1

The European regional Human Development and Human Poverty Indices

n° 02/2011

1.Introduction

ThisRegionalFocusanalysestheregionalHumanDevelopmentIndex(HDI)andHumanPoverty Index(HPI)aspublished intheFifthCohesionReport(EC2010).TheyarebasedonaslightvariationofthemethodologydevelopedbytheUnitedNationsDevelopmentProgramme(UNDP).

ThisRegionalFocusshowsthatahighlevelofhumandevelopmentisnoguaranteeofalowlevelofhumanpovertyorviceversa.Several regions in the UK, Spain, France, Belgium and Italycombineahighlevelofhumandevelopmentwithahighlevelofhumanpoverty.Estonia,Slovenia,theCzechRepublic,SlovakiaandPolandshowthatalowlevelofhumandevelopmentcanstillbecombinedwithalowlevelofhumanpoverty.ThethreeNordic Member States as well as Germany, Austria and theNetherlandsaremostsuccessfulatcombiningahighlevelofhumandevelopmentwithalowlevelofhumanpoverty.ManyregionsinPortugal,Spain,Italy,Greece,Romania,BulgariaandHungaryscoredpoorlyonbothindices(seeMap4).

Thispaperisstructuredasfollows:first,thedescriptionoftheUnitedNations(UN)nationalHDIandHPIindicators;second,theresultsoftheapplicationtoEuropeanregions;third,theconclusionswhichcanbedrawnfromthisanalysis.Thedetailedmethodologyandindicatordefinitionisincludedintheannex.

The data can be downloaded here: https://circabc.europa.eu/d/a/workspace/SpacesStore/16b1d1b9-5f7b-4a3b-b60a-6fa35187bed5/hdi_hpi.xls

2.TheUnitedNationsDevelopmentProgrammeApproach

2.1. Human Development Index

Since1990,theUnitedNationsDevelopmentProgramme(UNDP)hascalculatedtheHDIandincludeditinitsHumanDevelopmentReport.Itmeasurestheaverageachievementsinthreebasicdimensions:alongandhealthylife,accesstoknowledgeandadecentstandardofliving.ItunderlinesthemultidimensionalnatureofdevelopmentinthepolicydebatebygoingbeyondthetraditionaleconomicperspectivebasedonGDPorincome(SagarandNajam1998).TheHDIemphasisesthataspectsotherthaneconomicactivitiesandtheirgrowth(namelyGDPanditsgrowthrate)areimportantfordevelopment,includinglifeexpectancy,literacyandenrolmentrates.

Thisapproacharguesthatincome,commoditiesandwealtharemeanstoanend.Theydonotconstituteadirectmeasureofthelivingstandarditself.Developmentshouldbenefitpeople,andthereforeshouldconsiderthelifethatpeoplelead:theirachievements,freedomsandcapabilities(AnandandSen1994).GDPperheadfailstocapturesomebasicfeaturesofpeople'sstandardsofliving,aswellasthequalityandquantityofpublicandpublicly-providedgoodsandamenities.

However,lookingatHDIsinEuropefromaglobalperspectiverevealslittlevariationinEurope.Forexample,in2009,HDIs1forallEUMemberStateswereclassifiedas'highorveryhighhumandevelopment'2.Furthermore,withintheEU,theHDIishighlycorrelatedwithGDPperhead;primarilybecauseliteracy,enrolmentandlifeexpectancyareallhighcomparedtotherestoftheworld.

AseriesofshortpapersonregionalresearchandindicatorsproducedbytheDirectorate-GeneralforRegionalPolicy

By Rocco L. Bubbico and Lewis Dijkstra

1 Based on 2007 data.2 With the new 2010 methodology, all EU Member States are classified as 'very high human development' with the exception of Bulgaria ('high human development').

Regional FocusRegional Focus

Contents1. Introduction .........................................................................................................1

2. The United Nations Development Programme Approach ...............................................................................................................1

3. Regional Human Development and Poverty in Europe ..........3

4. Conclusions ..........................................................................................................8

5. Methodology ......................................................................................................8

References ..........................................................................................................10

2

3 For instance, the 2010 version introduces relevant changes in the calculation of HDI.

TheHDIhassparkedadebateontheappropriatemethodology,thelimitationsandsignificanceofasyntheticindicator,andthelackofrelevantdataanditstechnicalqualities(Noorbakhsh1998).Yet,theHDIhasbeenacceptedasarelevantindicatorofhumandevelopment,inpartthankstoconstantmethodologicalrefinements3.Nevertheless,asingleindexcannotsynthesisethefullspectrumofhumandevelopmentissues(Streeten1994),whichiswhyanothermeasuresuchastheHumanPovertyIndexwasadded.

2.2. Human Poverty Index

The1997HumanDevelopmentReportintroducedanindexofpoverty(HumanPovertyIndex-HPI).WhiletheHDIfocusesontheaverageofthreedimensionsofwell-being,theHPItargetsthedistributionofdimensionsofqualityoflife.Itcapturesthedisparitiesinasocietybyfocussingonpeoplewithalowerlifeexpectancy,lowerincome,loweducationandthelong-termunemployed.Furthermore,theformulausedtoaggregatethesefourissuespenalisescountrieswithhighvaluesforonedimension.Ifasimpleaveragewasused,acountrywith5%foreachofthefourdimensionswouldscorethesameasacountrywith37%forlowlifeexpectancy,but1%forthethreeotherdimensions.Theformulausedheregivehigherweightstoextremescores,givingthecountrywith5%forallfourdimensionsabetterscoreforhumanpovertythanthecountrywithanextremevalueforonedimension.

Map 1 - 2009 Human Development Index

3

3.RegionalHumanDevelopmentandPovertyinEurope

DespitetheveryhighHDIscoresinEurope,thereissignificantvariationbetweenEUcountriesandregionsintermsofhumandevelopment and poverty (see EC 2010). For instance, loweducationattainmentinEuropeanregionsrangesfrom3.3%to81.4%4;healthylifeexpectancyrangesbetween52and78years.TogainabetterperspectiveofregionaldisparitieswithintheEU,anEUregionalHDIandHPIwascalculatedwithamodifiedsetofindicators(seeannex).

3.1. EU regional human development

TheregionswithahighHDIareconcentratedinsouthernEngland,southernGermany,theNetherlands,ScotlandandSweden(seeMap1).AllFrenchregions,exceptPicardieandCorse,areabovetheEUaverage.MostEU-12regionsandthoseinPortugal,GreeceandItalyhaveanHDIbelowtheEUaverage,exceptPraha(CZ),Attiki(EL)andfourItalianregions.InSpain,eightregionsarebelowandeightabovetheEUaverage,withhighscoresinMadrid,NavarraandPaisVasco.

Thetop10HDIregionsincludefiveEnglishregions,thecapitalcityregionsofSwedenandFrance,andtworegionssurroundingBrussels(seeMap1).AlthoughtheEUregionalHDIiscorrelatedwithGDP,onlyafewregionsregisterhighlevelsofbothGDPperheadandHDI.Ofthetop10regions,onlythreeappearinthetop10basedonGDPperhead.ThefirstinbothrankingsofHDIandGDPpercapitaisInnerLondon.However,thesecondregionintheHDIranking(Surrey,EastandWestSussex,UK)isonly45thintherankingofGDPpercapitainEurope.Similarly,OuterLondonis8thforHDIand95thforGDP.

3.2. EU human poverty

Thehighestlevelsofhumanpoverty(HPI-2)areregisteredinsouthernEurope,inparticularinPortugal,Spain,Italy,GreeceandMalta(seeMap2).Thelowestlevelsofhumanpovertycanbefoundinhighly,moderatelyandlessdevelopedMemberStates–inparticularintheCzechRepublic,Sweden,Germany,SloveniaandSlovakia.Humanpovertyiscloselycorrelatedwithloweducationattainment,whichisthemostcommonsourceofalowHPI-2score.HPI-2isalsopositivelycorrelatedwithat-risk-of-povertyincome.

Thetop10HPI-2regionsincludesixPortugueseregions,Malta,twoSpanishregions(CeutaandExtremadura)andoneItalianregion (Campania). All these regions, except Campania, arealsointhetop10intermsofloweducationattainment.Ceuta,ExtremaduraandCampaniaarethetopthreeintermsofat-risk-of-povertyincome.

The10regionswiththelowestlevelofHPI-2includefiveCzechregions,threeSwedishregions,oneAustrianandoneGerman.Fouroftheseregionsareamongthosewiththe10lowestat-risk-of-povertyincomerates,whilethreeregionsareamongthosewiththe10lowestsharesofpeoplewithaloweducationattainment.

Long-termunemploymenthas,instead,nocorrelationwiththeperformanceinHPI-2.The10regionswiththehighestlong-termunemploymentconsistoftwoSlovakregions,sixGermanregions,BrusselsandCeuta(ES).Onlyoneoftheseregions(Ceuta)isalsoamongthe10regionswiththehighestHPI-2.

RegionsperformingsignificantlybetterinGDPperheadthaninHDIarecapitalregionssuchasPraha(CZ),Bratislava(SK)andBrussels(BE).Thismaybepartiallyduetocommuting,whichinflatesGDPperheadfigures.Ifthemoreaffluentprefertoliveoutsidethecapitalregion,thiswouldalsoexplainapartofthegapbetweenGDPperhead(basedonwherepeoplework)andnetadjustedhouseholdincome(basedonwherepeoplelive).

Thebottom10regionsforHDIcomprisesevenregionsinRomania,twoinHungaryandoneinBulgaria.Halfofthemarealsointhebottom10regionsintermsofGDPperhead.

4 Population aged 25-64, year 2007.

4

Map 2 - Human Development Index at NUTS 2 level

5

Map 3 - Human Poverty Index-2 at NUTS 2 level

6

3.3. Combining human development and poverty scores

TheregionalHPIandHDIareonlyweaklycorrelated.Onlyafewregionsscoreparticularlywellonboth,suchasStockholm,orpoorlyonboth,suchasAçores.Alargenumberofregionscombineahighscoreononeindexandalowoneontheother.SomeexamplesareBrussels,Luxembourg,NavarraandPaisVasco,wheretheHumanPovertyIndexismuchhigherthantheHumanDevelopmentIndexwouldimply.

Figure1reportsthedistributionofregionsbyHDIandHPI-2,showingtwocleardifferentiatedroadstotheidealsituationofhighdevelopmentandlowpoverty:a'highpovertyroad',withrelativelyhighHPI-2incomparisontoHDIlevels,anda'lowpovertyroad',withrelativelylowHPI-2incomparisontoHDI.Aclusteranalysishasbeenconductedtoidentifygroupswithcommoncharacteristicsanddifferentlevelsofpovertyanddevelopment5,alsomappedinmap4.

Map 4 - Regional HDI and HPI-2

7

ThegroupofhighdevelopmentandlowpovertyincludesarelativelylargenumberofregionslocatedinparticularinGermany,Austria,FinlandandSweden.Also,alargenumberofregionsintheUKandtheNetherlandsandafewinBelgium,DenmarkandFranceachievedthetargetoflowpovertyandhighdevelopment.

EasternEuropeanregionsarelocatedinthe'lowpovertyroad'todevelopment.Alargegroupofregionsregistermedium-lowhumandevelopment(lowHDI)andlowlevelsofhumanpoverty(lowHPI-2).Onaverage,theseregionsscorerelativelywellinbasiceducation,povertyandunemploymentbutregisterunsolvedchallengesinindicatorssuchaseconomicwealthandspecialisedhumancapital.ThisisthecaseofregionslocatedintheCzechRepublic,BalticStates,Poland,SlovakiaandSlovenia.OtherEasternEuropeanregionsinHungary,RomaniaandBulgariaalsoregisterlowlevelsofpovertymeasuresbutscoreverypoorlyintermsofhumandevelopment.

Incomparison,regionslocatedinNorthernSpainandNorthernItalyareonthe'highpovertyroad'todevelopment,achievinggoodlevelsofhumandevelopment(highHDI)butregisteringconsiderablelevelsofhumanpoverty(relativelyhighHPI-2).Theseregionsarethereforewelldeveloped,buttheirinternalinequalitiesareratherhigh.Asaconsequence,thechallengefortheseregionsistoincreasethebasiceducationattainmentandhealthcarequality,andtofightpovertyalsothroughemploymentpolicies.Toalesserextent,thisisthecaseoftheregionslocatedinIreland,France,areasoftheUKandDenmarkwhichfeaturemediumpovertyandhighdevelopment.Finally,southernEuroperegions,withtheexceptionofCyprus,registermedium-lowhumandevelopment(lowHDI)andatthesametimehighlevelsofhumanpoverty(highHPI-2).Thismeansthattheseregionshavenotachievedhighdevelopmentandarecharacterisedbyrelevantinternaldisparities,scoringrelativelypoorlyintermsofbasicwell-beingmeasures(at-risk-of-povertyincome,unemploymentandbasiceducation)anddevelopmentmeasures(economicwealth,highereducation,healthylife).TheregionswiththesecharacteristicsarelocatedmainlyinSpain,southernareasofItalyandSpain,Greeceand,toalargerextent,Portugal.

Figure 1 - Regional HPI-2 and Regional HDI

5 With k-means methodology.

8

OneofthemainreasonsforthedifferenceinthetwoindicesisthattheHDIisbasedonapercapitaaverageofanabsolutemeasureofincome(netadjustedhouseholdincome),whileHPI-2includesarelativemeasure(theportionofpopulationbelowthenationalpovertythreshold).Accordingly,aregionwithanunequaldistributionofahighlevelofincomecanhavebothahighaveragelevelofhumandevelopmentandahighlevelofpoverty.Aregionwithlowincome,butrelativelyequaldistributionofit,willhavealowHDIandalowHPI-2.

The increases in average income in the five less developedcountriesdid,infact,leadtohigherlevelsoflifesatisfactionandhappiness,despitetheat-risk-of-povertyrateremainingunchanged.Itcouldbeargued,therefore,thatimprovingwell-being,especiallyinlessdevelopedMemberStatesdependsonimprovingthefactorsbehindtheHDIandotherabsolutemeasuresofwell-being.Relativemeasuresofpovertyaddnuanceandcanguidepolicychoicesinsituationswherecircumstancesaresimilar.Forexample,inregionswithsimilarlevelsofHDI,averagewell-beingislikelytobehigherintheregionwithalowerHPI-2.Relativemeasures,however,aredifficulttocompareinradicallydifferentsituations.Forexample,StockholmandBratislavahaveaverysimilarHPI-2,yetresidentsinStockholmreportbeingmuchmoresatisfiedwiththeirlifeandhappierthanthoseinBratislava(EC2010).

4.Conclusions

ThispaperhashighlightedthewidevarietyofhumandevelopmentwithinEuropeanditsregions.Fromaglobalperspective,Europeappearstobeuniformlyhighlydeveloped.Bynarrowingdowntheanalysis,awidevarietyofhumandevelopmentlevelsemergeacrossandwithinEuropeancountries,targetingbothelementsforoverallwell-being(HDI)andthedistributionoftheseelementsacrosssociety(HPI-2).Fromapolicyperspective,theanalysisshowsthatmostEuropeanregionsfacechallengesintheareaofhumandevelopmentorpoverty.Inlessdevelopedregions,improvementsintheHDIcanhaveastrongimpactonwell-being,whileinthemoredevelopedregionsareductioninHPI-2andinequalitiesismorelikelytoincreasetheoverallwell-being.

5.Methodology

5.1. UNDP methodology for Human Development Index (HDI)

The 2009 methodology for HDI is to take the average ofthreenormalisedindices,oneineach'dimension'ofhumandevelopment.Theseindices(Life,EducationandIncome)measurethe achievements in each area, considering the followingindicators:

• Lifeexpectancyatbirth;

• Knowledgeandeducation(combiningadultliteracyrateandtotalgrossenrolmentratio);

• GDPpercapita(PPPUS$).

Theformulausedisthefollowing:

(1)

Wherethesub-indexofeducationisthecombinationofgrossenrolmentandliteracyrates,asshowninthefollowingformula:

= 1/3 ( ) + 2/3 ( ) (2)

Eachsub-index(I)isnormalisedtakingintoaccounttheminimumandmaximumvalueobserved:

(3)

TheincomeindicatorofGDPperheadistransformedthroughthenaturallogarithmoftheactualminimumandmaximumvaluesused.

Theindexhasavaluebetween0and100,where0isequaltolowlevelsofhumandevelopmentand100tohighlevelsofhumandevelopment.

5.2. EU methodology for regional HDI

Thisindexisthemeanofthenormaliseddimensionsub-indiceswhicharecalculatedthroughtheformula(6)andiscalculatedinthesamewayastheUNDPHDI:

(6)

TheindicatorsconsideredfortheregionalHDIare:yearsofhealthylifeexpectancy;netadjusteddisposablehouseholdincomepercapita6(asanindexofEU-27average);lowandhigheducationattainmentforpeopleaged25–64(%ofpopulation25-64withlowand%withhigheducationattainment). Inparticular, iscombinedasshownbelow.

= 1/3 (1- ) + 2/3 ( ) (7)

AllindicatorsareavailableatNUTS2levelfor2007andpublishedbyEurostatoraretheDirectorate-Generalforregionalpolicy(DGREGIO)estimatesbasedonEurostatdata.Thetwomain

6 These figures have been estimated using regional disposable household income figures and the difference between disposable household income and net adjusted disposable household income at the national level where available. For Romania, the difference between disposable and net adjusted disposable household income was not available. For Cyprus, Malta and Luxembourg, household income estimates from the 2nd European Quality of Life Survey have been used.

9

differenceswiththeUNDPindexare(1)theuseofthelowandhigheducationattainmentinsteadofliteracy,sincethelatterisnotavailableattheregionalleveland(2)theuseofnetadjustedhousehold income in the purchasing power consumptionstandard(PPCS)insteadofGDPinPPP.TheuseofnetadjustedhouseholdincomeinsteadofGDPperheadinPPSbringstheindicatorclosertothehouseholdperspectiveasadvocatedbytheStiglitz-Sen-Fitoussireport(2009).AsthedifferencesinnetadjustedhouseholdincomeintheEUareconsiderablysmallerthanGDPperheadworldwide,normalisationwithoutanaturallogarithmwasused.

Inordertoobtainvaluesbetween0and100,normalisationiscarriedoutfollowingtheformula:

(8)

5.3. UNDP methodology for Human Poverty Index 2 (HPI-2)

Theindexiscalculatedusingthefollowingformula7:

(9)

First, relatestosurvivalprobabilityandthelikelihoodofnotsurvivingtoage60.Second, correspondstobeing excluded from reading and communication, and ismeasuredbytheadultilliteracyrate.Finally, ismeasuredbythepercentageofthepopulationbelowtheincomepoverty line (50%ofmedianhouseholddisposableincome).Socialexclusionisthefourthdimensionoftheindex( LONG T UNEMPLOYMENT ) and is measured as the long-termunemploymentrate(over12months).Theindexisexpressedasapercentage,wherehigherpercentagesequalhigherlevelsofhumanpoverty.Inthe2010report,theHPIhasbeenreplacedbytheMultidimensionalPovertyIndex(MPI)whichincludesalargersetofstandardoflivingindicators.

5.4. EU methodology for regional HPI-2

TodeveloptheregionalHPI,thestartingpointhasbeentoadaptthemethodologybehindtheindexofdevelopedcountries(HPI-2).Inthiscase,thefollowingindicatorshavebeenconsidered:

• Probabilityofnotlivingto65atbirth( );

• Shareofpopulationaged25-64withloweducationattainment( );

• Shareofpopulationwithanat-risk-of-povertyincome,i.e.60% below the national median income after transfers( );

• Long-term unemployed as a share of the labour force( ).

AlltheindicatorsareavailableatNUTS2levelfortheyear2007andpublishedbyEurostat8.AlsointhiscasethedifferenceswiththeUNmethodologyarerelatedtothechoiceofavailableormoremeaningfulindicatorswithintheEUcontext.Inparticular,the

differencesare:ahigherthresholdforthelongevityprobabilityrate(65yearsinsteadof60);adifferentindicatorforloweducation(lowattainmentinsteadofilliteracy);andadifferentpovertyline(theat-risk-of-povertyincomeisequalto60%ofthenationalmedianinsteadof50%usedintheUNDPmethodology).TheformulaisidenticaltotheHPI-2formula(9).

Toobtainvaluesbetween0and100,normalisationiscarriedoutfollowingthesameformulausedfortheHDI(8).

Acknowledgements

The authors would like to thank Beatriz Torighelli, AngelaWinkelhorst,HugoPoelmanandCécileHuletfortheirsupportinthepreparationofthispaper.

References

Anand,S.,Sen,A.K.,Human Development Index: Methodology and Measurement,OccasionalPapers–HumanDevelopmentReportOffice,1994

Fifth report on economic, social and territorial cohesion,EuropeanCommission, 2010, available online at http://ec.europa.eu/regional_policy/sources/docoffic/official/reports/cohesion5/index_en.cfm

Sagar,A.D.,Najam,A.,'TheHumanDevelopmentIndex:acriticalreview',EcologicalEconomics,Volume25,Issue3,June1998,pp.249-264

Stiglitz,J.,Sen,A.K.,Fitoussi,J.,Report of the Commission on the measurement of economic performance and social progress,2009,availableonlineathttp://www.stiglitz-sen-fitoussi.fr/en/index.htm

Streeten,P., 'HumanDevelopment:Meansandends',American Economic Review,PapersandProceedings84,1994,pp.2132-2237

Human Development Report 1997,UNDP,1997,availableonlineathttp://hdr.undp.org/en/reports/global/hdr1997/

Human Development Report 2010,UNDP,2010,availableonlineathttp://hdr.undp.org/en/reports/global/hdr2010/chapters/en/

7 Calculating HPI is less complex than calculating HDI, since the indicators used are already expressed as percentages, so it is not necessary to create dimension indices for normalisation as for the HDI.8 The data is available for all NUTS 2 except French outermost regions;

Editor:EricVONBRESKA©EuropeanCommission,RegionalPolicyThetextsofthispublicationdonotbindtheCommission

Any questions, comments or contributions should be sent to the following address:

regio-papers@ec.europa.eu

Consult this website for further information:http://ec.europa.eu/regional_policy/index_en.htm