英文【世界銀行】The Worldwide Governance Indicators

PublicDisclosureAuthorizedPublicDisclosureAuthorized

PolicyResearchWorkingPaper10952

TheWorldwideGovernanceIndicatorsMethodologyand2024Update

DanielKaufmannAartKraay

WORLDBANKGROUP

ProsperityPracticeGroup

OfficeoftheChiefEconomistNovember2024

ReproducibleResearchRepository

Averifiedreproducibilitypackageforthispaperisavailableat

,clickherefordirectaccess.

PolicyResearchWorkingPaper10952

Abstract

ThispaperprovidesanoverviewofthedatasourcesandaggregationmethodologyfortheWorldwideGovernanceIndicators(WGI).TheWGIreportsixaggregategover-nanceindicatorsmeasuringVoiceandAccountability,PoliticalStabilityandAbsenceofViolence/Terrorism,Gov-ernmentEffectiveness,RegulatoryQuality,RuleofLaw,andControlofCorruptioninasampleof214economiesovertheperiod1996–2023.Theaggregateindicatorscom-bineinformationfrom35differentexistingdatasources,capturingsubjectiveperceptionsofthequalityofvarious

dimensionsofgovernancereportedbyexpertsandsurveyrespondentsworldwide.Thepaperbrieflydiscusseshowtousereportedmarginsoferrorwheninterpretingcross-coun-tryandover-timedifferencesintheaggregateindicators.ThepaperalsoupdatesandextendsearlieranalysisonthreekeyissuesrelatingtotheWGImethodology:(a)theeffectofcorrelatedperceptionerrors,(b)therobustnessoftheaggregateindicatorstoalternativeweightingschemes,and(c)theexistenceontrendsinglobalaveragesofgovernance.

ThispaperisaproductoftheOfficeoftheChiefEconomist,ProsperityPracticeGroup.ItispartofalargereffortbytheWorldBanktoprovideopenaccesstoitsresearchandmakeacontributiontodevelopmentpolicydiscussionsaroundtheworld.PolicyResearchWorkingPapersarealsopostedontheWebat

/prwp

.Theauthorsmaybecontactedatakraay@.Averifiedreproducibilitypackageforthispaperisavailableathttp://reproducibility.,clickherefordirectaccess.

R

Y

C

I

L

R

A

E

S

E

O

P

H

C

S

TRANSPARENT

P

E

R

W

O

R

K

I

ANALYSIS

A

NGP

ThePolicyResearchWorkingPaperSeriesdisseminatesthefindingsofworkinprogresstoencouragetheexchangeofideasaboutdevelopmentissues.Anobjectiveoftheseriesistogetthefindingsoutquickly,evenifthepresentationsarelessthanfullypolished.Thepaperscarrythenamesoftheauthorsandshouldbecitedaccordingly.Thefindings,interpretations,andconclusionsexpressedinthispaperareentirelythoseoftheauthors.TheydonotnecessarilyrepresenttheviewsoftheInternationalBankforReconstructionandDevelopment/WorldBankanditsaffiliatedorganizations,orthoseoftheExecutiveDirectorsoftheWorldBankorthegovernmentstheyrepresent.

ProducedbytheResearchSupportTeam

TheWorldwideGovernanceIndicators:Methodologyand2024Update

DanielKaufmann*andAartKraay**TheWorldBank

*PresidentEmeritus,NaturalResourceGovernanceInstitute,andSeniorFellowatResultsforDevelopmentandThe

BrookingsInstitution;**ChiefEconomist,ProsperityPracticeGroup,WorldBankakraay@.ThedatafortheWorldwideGovernanceIndicatorsareavailableat.Anoteonself-citation:This

paperdrawsextensivelyonourearlierjointworkoverthepast25yearstodesign,implement,andassesstheWGImethodology.Inadditiontothe2024updateoftheWGIdiscussedinthispaper,thelimitedneworupdated

analysisinthispaperisclearlyindicatedassuch.Theremainderofthepapershouldbeviewedasasummaryofourpreviousworkandnotoriginalresearch.AfulllistingofourearlierworkingpapersandjournalarticlesontheWGIisincludedinthereferencesofthispaper.Averifiedreproducibilitypackageforthispaperisavailableat

.Theviewsexpressedherearetheauthors’anddonotreflecttheofficialviewsoftheWorldBank,itsExecutiveDirectors,orthecountriestheyrepresent.

1

1.Introduction

Thispaperprovidesanoverviewofthedatasources,aggregationmethodology,andmainfindingsoftheWorldwideGovernanceIndicators(WGI).TheWGIarealong-runningdataproductfirstpublishedin1999(seeKaufmann,KraayandZoido-Lobaton(1999)forthefirstpresentationofthemethodologyandaggregateindicators),withsubsequentannualupdates.TheWGIreportdataonsixaggregateindicatorsofgovernance:VoiceandAccountability,PoliticalStabilityandAbsenceofViolence/Terrorism,

GovernmentEffectiveness,RegulatoryQuality,RuleofLaw,andControlofCorruption,covering214

economiesovertheperiod1996-2023.Theaggregateindicators,theunderlyingindividualindicatorson

whichtheyarebased,andextensivedocumentationofthemethodology,areavailableat.

TheWGIdonotinvolvenewprimarydatacollectiononourpartasauthorsoftheWGI.Rather,theaggregateindicatorsareconstructedbycombiningseveralhundredindividualindicatorsmeasuring

aspectsofgovernancetakenfrom35existingdatasourcesthatcaptureperceptionsandviewsofthe

qualityofgovernancereportedbymanyexpertsandsurveyrespondentsworldwide.TheWGIrelyonadiversesetofdatasources,includingseveralglobalandregionalsurveysofhouseholdsandfirms,aswellasmanyexpertassessmentsproducedbyarangeoforganizationsinthepublic,private,andNGO

sectors,oftenscoredbynetworksofexpertslivingandworkinginthecountriesorregionsthatthey

assess.FiveWGIdatasourcesareproducedbyorganizationsheadquarteredindevelopingcountries1,andafurthereightWGIdatasourcesarehouseholdorfirmsurveysdirectlycapturingtheviewsof

surveyrespondentsindevelopingaswellasadvancedeconomies.

TheWGIarebasedonperceptionsdataforfourreasons:(1)perceptionsmatter,becausehouseholdsandfirmsmakedecisionsbasedontheirviewsandperceptionsofthequalityofgovernance;(2)for

somedimensionsofgovernancesuchascorruptionthatdonotleavea“papertrail”,datameasuring

subjectiveperceptionsofcorruptionorself-reportedexperienceswithcorruptioncanprovidevaluableinsights;(3)datacapturingrespondents’perceptionsandviewscanprovidevaluableinformationonthegapbetweendejurerulesandtheirdefactoimplementation;and(4)unlikeobjectiveindicatorsthat

capturetheexistenceofspecificlaws,rulesandregulations,datacapturingsurveyandexpert

respondents’viewsarenotsusceptibleto“gaming”wherepolicymakerstargetreformstonarrowly

1TheAsianDevelopmentBank(basedinthePhilippines),theAfricanDevelopmentBank(basedinCoted’Ivoire),Afrobarometer(basedinGhana),theCenterforDemocracyandDevelopment(basedinGhana),andtheAfricanInstituteforDevelopmentPolicy(basedinKenyaandMalawi).

2

changespecificmeasuressimplybecausetheyhappentobeincludedinanaggregateindicatorthattheywishtoinfluence.Ofcourse,wedonotarguethatonlyperceptionsdataarerelevantformeasuring

governance.Rather,aswediscussinmoredetailinSection2.2below,ourpointissimplythatthistypeofdatabringsvaluableinsightsthatcanbeusedinconjunctionwithothertypesofinformationto

measuregovernanceacrosscountriesandovertime.

Theindividualindicatorsfromthe35datasourcesareassignedtothesixdimensionsofgovernanceandarecombinedintosixaggregateindicatorsusingastatisticalmethodologyknownasanUnobservedComponentsModel(UCM).Thestatisticalmethodologyconvertsthedatasourcesintocommonunits

andconstructsaweightedaveragethatcombinestheinformationineachofthedatasources.The

methodologyalsoproducesmarginsoferrorthatcapturetheunavoidableimprecisioninvolvedwith

measuringgovernanceacrosscountries.ThisimprecisionisnotuniquetotheWGI,butratherislikelytobepervasiveinanyeffortstomeasuregovernanceandinstitutionalqualityacrosscountries.Akey

attributeoftheWGIisthatthesemarginsoferrorareexplicitlyreportedalongsidetheestimatesof

governanceandshouldbeconsideredwhencomparingestimatesofgovernanceacrosscountriesandovertimeusingtheWGI.Incontrast,inmanyothermeasuresofgovernanceandinstitutionalquality,theyareleftimplicit.

TheWGIaredesignedtoenablebroadcross-countryandover-timecomparisonsofperceptionsof

governance,reflectingthesynthesisofviewsacrossmanyexistingdatasources,andwithdueregardformarginsoferror.Atthesametime,thereisawealthofinformationintheindividualdatasources

themselvesthatcanusefullybeexploredforunderstandingthefactorsbehindoverallpatterns

summarizedintheaggregateindicators.Forthisreason,thecomponentdataoftheWGIarereadily

availablethroughtheWGIwebsite,andusersareencouragedtoconsultthisdataalongsidethe

aggregateindicators.Inaddition,theWGIarenotdesignedtobeatooltoevaluatespecificgovernancereformsinindividualeconomies–forthispurpose,theWGIshouldbesupplementedwithmoregranularcountry-specificdataandanalysisthatcanshedlightonthelikelyimpactsofspecificpolicyand

institutionalreforms.

ThispaperalsoupdatesandexpandsearlierevidencewehaveprovidedonthreekeymethodologicalissuesrelatingtotheWGI:(1)thepossibilitythatdifferentdatasourceshavecorrelatedmeasurement

errors,withimplicationsfortheweightingschemeandprecisionoftheaggregateindicators;(2)the

possibilitythattheaggregateindicatorsmightchangesubstantiallywithalternativereasonableweightingschemes,withimplicationsfortherobustnessofthecross-countryandover-timepatternsinthesix

3

aggregateindicators;and(3)thepossibilityoftrendsinglobalaveragesofgovernance,withimplicationsfortheinterpretationofthebaselineWGIestimateswhereglobalmeansarenormalizedtozeroineveryperiod.Ourupdated–andinsomecasesexpanded–analysisisconsistentwithourearlierevidenceand

supportsourpreviousconclusionsthatthemethodologicalchoicesinthebaselineWGIestimatesareappropriate.Specifically,wedonotfindclearevidenceofcorrelatedperceptionserrorsacrossdata

sources,andwefindthatanequallyweighted(insteadofprecisionweighted)averageofthedata

sourcesleadstoverysimilarestimatesofgovernanceacrosscountries.Wealsofindnoevidenceof

significanttrendsinglobalaveragesofgovernanceasmeasuredbytheWGIdatasources.ThismeansthatthebaselineWGIchoiceofunitsthatnormalizestheglobalmeanofgovernancetobethesameineachperiodisappropriate,andthatchangesovertimeineconomies’relativepositionsonthesix

aggregateWGImeasuresbroadlycorrespondtoabsolutechangesascollectivelymeasuredbytheWGIdatasources.

TheWGIarewidelyusedforbroadcross-countryandover-timecomparisonsofgovernanceina

varietyofcontexts.Theyarefrequentlyusedinacademicandpolicyresearch,asevidencedbytheover25,000citationstothevariousWGImethodologypapersrecordedinGoogleScholar.2Theyareused–

togetherwithavarietyofotherindicators–bytheUnitedStatesMillenniumChallengeCorporationto

determinecountryeligibilityforitsaidprograms.3Recognizingthattheyhavepredictivepowerfordebtservicingdifficulties,theWGIareusedbymajorcommercialriskratingagenciesasoneofmanyinputstotheirmodelstoassesssovereignrisk,aswellasbytheInternationalMonetaryFundasakeyindicatorofgovernanceinitsDebtSustainabilityFrameworkforMarketAccessCountries.4Fourth,theyareused–togetherwithmanyotherindicators–toinformenvironmental,socialandgovernance(ESG)investmentstrategiesandcorporatesocialresponsibilitypoliciesbyfirmsintheprivatesector.5

Therestofthispaperisorganizedasfollows.Section2definesthesixgovernancedimensions,

providesanoverviewofthe35WGIdatasources,anddescribeshowtheindividualvariablesfromeach

2ThetopfivemostcitedWGImethodologypapers(withhyperlinkstotheirGoogleScholarlistings)areKaufmann,KraayandMastruzzi(2011),Kaufmann,KraayandMastruzzi(2009),Kaufmann,KraayandMastruzzi(2004),

Kaufmann,KraayandZoido-Lobaton(1999b)andKaufmann,KraayandZoido-Lobaton(1999a).

3ForinformationonMCCeligibilitycriteria,see

/who-we-select/scorecards/

.

4Forexample,forinformationontheFitchsovereignratingmethodology,clickhereandforMoody’s,clickhere.SeealsoIMF(2021)

5Forexample,theWGIareincludedintheWorldBank’sSovereignESGDataPortal,theyareusedintheMorganStanleyCapitalInternational(MSCI)ESGGovernmentRatingsMethodologyandtheFTSE-RussellSustainable

SovereignRiskMethodology,andtheyareusedbytheDisneyCorporationtoinformitsPermittedSourcingCountriespolicy.

4

ofthesedatasourcesareassignedtothesixgovernancedimensions.Annex1providesanannotatedsummaryoftheWGIdatasourceswithlinkstotheirmethodologyanddata,andAnnex2providesacompletelistingoftheassignmentofindividualindicatorstothesixaggregateindicators.Section3reviewstheaggregationmethodologyfortheWGI,andAnnex3providesfurthertechnicaldetails.

Section4providesabrieftourofthemostrecentaggregateindicators,withaparticularemphasisonhowtousemarginsoferrorwhencomparinggovernanceacrosscountriesandovertime.Section5updatesandexpandsonearlieranalysisofthreekeymethodologicalissuesintheWGI,andSection6offersbriefconcludingremarks.

2.Data

Inthissectionofthepaper,wefirstdefinethesixdimensionsofgovernancecorrespondingtothesixaggregateindicatorsreportedintheWGI.Wethendiscussthedatasourcesthemselves,howtheyare

selected,andhowtheindividualindicatorsfromthesemanydatasourcesareassignedtothesixaggregateindicators.

2.1SixGovernanceDimensions

Weorganizethemanyindividualindicatorsofgovernancedescribedbelowintosixgovernance

dimensions,basedonadefinitionofgovernanceas“thetraditionsandinstitutionsbywhichauthorityinacountryisexercised.Thisincludes(a)theprocessbywhichgovernmentsareselected,monitored,andreplaced;(b)thecapacityofthegovernmenttoeffectivelyformulateandimplementsoundpolicies;and(c)therespectofcitizensandthestatefortheinstitutionsthatgoverneconomicandsocialinteractionsamongthem”.Weconsidertwogovernancedimensionsineachofthesethreeareasforatotalofsix

dimensions.

(a)Theprocessbywhichgovernmentsareselected,monitored,andreplaced

1.VoiceandAccountability(VA)–capturingperceptionsandviewsoftheextenttowhicha

country'scitizenscanparticipateinselectingtheirgovernment,aswellasfreedomofexpression,freedomofassociation,andafreemedia.

2.PoliticalStabilityandAbsenceofViolence/Terrorism(PV)–capturingperceptionsandviewsofthelikelihoodthatthegovernmentwillbedestabilizedoroverthrownbyunconstitutionalorviolentmeans,includingpolitically-motivatedviolenceandterrorism.

(b)Thecapacityofthegovernmenttoeffectivelyformulateandimplementsoundpolicies

5

3.GovernmentEffectiveness(GE)–capturingperceptionsandviewsofthequalityofpublicservices,thequalityofthecivilserviceandthedegreeofitsindependencefrompolitical

pressures,thequalityofpolicyformulationandimplementation,andthecredibilityofthegovernment'scommitmenttosuchpolicies.

4.RegulatoryQuality(RQ)–capturingperceptionsandviewsoftheabilityofthegovernmenttoformulateandimplementsoundpoliciesandregulationsthatpermitandpromoteprivatesectordevelopment.

(c)Therespectofcitizensandthestatefortheinstitutionsthatgoverneconomicandsocialinteractionsamongthem:

5.RuleofLaw(RL)–capturingperceptionsandviewsoftheextenttowhichagentshave

confidenceinandabidebytherulesofsociety,inparticularthequalityofcontractenforcement,propertyrights,thepolice,andthecourts,aswellasthelikelihoodofcrimeandviolence.

6.ControlofCorruption(CC)–capturingperceptionsandviewsoftheextenttowhichpublic

powerisexercisedforprivategain,includingbothpettyandgrandformsofcorruption,aswellas"capture"ofthestatebyelitesandprivateinterests.

Whilethesesixdimensionsareconceptuallydistinct,thisdoesnotimplythattheyare

independentofoneanother.Forexample,betteraccountabilitymechanismsmayleadtolesscorruption,oramoreeffectivegovernmentismorelikelytoprovideabetterregulatoryenvironment,orrespectfortheruleoflawleadstofairerprocessesforselectingandreplacinggovernments.Giventhese

interrelationships,itisnotverysurprisingthatoursixcompositemeasuresofgovernancecorrespondingtothesesixdimensionsarestronglypositivelycorrelatedacrosseconomies.Werefrainhoweverfrom

combiningthedimensionsclustersintoasingleoverallcompositegovernanceindicator,forconceptualandstatisticalreasons.Conceptually,theresultingindicatorwouldbeextremelybroadanddifficultto

interpret.Statistically,wewouldfacetheadditionalchallengethatmanyofourunderlyingdatasourcesfeedintomorethanoneofthesixaggregateindicators.Thisraisestheconcernthatsomeofthe

observedcorrelationbetweenthesixaggregateindicatorsmaysimplyreflectthefactthattheydrawoncloselyrelatedindicatorsproducedbythesameorganization,ratherthantruepatternsintheunderlyingdimensionsofgovernance.Thisinturncomplicatestheconstructionofmarginsoferrorthatwouldbeessentialforinterpretingcross-countrydifferencesandover-timechangesinsuchanoverallaggregate

indicatorofgovernance.SeethediscussioninSection5.1belowformoredetailsonthisissue.

6

2.2SourcesofGovernanceData

The35WGIdatasourcesusedinthe2024updatearesummarized

inTabIe1,

withamoredetaiIeddescripⅥonofeachdatasourceinAnnex1.TheWGIdatasourcesfaIIintotwobroadcategories:surveysof?rmsandhousehoIds,andexpertassessments.TheformerincIudesseveraImajorsurveysof?rm

managers:theWorIdEconomicForum’sExecuⅥveOpinionSurveythatinformstheirannuaIGIobaI

CompeⅥⅥvenessReport,theInsⅥtuteforManagementandDeveIopment’sExecuⅥveOpinionSurvey

thatinformstheirannuaIWorIdCompeⅥⅥvenessReport,andtheWorIdBank’sEnterpriseSurvey

Program,whichhasrecentIybeengreatIyexpandedtoachievenear-gIobaIcoverageonarotaⅥngthree-yearcycIetoinformtheWorIdBank’snewBusinessReadyreportseries.TheWGIaIsoincIudedatafromseveraIregionaIhousehoIdsurveyprograms(Afrobarometer,LaⅥnobarometro,AmericasBarometer,andEuropeanQuaIityofGovernanceSurvey),asweIIasonegIobaIhousehoIdsurvey(theGaIIupWorIdPoII).InaddiⅥon,oneofourdatasources,theWorIdJusⅥceProjectRuIeofLawIndex,isahybridcombining

?ndingsfromanexpertassessmentwithacross-countryhousehoIdsurveywithnear-gIobaIcoverage.These?rmandhousehoIdsurveydatasourcesareparⅥcuIarIysaIientbecausetheydirectIycapture

surveyrespondents’percepⅥonsofthequaIityofgovernanceinthecountriesinwhichtheyIive.IntotaI,househoIdand?rmsurveysmakeup11ofthe35datasourcesintheWGI.

Theremaining24WGIdatasourcesareexpertassessments,relecⅥngthepercepⅥonsandviewsofexpertsa代IiatedwitharangeofNGO,privatesector,andgovernmentsectororganizaⅥons.AIIofthemshareabroadIysimiIarapproachinwhichexpertsempIoyedand/orrecruitedbytheorganizaⅥon

providenumericaIscoresofvariousdimensionsofgovernance,foIIowingastandardizedmethodoIogyandsetofde?niⅥonsdeveIopedbytheorganizaⅥonproducingthedatasource.Inmanycases,the

expertsprovidingtheassessmentsarebasedinthecountryorregiontheyassess.ThesedatasourcesbasedonexpertassessmentshaveadvantagesanddisadvantagesreIaⅥvetosurveys.OneadvantageisthattheyIendthemseIvesweIItocross-countrycomparisons,astheirmethodoIogiesareexpIicitIy

designedforthispurpose.ExpertassessmentscanaIsoprovidemoregranuIartechnicaIassessments,forexampIeonthequaIityofspeci?ctypesofpubIicinsⅥtuⅥons,thatwouIdbemoredi代cuItfora

typicaIhousehoIdor?rmsurveyrespondenttoprovideandinformedviewon.ExpertassessmentsaIsoareIessIikeIytobeaffectedbyrespondentreⅥcence,aconcerninhousehoIdand?rmsurveyswhere

7

respondentsmaybeunwillingtogivecandidresponsestosensitivequestionsaboutcorruptionorotherdimensionsofgovernance,particularlyincountrieswheregovernanceisweak.6

Ontheotherhand,ashortcomingofexpertassessmentsisthattheyreflecttheviewsofanarrowersetofrespondentsthanhouseholdorfirmsurveys.Italsoispossiblethattheratingsprovidedbyone

expertassessmenttosomeextentreflecttheviewsofotherexpertassessments,sothateach

assessmentdoesnotbringcompletelyindependentinformationontheunderlyinggovernanceconceptofinterest.Toguardagainstthis,wedonotuseexpertassessmentsthatareexplicitlybasedonotherexistingdatasources.7InSection5.1weprovidefurtheranalysisofthisissue.

TheWGIinclude12expertassessmentsproducedbyNGOs.Thedataprovidedbythese

organizationscoverarangeofspecifictopics,includingpressfreedoms(ReporterswithoutBorders),

politicalrightsandcivilliberties(FreedomHouse),humanrights(HumanRightsMeasurementInitiative),budgettransparency(OpenBudgetProject),andelectoralintegrity(AfricaElectoralIndex).Anumberofthesedatasourcescoverabroaderrangeoftopicsrelatingtogovernance,includingtheBertelsmann

TransformationIndexandtheVarietiesofDemocracyProject.Theorganizationsprovidingthedataaregeographicallydiverse,basedinEurope(VarietiesofDemocracy,ReporterswithBorders,BertelsmannFoundation),Africa(AfricaElectoralIndexandAfricaIntegrityIndicators),NewZealand(HumanRightsMeasurementInitiative),inadditiontotheUnitedStates(e.g.FreedomHouse,OpenBudgetProject,

WorldJusticeProject).

TheWGIalsoincludeeightdatasourcesprovideddirectlyorindirectlybypublicsector

organizations.Mostnotableamongthesearefourverysimilarexpertassessmentsofthequalityof

policiesandinstitutionsproducedbytheAfricanDevelopmentBank(CountryPolicyandInstitutional

Assessment),AsianDevelopmentBank(CountryPerformanceAssessment),WorldBank(CountryPolicyandInstitutionalAssessment),andtheInternationalFundforAgricultureandDevelopment(RuralSectorPerformanceAssessment).Theseassessmentsallcoverasimilarsetoftopicsrelatingtopolicyand

institutionalperformanceandarescoredbystaffoftheseorganizations,oftenlocatedinthecountriestheyassess,usingastructuredmethodologyandclearscoringbenchmarks.Allfourorganizationsuse

6SeeKraayandMurrell(2016)foradiscussionofrespondentreticenceandwaystomeasureit.

7Forexample,wedonotusedatafromtheTransparencyInternationalCorruptionPerceptionsIndexbecauseit

simplyisacompilationofexistingdatasources,inthesamewaythattheWGIare.Inaddition,wehaveovertimedroppedspecificquestionsandspecificdatasourcesfromtheWGIwhentheirmethodologieshavechangedtorelyexplicitlyonotherdatasources.ThisincludesselecteddimensionsoftheHeritageFoundation’sIndexofEconomicFreedom,aswellastheEuropeanBankforReconstructionandDevelopmentTransitionIndex.

8

theseassessmentstoallocateconcessionalresourcesacrosscountries.8Thiscategoryalsoincludestwodiscontinueddatasources,fromtheEuropeanBankforReconstructionandDevelopment,andthe

Frenchinternationaldevelopmentagency.Finally,twodatasourcesinthiscategoryarebasedondatareportedbytheUSDepartmentofState,onhumantrafficking(TraffickinginPersonsReport)andonhumanrightsviolations(ascodedbyacademicresearchersintheCingranelli-RichardsHumanRightsDataandthePoliticalTerrorScale).

Finally,fourWGIdatasourcescomefromcommercialbusinessinformationproviders.Threeof

themarelong-runningdataproductsprovidingsubscription-basedquantitativeratingsofabroadrangeofbusinessenvironmentandpoliticalrisks,producedbytheEconomistIntelligenceUnit(Country

Viewswire),PoliticalRiskServices(InternationalCountryRiskGuide)andS&PGlobal(CountryRisk

Service).ThefourthisCrisis24,aspecializedsecurityriskfirm,whichprovidesquantitativemeasuresofsecurityrisksfacingbusinesstravelersthatfeedsintothePoliticalInstabilityandAbsenceof

Violence/TerrorismmeasureintheWGI.Allfourorganizationsrelyonaglobalnetworkofexpertswhoprovidescoresfollowingastandardizedmethodologysubjecttocentralizedinternalreviewwithintheorganization.

ThesourcedatausedtoconstructtheWGIispubliclyavailableat

.

9Thedataareavailableintheiroriginalformasretrievedfromtheoriginalproviders.Wealsoreportthedatarescaledandreoriented(ifnecessary)torunfromzerotoonewithhighervaluescorrespondingtobettergovernanceoutcomes,usingtheminimumandmaximumpossiblescoresoneachvariabletodothis

rescaling.Therescaleddataiscomparablewithinagivendatasourceovertimeandacrosscountries.However,itisnotnecessarilycomparableacrossdifferentdatasources,becausedifferentdatasourcescoverdifferentsetsofcountrieswithdifferentunderlyingdistributionsofgovernance.Forexample,ascoreof6outof10mightmeansomethingdifferentinadatasourcethatcoversonlyadvanced

8Inaddition,theWorldBankandtheInternationalMonetaryFundusetheCountryPolicyandInstitutional

AssessmentdataasakeyindicatorintheirLow-IncomeCountryDebtSustainabilityFramework(LIC-DSF),

recognizingthatthereisastrongempiricalrelationshipbetweeninstitutionalqualityandtheriskofdebtservicingdifficulties(seeIMF(2017)).Inthesamevein,theInternationalMonetaryFundusescomponentsoftheWGIaskeymeasuresofinstitutionalqualityinitsDebtSustainabilityFrameworkforMarket-AccessCountriesbecause

theiranalysisshowsthattheWGIhasstatisticalpredictivepowerfordebtservicingdifficulties(seeIMF(2021)).

9Forasmallnumberofdatasourcesthatar