedonorexportflowsatthecountrylevel,inthiscasetoSouthAfrica.TheoverallresultssuggestthatODAingen-eral,andgrantsinparticular,mayhavesometradedistortingeffectthroughthedonor-recipientrelationship,thoughtheSouthAfricaanalysisisnotconclusive.
Aid modalities and tyingMostdonorsinSouthAfricastillmakeuseofthetraditionalprojectapproach.TheapparentlowusageofcountrysystemsissupportedbyanalysisintheSouthAfrican2006SurveyontheParisDeclaration:lessthanhalfofallODAmadeuseofSouthAfrica’sfinancialorprocurementsystems.Basedonin-
85 The study was undertaken by Yash Ramkolowan ([email protected]) and Matthew Stern ([email protected]) (Ramkolowan and Stern, 2009).
formationinQuistet al.(2008),just7%oftotaldonorfundsre-ceivedbySouthAfricainthe2007-2008financialyearwereintheformofbudgetsupport,andonlythreedonorshaveusedthebudgetsupportmodalitybetween2005and2008(theEU,NetherlandsandIreland).CountriessuchasJapan,theUnitedStates,theUKandGermany,thatreportahighpercentageoftotalODAasuntied(80%,45%,100%and72%respectivelyfor200686),makeverylowusageofSouthAfricanprocurementsystemsbasedonthe2006SurveyontheParisDeclaration(TableD.3).Inthesecases,thepossibilityofde factotyingismorelikely.Inthecountryreportonthe2008EvaluationoftheImplementationoftheParisDeclaration,somereasonsgivenbydonorsfornotusingSouthAfrica’ssystemsincludeconcernthattheuseofcountrysystemsdelaysthequickimplementationofprojectsandthatdonorHQshavedirectedlocalstaffnottouselocalsystems.
Table D.3 South Africa: donor use of country systems, 2005
Donor
Aid disbursed (US$ mn)
Public financial management
systems
Procurement systems
Australia 1 0% 20%
Belgium 6 75% 75%
Canada 0 100% 100%
Denmark 14 37% 75%
EC 139 42% 42%
Finland 9 85% 85%
France 1 2% 7%
Germany 21 33% 33%
Global Fund 10 67% 100%
Ireland 8 100% 100%
Italy 3 33% 0%
Japan 15 0% 0%
Netherlands 22 100% 100%
New Zealand 0 100% 100%
Norway 8 100% 100%
Sweden 7 42% 42%
Switzerland 0 88% 88%
UK 41 0% 29%
United Nations 2 81% 0%
USA 43 0% 4%
Total 351 38% 44%
Source: 2006 Survey on Monitoring the Paris Declaration: South Africa Country ChapterNotes: 1. Percentage figures reflect the baseline ratios for each donor.
86 Based on Clay et al. (2008).
Untying aid: Is it working? 81
Annex D
Project analysisThreedonorprojectswereusedascasestudiesfortheSouthAfricananalysis:twofromthehealthsector(bothrelatingtoHIV/AIDS)andonefromthewaterandsanitationsector.Allthreeprojectsweresimilarinthattheylargelyinvolvedtheprocurementofservices(especiallyrelatingtoconsultingservices)ratherthangoods,yetallthreeprojectshaddifferentcontracting,fundingandprocurementprocedures.ThisgavetheopportunitytocontrasttheprojectsandtoprovideamoreholisticviewofthewayuntiedaidworksinSouthAfrica.Thethreecasestudyprojectsaresummarisedasfollows:
1. TheRapidResponseHealthFund(RRHF)wasestab-lishedinlate2008asthefifthoutputoftheMulti-sectoralHIVandAIDSSupportProgramme(MSP),fundedbyDFID.TheRRHFwasinitiatedduringhighleveltalksbetweenSouthAfricanandUKGovernmentofficials.AtameetingbetweentheUKSecretaryofStateandthenewlyappointedSouthAfricanHealthMinister,itwasagreedthatahealthfundwouldbemadeavailabletotheMinister,inordertorapidlyimproveHIV/AIDSpolicyanddeliveryinSouthAfricaandtoassistwithanyunexpectedproblemsthatmightarise.GBP15mnhasbeencommittedtotheone-yearprogramme,whichrunsfrom2008to2009.TheheadcontractwasmanagedbyHLSP,thelo-calsubsidiaryofaBritishcompany,whichhadbeenawardedthemanagementoftheMSPunderaframeworkcontract.
2. TheExtendedPartnershipsfortheDeliveryofPrimaryHealthCareincludingHIVandAIDSProgramme(EPDPHCP)isanEUfundedprogrammethatbeganin2001.TheprogrammeaimstocontributetomoreaccessibleandaffordableprimaryhealthcareinSouthAfricabypartneringwithNotforProfitOrganisations(NPOs).Theprogrammehasevolvedovertheperiod,bothbyexpandingdeliveryofservicesfromfivetoallnineofSouthAfrica’sprovincesandbymovingfromEUtolocalsystemsandprocedures.Fortheperiod2001to2006,theEUcommittedEUR25mn.Intheperiod2007to2010theEUhascommittedafurtherEUR45mn.ThecontractingagencyfortheEPDPHCPistheSouthAfricanGovernment,throughtheNationalDepartmentofHealth.
3. TheStorm-waterDrainageandSanitationprojectwasimplementedin2005throughthemunicipaltwinningarrangementbetweenBuffaloCityMunicipality(SouthAfrica)andtheCityofLeiden(Netherlands),andwasfundedthroughtheCityofLeidenbyVNGInternational.VNGInter-nationalisfundedby,amongothers,theNetherlandsMin-istriesofForeignAffairsandHomeAffairs,theEU,theWorldBankandvariousUnitedNationsagencies.Fortheyear2009-10thebudgetapplicationwasEUR566,815,ofwhichEUR297,000wasprovidedbyLeidenintheformoftechnicalassistance.VNGInternationalrequiresthatthecontractingagencyisaDutchorganisation;assuchLeidenisseenasthecontractingagency.However,BuffaloCityislargelyrespon-sibleforthemanagementandexecutionofthestorm-waterproject.Thesethreeprojectsraisetwoverydifferentsetsofconcerns:
• Firstly,theuseofframeworkcontractscaneffectivelyexcludesmallerandlessexperiencedSouthAfricanfirmsfromparticipatinginlucrativeheadcontracts.Forexam-ple,theRRHFcontractwasawardedtoaUK-ownedcom-pany,whichalreadyheldaframeworkcontractwithDFID.Thismayhavebeennecessaryforexpediencyreasons,butitdideffectivelyexcludeSouthAfricanhealthcompaniesfromtenderingforthisspecificcontract.
• Secondly,thereissomeevidencethatthehighleveloftechnicalassistanceandco-operationagreementsinSouthAfricamakesODAlesstransparentandincreasesthepotentialfortying.Forexample,inthecaseoftheBuffaloCitystorm-waterproject,asignificantamountofbudgetassistancewasintheformoftechnicalassistancefromLeiden,andfortravelandaccommodationofbothLeidenandBuffaloCityofficials.BuffaloCitywasalsorequiredtocounter-fundasubstantialamountfortheprocurementofconsultingservicesrelatingtotheproject,withthiscounter-fundingamountbelievedtobesubstan-tiallymorethanthefinancialcontributionbyVNG.
Cost and developmental effectivenessIntheactualprocurementofgoodsandservicesforallthreeprojects,thereislittleevidenceofimplicitorde factotyingbydonors,withallthreeprojectsmakingextensiveuseoflocalsuppliersforbothservicesandworkscontracts.Theextenttowhichthishascontributedtoneworsustainabledomesticcapacityislessclear.FortheRRHF,theuseoflocalconsultantsandcompaniescer-tainlyraisesthepotentialfordomesticcapacitydevelopment,butmanyoftheconsultantscontractedwerealreadyhighly-sought,withspecialisedskills.TheEPDPHCPhasamuchclearerdevelopmentimpactandhasexplicitlycontributedtotheestablishmentoftechnicalandadministrativecapacitywithinthenationalandprovincialhealthdepartments.TheBuffaloCitystorm-waterprojecthasusedaSouthAfricancon-sortiumforthemajorconsultingworkandlocal(basedinthemunicipality)workscompaniesfortheminimalconstructionactivities.Thiswouldhavehadpositivespin-offswithinthelocalcommunity.However,fromabudgetperspective,techni-calassistanceprovidedbyLeiden,travelandaccommoda-tionaccountforoverhalfthetotalinternationalfunding.ThefactthatBuffaloCityhascounter-fundedoverhalftheactualamountspentontheprocurementofworksandservices(out-sideoftravelandaccommodation)reducesthepotentialben-efitofaidprovided.Inaddition,itisnotclearthatthetwinningarrangement(i.e.informationandknowledgeexchange)hasbeenusedefficiently,soVNGInternational’sactualcontribu-tiontolong-lastingcapacityandinfrastructuredevelopmentmaybesmall.Cost-effectivenessisoftencitedasakeyreasonfortheunty-ingofaid.Assessingthecost-effectivenessofconsultingserv-icesisverydifficult,giventhevaryingdegreesofqualityandspecialisation.Professionalfeesaredifficulttocompareandareusuallytreatedascommerciallysensitive.Thisstudyhasrevealedfourissuesthatdonorsneedtoconsiderinanuntied
Untying aid: Is it working?82
Annex D
environmentingeneraland,specifically,whenlocalsystemsandproceduresareused:
• localrulesandguidelinesforconsultingratesmaydifferfromtheinternationalnorm
• preferentialprocurementpoliciesmayraisethecostofcertaingoodsandservices
• theadministrativeburdenoflocalprocurementmaybehigherormoretimeconsuming
• alackofcompetitioninsomedomesticmarketsmaycontributetopremiumratesoncertainskills.
ConclusionSouthAfricaisexceptionalasarecipientofaid,largelybe-cause,ingeneral,itisnotverydependentonaidasarevenuestream,andalsobecauseithasthenecessaryinstitutionsandcapacitytoco-ordinatedonoractivitiesandmanagedonorfunds.Thisisconfirmedbytheeconometricresults:donorflowsimposenosignificantdistortionsontheSouthAfricaneconomy.Thepresumptionisthatthisshould,inpractice,con-tributetothemoreeffectiveandefficientuseofODA.Thecasestudiesrevealthreereasonswhythisisnotalwaysthecase:
• Firstly,intwoofthethreecases,thereremainssomeevi-denceofde factotying,withsourcecountryconsultantsorcompaniesbroughtintodoworkthatcouldpotentiallybedonedomestically.
• Secondly,althoughalargeproportionofreportedODAisuntied,theriskofnon-reportingmayincreaseinmorede-velopedcountries.Thisislargelybecause,incountrieslikeSouthAfrica,areasonablylargeproportionofoverallODAisprovidedasad hoc orshort-termtechnicalco-operation.Thisformofassistanceisgenerallylesstransparentthanmoreconventionallong-termdevelopmentprojects.
• Thirdly,inanuntiedenvironment,donorsaremorelikelytomakeuseofrecipientcountryprocedures.Thecasestudiesshowthattheuseofrecipientcountryprocure-mentsystemsenablesGovernmentstopursuedomesticpolicyprioritiesandimprovealignment.But,insomein-stances,thiscanraisethecostanddiminishtheeffective-nessofspecificdonorprogrammes.
InthecaseofSouthAfrica,itwouldappearthatuntyinghascontributedtoimprovementsinthealignmentanddevelop-menteffectivenessofaid.Moreover,althoughdonorsareminorfundersinSouthAfrica,theycanplayacriticalroleindeployingtechnicalandfinancialresourcesquickly.Theexam-plesexploredinthisreporthighlightthesignificantcontribu-tionmadebyjustthreeprojects:rescuingvariousnationalHIV/AIDSprogrammes;providinghigh-leveladvicetoanewCabinetMinister;andprotectinganimpoverishedcommu-nityfromsevereflooding.Couldmorehavebeenachievedif
thesedonorshadadoptedamoreopenprocurementprocess,ormadegreateruseoftheSouthAfricansystem?Certainly,alignmentanddevelopmenteffectivenesscouldbefurtherenhanced,butthesegainsmightbepartiallyoffsetbytheintended(policy-related)andunintended(inefficiencies)costsofalternativeprocurementprocedures.
Untying aid: Is it working? 83
Annex D
d.4 Zambia Country study87
IntroductionThisstudyaimstohighlightkeyissuesthroughashortstatisti-calreviewoftheavailabledataandthroughtheassessmentofalimitednumberofprojectcasestudies.Thefocusofthecasestudieswasontheprocurementpracticesofthedonors,governmentalexecutingagenciesandimplementingagents,thatmayhaveasignificantimpacthowgoodsandservicesarepurchased,withimplicationsforaideffectiveness.
Thefourprojectsselectedascasestudieswere:
• theWaterSectorSupportinNorthernProvince(WSSNP)supportedbyIrishAid
• theSouthernWaterandSanitationProject(SWSP)fundedbyGermany
• theWaterSectorPerformanceImprovementProgramme(WSPIP)financedbytheWorldBank
• thedistrictfocusedagriculturaldevelopmentProduc-tion,FinanceandTechnologyProject(PROFIT)fundedbyUSAID.
Statistical and econometric analysisZambiahasreceivedsignificantlevelsofexternalassist-anceandcurrentlyranksamongthetop10AfricancountryrecipientsofODA.Aidflows,asashareofoverallpublicex-penditure,peakedat53%in2001(i.e.theyearwhentheDACRecommendationonaiduntyingwaspassed).Sincethen,thecontributionofODAtopublicexpenditurehasdeclinedsteadily,toaround24%inthe2008budget.Inabsoluteterms,totalODAcommitmentshalvedbetween2005and2007.ThefallinthecontributionofaidflowstoZambiacanlargelybeexplainedbyimprovedexternalconditions,arisingfrombothdebtreliefandrisingcopperprices.
AccordingtoOECD/DACstatisticsfrom2005to2007,bilateralaidaccountedforjustover90%oftotalODAcommitments.AlmostallofthisbilateralODAtakestheformofgrants,withnonewloancommitmentsmadein2007.Thetopbilateraldonors,intermsoftheircommitmentstoZambia,areJapan,theUSA,Germany,theECandtheUK(inthatorder).Together,thesefivebilateraldonorsaccountedfor70%ofthetotalbilateralODAtoZambiaoverthisperiod.
ThevastmajorityofbilateralODAtoZambiaisreportedasuntied(86%),witharound4%ofbilateralODAreportedastiedand9%aspartiallytied.Buttherearelargediscrepanciesinapproachbetweenthedifferentdonors.WhereastheUK,Swe-den,Austria,IrelandandLuxembourgreportedalloftheiraidasuntied,Switzerlandreportedthat46.1%ofitsODAistied.Amongothermajorbilateraldonors,Japanreported94.2%of
87 The study was undertaken by by Oliver Saasa ([email protected]) and Chrispin Mphuka ([email protected]) (Saasa and Mphuka, 2009).
itsaidtoZambiaasuntied;theUSA81.8%,Germany94.5%;andtheECjust17.7%untied.Thelargestproportion(82.3%)ofECaidisreportedaspartiallyuntied.
Aid modalities and tyingDonorscurrentlyadoptfourdifferentmechanismsinoperat-inginZambia:ProjectSupport,IntegratedProjectSupport’,Sector-WideApproach(SWAp)andGeneralBudgetSupport(GBS).
Thereisnoofficialdataontheuseofthesedifferentmecha-nismsbythemajordonors,butasurveyundertakenaspartofthiscountrystudyprovidessomeinformationontheprefer-encesofthetenparticipatingdonorsbasedon2007disburse-ments(TableD.4).Sevenofthetendonorsusegeneralbudgetsupport(GBS)and/ornon-earmarkedsectorbudgetsupport(SBS)todisburseODAtoZambia,withtheEUdisbursingtwo-thirdsofitsaidthroughthischannel.Amongsttherespond-ents,Canada,JapanandtheUSAdisbursenoODAthroughthischannel.EarmarkedbudgetsupportisnotwidelyusedinZambia.
TheWorldBank,Germany,Norway,Sweden,Japan,CanadaandtheUKgenerallyadoptaSector-WideApproachandusetheZambianprocurementsystemtofundnon-pooledprojects.Morethan70%ofWorldBankdisbursementsand44%ofUKaidischannelledthroughthismodality.AlmostallUSAODAgoestonon-pooledprojects,butthesedisburse-mentsfollowUSrulesandguidelines.Anumberofsmalldo-nors,includingCanada,FinlandandSweden,domakeuseofpooledprojects,withCanadadistributing83%ofitstotalODAthroughthischannel.Japanadoptsaverydifferentapproach,withmorethan97%ofallODArestrictedtospecificprojectsandJapaneseprocurement.TheEU,Germany,SwedenandtheUKalsousetheirownprocurementsystemstoprovideprojectsupport,buttheproportionofoverallODAdisbursedthroughthismodalityismuchlower,at28%,18%,17%and13%respectively.
Untying aid: Is it working?84
Annex D
Anumberofimplicationscanbedrawnfromthissurvey.Firstly,theresultsconfirmthedatareportedtotheOECD/DAC.MostdonorsusetheGBSand/ornon-earmarkedsectorbudgetsupportmodalityandtheZambianprocurementsys-tem.Suchsupportisde jureuntied.Secondly,thesurveyraisessomequestionsaboutthereportingofsomespecificdonors.Forexample,althoughtheUKreportsallofitsODAtoZambiaasuntied,alargeproportionofthisfundingisdisbursedthroughhostcountryandhead-officeprocurementsystems.Thisrequiresfurtheranalysistoexplorewhether,infact,thisODAistrulyuntied.Finally,thesurveyhighlightsanumberofcountriesthatappeartoavoidtheuseofZambianprocure-mentsystems.Again,thiswouldappeartoraisetheriskofde factotying.Twoofthesecountries(JapanandtheUSA)areamongthelargestdonors.
Project analysisFourprojectswereselectedascasestudies:threewaterandsanitationprojectsfundedbyIreland,GermanyandtheWorldBank;andaUSAID-fundedprivatesectorcompetitivenessproject.
1. WSSNPhasitsoriginsintheIrishGovernment’ssupporttowatersupplyandsanitationactivitiesintheNorthernProv-inceofZambiasince1983.Initially,theprogrammewasimple-menteddirectlybyDevelopmentCooperationIreland’s(DCI)ownstaff.Afterareviewinthemid1990s,theProgrammebe-
ganitsfullintegrationintotheZambianGovernment’sWater,SanitationandHygieneEducation(WASHE)systemin1998,thoughfullintegrationwasonlyachievedin2002.DCIstaffhavereducedtheirdirectparticipationinProvincialWASHE(P-WASHE)meetingsandcurrentlyattendonlyattheinvita-tionoftheChairoftheCommittee.TheProgrammecurrentlyprovidesbothfinancialandtechnicalsupporttofourdistrictsintheNorthernProvince:Mbala,Isoka,LuwinguandMpika.
2. SWSPwasinitiatedin1998throughthesigningofafinancialagreementworthDM27mnbetweentheZambianandGermanGovernments.Theagreementcovered8townsintheSouthernProvince,excludingLivingstone.Laterin2000,anadditionalagreementforLivingstoneworthDM13mnwassigned.Theobjectiveoftheprojectwastoprovideacontinu-ousandsufficientsupplyofhygienicallysafedrinkingwaterforthemajorityofthepopulationandindustryintheninetownsoftheSouthernprovince,throughtherepairofexistingwatersupplysystemsandtheexpansionofnewnetworks.TheProjectalsoprovidedsupporttotheprovisionofadequatesewageandhumanwastedisposalservicesinthosetownsthathavesewagesystems.SWSPwasfundedthroughGermanFinancialCoopera-tion(KfW)andrequiredasmallcounterpartfundingfromtheGovernmentoftheRepublicofZambia.AlthoughtheProjectwasoriginallyscheduledtocommenceinJuly2000andendbySeptember2003,anumberofdelaysresultedinitsextensiontoFebruary2009.
Table D.4: Zambia: disbursements of ODA by selected donors in 2007/8
Donor
Total Disburse-ments
GBS & Unear-marked SBS
Earmarked SBS
Non-Pooled Projects Pooled Funding
Project Funding involving agency’s own procurement
using zambian System
using non-state pro-cedures other than agency’s
at local office
through HQ
US$ million (%) (%) (%) (%) (%) (%) (%)
WB 54.4 18.4 0.0 71.3 0.0 10.1 0.0 0.0
Canada 15.7 0.0 0.0 1.5 0.3 82.7 5.9 9.6
Japan 34.4 0.0 0.0 2.6 0.0 0.0 97.4 0.0
EU 58.0 66.7 0.0 0.0 0.0 5.7 27.6 0.0
Germany 29.4 23.1 0.0 44.2 3.4 11.2 18.4 0.0
USA 212.9 0.0 0.0 0.0 99.9 0.1 0.0 0.0
Finland 17.7 39.0 0.0 0.0 14.7 43.5 0.0 0.0
UK 72.9 60.6 10.8 0.5 0.0 13.2 1.9 12.9
Sweden 52.2 27.2 0.0 8.2 4.6 42.7 17.2 0.0
Norway 72.7 24.9 22.7 10.3 36.7 5.4 0.0 0.0
Source: Data collected directly from donors through questionnaire administered by Premier Consult Limited in May 2009.
Untying aid: Is it working? 85
Annex D
3. WSPIPwasestablishedin2007underaconcessionalloanagreementfromtheWorldBankforUS$23mn.Itwasinitiallyplannedtobecompletedwithinaperiodofthreeyears,buthassincebeenextendedto2012,withanadditionalinjectionofUS$10mn.TheobjectiveoftheProjectistosup-porttheGovernment’son-goingcommitmenttourbanandruralwatersectorreforms,andincreasedaccesstoservices,throughtheexpansionofexistingfacilities.ItincludessupporttotheGovernmentanditsutilitiestoimplementacompre-hensiveinstitutionalstructurewhichwillleadtoabetter,moresustainedandco-ordinatedapproachtothepublicandprivateprovisionofwaterandsanitationservices.
4. PROFITisafive-yearUSAIDprojectthatbeganinJune2005.TheprojectisimplementedbytheCo-operativeLeagueofUSA(CLUSA)inpartnershipwiththeglobaldevelopmentconsultingcompany,theEmergingMarketGroup(EMG).FundingoftheProjectwasobligatedinthebilateralStrate-gicAgreementforEconomicGrowthbetweentheZambianGovernmentandtheGovernmentoftheUnitedStates,andthetotalfinanceforPROFITisUS$17.5mn.PROFITseekstoimprovethecompetitivenessofselectedindustriesinwhichlargenumbersofmediumandsmallenterprisesparticipate.Thiswillenablefirmsandindustriestorespondtomarketdemands,bothintheshortandlongterm,andtoraisethebenefitsthatthesefirmsprovidetomarketparticipants.
Therulesandregulationsforprocurementdifferacrossthefourdonors/projects.InthecaseoftheIrishAID-supportedWSSNP,theProjectiscompletelyuntied,useslocalprocure-mentrules,regulationsandsystems,andhasnorestrictiononprocurementtoaparticulargeographicregion.Theotherthreeprojectsusedonorrulesandguidelinesforprocure-ment.Althoughtheserulesandregulationsdonotusuallyrestrictsourcingofgoodsandservicestothedonorcountry,itwasfoundthattherearecertainspecificationsthatmayre-quirethesourcingofbothequipmentandconsultantsfromthedonorcountry.Furthermore,mostZambianlocalfirmsdonothavethefinancialortechnicalcapacitytomeetthetenderrequirementsoflargedonorprojects.Asaresult,intheGerman-fundedSWSP,about85%oftheprojectmoneywenttoGermanfirms,whereasaround29%ofthetotalfundingfromtheUSAID-fundedPROFITwenttoUSorganisa-tions.
Cost and developmental effectivenessAllfourprojectsexamineddemonstratedtheirdevelopmentalrelevancetotheextentthattheyarealignedtotheFifthNa-tionalDevelopment(FNDP)priorities.Partlyasaresultofunty-ing,donorsareseentobeworkingtowardsharmonisingtheirprogrammesthroughtheJointAssistanceStrategyforZambia(JASZ).Inaddition,moreofthedonorscurrentlyuseGBSandSBSasaidapproaches(ratherthanprojectspecificfunding).Delegatedco-operationisalsobeingemployedinZambia.
Duetothelimitednumberofprojectsevaluatedinthisstudy,ithasnotbeenpossibletoassessfullywhethertheuntyingof
aidleadstoimproveddevelopmentaleffectiveness.Neverthe-less,theevidenceprovidedbythesefourprojectssuggeststhat,ingeneral,theextensiveuseoflocalsub-contractors,whichismadeeasierinanuntiedenvironment,hascon-tributedtothecreationofemploymentopportunitiesandatransferofskills.ThisiscertainlytrueoftheWSSNPproject,inwhichallworksarenowhandledbyZambianfirms.Inothercases,however,suchastheSWSPproject,therewasnosub-contractinginvolved(exceptforonejoint-venturebetweenaZambianandaGermancompany)andthedevelopmenteffectivenessis,therefore,likelytobeconsiderablylower.Thesizeandtechnicalrequirementsoftheprojectclearlyhaveastrongbearingontheextenttowhichgoodsandservicescanbeprovidedbylocalfirms.
Anotherspin-offbenefitofuntying,intheZambiancontext,isthatithasprobablycontributedtothestrengtheningoflocalprocurementsystems.Forexample,theIrish-fundedWSSNPisimplementedatthedistrictlevel,andtheprojecthashadtoinvestinimprovingplanning,procurement,accounting,andmonitoringandevaluationwithcouncils,toenablethemtocontractlocalsuppliers.Similarly,inthecaseoftheGerman-fundedSWSP,theuseoflocalprocurementsystemshelpedtoinstilasenseoflocalownership.
Measuringthecosteffectivenessofuntiedaidproveddifficult,andwaslimitedtoasetofitemsinonecasestudy.Inthisexample,thepriceofthegoodsprocuredusinguntiedaidwaslowerbyatleast6%,relativetotheimportparityprice.Therewasnootherevidencetosuggestthatuntyinghadraisedcosts,inanyofthestudies,thoughthismayrequiremorecomprehensiveanalysis.Suchanalysiswouldneedtolookbeyondthemonetarycostofuntyingandalsoconsidertheexpectedtimebenefitsofprocuringlocally.
ConclusionZambiaisamajorbeneficiaryofODAandaidflowsconstituteasubstantialshareofthecountry’sbudgetandGDP.Improve-mentsinthedevelopmentandcosteffectivenessofODAare,therefore,likelytohaveasignificantimpactonindividualbeneficiariesandthecountryasawhole.ThefourcasestudiesconfirmthesignificantcontributionmadebythesespecificprojectsinZambiaandshowhowuntyingingeneral,andtheuseoflocalprocurementsystemsandsub-contractorsinparticular,cancontributetoemploymentcreationandskillstransfers.
Intheactualprocurementofgoodsandservices,thereissomeevidenceofimplicitorde factotyingbydonors.InthecaseofPROFIT,nolocalfirmswereinvolvedintheinitialtender,andparticipationbyotherinternationalorganisationswaspoor.IntheawardofSWSP,theGovernmentindicatedthatitwasaskedtosuggestfirmsforconsiderationandthatGermanfirmsshouldbeamongstthoserecommended.Thisprojecthasmadenouseoflocalsub-contractors(exceptinonein-stance,asajointventurewithaGermancompany).
Untying aid: Is it working?86
Annex D
Therearetwomainreasonsfortheuseofhostcountrycompa-nies,consultantsandmaterialsinZambia:
1. Althoughrulesandregulationsdonotrestrictthesourc-ingofgoodsandservices,therearecertainspecificationsthatmayrequiresourcingfromthedonorcountry.Forexample,inSWSP,replacementpartshadtobepurchasedfromGermanybecausethisiswheretheexistingwaterpumpscamefrom.USAIDimposesrestrictionsontheprocurementofspecificgoodsandservices,suchasmo-torvehiclesandairtravel(althoughthesearenotalwaysenforced).Inprojectsthatrequireaconsultantoranimplementingagent,theprimecontractororconsultantusuallycomesfromthedonorcountry.
2. MostZambianfirmsdonothavethefinancialcapacityortechnicalexperiencetotenderforlargedonorprojects,re-gardlessofwhetherornottheprojectistied.Forexample,inthecaseofWSSIP,postqualificationrulesrequiredthatthebiddingcompanyshouldhavedeliveredatleasttwocontractsofatleastthesamevalueinthelastthreeyears,thathavebeensuccessfullyandsubstantiallycompletedandthataresimilartotheproposedworks.ThiscriterioneffectivelydisqualifiedalmostallZambian-basedfirms.
Inadditiontotheseexamples,theincreasinguseofverticalfundingarrangementssuchasthePEPFARandtheGlobalFund,whicharebetterresourcedthatmostotherprogrammesinZambiaandremainstrictlytied,maydiminishthegainsfromuntying.Finally,interviewswithvariousdonorsandthestatisticalanalysisrevealedthatuntyinghasfacilitatedtheincreaseduseofGBSandSWAPsinZambia.Thisimprovesthetransparencyandalignmentofaid,butalsoraisesnewchallenges.ForZam-bia,procurementcapacityandsystemsneedtobeimprovedtomanageadditionalfundswellandprovidedonorswithsufficientconfidence.Forthedonors,thereisneedtoworktowardsintegratingtheirsupportedprojectsintothenationalsystem,andtoreducethetransactioncostsemanatingfromtheparallelstructuresthatoftenthreatenthesustainabilityofproject/programmeoutcomes.
Untying aid: Is it working? 87
Annex D
d.5 lao pdr Country study88
IntroductionInthe2008PhaseOnestudyTheDevelopmentalEffectivenessofUntiedAid,itwasfoundthataidtoLDCsisnowoverwhelm-inglyuntied(excludingexemptcategories,technicalco-operationandfoodaid).However,itwasalsofoundthatthereexistedverylittleformalanalysisofboththedonorpracticesinpartnercountriesandtheactualconsequencesofuntyingaidatacountrylevel.ThePhaseTwostudyisinresponsetotheproposalsoftheWorkingPartyonAidEffectivenessandtheDevelopmentAssistanceCommittee(DAC)NetworkonDevelopmentEvaluationandthereportingrequirementinthe2001DACRecommendationonuntyingODAtoLDCs.
PhaseTwooftheprojectisintendedtoprovideevidence-basedconclusionsabouttheimplicationsofuntyingaidanditsimpactonaideffectivenessinselectedcase-studycoun-tries,forpresentationtotheDACbyDecember2009.ThisreportprovidestheresultsoftheLaoPDRCountryStudy.
ThisstudyaimstohighlightthekeyissueoftheimplicationsofuntyingaidforaideffectivenessinthecontextofLaoPDR,throughashortstatisticalandeconometricreviewoftheavail-abledataandthroughananalysisoffourprojectcasestudies.Thefocusofthecasestudieswasplacedonprocurementpracticesofthedonors,governmentalexecutingagenciesandimplementingagents,thatmayhaveasignificantimpactonhowgoodsandservicesarepurchased,withimplicationsforaideffectiveness.
Thefourprojectsselectedascasestudieswere:• SIDAfundedroadproject• EUfundedruraldevelopmentprojectwithawatersector
component• AFDpart-fundedhydroelectricpowerproject• JICAfundedwatersupplyproject.
Statistical and econometric analysisBilateralODAtoLaoPDRoriginatedintheearly1960s,andmultilateralODAwasfirstrecordedinthelate1970s.Sincethen,totalODAhasshownastrongincreasingtrenduptothepresent,althoughtherehasbeensomefluctuationyearonyear,withbilateraldonorstraditionallycontributingthelargerproportionofODA.TotalODAin2007stoodatUS$280mn,whichisapproximately7%ofGDP–alargepercentagerela-tivetootheraidrecipients.Japan,France,andSwedenareLaoPDR’slargestbilateraldonors,contributingrespectively40.1%,16.8%and11.8%oftotalbilateralaidbetween2005and2007.TheaimoftheeconometricanalysisistodeterminewhetherODA,thetyingstatusandtheinstrumentsbywhichaidisprovided(loansandgrants)haveanysignificantimpactonaggregatedonorexportflowstotherecipient,inthiscasetoLaoPDR.Overall,theresultsshowthataggregateODA,andgrantsinparticular,havesignificanttradedistortingeffects
88 The study was undertaken by Adam McCarty and Alex Julian (2009) ([email protected]).
throughtheincreaseindonor-recipientexports.Thisempiri-calevidencesuggeststhataidflowscouldbeinformallyorde factotied,whenanalysisofdatafromacrosssectionofdonorstoLaoPDRisperformed.However,asoutlinedintheecono-metricsinvestigationsthisanalysishassomecaveats.
Aid modalities and tyingLaoPDRisclassifiedasanLDCandis,therefore,coveredbythe2001DACRecommendationtountieaid.Between2005and2007DACdonorcountriesformallyuntiedover70%oftheirODAcommitmentstoLaoPDR(CRS).Meanwhile,forthesameperiod,22%ofbilateralODAremainedunreportedwithregardtotyingstatus.Theprocessofuntyingaidhasimprovedoverthelastdecade,despitethehesitationofsomedonorstomovetonon-projectbasedaidmodalities,suchasbudgetsupportandpooledfunding,whichareoftenassociatedwithuntying.LaoPDRreceiveslargelygrantaidasaninstrument(98%ofDACODAin2007)andprojectbasedaidmodalities.Severaldonorsalsoprovideprojectbasedorfree-standingtechnicalco-operation,inmanycasesalongsidegrantfundingforprojects,whichisoftentied.Forexample,grantswhichinvolvednofree-standingtechnicalcooperation(FTC)reporteda96.8%untiedshare,whereasgrantswhichhadawholeFTCcomponentreportedanuntiedshareofjust27.8%(CRS,2005-07).ThethreelargestdonorstoLaoPDR,Japan,FranceandSweden,allreportedahighproportionoftheirODAasuntiedfor2007:68.9%,62.3%and98.5%respec-tively.LaoPDRalsoreceivessubstantialaidfromnon-DACdonors,thatisconsideredtobelargelytiedandnotnecessar-ilyconformingtoOECDdefinitionsofODA(TableD.5).
Untying aid: Is it working?88
Annex D
Project analysisFourprojectswereselectedfromthetopactivebilateraldonorstoLaoPDR,tobeusedascasestudiesfortheprojectanalysis.Althoughwaterandsanitationwasthetargetsec-tor,exceptionshadtobemadeinthiscountrystudyduetolowactivityinthatsector.Anin-depthreviewofthefourinfrastructureprojectswasundertaken,torevealmethodsofaiddeliveryanddonorpracticeswithrespecttountying.Theapproachusedinvolvedtrackingdonorfundsforeachproject,fromcommitmentbythedonortodisbursementbytheleadimplementingagencydowntogroundlevel.Fromthisaccount,procurementandfundingprocedureswerecloselyinvestigatedtodeterminetheexactoriginofgoodsandservicesemployedfortheproject,andwhetheranysourc-
ingrestrictionsonthebasisofnationalityoccurred.Thecasestudiesalsoattemptedtoassesstheimpactofthesourcingofcontractsonthecosteffectivenessanddevelopmentalef-fectivenessoftheoriginalaiddisbursement.Allfourprojectswereimplementedusingdifferentfundingchannels,practicesandimplementationstrategies,allowingasuitablerangeoffactorstobeinvestigated.Thefourcasestudyprojectsaresummarisedasfollows:
1. TheLao-SwedishRoadSectorProjectPhaseIII(LSRSP3)isaSidafundedproject,designedtoprovidebasicroadaccesstoremotevillagecommunitiesin26ofthepoorestdistrictsinLaoPDR.ThisprojectphasecommencedinOctober2005,af-terthegovernment,throughtheMinistryofPublicWorksandTransport(MPWT),requestedcontinuedsupportfromSidafortheroadsector.ThetotalfundingforthefouryearprojectperiodisaUS$20mngrantfromSida(70%)andaUS$9mncontributionfromGovernmentofLaoPDR(30%).TheMPWTisresponsibleforthemanagementandprocurementontheproject,whereprocurementwasconductedusingcountrysystems.
2. TheIntegratedCommunityBasedRuralDevelopmentProjectinNamorandXayDistricts,OudomxayProvinceisathreeyearECprojectwhichstartedinJanuary2007,withanoverallobjectiveofmakingapositiveimpactontheimprove-mentofthefoodsecurityandsocio-economicsituationofthetargeted20communities.TheprojectwasimplementedbyaGermanInternationalNGO,whichalsocontributed25%ofthetotalEUR1,098,678grantfunds(75%fromEC).TheINGOwasselectedbyinternationalcompetitiveselectionbytheEC,andallprojectprocurementrequiredECrules.
3. TheNamTheun2HydropowerProjectisoneofthelargesthydropowerprojectsinLaoPDR,withatotalvalueofUS$1.5billion,whichbegantheconstructionphaseinJune2005andhasanexpectedcompletiondateofDecember2009.Thefundingarrangementsoftheprojectinvolvebothprivateandpublicsources,whichincludepartloanandgrantfundingbyAFD,WorldBankandADB.
4. TheProjectforVientianeWaterSupplyDevelopmentisanurbanwatersupplyproject,undertakenintheLaoPDRcapitalfromFebruary2006toMarch2009,withthemainob-jectivestoattainastablewatersupplyandimprovethewatersupplyserviceratioinVientianeCapital.JICAfundedUS$30.2mnfortheproject,intheformofgrantaid.TheexecutingagencywastheMPWT,andtheprojectutilisedthedonor’sprocurementguidelines.
Investigationofthesefourprojectcasestudiesproducedthefollowingkeydescriptiveresults:
• Overall,contractsintheseprojectswerefoundtobecom-petitivelysourcedthroughstandardbiddingprocedures,withthemajorityof(butnotall)contractsforconstruc-tionwork,groundlevelworkersandmaterialsusedfor
Table D.5 Lao PDR: bilateral ODA disbursements by source, 2005-061
Donor
CRS data Local data
US$ mn % US$ mn %
DAC bilateral donors (including EC)
Australia 12.8 6.8 17.2 7.7
Belgium 2.7 1.4 5.3 2.4
Canada 1.9 1.0
Denmark 3.4 1.8
EC 8.8 4.7
Finland 2.0 1.1 1.7 0.8
France 15.9 8.5 17.7 7.9
Germany 18.3 9.8 6.3 2.8
Ireland 2.4 1.3
Japan 66.2 35.2 62.2 27.8
Luxembourg 6.8 3.6 6.3 2.8
Netherlands 0.1 0.0
New Zealand
2.5 1.3 1.8 0.9
Norway 11.4 6.1 14.7 6.6
Sweden 23.7 12.6 19.8 8.9
Switzerland 4.5 2.4
UK 0.3 0.1
USA 4.3 2.3
Non - DAC bilateral donors
Vietnam 24.9 11.1
China 21.2 9.5
Thailand 18.8 8.4
Korea 5.9 2.6
Total 187.8 100.0 223.5 100.0
Source: OECD CRS database and Lao PDR Ministry of Planning and InvestmentNotes: 1. Aid from non-DAC donors does not necessarily satisfy DAC criteria to qualify as ODA.
Untying aid: Is it working? 89
Annex D
constructionoriginatingfromLaoPDR–thelowestcostandeasiestavailablesource.
• Someotherprojectcomponentsremainedsourcedfromthedonorcountry/regioninpractice,evenwhensourcedcompetitively,suchastechnicalassistance,projectcon-sultants,largeheadcontracts,andprojectimplementingINGOs.
Cost and developmental effectivenessTheimplicationsofdonorpracticesonthefourcasestudyprojectsforcost,aidanddevelopmentaleffectivenesscanbesummarisedasfollows:
• Byreviewingandground-truthinguntiedprojects,infor-mationandpricedataobtained(whereavailable)showedthatcostsavingsweremadeonthepurchaseofequip-mentitemsandconstructionmaterialsrelativetomarketreferenceprices.Thisimpliesthatuntyingledtomorecost-effectiveuseofaidfunds,atleastforequipment.
• UntyinghasbeenassociatedwithashiftawayfromdonorprioritiesandtowardstheLaoGovernment’sneedsandobjectives(alignment),supportedbythefactthatmoregovernmentagenciesarenowresponsibleforprojectimplementationandmanagement.
• Localsourcingofcontracts,consultants,workersandmaterialsasaresultofuntyingcanhavepositivedevel-opmentaleffects,byprovidingincreasedemploymentopportunitiesandrevenuesforlocalmanufacturersandsuppliers.However,asdiscoveredfromtheaccountofprojectpractices,itisnotalwayspracticaltoprocureallgoodsandservicesdomestically.Severalequipmentitemsorskillsareunavailablelocally,somustbeimported.Qual-itystandardsmayalsorequiregoodsandservicestobeimported,fortheendproducttobesufficientlyeffective.
Conclusions and recommendationsTheLaoPDRuntiedaidcountrystudyprovidesvaluablein-sightintothepracticalfunctioningandprocessesofdifferentdonorfundedinfrastructureprojectsinthecontextofaSouthEastAsianLDC–ausefulcontributiontoanareawithverylit-tlepreviousinvestigation.Thebenefitofthestudyisthatithasenabledtheidentificationandseparationofthecomponentsofaprojectwherebothuntyingandtyingareoccurring.Asoneofsixcountrystudies,theLaoPDRstudywillcontributetoacrosscountrycomparisonofdonorpracticeswithrespecttountyingandthecorrespondingimpactonaideffectiveness,whichwillassistintheunderstandingoftheinternationaluntyingprocessandrecommendationstofurtherimproveuntiedaiddeliveryandaideffectiveness.
Fromthein-depthprojectcasestudyprojects,threeoutoffourwerefoundtobelargelyuntied.Someareasofde factotyingwerestilluncovered,includingtheareasofTA,headcontractorsandequipment.Interestingly,noequipmentor
materialsintheEU,SidaandAFDprojectswerefoundtobetied–atypicalareaoftyinginthepast.Theanalysisshowedthatprojectcasestudyprojectsthatwereoveralluntiedalsodisplayedcharacteristicsofbeingcostanddevelopmentallyeffective,therebyprovidingevidencesupportingthebenefitsofuntiedaid.
ToimprovetheuntyingprocessfurtherinLaoPDR,andtoin-creaseaideffectivenessingeneral,itisrecommendedonthebasisofthisstudythatfurtherstepsaretakentoencouragetheuseof‘newer’aidmodalitiesamongdonors(e.g.budgetsupportandpooledfunding)andincreaseduseofcountryprocurementsystems.FurtheractionshouldbetakentountietheservicecomponentsandTAofprojects,bydevelopingtheskilledlaboursectorthrougheducationandtrainingactivities,andbyincreasingthesupplyofqualifiedconsultancyservicesinLaoPDR.Encouragingparticipationofnon-DACdonorsinuntiedaidpoliciesisanotherpotentialchanneltoexplore,forachievingfurtherprogressinuntyingaidwithrespecttobothLaoPDRandinternationally.
Untying aid: Is it working?90
Annex D
d.5 vietnam Country study89
IntroductionInthe2008PhaseOnestudyTheDevelopmentalEffectivenessofUntiedAid,itwasfoundthataidtoLDCsisnowoverwhelm-inglyuntied(excludingexemptcategories,technicalco-operationandfoodaid).However,itwasalsofoundthatthereexistedverylittleformalanalysisofboththedonorpracticesinpartnercountriesandtheactualconsequencesofuntyingaidatacountrylevel.ThePhaseTwostudyisinresponsetotheproposalsoftheWorkingPartyonAidEffectivenessandtheDevelopmentAssistanceCommittee(DAC)NetworkonDevelopmentEvaluationandthereportingrequirementinthe2001DACRecommendationonuntyingODAtoLDCs.
PhaseTwooftheprojectisintendedtoprovideevidence-basedconclusionsabouttheimplicationsofuntyingaidanditsimpactonaideffectivenessinselectedcase-studycoun-tries,forpresentationtotheDACbyDecember2009.ThisreportprovidestheresultsoftheVietnamCountryStudy.
ThisreportprovidestheresultsoftheVietnamCountryStudyandispartoftheLDCSouthEastAsianLaoPDR-VietnamCountrypair.Usingasampleoffourprojectcase-studies,combinedwithabriefstatisticalandeconometricreview,thisstudyaimstocontextualisethekeyissuesofaideffective-nessandimplicationsofuntyingaidinVietnam.Theselectedprojectcase-studiesfocusonprocurementpracticesofthedo-norsandgovernmentalexecutingagencies,whichmayhaveasignificantimpactonhowgoodsandservicesarepurchased,withimplicationsforaideffectiveness.
Thefourselectedcasestudieswere:• twoirrigationprojectsfundedrespectivelybyLux-Devel-
opmentandJICA• awatersupplyandsanitationprogrammefundedby
AusAID,DanidaandtheNetherlands• afloodpreventionconstructionprojectfundedbyAfD.
Statistical and econometric analysis ODAcommitmentstoVietnamhaveincreasedsteadilysince1993,withlargercontributionsinthepastdecade.InanoverallreviewofOECDODAtoVietnam,donorcountrieshadby2007formallyuntiedover70%oftheirODAcommitments.From2005to2007,Vietnamreceivedaidthroughbothgrants(41%)andloans(59%),withdonorsprovidingasignificantamountoftechnicalassistance.Approximately65%ofallODAcommitmenttoVietnamwasuntiedfrom2005to2007.AlargeproportionofdonorODAtoVietnamisintheformoftechnicalco-operation(TC)(TableD.6),andthetyingstatusofmuchofthisformofODAisnotreportedtotheDAC.
TheaimoftheeconometricanalysisistodeterminewhetherODA,thetyingstatusandthedifferentmodalitiesthroughwhichaidisprovided(loansandgrants)haveanysignificant
89 The study was undertaken by Adam McCarty et al. (2009) ([email protected]).
impactonaggregatedonorexportflowsatthecountrylevel,inthiscasetoVietnam.Theresultssuggestthatgrantsfromdonorcountrieshavetrade-distortingeffectsondonorexports,andpossiblyalludetoinformalorde factotyingofgrants.However,theanalysisforVietnamisnotconclusive.
Aid modalities and tyingAlthoughthemajorityofdonorsinVietnamcontinuetoprovideODAthroughaprojectbasedapproach,thereisanin-creasingtrendtomovetowardsnon-projectbasedmodalitiesandusecountrysystems,throughgeneralorsectorbudgetarysupportandpooledfunding.DonorsprovidingODAthroughprojectbasedapproachesareprogressivelytailoringprojectstoensureuntyingstatusthroughgreateruseofcountrysys-tems,bydecentralisingprojectimplementationandexecutionatsub-nationallevelsand,insomecases,ensuringgovern-mentagenciesremainthemaincontractingauthorityforprocurement.Asof2008,‘untiedaid’represented74.3%ofallcommittedaidreceivedbyVietnamfromDACcountries.
Table D.6 Vietnam: aid commitments and disburse-ments by type of assistance 2006-07
Type of Assistance Commit-ments
Disburse-ments
US$ mn
% US$ mn
%
Emergency and Relief Assistance 167 3.2 61 1.7
Free-standing Technical Cooperation 683 13.0 697 19.1
Investment Project Assistance 1841 35.0 1252 34.3
Investment Project Assistance with a TC Component
1538 29.2 1240 34.0
Investment-related Technical Cooperation
36 0.7 30 0.8
Programme/Budgetary Aid or Balance of Payments Support
467 8.9 308 8.5
Unspecified 533 10.1 58 1.6
Total 5264 100.0 3646 100.0
Source: DAC database Vietnam
Project analysisThefourselectedcasestudiesconsistedofthreeactiveprojectsandoneprogramme,alllocatedwithinthewatersupplyandsanitationsector.Aidmodalitiesvariedacrossthecasestudies,yetallfourweresimilarinthattheylargelyinvolvedprocurementofservices,whiletheprocurementofequipmentandmaterialswasrelativelysimilar,allowingforcross-comparison.
1. TheSaigonRiverLowlandsFloodPreventionProjectwashandedoverforfinancingtoAfDin2007,aftertheGov-
Untying aid: Is it working? 91
Annex D
ernmentofVietnamdecideditwasnolongerabletocontinuefundingtheproject.AfDiscontributingUS$11.8mn,throughgeneralbudgetsupport.TheexecutingagencythroughMARDistheHydraulicProjectInvestmentandConstructionManage-mentBoard9(ICMB-9),responsiblefortheprojectsalongtherightbank,andtheHoChiMinhCityPMUforprojectsalongtheleftsideoftheSaigonRiver.
2. TheRuralWaterSupplyandSanitationNationalTar-getedProgrammeII(RWSSNTPII)isafollowupprogrammetotheRuralWaterSupplyandSanitationprogramme(RWSSNTPI),whichstartedintheearly1990sinco-operationwithAustraliaandDenmark,withthegovernment.TheRWSSNTPIIisapooledfundingprogrammebetweenDenmark,Australia,theNetherlandsandthegovernment,anditisexpectedtorunfrom2006to2011.TotalfundingfortheprojectisUS$315mn,ofwhichAusAIDhascommittedUS$34mn.DanidacontractedtheDanishcompanyMercerEwertoassistintheprocurementoftheinternationalTAs,whileAusAIDcontractedtheAustraliancompanyURStoprocurenationalconsultantsandTAsforthisprogramme.
3. TheNam-TuanIrrigationprojectisasingle-donorprojectwhichcommencedinApril2006,andisexpectedtocompleteinApril2010.LuxembourgDevelopmentissupport-ingtheproject,attherequestofthegovernment,throughanon-reimbursablegrantofEUR3,293,100.TheinternationalTCisofBritishnationality,directlyappointedbyLux-Dev.Lux-Devremainsthecontractingauthorityforallservicesandconstruc-tionworkscontracts,buttheexecutingagencyresponsibleforprocurementistheDARD,specificallythePMUoftheNam-TuanIrrigationproject,ledbytheinternationalTC.
4. ThePhanRi-PhanTietIrrigationprojectisbeingfundedbyJICAthroughaloanagreement,andwasintendedtobeim-plementedfromMarch2006untilDecember2012.JICAisthesoledonor,committingatotalODAloanofUS$51.9mntothegovernment.Theaidmodalityisprojectbasedaidassistance,andtheprojectisexecutedthroughthe‘CentralProjectOffice’(CPO)underMARD,whichisresponsibleforitsmanagement,includingholdingresponsibilityforprocurementofcontrac-torsandtheemploymentofconsultants.BCEOMCompany(France)wasawardedthecontracttoprovideconsultancyservicesforprojectimplementation,andwasprocuredthroughshortlistedmethod.AllconstructioncontractswereawardedtoVietnamesefirms.
Cost and developmental effectivenessAtthesub-nationalandnationallevel,developmenteffec-tivenesswasimprovedthroughknowledgetransfers,firstlybetweendonorsandthegovernment,andsecondlybetweentheinternationalTAsandnationalconsultantsorprojectowners.LocallabourandlocalsourcingofgoodsincreasedtheincomesandemploymentofsomeofVietnam’spoorestprovinces,wherethesedevelopmentinitiativestakeplace.Yetproblemsofseasonalortemporaryemploymentquestionthesustainablelong-termdevelopmentimpact.Difficultiesalso
lieindeterminingwhethertheeffectonlocalemploymentisadirecteffectoftheuntiednatureoftheprojects,i.e.thereislittleevidencetoassumethatiftheseprojectsweretiedthattheywouldsourcelabourfromforeignmarkets.Intermsofcosteffectiveness,certainprojectsindicatedacleartrade-offbetweentime-efficiencyandcosteffectiveness,suchastheNamTuanIrrigationProjectandtheAfDSaigonRiverFloodPrevention,bothofwhichusedVietnamesegovernmentagenciestoprocureconsultingservicesandconstructionworks.However,theprojectssuffereddelaysduetobureau-craticproceduresoroftenineffortstopreventcollusion.Bycomparison,AusAIDoutsourcedtheprocurementofnationalconsultantstoanAustralianfirm,whichledtothetimelypro-curementandawardingofnationalconsultancycontracts.ThefourcasestudiesindicatedthatpricespaidforequipmentandmaterialsprocuredthroughtheGoVprojectimplementa-tionunitsislikelytohavebeenaboveorequaltothemarketpriceduetotheeffectofdirectcontracting,localsupplyconstraintsandsocialnetworks.
Conclusion and recommendations Inactualpractice,theprocurementofgoodsandservicesforallfourprojectssuggestslittleevidenceofimplicitorde factotyingofaidbydonors,sincethedecentralisationofallfourprojectsensuredextensiveuseoflocalsuppliersforbothservicesandworkscontracts.ThebiggestobstacletountyingaidcompletelyinVietnamisthegovernment’scurrentcapacitytoprocurequalitycontractsinatimely,effectiveandtransparentmanner.EffortshavebeenmadebyallfourdonorstoenhancetheVietnameseProcure-mentLaw(VPL)invariouswaystoimprovethis.However,thiscanleadtoconfusionandoverlapping,inevitablyresultinginanineffectiveprocurementsystem.FurthereffortsneedtobemadetoensureanimprovedVPLisappliedtoalldonorprojects.Furthermore,thestrongdonorfocusonmonitoringandevalu-ationconstrainsODAfrombeingcompletelyuntied.Basedonthefindingsfromthecasestudies,policieswhichstrengthengovernment’scapacityininternationalprocurementandenhancethegovernment’sfinancialandprojectmonitoringsystemswouldperhapsbringaboutsomeimprovement.
ICBisthepreferredmethodofselection,sinceitisconsideredthemostopenandmostcompetitive,andisstronglyrelatedtountying.IfICBisnotpractical,thisencouragesuncom-petitiveselectionanddirecthire.Therefore,asinternationalcompetitorsincreasetheirpresenceinVietnamandhavegreatermarketaccess,thiswillalsoenablemoreICBselectionproceduresandgreateruntying.Regardlessoftheaidmodality,fromthecasesreviewed,asig-nificantefforthasbeenmadetoensurethatownershipoverODAliesinthehandsofthegovernment,throughtheuseofcountrysystems,nationalTAsandthegreaterinvolvementofgovernmentexecutingagencies.Althoughcounterfactualsarenotavailableforcomparisonforthisreport,untyingaidisjudgedtohavecontributedtoimprovingaideffectivenessondevelopmentinVietnam.Furthermore,despiteVietnam’sfast
Untying aid: Is it working?92
Annex D
approachingmiddle-incomecountrystatus,donorprovisionofODAcontinuestoplayasignificantroleinthecountry’seconomicandsocialdevelopment,whichcouldbeacceler-atedbycontinuingtoopentheprocurementprocessandenablemorerecipientcountryownership.
Untying aid: Is it working? 93
Annex E
Abildsnes,K.J.(2007)‘CreditorresponsibilityandtheNorwe-gianShipExportCampaign’.Unpublished.
ActionAid(2003)‘Towardseffectivepartnership–Untieaid.’London.
ActionAidInternational(2005)RealAid:An agenda for making aid work,Johannesburg,ActionAidInternational.
ActionAidInternational(2006)RealAid:Making Technical As-sistance work,Johannesburg,ActionAidInternational.
Alonso,J.A.(1999)‘Instrumentos:loscréditosFAD’Estrategia para la Cooperación Española,Madrid:MAE-SECIPI.
Amegashie,J.A.,Ouattara,B.,andStrobl,E.(2007)‘Moralhazardandthecompositionoftransfers:theorywithanapplicationtoforeignaid’,CESifoWorkingPaperSeriesNo.1996.
Arvin,M.,Choudhry,S.,andDunlop,S.(1996)‘Lingeringeffectofuntiedforeignaidonexports’,Atlantic Economic Journal24(2):180.
Arvin,M.andChoudhry,S.(1997)‘UntiedAidandExports–Dountiedaiddisbursementscreategoodwillfordonorexports?’,Canadian Journal of Development Studies18(1997):9-22.
Arvin,M.andBaum,C.(1997)‘Tiedanduntiedaid:atheoreti-calandempiricalanalysis’,Keio Economic Studies34(1997):71-79.
Arvin,M.,Cater,B.,andChoudhry,S.(2000)‘Acausalityanalysisofuntiedforeignassistanceandexportperformance:thecaseofGermany’,Applied Economic Letters7:315-319.
Arvin,M.,Piretti,A.andLew,B.(2003):“ForeignAidandExportExpansion:TestingforCausalityandCointegration,,RivistaInternazionaleDiScienzeSociali,111,459-492
Aryeetey,E.,Osei,B.,andQuartey,P.(2003)Does Tying Aid make it More Costly?AGhanaianCaseStudy.PaperpresentedattheworkshoponQuantifyingtheImpactofRichCountries’PoliciesonPoorCountries,23–24October.Washington,D.C.,CenterforGlobalDevelopmentandtheGlobalDevelopmentNetwork.
AusAID(2006)Australian aid: promoting growth and stabil-ity: a White Paper on the Australian Government’s Overseas Aid Program.Canberra,GovernmentofAustralia.
AusAID[n.d.]‘ChiefExecutiveInstructions:2.4Procurementofgoodsandservices.’GovernmentofAustralia.Unpublished.
Australia,Government,DepartmentofFinanceandAdminis-tration,(2004)Commonwealth Procurement Guidelines – Janu-ary 2005.(FinancialManagementGuidanceNo1).Canberra.
Barrett,C.B.(1998)‘Foodaid:Isitdevelopmentassistance,tradepromotion,both,orneither?’,American Journal of Agri-cultural Economics80(3):566.
Barrett,C.B.(1999)‘ThedynamiceffectsofU.S.foodaid’,Eco-nomic Inquiry37(4):647-656.
Barrett,C.B.andMaxwell,D.(2005)Food aid after fifty years.London,Routledge.
Belfrage,C.J.(2007)‘ComparativeeffectsofEUandUSfoodaidonlocalproductionandcommercialtrade’inBourdetet al.(eds)The European Union and developing countries: trade, aid and growth in an integrating world. Cheltenham,EdwardElgar.
Bhagwati,J.N.(1985)‘Thetyingofaid’,ch.12inBhagwati,J.N.(ed.)Dependence and interdependence.Oxford,Blackwell.
Bhagwati,J.N.,Brecher,R.A.,andHatta,T.(1983)‘TheGeneral-izedTheoryofTransfersandWelfare:bilateraltransfersinamultilateralworld’,American Economic Review 73(4):606.
AnnEX E: rEfErEnCEs
Untying aid: Is it working?94
Annex E
Brakman,S.andvanMarrewijk,C.(1995)‘Transfers,returnstoscale,tiedaidandmonopolisticcompetition’,Journal of Devel-opment Economics47(2):333.
Chatterjee,S.,Sakoulis,G.,andTurnovsky,S.J.(2003)‘Unilater-alcapitaltransfers,publicinvestment,andeconomicgrowth’,European Economic Review47(6):1077-1103.
Chatterjee,S.andTurnovsky,S.J.(2004)‘Tiedversusuntiedforeignaid:consequencesforagrowingeconomy’,Computing in Economics and Finance2004(8),SocietyforComputationalEconomics.
Chatterjee,S.andTurnovsky,S.J.(2007)‘ForeignAidandEconomicGrowth:Theroleofflexiblelaboursupply’,Journal of Development Economics84,507-533.
Chichilnisky,G.(1983)‘Thetransferproblemwiththreeagentsonceagain’,Journal of Development Economics13:237-248.
Chinnock,J.(1998)In whose benefit? The case for untying aid.London,ActionAid.
Chinnock,J.andCollinson,S.(1999)PurchasingPower–AidUntying,targetedprocurementandpovertyreduction.Lon-don,ActionAid.
Clay,E.J.,Geddes,M.,Natali,L.,Mensah,E.J.,andQuartey,P.(2009)OECDDAC–PDEThematicStudyonUntiedAid:Ghanacountrystudy,ODI,London.
Clay,E.J.,Geddes,M.,Natali,L.andteVelde,D.W.(2008)The-maticStudy,TheDevelopmentalEffectivenessofUntiedAid:EvaluationoftheImplementationoftheParisDeclarationandofthe2001DACRecommendationonUntyingODAToTheLDCs,PhaseOneReport.Copenhagen.
Clay,E.J.,Geddes,M.,Natali,L.,Mensah,E.J.andQuartey,P.(2009) ‘Thetradeimplicationsofaidinstrumentsandtyingpractices:Ghanacountrystudy.’ReporttoKfWEntwick-lungsbank, OverseasDevelopmentInstitute,London,July2009
Clay,E.J.andTurner,E.(2007)‘ImpactofOverseasDevelop-mentAssistance(GrantsandLoans)onTrade’ReporttoKfWEntwicklungsbank, OverseasDevelopmentInstitute,London,April2007.
Colding,B.andPinstrup-Andersen,P.(1998)Food aid as a development assistance instrument: past, present and future.Washington,D.C.:InternationalFoodandPolicyResearchInstitute(draft).
Collier,P.(2008)‘ImplicationsofChangedInternationalCondi-tionsforEITI’,EITIInternationalSecretariat,BoardPaper6-8.Availableat:http://eiti.org/node/796.
Coulter,J.,Walker,D.J.andHodges,R.(2007)‘LocalandregionalprocurementoffoodaidinAfrica:impactandpolicyissues’Journal of Humanitarian Assistance,28/10/2007.
Danida(2002)Evaluation of Mixed Credit Programme.Report2002/04.Copenhagen,GovernmentofDenmark,MinistryofForeignAffairs.
Danida(2004)Meta-evaluation: private business sector develop-ment interventions,2004/6.Copenhagen,GovernmentofDenmark,MinistryofForeignAffairs.
Devarajan,S.,Rajkumar,A.S.,andSwaroop,V.(1999)‘WhatdoesaidtoAfricafinance?’,World Bank Policy Research Working Paper.
Ensign,M.(1992)Tied aid and its outcome: doing good or doing well? Japan’s foreign aid program,NewYork,ColumbiaUniversity.
FAO(2006)The State of Food and Agriculture(SOFA).Rome,FoodandAgricultureOrganisation.
FerreiraDuarte,R.andMetz,M.(1996)Cape Verde: an extended study.(JointEvaluationofEUProgrammeFoodAidBack-groundPaper04)London,ODI,1996.
Feyzioglu,T.,Swaroop,V.,andZhu,M.(1998)‘APanelDataAnalysisoftheFungibilityofForeignAid’,World Bank Economic Review12(1):29-58.
GATT(1979)Agreement of Government Procurement.(doneatGenevaon12April1979).Geneva.
Geddes,M.,Mensah,E.J.NataliL.andQuarteyP.(2009)‘OECDDAC-PDEThematicStudyonUntiedAid:GhanaCountryStudy,OverseasDevelopmentInstitute(ODI),London,LondonandISSER,UniversityofGhana,Lagon.
Gill,B.andReilly,J.(2007)‘TheTenuousHoldofChinaInc.inAfrica’,Washington Quarterly30(3):37-52.
Hatzipanayotou,P.andMichael,M.S.(1995)‘Foreignaidandpublicgoods’,Journal of Development Economics47(2):455-467.
Hejazi, W. and A. E. Safarian (2001) ‘The Complementarity be-tween U.S. Foreign Direct Investment Stock and Trade’, Atlan-tic Economic Journal, 29, pp. 420–437.
Helleiner,G.(2000)‘Towardsbalanceinaidrelationships:donorperformancemonitoringinlow-incomedevelopingcountries’,CISWorkingPaper,UniversityofToronto.
Hendra,J.(1987)‘Onlyfittobetied:AComparisonoftheCanadianTiedAidPolicywiththeTiedAidPoliciesofSweden,NorwayandDenmark’,Canadian Journal of Development Stud-ies8(2):261-281.
Untying aid: Is it working? 95
Annex E
Hopkin,A.(1970)‘Aidandthebalanceofpayments’,Economic Journal80(370):1-22.
IDD,&Associates.(2006)Evaluation of General Budget Sup-port: Synthesis Report. Birmingham:UniversityofBirmingham,InternationalDevelopmentDepartment.www.oecd.org/da-taoecd/25/43/37426676.pdf
Jepma,C.J.anddeHaan,H.(1984)‘TyingoftheNetherlands’developmentaid’inJepma,C.J.et al.(eds)Development aid and the Netherlands’ business community.Leiden.
Jepma,C.J.(1988)‘TheimpactofuntyingaidoftheEuropeanCommunityCountries’,World Development16(7):797-805.
Jepma,C.J.(1991)The tying of aid,Paris,OECD
Jepma,C.J.(1994) Inter-nation policy co-ordination and untying of aid.Brookfield(USA)Avebury.
Johansson,L.,Nannyonjo,J.,Pettersson,J.andStavlota,U.(2006)Uganda:Aidandtrade.CasestudyfortheOECDPolicyDialogueonAidforTrade,6-7November,Qatar.
Johansson,L.M.,andPettersson,J.(2009)TiedAid,Trade-FacilitatingAidorTrade-DivertingAid?Working Papern.5,UniversityofUpsala.
Kemp,M.C.andKojima,S.(1985)‘Tiedaidandtheparadoxesofdonor-enrichmentandrecipient-impoverishment’,Interna-tional Economic Review26(3):721.
Khilji,N.M.andZampelli,E.M.(1991)‘ThefungibilityofUSas-sistancetodevelopingcountriesandtheimpactonrecipientexpenditures:acasestudyofPakistan’,World Development19(8):1095-1105.
LaChimia,A.andArrowsmith,S.(2009)‘AddressingTiedAid:TowardsaMoreDevelopment-OrientedWTO?’Journal of Inter-national Economic Law12(3):707-747.
Lahiri,S.andRaimondos,P.(1995)‘Welfareeffectsofaidunderquantitativetraderestrictions’,Journal of International Econom-ics39(3/4):297-315.
Land,T.(2007)Joint evaluation study of provision of technical assistance personnel: what can we learn from promising experi-ences?SynthesisReport.DiscussionPaper,78.Maastricht,EuropeanCentreforDevelopmentPolicyManagement.
Lanser,P.,Mathijssen,J.,Pare,E.(2009)OECDDAC–PDEThematicStudyonUntiedAid:BurkinaFasocountrystudy,ECORYS,NederlandBV.
Larrú,J.M.(2003)‘LaAyudaLigada:¿UnMalInstrumentodelaPolíticaEspañoladeCooperaciónparaelDesarrollo?’inGonzálezLaxe,F.andSequeirosTizón,J.(eds)Orden Económi-
co Mundial. Globalización y Desarrollo,ACoruña:SociedaddeEconomíaMundial.
Larrú,J.M.andGonzalez,M.(2004)‘¿AquiénbeneficianloscréditosFAD?Losefectosdelaayudaligadasobrelaeconomíaespañola’,Estudios Economicos de Desarrollo Interna-cional 4(1).
Levitt,M.S.(1970)‘AidandtheBalanceofPayments:thescopeforuntyingandforincreasingaid’,The Manchester School38(3):247-257.
Lloyd,T.,McGillivray,M.,andMorrissey,O.(2000)‘Doesaidcreatetrade?AninvestigationforEuropeandonorsandAfricanrecipi-ents’,European Journal of Development Research12(1):107-123.
Lloyd,T.,Morrissey,O.andOsei,R.(2001),“ProblemswithPoolinginPanelDataAnalysisforDevelopingCountries:TheCaseofAidandTradeRelationships”,UniversityofNottinghamCREDIT Research Paper01/14
Love,J.andDunlop,S.(1990)‘Thedomesticimpactofover-seasaid:acasestudyoftheAidandTradeProvision’, National Westminster Bank Quarterly ReviewAug.:54-68.
MacQuaide,C.andToye,J.(1986)‘ATPProjectsEvaluation:AstudyofsixprojectsfinancedundertheAidandTradeContingencyProvisionbetween1978and1980.ReporttotheOverseasDevelopmentAdministration’,Ev.316,London:ODA
Malek,M.H.,May,R.S.,andSchumacher,D.(1990)‘ACompara-tiveAssessmentoftheEmploymentConsequencesofUKandWestGermanAid’,International Review of Applied Economics,(4):1-27.
Manning,R.(2006)‘Will‘EmergingDonors’changethefaceofinternationalco-operation?’,Development Policy Review 24(4):371-385.
MAPConsult.(2007)‘AnalysisofbaselineresearchonuntyingbilateralaidtoGhana.’RevisedFinalReport.Accra,31January2007.
Marchione,T.(2002)‘FoodsprovidedthroughtheUSGovern-mentEmergencyFoodAidProgrammes:policiesandcustomsgoverningtheirformulation,selectionanddistribution’,Jour-nal of Nutrition132:2194S-2111S.
Martínez-Zarzoso,I.,Nowak-Lehmann,D.F.,andKlasen,S.(2008)‘DoesGermandevelopmentaidpromoteGermanexports?’,DiscussionPaper,170,Goettingen:Ibero-AmericaInstituteofEconomicResearch.
Massa,I.andteVelde,D.W.(2009)‘Thetradedistortionim-plicationsofloansandgrants:aneconometricexamination.’ReporttoKfWEntwicklungsbank. OverseasDevelopmentInstitute,London,May2009.
Untying aid: Is it working?96
Annex E
McCarty,A.andJulian,A.(2009)‘OECDDAC-PDEThematicStudyonUntiedAid:LaoPDRCountryStudy’,MekongEco-nomics,Hanoi.
McCarty,A.,Smart,K.andBanerjee,D.(2009)‘OECDDAC-PDEThematicStudyonUntiedAid:VietnamPDRCountryStudy’,MekongEconomics,Hanoi.
McCormick,D.(2008)‘China&IndiaasAfrica’snewdonors:theimpactofaidondevelopment’,Review of African Political Economy35(115):73-92.
Michael,M.S.andvanMarrewijk,C.(1998)‘Tiedtocapitaloruntiedforeignaid?’,Review of Development Economics2(1):61.
Miquel-Florensa,J.M.(2007)‘Aideffectiveness:acomparisonoftiedanduntiedaid’.YorkUniversity.Unpublished.
Morrissey,O.(1989)Donorbenefitsfromtiedaid:somereflec-tionsonstrategictradepolicyfortheUK.CreditResearchPaper,47460/89/4.Nottingham,UniversityofNottingham.
Morrissey,O.(1990)‘Theimpactofmultilateralandtiedbilat-eralaidontheUKEconomy’Journal of International Develop-ment2(1):60-76.
Morrissey,O.(1991)‘AnEvaluationoftheeconomiceffectsoftheAidandTradeProvision’,The Journal of Development Stud-ies28(1):104-129.
Morrissey,O.andWhite,H.(1996)‘Evaluatingtheconcessional-ityoftiedaid’, Manchester School of Economic and Social Studies 64(2):208-26.
Morrissey,O.andRudaheranwa,N.(1998)Research consultancy report on tied aid.London,ActionAid.
Naschold,F.andBooth,D.(2002)GeneralBudgetSup-portEvaluabilityStudy:LiteratureReview,ODI,London.Availableat:http://www.odi.org.uk/resources/details.asp?id=3484&title=general-budget-support-evaluability
Natali,N.,Geddes,M.,andClayE.J.(2009)“Econometricanalysisoftheimplicationsofbilateralaidfordonorexports:theroleoftyingstatusandotherinfluences.”(BackgroundPaperforPDE-DACThematicStudyonAidUntying.)London:OverseasDevelopmentInstitute(ODI),London.Draft.
Nilsson,L.(1997)‘AidandDonorExports:TheCaseoftheEUCountries’inNilsson,L.(ed.)Essays on North-South Trade,Lund:LundEconomicStudies,N.70.
Nilsson,M.(2004)EffectsofBudgetSupport–ADiscussionofEarlyEvidenceUTVWorkingPaper2004:4
Norad(2007a)Evaluation of the Norwegian petroleum-related assistance: case studies regarding Mozambique, Bangladesh, East
Timor and Angola.EvaluationReport,1/2007.Oslo,NorwegianAgencyforDevelopmentCooperation.
Norad(2007b)Oil for development.Oslo,NorwegianAgencyforDevelopmentCooperation.
Norad(2008)‘EvaluationoftheNorwegianEmergencyPre-parednessSystem(NOREPS)’.EvaluationReport1/2008.Oslo,NorwegianAgencyforDevelopmentCooperation.
Norway,Government,MinistryofForeignAffairs(2007)Annual budget 2008.StatePaper1(2007-2008)Oslo.
Nowak-Lehmann,D.F.,Martínez-Zarzoso,andI.,Klasen,S.(2008)‘AidandTrade–ADonor’sPerspective’,DiscussionPaper,171,Göttingen:Ibero-AmericaInstituteofEconomicResearch.
NRI(2005)Contributions to rural development by local purchase and regional procurement of food.ReporttoEuropeanCommis-sion.London,NaturalResourcesInstitute.
ODA(1996)‘AReviewofUKAidTyingPolicy’,London,OverseasDevelopmentAdministration.
OECD(1991)‘TheArrangementonguidelinesforofficiallysup-portedexportcredits’.(HelsinkiArrangement)Paris,OECD.
OECD(2006a)The developmental effectiveness of food aid. Does tying matter?Paris.
OECD(2006b)2006.SurveyonMonitoringtheParisDeclara-tion,CountryChapters,SouthAfrica.Availableonline:http://www.oecd.org/dataoecd/50/5/38939834.pdf
OECD(2008a)TheExportCreditsArrangement1978-2008.AchievementsandChallenges–Continued!Paris,OECD.
OECD(2008b)2008.SurveyonMonitoringtheParisDeclara-tion,CountryChapters,BurkinaFaso.Availableonline: www.oecd.org/dataoecd/26/17/42541095.pdf
OECD(2008c)2008.SurveyonMonitoringtheParisDeclara-tion,CountryChapters,Ghana.Availableonline:www.oecd.org/dataoecd/26/2/41951381.pdf
OECD(2008d)2008.SurveyonMonitoringtheParisDeclara-tion,CountryChapters,Zambia.Availableonline:www.oecd.org/dataoecd/25/39/42056851.pdf
OECD(2008e)2008.SurveyonMonitoringtheParisDeclara-tion,CountryChapters,LaoPDR.Availableonline:www.oecd.org/dataoecd/25/37/41951420.pdf
OECD(2008g).Development Co-operation Report 2007.Paris
Untying aid: Is it working? 97
Annex E
OECD(2008f )2008.SurveyonMonitoringtheParisDeclara-tion,CountryChapters,Vietnam.Availableonline:www.oecd.org/dataoecd/55/34/42056862.pdf
OECDDAC(1987)‘DACGuidingPrinciplesforAssociatedFinancingandTiedandPartiallyUntiedOfficialDevelopmentAssistance.’Paris.
OECDDAC(2001)‘DACRecommendationonUntyingOfficialDevelopmentAssistancetotheLeastDevelopedCountries.’Paris:DevelopmentAssistanceCommitteeoftheOrganizationforEconomicCooperationandDevelopment.Paris.
OECDDAC(2005)‘ParisDeclarationonaideffectiveness.’Paris.
OECDDAC(2006a)‘DACRecommendationonuntyingOfficialDevelopmentAssistancetotheleastdevelopedcountries.Amended.’Paris.
OECDDAC(2006b),BudgetSupport,SectorWideApproachesandCapacityDevelopmentinPublicFinancialManage-ment,Paris:OECD,Internet:http://www.oecd.org/ datao-ecd/53/7/34583142.pdf#search=%22DAC%20Budget%20sup-port%20Good%20Practices%22.
OECDDAC(2007)‘Aid untying to non-LDC HPCS’Paris,Develop-mentAssistanceCommitteeoftheOrganizationforEconomicCooperationandDevelopment.
OECDDAC(2008a)‘Implementing the 2001 DAC Recom-mendation on untying Official Development Assistance to the Least Developed countries. Progress Report 2008 – Draft’.DCD/DAC(2008)21Paris.
OECDDAC(2008b)‘Promotinglocalandregionalprocurementindevelopingcountries’DCD/DAC(2008)21,Paris.
OECDDAC(2009)‘Implementing the 2001 DAC Recommenda-tion on untying Official Development Assistance to the Least Developed countries. Progress Report 2009 – Draft’.DCD/DAC(2009)21/REV2Paris.
Osei,B.(2003)‘How tied aid affects the cost of aid-funded projects in Ghana’,AfricanEconomicResearchConsortium.AERCResearchPaper137;Nairobi,AERC.
Osei,B.(2004)‘ThecostofaidtyingtoGhana’,AfricanEconomicResearchConsortium.AERCresearchpaper144;Nairobi,AERC.
Osei,R.,Morrissey,O.,andLloyd,T.(2004)‘TheNatureofAidandTradeRelationships’, European Journal of Development Research16(2):354-374.
Pack,H.andPack,J.R.(1993)‘Foreignaidandthequestionoffungibility’,Review of Economics & Statistics75(2):258-265.
Paulo,S.andReisen,H.(2009)‘Oldhabits,newdonors’,D+C,2009/10,Tribune,Page386-387.Availableat:http://www.inwent.org/ez/articles/157339/index.en.shtml.
Pehnelt,G.(2007)‘ThePoliticalEconomyofChina’sAidPolicyinAfrica’.Unpublished.
Putterside,K.,VazquezMedina,B.,LaChimia,A.,andSchnei-der,A(2004)‘AnindependentstudyonfurtheruntyingofEuropeanaid’,OECD(MCIPS)RoomDocumentN.4.Paris,OECDQuartey,P.(2005)‘InnovativeWaysofMakingAidEffectiveinGhana:TiedAidversusDirectBudgetarySupport’,Journal of International Development17(18):1077-1092.
Ramkolowan,Y.andStern,M.(2009)‘OECDDACPDEThe-maticStudyonUntiedAid:SouthAfricaCountryStudy’,DNA,Pretoria.
Ray,J.(1995)Managing official export credits: the quest for a global regime.WashingtonD.C.,InstituteforInternationalEconomics.
Riddell,R.C.andStevens,S.(1997)Evaluation of Common-wealth Fund for Technical Co-operation (CFTC) to Uganda,Lon-don:StrategicPlanningandEvaluationUnit,CommonwealthSecretariat.
Saasa,O.S.andMphuka,C.(2009)‘OECDDACPDEThematicStudyonUntiedAid:ZambiaCountryStudy’,PremierConsult,Lusaka.
Schweinberger,A.G.(1990)‘Onthewelfareeffectsoftiedaid’,International Economic Review 31(2):457.
Schönherr,S.andVogler-Ludwig,K.(2002)“ForeignAidandDonorExportsPerformance:TheCaseofGermany”,in:Arvin,B.M.(ed.),New Perspectives on Foreign Aid and Economic Develop-ment,Westport(CT.),PraegerPublishers.
Stokke,O.(1989)‘ThedeterminantsofNorwegianaidpolicy’inStokke,O.(ed.)Westernmiddlepowersandglobalpoverty.Uppsala,ScandinavianInstituteofAfricanStudies,1989:159-229.
Stokke,O.(2005)‘Norwegianaidpolicy:continuityandchangeinthe1990s’inHoebink,P.andStokke,O.(eds.)Perspectives on European development co-operation.London,Routledge,2005:448-492.
Suwa-Eisenmann,A.andVerdier,T.(2006),“TheCoherenceofTradeFlowsandTradePolicieswithAidandInvestmentFlows”,OECDDevelopmentCentreWorkingPaperNo.254,Paris.Availableat:www.oecd.org/dataoecd/9/63/37450734.pdf
Switzerland,Government(1994)Loifédéralesurlesmarchéspublicsdu16décembre1994.L’AssembléefédéraledelaConfédérationsuisse,172.056.1.Bern.
Untying aid: Is it working?98
Annex E
Switzerland,Government(1995)LeConseilfédéralsuisse,1995Ordonnancesurlesmarchéspublics(OMP)du11décem-bre1995172.056.11.Bern
Switzerland,Government(2007a)Départementfédéraldel’économie2007OrdonnanceduDFEsurl’adaptationdesva-leursseuilsdesmarchéspublicspourl’année2008.172.056.12Bern.
Switzerland,Government(2007b)Switzerland’s international cooperation.SwissAgencyforDevelopmentandCooperation(SDC)andStateSecretariatforEconomicAffairs(SECO)AnnualReport,2006.Bern.
Tajoli,L.(1999)‘Theimpactoftiedaidontradeflowsbetweendonorandrecipientcountries’,Journal of International Trade & Economic Development8(4):373.
Tinbergen,J.(1962)ShapingtheWorldEconomy.NewYork:TwentiethCenturyFund.U.S.Congress,OfficeofTechnologyAssessment,(1993)DevelopmentAssistance,ExportPromo-tion,andEnvironmentalTechnology--BackgroundPaper,OTA-BP-ITE-1O7(Washington,DC:U.S.GovernmentPrintingOffice).
Tschirley,D.anddelCastillo,A.M.(2006)‘Localandregionalfoodaidprocurement:anassessmentofexperienceinAfricaandelementsofgooddonorpractice.’PolicySynthesisforCooperatingUSAIDOfficesandCountryMissions,79EastLansing,MichiganStateUniversity.
Tschirley,D.(2008)‘LocalandregionalfoodaidprocurementandtheU.S.FarmBill.’MSUBrownBag,June12,2008www.aec.msu.edu/fs2/outreach/LRP-June08BB-MSU.pdf
Toye,J.andClark,J.(1986)‘TheAidandTradeProvision:origins,dimensionsandpossiblereforms’,Development Policy Review4:291-313.
USGovernmentAccountabilityOffice,(2007)Foreign assist-ance: various challenges impede the efficiency and effectiveness of U.S. food aid,GAO-07-560,April13,2007.Washington,D.C.
Wagner,D.(2003)‘Aidandtrade–Anempiricalstudy’,Journal of the Japanese & International Economies17(2):153.
Williams,G.,Jones,S.,andAmber,V.(2003)A vision for the fu-ture of Technical Assistance in the International System,Oxford:OxfordPolicyManagement.
WFP(2006)Food procurement in developing countries.PolicyIssues,AgendaItem5WFP/EB.1/2006/5C,30January2006.Rome,WorldFoodProgramme.
WFP(2008)Food aid flows, 2007.Rome,WorldFoodPro-gramme,Interfais.
Wood,B.,Kabell,D.,Muwanga,N.andSagasti,F.(2008)Evalu-ation of the implementation of the Paris Declaration, Phase One.SynthesisReport.GovernmentofDenmark,MinistryofForeignAffairs.
WTO(1995)AgreementonSubsidiesandCountervail-ingMeasures.WTO,http://www.wto.org/english/docs_e/legal_e/24-scm.pdf
WTO(2008)RevisedDraftModalitiesForAgriculture,Tn/Ag/W/4/Rev.3,AnnexL,PossibleNewArticle10.4toReplaceTheCurrentArticle10.4oftheAgreementonAgriculture,InternationalFoodAid.Geneva,WorldTradeOrganisation.
Yassin,I.H.(1991)‘Aid-tyingandtherealvalueofforeignas-sistance:thecaseofSudan’,Pakistan Development Review30(2):189-206.
Yeats,A.J.(1990)‘DoAfricancountriespaymoreforimports:Yes’,EconomicReview4(1).
Zarin-Nejadan,M.,Monteiro,J.A.andNoormamode,S.(2008)‘TheImpactofOfficialDevelopmentAssistanceonDonorCountryExports:SomeEmpiricalEvidenceforSwitzerland’,TechnicalWorkingPaper,08-01,UniversityofNeuchatel.
Evaluation
of the implementation of the
Paris Declaration
Thematic Study
The Developmental Effectiveness of Untied Aid:
Evaluation of the Implementation of the Paris Declaration
and of the 2001 DAC Recommendation on Untying
ODA to the LDCs
Ownership, Alignment, Harmonisation, Results and Accountability
This report presents the results of a thematic study undertaken within the framework of the Evaluation of the Paris Declaration and in response to the request by the DAC for an evaluation of the effects of untying of aid to LDCs.The report builds on extensive research carried out in a first phase, pub-lished in October 2008 and subsequent case studies in Burkina Faso, Ghana, Laos, South Africa, Vietnam and Zambia.
Untying aid: Is it working?
Evaluation of the Paris Declaration
Untying aid
: Is it wo
rking?D
ecemb
er 2009