Namibia Poverty Mapping Macroeconomic Planning Department

REPUBLIC OF NAMIBIANATIONAL PLANNING COMMISSION

Namibia Poverty Mapping

Macroeconomic Planning Department

National Planning Commission | Namibia Index of Multiple Deprivation2

National Planning CommissionGovernment Office ParkLuther StreetPrivate Bag 13356WindhoekTel.: +264 61 283 4111Website: www.npc.gov.na

Republic of Namibia

National Planning Commission | Namibia Index of Multiple Deprivation 3

TABLE OF CONTENTS

PREFACE ......................................................................................................................8

1 INTRODUCTION ...................................................................................................................9

1.1 About Namibia and Overview of Development Challenges .......................................9

1.2 Understanding Poverty ..............................................................................................9

1.3 IntroductiontoPovertyMappinginNamibia ..........................................................10

2 METHODOLOGY ................................................................................................................11

3 POVERTY PATTERNS AND TRENDS ....................................................................................12

3.1 RegionalPovertyPatternsandTrends .....................................................................12

3.2 ConstituencyPovertyPatternsandTrends ..............................................................17

3.2.1 Overview of constituency poverty patterns and profiles ............................17

3.2.2 Zambezi region ...........................................................................................18

3.3.3 Erongo region .............................................................................................21

3.3.4 Hardap region ............................................................................................24

3.3.5 Karas region ...............................................................................................26

3.3.6 Kavango region ..........................................................................................28

3.3.7 Khomas region ............................................................................................31

3.3.8 Kunene region ............................................................................................34

3.3.9 Ohangwena region .....................................................................................36

3.3.10 Omaheke region .........................................................................................38

3.3.11 Omusati region ...........................................................................................40

3.3.12 Oshana region ............................................................................................43

3.3.13 Oshikoto region ..........................................................................................46

3.3.14 Otjozondjupa region ...................................................................................48

4 CONCLUSIONS AND RECCOMENDATIONS ........................................................................51

4.1 Conclusions ..............................................................................................................51

4.2 PolicyRecommendations .........................................................................................52

References ....................................................................................................................53

Annex 1: Detailed Headcount Poverty using the Upper Poverty Line 2001 - 2011 ...54 Annex2: TechnicalNotes ..........................................................................................59


LIST OF TABLES

Table 1: TrendsinPovertyHeadcountRate,2001-2011(upperboundpovertyline) .........12

Table 2: TrendsinPovertyHeadcountRate,2001-2011(lowerboundpovertyline) .........16

Table 3: Proportionofconstituencieswithmorethan30%and50%ofthe populationclassifiedaspoor(upperboundpovertyline),2011 ..........................18

Table 4: ZambeziRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................19

Table 5: Zambezi Region Poverty Headcount Rate Scores and Values (lowerboundpovertyline) .....................................................................................20

Table 6: Erongo Region PovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................21

Table 7: Erongo Region PovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................23

Table 8: HardapRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................24

Table 9: HardapRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................26

Table 10: KarasRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................26

Table 11: KarasRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................28

Table 12: KavangoRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................29

Table 13: KavangoRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................30

Table 14: KhomasRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................31

Table 15: KhomasRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................33

Table 16: KuneneRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................34

Table 17: KuneneRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................36


Table 18: OhangwenaRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................36

Table 19: OhangwenaRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................38

Table 20: OmahekeRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................39

Table 21: OmahekeRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................40

Table 22: OmusatiRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................41

Table 23: OmusatiRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................43

Table 24: OshanaRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................43

Table 25: OshanaRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................45

Table 26: OshikotoRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................46

Table 27: OshikotoRegionPovertyHeadcountRateScoresandValues,2001-2011 (lowerboundpovertyline) .....................................................................................48

Table 28: OtjozondjupaRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................49

Table 29: OtjozondjupaRegionPovertyHeadcountRateScoresandValues,2001-2011 (upperboundpovertyline) .....................................................................................50


LIST OF MAPS

Map 1: NamibiaPovertyHeadcountRate,2011(upperboundpovertyline) ................... 13

Map 2: ChangeinNamibiaPovertyHeadcountRate,2011-2001 (upperboundpovertyline,percentagepoints) ..................................................... 15

Map 3: Zambezi Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 19

Map 4: ZambeziRegionChangeinPovertyHeadcountRate,2001-2011 (upperboundpovertyline) ................................................................................... 20

Map 5: Erongo Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 22

Map 6: ErongoRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 23

Map 7: Hardap Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 25

Map 8: HardapRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 25

Map 9: Karas Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 27

Map 10: KarasRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 27

Map 11: Kavango Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 29

Map 12: KavangoRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 30

Map 13: Khomas Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 32

Map 14: KhomasRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 33

Map 15: Kunene Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 35

Map 16: KuneneRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline). .................................................................................. 35

Map 17: Ohangwena Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 37


Map 18: OhangwenaRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 38

Map 19: Omaheke Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 39

Map 20: OmahekeRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 40

Map 21: OmusatiRegionPovertyHeadcountRatein2011 (upperboundpovertyline) ................................................................................... 41

Map 22: OmusatiRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 42

Map 23: Oshana Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 44

Map 24: OshanaRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 45

Map 25: Oshikoto Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 46

Map 26: OshikotoRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 47

Map 27: Otjozondjupa Region Poverty Headcount Rate in 2011 (upperboundpovertyline) ................................................................................... 49

Map 28: OtjozondjupaRegionChangeinPovertyHeadcountRate,2011-2001 (upperboundpovertyline) ................................................................................... 50

LIST OF FIGURES

Figure 1: NationalPovertyHeadcountShares,2011(upperboundpovertyline) 17


PREFACE


1. INTRODUCTION

1.1 About Namibia and Overview of Development Challenges

Namibiahasapopulationof2113077people,57percentofwhom live in ruralareas.Overthe2001to2011period,thepopulationgrowthratedeclinedfrom2.6percentperannumto1.4percent,whilethefertilityratedeclinedfrom4.1childrenperwomanto3.6childrenperwoman.Namibia is classified as an uppermiddle income country,with an estimated annualGrossNationalIncome(GNI)percapitaofUS$5693.Sixtyfivepercentofthetotalpopulationfallswithintheagecategory15yearsandabove.Ofthese,71percentcomprisethelabourforce,withtheunemploymentrateestimatedat29.6percentofthetotallabourforce.

Sinceindependence,theGovernmentoftheRepublicofNamibiahasconsistentlyformulatedpolicies and programmes to address developmental challenges. The current fourth NationalDevelopmentPlan(NDP4)outlinesthedevelopmentobjectivesandpriorityprogrammestobeimplementedover thefiscal period 2012/13 to 2016/17. The threeoverarching goals of theNDP4aretoachievehighandsustainedeconomicgrowth,employmentcreation,andincreasedincomeequality.TheGovernmentisalsocommittedtoachievingtheMillenniumDevelopmentGoals(MDGs)andotherinternationaldevelopmentgoalsandobjectives,suchastheSustainableDevelopmentGoals(SDGs)currentlyunderdiscussionsattheintergovernmentallevel.

Acoreobjectiveofthenationalpolicyformulationandplanningprocess,andconsonancewiththeaspirationofachievingtheMDGs,iseradicationofpoverty.In1998,theGovernmentadoptedthePovertyReductionStrategyanditsActionPlan,whilemorerecentlyin2012,theNationalRuralDevelopmentPolicywasalsoadopted.Theaimofthispolicyistopromotesystematicandcoordinated development planning, and respond to the plethora of development challengesfacing rural populations. The central objective of the Rural Development Policy, which wasdeveloped in furtheranceof theDecentralisationPolicy, is topromoteservicedeliverywithinthedecentralisedlevelsofgovernance–regionsandconstituencies.Todriveeconomicgrowthand, importantly, create jobs and thus address poverty, the Government has prioritised theagricultural,education,healthandhousingsectorsforpublicinvestmentssinceindependence.

1.2 Understanding Poverty

Povertyisamultidimensionalconceptrelatingtoalackofresourceswithwhichtoacquireasetofbasicgoodsandservices.Conceptually,povertycanbeviewedasastateofdeprivationandcanbedefinedinbothabsoluteandrelativeterms.Absolutepovertycanbeseenastheinabilitytoaffordcertainbasicgoodsandservices.Delineationofthoselivinginabsolutepoverty,therefore,aims to determine the number of people living below a certain income threshold or the number ofhouseholdsunabletoaffordbasicgoodsandservices. Ineverycountry,thepoverty line issettomeasurepovertyinaccordancewiththeexpectationofthecostofmeetingbasichumanneeds.


Relativepoverty,ontheotherhand,referstoastandardoflivingthatisdefinedintermsoftheexpectationsofthewidersocietyinwhichanindividuallives,andisacomparativemeasureofpoverty.Thusanindividualmaybenon-poorinabsolutetermsbutmaystillbeconsideredpoorrelativetoothermembersofhisorhersociety.

Thepovertylinesbasedonthesamplesurveywerederivedusingtheabsolutepovertymeasure,basedontheestimatesofcostofbasicneedsasadoptedbyNamibiaStatisticsAgency(thenCensusBureauofStatistics)in2004.Atwostageestimationprocesswasadoptedforderivingthispovertyline.First,estimatesofthecostofbasicfoodneeds,thatis,thecostofanutritionalbasketoffoodconsideredminimalforthehealthysurvivalofatypicalhousehold,wasusedtodefinealowerboundor‘severe’povertyline.Second,anestimateofthevalueofabundleofnon-food items consistent with the spending of the poor was added to the lower bound or food poverty/severepovertytodeterminetheupperboundpovertyline.

Thus,povertyisdefinedasthepercentageofpeopleinaspecificareawhoseannualperadultequivalentconsumptionisbelowthepovertyline.In2003/2004thepovertylinesofannualisedperadultequivalentexpenditurewere:lowerbound-N$2217.72andupperbound-N$3149.40.In 2010 the poverty line of annualised per adult equivalent expenditure, after adjusting forinflation,were:lowerbound-N$3330.48andupperbound-N$4535.52.Whentheannualperadultequivalentconsumptionisbelowtheupperboundpovertyline,anindividualisconsideredtobepoor,andwhenitisbelowthelowerboundpovertylinetheindividualisconsideredtobeseverelypoor.Thepovertyheadcount(incidenceofpoverty)istheproportionofthepopulationwhoseconsumptionisbelowthepovertyline.

1.3 Introduction to Poverty Mapping in Namibia

The present report presents the results of povertymapping inNamibia. Povertymapping isconsidered importantbecause itprovidesadetaileddescriptionof thespatialdistributionoftrends inpovertyat regionalandconstituency levels. This report combines the2003/04and2009/10NamibiaHouseholdIncomeandExpenditureSurvey(NHIES)data,andthe2001and2011NamibiaPopulationandHousingCensusdata,with theobjectiveofestimatingpovertylevelsforthethirteenregionsand107constituenciesofNamibia.Inthepast,povertyestimateshave been done using the NHIES data alone.

However,duetothelowstatisticalpowerresultingfromthesmallsamplesizeassociatedwithsuch surveys (approximately 10 000 households only), it has not been possible to estimatepovertymeasuresatconstituencylevelinNamibiaandearlierestimateshaveonlybeendoneatregionallevels.UsingeconometrictechniquesthatcombinetheNHIESandCensusdata,thestudyprovidespovertymeasuresatregionalandconstituencylevelsattwotimepoints–2001and2011–theyearsinwhichNamibiaHousingandPopulationCensuseswereconducted.Thustheresultsarebasedontheentirepopulationwithoutasamplebias.Themajorlimitationofthestudy,however,isthattheestimationprocessisbasedonthegeneralisedassumptionthatcharacteristicsofpoorindividualsorhouseholdsinthesamplesurvey(NHIES)definethepoorindividualsand/orhouseholdsintheentirepopulation.


Thestudycoversthethirteenregionsand107constituentsthatwere inexistencebeforetherecent boundary reviews by the Delimitation Commission. These are the geographic areasthat formed the basis of both the surveys and censuses. For consistency in the applicationand interpretation of data, and especially in order to avoid any possiblemisapplication andmisinterpretationofdata,itwasdeemednecessarytoanalyseandpresentdataonthebasisofthe boundaries that existed during the surveys and censuses.

Thepurposeofthisexerciseistoprovideanadditionalbodyofdataandinformationonpovertydynamics inNamibia. The added value of the present exercise is that the analysis has beenundertakenandresultspresentedformuchsmallergeographicunits–constituencies.Althoughan attempthasbeenmade to identify thepossible causesof andexplanations for observedpovertytrends,thisisbynomeansexhaustiveandfurtheranalysiswillberequiredtodeepentheunderstandingofthecausesofpoverty inNamibiaatnational,regionalandconstituencylevels.Afterthisintroduction,Section2describesthemethodologyappliedinthisstudy,Section3elaboratesonthefindingsofthestudy,whileSection4drawssomeconclusionsandpolicyrecommendations.

2 METHODOLOGY

ThisreportpresentstheincidenceofpovertyinNamibiaattheconstituencylevel.Ideallythisshouldbedoneusingasingledataset.However,todosowouldrequireadatasetthatnotonlycontainsenoughhouseholdinformation,butalsohasenoughobservationsforeachconstituencyto allow for the accurate measurement of poverty at a local level. No such dataset currently existsinNamibia.Infact,veryfewcountriesintheworldhavedetailedhouseholdsurveyswithsuchlargesamplesthataccurateestimatesofpovertycanbedeterminedforgeographicareaswithsmallpopulations.

Instead the report combines two sources of data: the Namibia Household Income and ExpenditureSurvey(NHIES)andtheNamibiaPopulationandHousingCensus.The2003/04and2009/10NHIESdatasetscontainaccurateincomeandexpendituredata,buttoofewhouseholdsare sampled in each constituency for poverty estimates at constituency level. The 2001 and2011Censusescontainnoincomeorexpendituredata,buthaveampleobservations.SincetheCensusdoesnotcontainanyexpenditureinformation,theperadultexpenditurelevelforeachhouseholdwasestimatedusingapovertymappingmodel.Amoretechnicalexplanationofthemethodology followed isprovided for specialist readers in theAnnex2,while theparagraphbelow provides a broad overview.

ThemodelfollowstheimputationapproachofElbersetal.(2003).Theseauthorssuggest,first,choosing a set of household characteristics found in both datasets. Next, using the smallerdatasetthathasaccurateexpendituredata(theNHIESinthiscase),itispossibletoderivetherelationshipbetweenthechosensetofhouseholdcharacteristicsandhouseholdexpenditure.ThisrelationshipcanbeusedtopredicttheexpectedlevelofexpenditureforeachhouseholdintheCensus,sincethesamesetofhouseholdcharacteristicsispresentintheCensus.


Not every household with the same characteristics will have exactly the same expenditurelevel.Forthatreason,themodelalsogeneratesasetofexpecteddeviationsfromtheaveragethrough a Monte Carlo process that also considers that households in the same survey cluster aresomewhatmorealikethanotherhouseholds.Averagepovertyratesarethenestimatedforeachconstituency.

3 POVERTY PATTERNS AND TRENDS

Inthissection,theresultsofthepovertymappingexercisearepresented.Foreachregion,themajordefiningcharacteristics intermsofgeographicarea,populationsizeanddensity,majorphysicalfeatures,andresourceendowmentsareoutlined.Thisisfollowedbyadiscussionofthespatialdistributionofpovertytrendsoverthe2001to2011period.Pooreducationliesattherootofmuchofthepoverty,thusthereportoftenreferstotheeducationsituationindifferentareas.People’smovementswithinandbetweenregionsareoftendrivenbyeconomicopportunities,thereforepopulationgrowthisdiscussed.Furthermore,povertyiscloselylinkedtootherformsofdeprivation,makingserviceprovisionamajorfactorinaddressingpoverty.

3.1 Regional Poverty Patterns and Trends

AscanbeseenfromTable1,Namibiaregisteredageneraldeclineintheincidenceofpovertyof11percentagepointsoverthe2001to2011period,withthenationalincidenceofpovertydeclining from 37.9 percent to 26.9 percent over this period. Currently about 568 418 people areestimatedtobepoor.Thisindicatesatotalnumberof125277fewerpeoplelivinginpovertyat the end of this period of ten years than would have been the case if the poverty rate had remained unchanged.

ThegreatestdeclineswereregisteredinthenorthernregionsofOhangwena,Omusati,Kuneneand Oshikoto, as well the eastern region of Omaheke. However, two regions (Zambezi andKhomas)registeredincreasesof7.2percentagepointsand1.2percentagepoints,respectively.In2011,outofthethirteenregions,sevenregions(Otjozondjupa,Oshikoto,Omusati,Ohangwena,Kunene,ZambeziandKavango)hadpovertyincidencesthatwereabovethenationalrateof26.9percent.Thesedeviationsfromthegeneraldeclinewillbediscussedinmoredetaillaterinthisreport.

Table 1: Trends in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

RegionPoverty Headcount Rate

2001 2011 ChangeZambezi 32.0 39.3 7.2Erongo 9.3 6.3 -3.0Hardap 20.4 17.2 -3.2Karas 18.0 14.5 -3.4Kavango 57.9 53.2 -4.8


Khomas 3.4 4.6 1.2Kunene 53.7 38.9 -14.8Ohangwena 62.8 35.3 -27.5Omaheke 41.6 26.2 -15.5Omusati 50.9 28.6 -22.2Oshana 28.3 21.1 -7.1Oshikoto 57.3 42.6 -14.7Otjozondjupa 30.4 27.5 -2.9Namibia 37.9 26.9 -11.0

Map1,below,givesthespatialdistributionoftheincidenceofpovertybyregionin2011.Itcanbeseenthatinboth2011and2001,Ohangwena,Kunene,Zambezi,OshikotoandKavangohadmorethanonethirdoftheirpopulationclassifiedaspoor.PovertyinNamibiastillbearsadistinctruralface,withthepoorestregionsbeingthoseinwhichthemajorityofthepopulationlivesinruralareas.Theregionswiththelowestincidencesofpoverty(KhomasandErongo)havelargelyurbanpopulationsandaretheeconomichubsofthecountry,withrelativelymoreemploymentopportunities.

Although,asinZambeziregion,thepovertyincidenceinKhomasincreasedbetween2001and2011,theregionstillhasthelowestincidenceofpovertywithonly5percentofitspopulationlivingbelow thepoverty line. Erongo, Karas,Hardap andOshana also reported low levels ofpoverty.KhomasregionishometoWindhoek,thepoliticalandeconomiccapitalofthecountry.ErongoregionnotonlyhasmostoftheexistingminesbutalsoborderstheAtlanticOceanwhichproducesfish,amajorexportcommodityforNamibia.ThisregionalsohastheNamibDesert,animportanttouristdestination.Indeedin2011theregionrecordedthesecondhighesttouristarrivalsinthecountry,withabout345000visitors.

Map 1: Namibia Poverty Headcount Rate, 2011 (upper bound poverty line)


Although therewasageneraldecline in the incidenceofpovertyat thenational level, thereweremarkeddifferencesintherecordedchangesintheincidenceofpovertyacrosstheregions(seeMap2).AscanbeseenfromTable1above,in2001thepoorestregionwasOhangwenafollowedbyKavango,Oshikoto,KuneneandOmusati,withmore thanhalfof thepopulationbeingclassifiedaspoorintheseregions.By2011,however,thesituationhadchangedwithonlyKavango(at53percent)havingmorethanhalfofitspopulationclassifiedaspoor.Thiswidely-baseddeclineinpovertyisareflectionofimportanteconomic,socialandpolicyprogressthathasbeenmade,andisthemostnotabletrendbetweenthetwocensusyears.

Intermsofregionalranking,thesituationhaschanged,withKavangobeingthepoorestregionfollowedbyOshikoto,Zambezi,KuneneandOhangwena.Importantly,Omusatiregionhadfallenoutofthefivehighestpovertyheadcountrateregions,whileZambezihadjoinedthisgroup.Overthe2001to2011period,Omusatiregionexperiencedareductionof22percentagepointsintheincidenceofpoverty,fromahighof51percentin2001toalowof29percentin2011.AccordingtotheTownCouncilauthorities,thisremarkableprogresscanbeattributedtoincreasedprivateinvestmentduringtheperiod,asexemplifiedbyrapidgrowthofshoppingcomplexesandotherbusinessactivitiesinthetownofOutapiafteritsproclamationasatownwithanautonomousTownCouncilin2002.Theincreaseinbusinessactivitiesledtojobcreationandavailabilityofcriticalservicesinthearea.

ThedeclineinthepovertyheadcountratewasnotlimitedtoOmusatiregion,however.Theratedeclined in nearly all regions,withOhangwena, KuneneandOshikoto andOmaheke regionsregisteringthegreatestdeclines.Forinstance,Ohangwenaregion,whichwasthepoorestregionin2001,recordedaremarkablereductioninthepovertyheadcountrateof28percentagepointsduring the period under consideration. According to the Ohangwena Regional Council, thisisattributable to increasedeconomicactivity in that region,stimulatedbypublicandprivateinvestments which boosted the regional economy.

Inaddition, thepastdecadehaswitnessed the successful completionofmany infrastructureprojects, including road networks, sanitation in rural areas and the construction of publicinfrastructure–schools,earlychildhooddevelopmentcentres,shoppingcomplexes,smallandmedium enterprise (SME) parks and health facilities. The region has also benefitted from asuccessfulroll-outofantiretroviraltherapy(ART),aswellascross-bordertradewithneighbouringAngola,mainlycarriedoutthroughthebordertownofOshikango.


Map 2: Namibia Change in Poverty Headcount Rate, 2011 - 2001 (upper bound poverty line, percentage points)

Two regions (Zambezi and Khomas) recorded increases in the incidence of poverty over the2001to2011period,withtheincidenceofpovertyintheseregionsincreasingby7.3percentand1.2percent,respectively.AlthoughKhomaswastheleastpoorregionatboththe2001and2011timepoints,itspovertylevelsincreasedslightlybetweenthesetwopoints.Thiscouldbeattributedtothehighrateofruraltourbanmigration,withmostofthemigrantsbeingyoungpeoplefromother,oftenmuchpoorer,regions.ThepopulationofKhomasincreasedbyalmost92000orabout37percentoverthedecade,morethantwicetheNamibianrateofpopulationgrowth. For most young migrants, Khomas region, especially Windhoek, is their preferreddestination.Many,however,areillequippedforthejobmarketandenduplivingindeplorableconditionswithoutjobs.

While in 2001 the incidence of poverty in the Zambezi region was comparable to that inOtjozondjupa, by 2011 the situation in these two regions had changed drastically, with theincidence of poverty in Otjozondjupa having declined by about 3 percentage points while it hadincreasedinZambeziby7percentagepoints.Indeedby2011,theincidenceofpovertyinZambeziwascomparabletothereportedpovertyincidenceinOhangwenaandOshikoto,whilein 2001 poverty headcount in Zambezi had been just under half that in Ohangwena and Oshikoto regions.


WhilethenorthernregionsandOmahekeregionintheeastregisteredsignificantreductionsintheincidenceofpoverty,mostoftheregionsinthecentralandsouthernpartsofthecountrydidnotregistersimilardeclinesinthepovertyheadcountoverthe2001to2011period.Thiscouldbebecauseitisusuallydifficulttofurtherreduceanalreadylowlevelofpoverty.Theseregionshavehugeeconomicpotentialintheagriculturalandextractivesectors.

Mostof theexistingminesare located inErongoandKaras,whileOtjozondjupa,HardapandKarasarecharacterisedbylargecommercialfarms,whichformthebasisofNamibia’sagriculturalexportstoexternalmarketssuchastheEuropeanUnion.However,thereisuntappedpotentialinvalueaddition,especiallyindiamondpolishingandprocessingofagriculturalproducts,tocreatejobs,spureconomicgrowthandultimatelyleadtopovertyreduction.

Table 2: Trends in Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

RegionPoverty Headcount Rate

2001 2011 ChangeZambezi 17.3 22.8 5.4

Erongo 4.4 2.4 -1.9

Hardap 10.5 7.8 -2.7

Karas 9.2 6.7 -2.6

Kavango 39.4 34.4 -5.0

Khomas 1.0 1.6 0.6

Kunene 37.5 24.8 -12.7

Ohangwena 40.7 18.6 -22.2

Omaheke 26.3 13.5 -12.8

Omusati 31.6 14.1 -17.5

Oshana 15.1 10.1 -5.0

Oshikoto 38.8 26.5 -12.1

Otjozondjupa 17.9 14.9 -3.0

Namibia 23.8 15.0 -8.8

Figure1,illustratesthecontributionofeachregiontotheoverallpovertyofthecountry.Kavangoregion,withapopulationshareof11percentandapovertyheadcount rateof53.2percentaccounts for 21percentof total poverty inNamibia. The Figure indicates that 15percentofall thepoor live inOhangwena,and14percentand12percent respectively inOshikotoandOmusatiregions.Only2percentofthetotalpoorliveinErongo,HardapandKarasregionseach.Similarly,KhomasandOmahekeregionsaccountforthreepercentofthetotalpooreach.


Figure 1: National poverty headcount shares, 2011 (upper-bound poverty line)

3.2 Constituency Poverty Patterns and Trends

3.2.1 Overview of constituency poverty patterns and profiles

Regional poverty aggregates, as presented above, often mask wide intraregional variations.Beyondtheregions, thereexistwidevariations in reportedpoverty incidenceacross the107constituenciesofNamibia.While,at the regional level, thehighest incidenceofpovertywasreportedinKavangoregion(53percent),atconstituencylevel,thehighestincidenceofpovertywas reported in Epupa constituency in Kunene region, with 69 percent of the populationclassifiedaspoor,while the lowest incidencewasreported inWindhoekEastconstituency inKhomasregion,withonly0.1percentofthepopulationbeingclassifiedaspoor.

Therearealsowidevariationsinthereductioninthepovertyheadcountrateoverthe2001to2011periodacrossthe107constituencies.Thebiggestreduction,intermsofpercentagepoints,wasregisteredinthenorthernregionsofOhangwenaandOmusati,whilethebiggestincreasewas in the north-eastern Zambezi region. Eenhana, Endola, Engela, Okongo and OngengaconstituenciesinOhangwenaregionandOshikukuconstituencyinOmusatiregionallregistereda reduction in thepovertyheadcount rateofmore than30percentagepoints,whileKatimaMuliloUrbanandKongolaconstituenciesinZambeziregionregisteredanincreaseofmorethan10 percentage points over the 2001 to 2011 period.

In2011,sixofthethirteenregionshadoneormoreoftheirconstituencieswheremorethan50 percent of the population was classified as poor, while nine regions had one or moreconstituencies inwhichmorethan30percentof thepopulationwasclassifiedaspoor.Table3showstheproportionofconstituencieswith30or50percentofthepopulationclassifiedaspoor.


Table 3: Proportion of constituencies with more than 30% and 50% of the population classified as poor (upper bound poverty line), 20111

Regions % of constituencies with at least 30% of population poor

% of constituencies with at least 50 % of population poor

Zambezi 83 33

Kavango 89 78

Kunene 50 17

Ohangwena 82 1

Omaheke 43 0

Omusati 33 0

Oshana 40 0

Oshikoto 90 30

Otjzondjupa 43 14

3.2.2 Zambezi region

Zambeziregion(formerlyCaprivi),withalandareaof14528km2andatotalpopulationof90596,liesinthenorth-easternpartofNamibia,borderingBotswana,Zimbabwe,ZambiaandAngola.ItalsobordersKavangoregionintheeast.Givenitsgeographiclocation,theregionisanimportantlogistical centre and serves as the gateway to the SouthernAfricaDevelopment Community(SADC)region.Thepopulationis69percentrural.Theregionreceivesanaverageannualrainfallof about 735mm. It is home to threeperennial rivers – Kwando, Chobe and Zambezi.Giventhe nature of the terrain and soil types, however, these rivers often causeflooding inmanypartsoftheregion.Zambeziregionalsohasmanynationalparkswithabundantwildlife.Thusitpossesseshugepotentialintheagricultural,tourism,andtransportandlogisticssectorsaskeydrivers of economic growth and development.

In2001,povertyincidenceinZambeziwasestimatedat32percent,withnosingleconstituencyhavingmorethanhalfofitspopulationlivinginpoverty.By2011,theregionalpovertyincidencehadincreasedby7.2percentagepoints.Thismeansthatin2011,10060morepeoplewerelivinginpoverty,whilethenumberofnon-poorhadincreasedbyjust710people.PovertyishighestinKongolaandSibbindaconstituenciesat58percentand55percent,respectively,andlowestinKatimaMuliloUrbanatonly17percent.

Intermsofpercentagechange,however,thehighestincrease,of11percentagepoints,intheincidenceofpovertyoverthe2001to2011periodwasrecordedinKatimaMuliloUrbanandKongolaconstituencies.Despiteitslowpovertyrate,KatimaMuliloUrbancontributedaboutonethird(34percent)oftheincreaseinpoverty,withanincreaseof3425poorpeople,whileLinyatiaccountsfor19percent,KatimaMuliloRural15percentandSibbinda14percentoftheincrease.

1Theregionsnotlistedhere(Erongo,Hardap,KarasandKhomas)hadnoconstituencieswith30percentormoreofthepopulationclassifiedaspoor.


Table 4: Zambezi Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Zambezi Region 2001 Rank 2011 Rank Change

Kabbe 42.2 3 49.1 4 7.0

Katima Mulilo Rural 38.3 5 43.0 5 4.7

Katima Mulilo Urban 6.4 6 17.2 6 10.8

Kongola 47.4 1 58.1 1 10.7

Linyanti 41.2 4 49.4 3 8.1

Sibbinda 45.8 2 55.0 2 9.2

Regional rate 32.1 39.3 7.2

Map3presentscolour-codedpovertylevels,withthedarkercolourindicatinghigherincidenceofpoverty.Asisevidentfromthemap,thetwopoorestconstituenciesareKongolaandSibbinda.The populations in these constituencies are largely rural, eking a living from subsistenceagriculture(livestockrearingandcropfarming).Theyalsorelyheavilyonsocialtransfers,mainlyintheformofoldagepension.Withonly1.1percentofthepeopleaged15yearsandabovehavingneverattendedschool,educationalattainmentinZambeziregionisrelativelyhigh.

Theliteracyrateforthepopulationaged15yearsandaboveisestimatedat84percent,whilethe youth literacy rate is estimated at 93 percent. Notwithstanding the good educationalattainment,more than one third (38 percent) of the economically active population (labourforce) is unemployed. With the exception of Sibbinda, with an estimated unemploymentrateof29percent,morethanhalfoftheeconomicallyactivepopulationisunemployedinallconstituencies. The agriculture sector is themain employer in the region, accounting for 42percentofemployment. It isfollowedcloselybythepublicsectorat22percent.Thetourismsectorcontributesonlyabout3percentoftheemployedpopulation.

About14percentofhouseholdsuseelectricityforcooking,whilearoundonethird(32percent)useitfor lighting.About73percentofthepopulationhasaccesstosafewater. InKabbeandKatimaMuliloRural,only25and55percent, respectively,ofhouseholdshaveaccess to safedrinkingwaterbut,intherestoftheconstituencies,morethanthreequarters(75percent)ofthe households have access to safe drinking water.

Map 3: Zambezi Region Poverty Headcount Rate in 2011 (upper bound poverty line)


Map 4, shows the changes in poverty incidence between 2001 and 2011. Although povertyincreased inallconstituenciesduringthisperiod, themap indicatesthatKongolaandKatimaMulilo Urban registered increases of more than 10 percentage points in poverty headcount.

Map 4: Zambezi Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

In2011,theincidenceofseverepovertywasestimatedat23percent,representinganincreaseof5percentagepointsfromthe2001figureof17percentand8percentagepointsabovethenationalaverageof15percent.Aswithpovertylevels,theincidenceofseverepovertyishighestinKongolaandSibbinda,withmorethanonethirdofthepopulationbeingseverelypoorintheseconstituencies.KatimaMuliloUrbanhasthelowestincidenceofseverepovertyat7percent.

Thepovertyliterature(e.g.Cage,2009)arguesthatsecurityandpoliticalstabilityarefundamentaltoeconomicgrowth,employmentcreationandpovertyreduction.Zambeziregionexperiencedpolitical instability in 1999 which had a negative impact on investment by both local andinternationalinvestors.In2001,about83percentofthelabourforcewasemployedcomparedto62percentin2011.Thisindicates,thatoverthisperiod,instabilitydiscouragedinvestmentwithnegative impactsonemploymentcreationandpovertyreduction.Furthermore,politicalinstability as one of themigration push factors, could have led to the higher outmigrationexperiencedbetween1991and2011,aminimalpopulationincreasefrom90422to90596overaperiodoftwentyyears.Thistendstoincreasepovertyasitismostlytheeconomicallyactivewho migrate.

Table 5: Zambezi Region Poverty Headcount Rate Scores and Values, 2001 - 2011 (lower bound poverty line)

Zambezi Region 2001 Rank 2011 Rank Change

Kabbe 23.6 3 29.3 4 5.7

Katima Mulilo Rural 20.9 5 24.7 5 3.8

Katima Mulilo Urban 2.0 6 7.2 6 5.3

Kongola 29.3 1 36.6 1 7.4

Linyanti 22.2 4 29.6 3 7.3

Sibbinda 25.1 2 35.6 2 10.5



3.3.3 Erongo region

ErongoisthesecondmosturbanisedregioninNamibiaafterKhomas,with87percentoftheinhabitants living inurbanareas. The regionhas a total land areaof 63586 km2 accountingfor7.7percentofthecountry’slandsurface.Theregion,whichborderstheAtlanticOcean,isamajortouristdestinationandishometoWalvisBayharbour,thelargestportinthecountryandanimportantgatewaytomanySADCcountriessuchasAngola,Botswana,theDemocraticRepublicofCongo,ZambiaandZimbabwe.ThusErongohasgreatpotentialtobeanimportanttransportandlogisticalhubfortheSADCregion.Othermajoreconomicactivitiesintheregionincludetourism,mining,fishingandmanufacturing.

Overthe2001to2011period,therewasathreepercentagepointreductioninpovertyinErongoregion. Fouroutof the sevenconstituencies inErongo recordeddeclines in the incidenceofpovertyoverthisperiod.Thelargestdecline,of18percentagepoints,wasrecordedinDauresconstituency.About1659fewerpeoplearepoorthanin2011.Thispositivechangecouldbeattributedtosmallminingandconservancyactivitiesintheconstituency.DespitethedeclineinpovertyinDauresconstituency,itstillhasthehighestincidenceofpoverty,at20percentofthepopulation,althoughthenumberofpeoplelivinginpovertyonlyamountsto2281people.

Three constituencies recorded increases in the incidence of poverty, the largest being 2percentagepointsrecordedinOmaruruconstituencywhere317morepeoplearepoorthanin2001,whilethenumberofnon-poorpeopleincreasedby1104.Erongoregionhasexperiencedrapidpopulationgrowthoverthepasttenyears,withthepopulationgrowingatanaveragerateof3.4percentperannum,twopercentagepointsabovethenationalaverageof1.4percent.Indeedtwooftheconstituencies,SwakopmundandWalvisBayRural,registeredapopulationgrowthrateof5percentperannumoverthe2001to2011period.EducationallevelsinErongoarehigh,withliteracyrateforthoseaged15yearsandaboveestimatedat97percent,asalmosteveryoneintheregionhasattainedsomeformaleducation.

Table 6: Erongo Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Erongo Region 2001 Rank 2011 Rank Change

Arandis 5.4 4 6.3 4 0.9

Daures 38.3 1 20.1 1 -18.2

Karibib 16.9 2 15.4 2 -1.5

Omaruru 9.6 3 11.7 3 2.1

Swakopmund 4.4 5 3.6 6 -0.8

Walvis Bay Rural 3.4 6 3.7 5 0.2

Walvis Bay Urban 3.4 6 2.4 7 -1.0

Regional rate 9.3 6.3 -3.0

Themainsourceofincomefortheregionissalariesandwages,at73percent.Dauresconstituencyistheonlyexceptiontothispattern,with28percentofhouseholdsinthisconstituencycitingsalariesandwagesastheirmainsourceofincomewhileaboutaquarter(24percent)indicatedsubsistencefarmingandthesameproportionoldagepensionsastheirmainsourceofincome.


About79percentofthose15yearsandaboveareintheeconomicallyactivecategory,withtheunemploymentratewithinthisgroupestimatedtobe30percent.Themainemployersarethefishing,miningandmanufacturingsectors,eachofwhichaccountsformorethan10percentoftotalemployment.Thetourismsectorcontributesonlyabout5percentoftotalemploymentinthe region.

Although Erongo has the second lowest poverty headcount in the country, Map 5, below,illustratesthatthepovertyheadcountisstillhighinDaures,KaribibandOmaruruconstituencies.AccesstoelectricityinErongoishigh,with76percentofhouseholdsusingelectricityforcooking,althoughinDauresconstituencyonly13percentofhouseholdsdoso.Accesstosafewaterisestimatedat96percent.HereagainDaureslagsbehind,withonly65percenthavingaccesstosafe water.

Map 5: Erongo Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Map6showschangesinpovertylevelsoverthe2001to2011period.ThemapindicatesthatthegreatestdeclineinthepovertyheadcountratewasregisteredinDauresconstituency,whiletheincidenceofpovertyinOmaruruconstituencyincreasedbytwopercentagepoints.


Map 6: Erongo Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

At 2 percent, the region registered a very low incidence of severe poverty. This is a twopercentagepointreduction inthe incidenceofseverepovertyoverthe2001to2011period,i.e.theseverepovertyrate,whichwasalreadyextremelylowin2001,halved.Aswasthecasewiththeincidenceofpoverty,althoughDauresconstituencyregisteredthegreatestdeclineintheincidenceofseverepovertyof13percentagepoints,theconstituencystillhasthehighestincidenceofseverepoverty,at9percent.Againaswiththepovertyheadcountrate,DauresisfollowedcloselybyKaribibconstituencyat7percent.

Table 7: Erongo Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Erongo Region 2001 Rank 2011 Rank Change

Arandis 1.6 4 2.1 4 0.5

Daures 22.5 1 9.4 1 -13.1

Karibib 8.5 2 7.2 2 -1.3

Omaruru 4.1 3 4.5 3 0.4

Swakopmund 1.4 5 1.1 5 -0.3

Walvis Bay Rural 1.0 6 1.0 6 0.0

Walvis Bay Urban 1.0 7 0.7 7 -0.3



3.3.4 Hardap region

Hardapisoneofthesouthernregionsofthecountry.ItborderstheAtlanticOceantothewestandBotswanatotheeast,andcoversatotallandareaof109659km2.Withatotalpopulationof79705,theregionisoneoftheleastdenselypopulatedareasofNamibia,withapopulationdensityof0.7peoplepersquarekilometre.Sixtypercentofthepopulationlivesinurbanareas.HardapisoneofthedriestregionsinNamibia,withanaveragerainfallrangingbetween71and91mm,comparedtothenationalaveragelevelof800mm.NaukluftParkandFishRiverGrandCanyon(thesecondlargestcanyonintheworld)aremajortouristattractionsintheregion.

At the regional level, the incidence of poverty is estimated at 17 percent (13 675 people),havingdeclinedby3percentagepointsbetween2001and2011.PovertyishighestinGibeon,RehobothRuralandMarientalRuralconstituencies,withnearlyaquarterofthepopulationintheseconstituenciesclassifiedasbeingpoor,andlowestinRehobothUrbanWest,wheretheincidenceofpovertyisestimatedat4percent.

Over the2001 to2011period the incidenceof povertydeclined inGibeon,MarientalRural,RehobothUrbanWestandRehobothRuralconstituencies,whileMarientalUrbanandRehobothUrbanEastrecordedmarginalincreases,with408moreand813morepeoplerespectivelylivinginpoverty.Notwithstandingthemarginalincreaseinthenumbersofpoorpeople,thenumberofnon-poorpeopleinthesetwoconstituenciesincreasedby2040and4331respectively.Overthepast tenyears, theregionexperiencedpopulationgrowthof1.5percent,with theurbanpopulation growing by 4.3 percent,while the rural areas experienced a negative populationgrowth of 1.5 percent.

Table 8: Hardap Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Hardap Region 2001 Rank 2011 Rank Change

Gibeon 30.2 1 24.8 1 -5.4

Mariental Rural 27.5 2 23.3 3 -4.3

Mariental Urban 15.4 4 15.6 4 0.2

Rehoboth Rural 27.1 3 24.2 2 -2.9

Rehoboth Urban East 13.0 5 13.8 5 0.8

Rehoboth Urban 9.0 6 4.0 6 -5.0


Map7showsthe incidenceofpoverty in theconstituenciesofHardapregion.The incidenceofpoverty is highest inGibeon, at 25percent, and lowest inRehobothUrban, at 4percent.Generally,theregionhasahighlevelofeducationalattainment,withaliteracyrateof96percentandonly10percentofthepeopleaged15yearsandabovehavingneverattendedschool.Themajorityof thosewhohaveneverattended school are concentrated in thepoorerand ruralconstituencies ofMariental Rural and Gibeon. Agriculture, construction, and wholesale andretailtradearethemaineconomicactivities,employingabouthalfoftheeconomicallyactivepopulation.Miningaccountsforonly2percent,whilemanufacturingandtourismaccountforabout 4 percent of employment in the region.


Map 7: Hardap Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Map8illustrateschangesinpovertylevelsoverthelasttenyears.ThemapshowsthatthehighestreductionwasregisteredinGibeonconstituencyfollowedbyMarientalRuralconstituency.

Map 8: Hardap Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Fromahighof11percent in2001,anestimatedeight8percentoftheHardappopulation iscurrentlyclassifiedasbeingseverelypoor,adeclineof3percentagepoints.GibeonandRehobothRuralconstituencieshavemorethan10percentoftheirpopulationsclassifiedasseverelypoor.


Table 9: Hardap Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Hardap Region 2001 Rank 2011 Rank Change

Gibeon 17.3 1 12.1 2 -5.2

Mariental Rural 15.0 2 11.0 3 -4.0

Mariental Urban 7.1 4 6.6 4 -0.5

Rehoboth Rural 14.9 3 12.6 1 -2.3

Rehoboth Urban East 5.1 5 5.4 5 0.3

Rehoboth Urban West 3.9 6 1.3 6 -2.6


3.3.5 Karas region

Karas region is the driest and southernmost region of Namibia. It covers a total land area of 161 086 km2,whichrepresents19.6percentofthecountry’s landsurface.Withapopulationof77421,theregionaccountsforanestimated3.8percentofthenationalpopulation.Karasis characterised by low rainfall, high evaporation rates and sparse vegetation. The region is,however,endowedwithplenteousnaturalresources,suchasalluvialgold,diamonds,ironandzinc,andishometothecountry’slargestminingactivities.TheregionisalsoendowedwiththeperennialOranjeRiveralongtheborderwithSouthAfricaandNauteDamwhichofferspotentialfor irrigated agriculture.

PovertyincidenceinKarasregionisestimatedat14percent(11226people),havingdecreasedby3.4percentagepointsoverthepasttenyears.ThegreatestchangewasregisteredinBersebaconstituency,wheretheincidenceofpovertydeclinedby11percentagepointsoverthe2001to2011period.Bersebais,however,stilltheconstituencywiththehighestincidenceofpovertyintheregion,with27percentofthepopulationclassifiedasbeingpoor(2880people).PovertyhasalsodeclinedinOranjemund,LuderitzandKeetmanshoopRuralconstituencies.

Table 10: Karas Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Karas Region 2001 Rank 2011 Rank Change

Berseba 38.5 1 27.2 1 -11.3

Karasburg 21.6 3 20.8 3 -0.8

Keetmanshoop Rural 25.8 2 23.0 2 -2.8

Keetmanshoop Urban 9.8 4 9.9 4 0.1

Luderitz 9.7 5 7.0 5 -2.7

Oranjemund 7.6 6 2.9 6 -4.7



AsshowninMap9,Berseba,KeetmanshoopRuralandKarasburgconstituencieshavemorethan20percentoftheirpopulationclassifiedaspoor.Aseducationisknowntohaveanamelioratingimpactonpoverty, it isnotsurprisingthatthefirst twoof theseconstituencies,BersebaandKeetmanshopRural,havethehighestpercentageofpeoplewithnoformaleducationat9percentand8percent,respectively.Incontrast,theregionasawholehasaliteracyrateof97percentwithonly5percentofthepopulationhavingneverbeentoschool.Anestimated68percentofthepopulationiseconomicallyactive.

UnemploymentishighestinBersebaandKarasburg,at38percentand29percent,respectively.Agriculture,miningandconstructionarethemainemployers,whilemanufacturingandtourismaccountfor,respectively,6percentand3percentoftotalemployment.Theregionhasapotentialforgreenscheme(irrigation)projects,whichcouldhaveapovertyreducingimpact,especiallyinKarasburg,KeetmashoopRuralandBersebaconstituencies.

Map 9: Karas Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Map 10: Karas Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)


Map 10, above, illustrates the reduction in poverty levels over the past ten years. Themapshowsthatoverthe2001to2011period,thelargestdeclineinthepovertyheadcountratewasrecordedinBersebaandOranjemundconstituencies.ThedeclineinthepovertyrateinBersebais especially welcome given its high poverty incidence.

About7percentoftheKaraspopulationisestimatedtobeseverelypoor,havingdeclinedby2.6percentagepointsfrom9.2percent in2001.Similartothepatternforpoverty levels, theincidenceofseverepovertyishighestinBerseba,at14percent,andlowestinOranjemund,atless than 1 percent.

Table 11: Karas Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Karas Region 2001 Rank 2011 Rank Change

Berseba 23.2 1 13.8 1 -9.4

Karasburg 11.0 3 9.9 3 -1.1

Keetmanshoop Rural 14.4 2 11.2 2 -3.2

Keetmanshoop Urban 3.6 5 3.8 4 0.1

Luderitz 4.0 4 2.6 5 -1.4

Oranjemund 3.4 6 0.9 6 -2.5


3.3.6 Kavango region Kavango is the fourth most populous region in the country with a population of 223 352,accountingfor11percentofthetotalnationalpopulation.Between2001and2011,theregionalpopulationgrewby1percentperannum,thatis,moreslowlythanthenationalrate.Theregionhasapopulationdensityof4.6peopleperkm2.ThemainhydrologicalfeatureoftheKavangoregionistheOkavangoRiver,whichpresentshugepotentialforirrigationandartisanalfishing.Thepastdecadehaswitnessedanincreaseininvestmentingreenschemeprojects,mainlyalongtheOkavango, leadingto increasedagriculturalproductionandproductivity intheregion.Anestimated71percentofthepopulationlivesinruralareas.Theregionrecordedanetoutflowofmigrants both between 1996 and 2001 and between 2001 and 2011.

In2011,Kavangoregionhadthehighestincidenceofpovertyofallregionsat,53percent(118823people),representingadeclineof5percentagepointsfromthe2001figureof58percent.WiththeexceptionofRunduUrbanconstituency,allconstituenciesinKavangoregionhavepovertyincidenceabovethenationalaverageof27percent.ThehighestpovertywasreportedinKapakoconstituency(63percentor16891people),whilethelowestincidencewasreportedinRunduUrban(19percent).Kahenge,Kapako,MashareandMpunguconstituenciesallhave60percentormoreoftheirpopulationclassifiedaspoor.

Intermsofchangesintheincidenceofpovertyovertime,thegreatestdeclinewasreportedinMashare,NdiyonaandRunduUrbanconstituencies,which recorded reductionsof14.7,12.8and11.8percentagepoints,respectivelybetween2001and2011.Overthe2001to2011period,thepovertyheadcountratedeclinedinalloftheconstituencies,withtheexceptionofKahengeand Kapako. Kapako recorded an increase of about 7 percentage points in poverty headcount.


Table 12: Kavango Region Poverty Headcount Rate, 2001 - 2011 (upperbound poverty line)

Kavango Region 2001 Rank 2011 Rank Change

Kahenge 60.3 6 60.6 2 0.3

Kapako 55.8 7 62.6 1 6.8

Mashare 75.3 1 60.5 3 -14.8

Mpungu 60.8 5 60.3 4 -0.6

Mukwe 65.2 3 58.2 5 -7.0

Ndiyona 69.3 2 56.6 6 -12.7

Rundu Rural West 46.8 8 45.1 8 -1.7

Rundu Urban 30.3 9 18.6 9 -11.8

Rundu Rural East 61.5 4 56.3 7 -5.2


Map 11, below, shows that, with the exception of Rundu Urban and Rundu Rural Westconstituencies,inalltheconstituenciesinKavangoregionmorethanhalfofthepopulationispoor.Althoughtheliteracyrateishigh,at79percent,18percentofthepopulation6yearsandabovehaveneverenteredformaleducationwhilemorethanonethird(35percent)of thoseaged15yearsandabovehavenotcompletedprimaryeducation.About61percentofthoseaged15yearsandaboveareintheeconomicallyactivecategory.However,onlyhalfareemployed,resultinginanunemploymentrateof50percent.

The agricultural sector is themain source of employment, accounting for 60 percent of theemploymentintheregion.UnemploymentishighestinRunduRuralEast,KapakoandMashareconstituencies.Subsistencefarmingisthemainsourceofincome,involving43percentofthehouseholds in the region. The only exception to this is RunduRuralWest andRunduUrbanconstituencieswhere39percentand55percent,respectively,ofthepopulationcitedsalaryandwages as the main source of income.

Map 11: Kavango Region Poverty Headcount Rate in 2011 (upper bound poverty line)


Map12indicatesareductioninpovertylevelsovertheperiodoftenyears(from2001to2011).ThemapindicatesthatthegreatestdeclineinpovertyoccurredinMashare,NdiyonaandMukweconstituencies.

Map 12: Kavango Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Morethanonethird(34.4percent)ofthepopulationinKavangoregionisseverelypoor.Theincidenceofseverepovertyreducedbyfivepercentagepointsoveradecade,withthegreatestdecline registered inMashare and Ndiyona constituencies. The incidence of severe povertyincreased in Kapako by 7 percentage points. Targeted poverty interventions are required toreduce poverty levels.

Table 13: Kavango Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Kavango Region 2001 Rank 2011 Rank Change

Kahenge 41.7 5 39.7 4 -2.1

Kapako 36.0 7 42.5 1 6.5

Mashare 56.3 1 40.5 3 -15.8

Mpungu 42.5 4 40.6 2 -1.9

Mukwe 46.4 3 39.1 5 -7.3

Ndiyona 50.9 2 35.5 7 -15.4

Rundu Rural West 29.2 8 27.4 8 -1.7

Rundu Urban 15.9 9 8.5 9 -7.4

Rundu Rural East 40.1 6 36.6 6 -3.4



3.3.7 Khomas region

KhomasRegionishometoWindhoek,thecommercialhubandcapitalofthecountry.Theregionhas a populationof 342 141 accounting for about 16.2 percent of the total population. TheregionispredominantlyurbanandsixoutofthesevenconstituenciesareurbanconstituencieslocatedinWindhoek.Theregionisanetrecipientofmigrantsfromotherpartsofthecountryduetoitsstrongeconomicpullaswellasthepushfactors(povertyandunemployment)presentin some rural parts of the country.

Table14indicatesmixedresultswithregardtoreductionintheincidenceofpoverty,withfourout of the ten constituencies having registered increases,while six constituencies registereddeclines in the poverty headcount rate over the 2001 to 2011 period.Windhoek Rural andKatuturaCentralregisteredthe largestdeclines inpoverty incidence. Withpoverty incidenceof5percent(15738people),KhomasistheleastpoorregioninNamibia.Notwithstandingthisrelativelylowpovertylevel,thereexistwidevariationsbetweenthetenconstituenciesoftheregion,withthosecharacterisedbyinformalsettlementsrecordinghigherlevelsofpoverty.WhilethereisvirtuallynopovertyincidenceinWindhoekEastconstituency(0.1percentincidence),inTobiasHainyekotheincidenceofpovertystandsat10percent,andinbothMosesGaroebandWindhoekRural,povertystandsat8percent.Overall,therewasa1percentagepointincreaseinpovertyinKhomasoverthe2001to2011period,meaningthat7230morepeoplearelivingin poverty than in 2001.

Table 14: Khomas Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Khomas Region 2001 Rank 2011 Rank Change

Tobias Hainyeko 4.8 3 9.6 1 4.8

Katutura Central 6.2 2 4.0 6 -2.2

Katutura East 4.5 4 4.1 5 -0.4

Khomasdal North 1.6 7 2.4 7 0.7

Soweto 2.9 6 2.1 8 -0.8

Samora Machel 3.3 5 4.3 4 0.9

Windhoek East 0.2 10 0.1 10 -0.1

Windhoek Rural 11.3 1 7.7 3 -3.6

Windhoek West 0.5 9 0.4 9 -0.2

Moses Garoeb 1.4 8 8.4 2 7.0


TherelativelyhighincidenceofpovertyinTobiasHainyekoandMosesGaroebconstituenciescanbeattributedtorapidpopulationgrowthduetoaninflowofmigrants.Whiletheregionhadapopulationgrowthof3percentperannumoverthe2001to2011period,MosesGaroeb,SamoraMachelandKhomasdalNorthconstituencieshadpopulationgrowthratesof5percentorhigher.Itisnoteworthythatmostofthemigrantsintotheseconstituencieslackthenecessaryskillsandeducationtobeeasilyabsorbed inthe jobmarket, leadingtohighratesofunemployment intheseconstituencies.


Anestimated74percentofthepopulationage15yearsandaboveiseconomicallyactive,withunemployment estimated at 21 percent. Unemployment is highest inMoses Garoeb, at 30percentof the labour force, followedbyTobiasHainyeko (29percent), andKatuturaCentral,KatuturaEastandSamoraMachel(allat28percent).

Map 13, below, depicts spatial patterns of poverty in Khomas at the 2011 time point. Theeducation level in Khomas is very high,with the literacy rate estimated at 97 percentwhileanestimated5percentofthepopulationhasneverattendedschool.WindhoekRural,TobiasHainyekoandMosesGaroebconstituencieshavehigherproportionsofpeoplewhohaveneverattendedschool,at13percent,9percentand7percent,respectively.

Map 13: Khomas Region Poverty Headcount Rate in 2011 (upper bound poverty line)

AsstatedaboveandshowninMap14below,thelargestincreaseintheincidenceofpovertywasrecordedinMosesGaroebandTobiasHainyekoconstituencies.


Map 14: Khomas Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Theincidenceofseverepovertyisestimatedatanextremelylow1.6percent,indicatingthatinthisurbansettingevenmostoftheunemployedareabletoavoidpoverty,andparticularlyseverepoverty.SeverepovertyisfoundinTobiasHainyeko,MosesGaroebandWindhoekRural.IthasactuallyincreasedinbothTobiasHainyekoandMosesGaroebby2and3percentagepointsrespectively,whileithasdeclinedinWindhoekruralbyabout1.6percentagepoints.AddressingthemigrationpushfactorsanddecentralisationcouldhelpreducepovertyinKhomasregion.

Table 15: Khomas Region Poverty Headcount Rate Scores and Values, 2001 - 2011 (lower bound poverty line)

Khomas Region 2001 Rank 2011 Rank Change

Tobias Hainyeko 1.4 3 3.6 1 2.1

Katutura Central 1.6 2 1.1 6 -0.5

Katutura East 0.9 4 1.2 5 0.3

Khomasdal North 0.4 7 0.7 7 0.3

Soweto 0.9 4 0.5 8 -0.4

Samora Machel 0.7 6 1.3 4 0.6

Windhoek East 0.0 10 0.0 10 0.0

Windhoek Rural 4.5 1 2.9 3 -1.6

Windhoek West 0.1 8 0.1 9 0.0

Moses/Garoëb 0.2 8 3.0 2 2.9



3.3.8 Kunene region

Kunene region derives its name from the Kunene River, which formsNamibia’s borderwithAngolaandisthesecondlargestriverinthecountry.Theregionischaracterisedbyrockyandbaremountains.KunenereceiveslowandunreliablerainfallandtheclimateisgreatlyinfluencedbytheSouthAtlanticandBenguelacurrents.Nomadicpastoralismisthemaineconomicactivityasthepotentialforirrigatedagricultureremainslargelyuntapped.TheregionishometooneofNamibia’smajorinternationaltouristattractions,theEpupafalls.Ithasapopulationof86856people of whom 74 percent live in rural areas.

In 2011, Kunene region,with a headcount poverty rate of 39 percent (33 787 people),wasthefourthpoorestregioninthecountryafterKavango,OshikotoandZambezi.Between2001and2011, theregionregistereda15percentagepoint reduction in the incidenceofpoverty.Reductions were registered in all the constituencies, with the highest reductions being inSesfontein(29percentagepoints),followedbyOpuwo(21percentagepoints)andKamanjab(17percentagepoints).

Table 16: Kunene Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Kunene Region 2001 Rank 2011 Rank Change

Epupa 76.8 1 69.2 1 -7.7

Kamanjab 37.0 4 19.9 4 -17.1

Khorixas 34.3 5 18.8 5 -15.5

Opuwo 65.2 3 44.1 2 -21.1

Outjo 22.7 6 18.0 6 -4.7

Sesfontein 69.0 2 40.0 3 -29.0


Despiteregisteringareductionintheincidenceofpovertyof8percentagepoints,Epupa,withpovertyheadcountof69percent,isstillthepoorestconstituencyinKunene.ItisfollowedbyOpuwo(44percent)andSesfontein(40percent).ThisisshowninMap15,below.Thepopulationoftheregiongrewby2.3percentperannum,withEpupaandOutjoconstituenciesregisteringthehighestgrowthratesat3percentand3.1percent,respectively.Theregionhasaliteracyrateof65percent,whilemorethanonethird(36percent)ofpeopleaged6yearsandabovehaveneverattendedschool.Epupa(68percent)andOpuwo(42percent)havethehighestratesofpeoplewhohaveneverattendedschool.

Theeconomicallyactivepopulationisestimatedat67percentofthepopulation.Ofthese,36percent are unemployed.More than half (56 percent) of the employed population is in theagriculturalsector,withtourismandmanufacturingaccountingforabout4.2and4.3percent,respectively, of employment in the region. Apart from Outjo, Kamanjab and Sesfontein, allconstituenciesinKuneneregiondependonsubsistencefarmingastheirmainsourceofincome.


Map 15: Kunene Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Map 16 illustrates the changes in poverty incidence over the 2001 to 2011 period. As can be seenfromthemap,withtheexceptionofEpupa(8percentagepoints)andOutjo(5percentagepoints), all constituencies registered a poverty reductionofmore than15percentagepointsbetween 2001 and 2011.

Map 16: Kunene Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)


Despiteareductionof13percentagepointsbetween2001and2011,anestimatedonequarteroftheKunenepopulationisstillclassifiedasseverelypoor. InEpupaconstituency,morethanhalf(51percent)ofthepopulationisclassifiedasseverelypoor,whilealmostathird(28percent)of thepopulation inOpuwo is severelypoor.The regionhas thepotential to reducepovertythroughagriculture,tourismandlogistics.

Table 17: Kunene Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Kunene Region 2001 Rank 2011 Rank Change

Epupa 56.5 1 50.9 1 -5.6

Kamanjab 21.4 4 9.5 4 -12

Khorixas 20.6 5 8.8 5 -11.7

Opuwo 49.2 2 28.2 2 -21.1

Outjo 10.9 6 8.4 6 -2.5

Sesfontein 48.8 3 23.7 3 -25.1


3.3.9 Ohangwena region

Ohangwena region borders Cunene Province in Angola to the north and Kavango,Oshikoto,OshanaandOmusatiregionsinNamibia.Theregionhasapopulationof245446,whichis11.6percentofthenationalpopulation.At23peoplepersquarekilometre,theregionhasthehighestpopulationdensityinthecountry.Anestimated90percentofthepopulationlivesinruralareas.Between2001and2011,theregionregisteredthegreatestdeclineintheincidenceofpoverty,from63percentto35percent.Asaresult,56783fewerpeoplearelivinginpovertythanwasthe case in 2001.

Thedeclineinpovertyincidencewasobservedinallthe11constituencies.WiththeexceptionofOndombeconstituencywhere the incidenceofpovertydeclinedby11percentagepoints,allconstituenciesinOhangwenaexperiencedapovertyreductionofmorethan20percentagepoints,withthehighestreductionof34percentagepoints(8290people),beingregisteredinEndolaconstituency.Thisdeclinenotwithstanding,Ohangwenaremainsamongthefivepoorestregions in the country.

Table 18: Ohangwena Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Ohangwena Region 2001 Rank 2011 Rank Change

Eenhana 62.6 6 31.1 8 -31.4

Endola 62.0 7 27.7 10 -34.2

Engela 59.2 8 26.0 11 -33.2

Epembe 72.2 3 48.4 2 -23.8

Ohangwena 57.8 10 29.9 9 -27.9

Okongo 73.7 2 41.0 3 -32.7


Omundaungilo 76.1 1 51.8 1 -24.3

Ondobe 51.0 11 39.8 5 -11.2

Ongenga 65.3 4 32.3 6 -33.0

Oshikango 58.9 9 31.6 7 -27.2

Omulonga 63.1 5 40.4 4 -22.7


FromTable18,above,andMap17,below,itcanbeseenthattheincidenceofpovertyishighestinOmundaungiloat52percent,followedbyEpembeat48percent,Okongoat41percent,andOmulongaandOndobeeachataround40percent.Overthepastdecade,theregionexperienceda population growth of about 0.7 percent per annum.Ohangwena has a literacy rate of 86percent,while14percentof thoseaged6yearsandabovehaveneverattendedschool.Thepoorer constituencies have high proportions of people who have never attended school –Omundaungilo (19.4percent),Okongo(17.3percent),Epembe(16.4percent),andOmulongaandOndobe(15.2percent).

Theeconomicallyactivepopulationisestimatedat49percentoftheregionalpopulationand43percentoftheseareunemployed.Theagriculturalsector isthemainemployerwithmorethan half (51 percent) of the employed population engaged in this sector. It is followed bythepublic sectorandwholesaleandtrade.Tourismandmanufacturingsectorsaccount for4percentand3percentoftheemployed,respectively,whileconstructionaccountsfor5percentofemployment.Accesstosafewaterisestimatedat56percentofthepopulation.Again,thepoorerconstituencieshave lowerpercentagesofpopulationwithanestimated22percentofthepopulationinOmundaungiloconstituencyhasaccesstosafewater,withthecorrespondingfigures for Epembe, Omulonga and Ondobe being 23 percent, 44 percent and 43 percent,respectively.

Map 17: Ohangwena Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Map18, illustrates thereduction inpoverty incidence for theperiod2001to2011.With theexceptionofOndobe,alltheconstituenciesinOhangwenaregionregisteredreductionsintheincidence of poverty of more than 20 percentage points.


Map 18: Ohangwena Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Table19presentstheincidenceofseverepoorindividualsinOhangwenaregion.Thetableshowsthattheincidenceofseverepovertyreducedsignificantlyby22percentagepointsoverthe2001to2011period.TheincidenceofseverepovertyishighestinOmundaungilo,at31percent,andlowestinEngela,at12percent.

Table 19: Ohangwena Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Ohangwena Region 2001 Rank 2011 Rank Change

Eenhana 38.3 9 15.7 8 -22.7

Endola 39.1 8 12.9 10 -26.2

Engela 39.6 6 12.1 11 -27.5

Epembe 48.3 3 28.4 2 -19.9

Ohangwena 37.0 10 15.1 9 -22.0

Okongo 51.8 1 22.5 3 -29.2

Omundaungilo 51.4 2 31.3 1 -20.2

Ondobe 33.5 11 21.3 5 -12.2

Ongenga 43.1 4 16.7 6 -26.4

Oshikango 36.7 5 15.9 7 -20.8

Omulonga 39.2 7 22.1 4 -17.2


3.3.10 Omaheke region

Omahekeregion,withapopulationof71233people,liesinthecentraleasternpartofNamibia.ItbordersBotswanatotheeastandtheHardap,KhomasandOtjozondjuparegions.TheTrans-Kalaharihighway,which linksNamibiawithBotswana, SouthAfricaandZimbabwe, traversestheregion,thuspresentingopportunitiesfortransportandlogistics,andrelatedactivities.Theregion iswell known for its largecommercial cattle ranches.Anestimated30percentof thepopulationlivesinurbanareas.


Fromahighof42percent in2001, the incidenceofpovertydeclined to26percent (18663people)in2011,a16percentagepointreduction.Overthe2001to2011period,theincidenceofpovertydeclinedinallconstituenciesexceptGobabis.ThegreatestdeclinewasregisteredinAminius(23percentagepoints),followedbyOtjinene(22percentagepoints)andEpukiroandOtjombinde(21percentagepointseach).Otjombindeconstituencyhasthehighestincidenceofpoverty,at37percent,whileGobabishasthelowestat17percent.

Table 20: Omaheke Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Omaheke Region 2001 Rank 2011 Rank Change

Aminius 50.9 4 28.4 4 -22.5

Gobabis 16.9 7 17.1 7 0.2

Kalahari 44.9 5 27.1 5 -17.8

Otjinene 55.1 2 33.2 2 -21.9

Otjombinde 57.0 1 36.3 1 -20.7

Steinhausen 35.5 6 26.1 6 -9.4

Epukiro 52.5 3 31.2 3 -21.2


Map19showsthespatialdistributionofpoverty incidenceinOmaheke.Poverty ishighest inOtjombinde and lowest in Gobabis, the region’s commercial and administrative capital. Theregionrecordedapopulationgrowthrateof0.5percentperannumoverthe2001to2011period,withGobabisconstituencyrecording3.3percentperannum.Omahekehasaliteracyrateof73percent,while25percenthasneverattainedformaleducation.Morethan30percentof thepopulationaged6yearsandaboveinKalahari,OtjombindeandSteinhausenhadneverenteredformaleducation.Theeconomicallyactivepopulationisestimatedat65percent,40percentofwhichisunemployed.About45percentoftheemployedpopulationisintheagriculturesector.Tourismaccountsfor5percentoftheemployedpopulationintheregion,whilemanufacturingandlogisticseachaccountfor2percent.Constructionisakeysector,yieldingabout7percentoftheregion’semployment.

Map 19: Omaheke Region Poverty Headcount Rate in 2011 (upper bound poverty line)


Map 20 shows that, with the exception of Gobabis, all constituencies registered significantreductionsintheincidenceofpovertyoverthe2001to2011period.

Map 20: Omaheke Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Between2001and2011,theproportionofseverelypoorindividualsintheregionwasreducedbyalmosthalf.Fromahighof26percentin2001,theproportionofthoseclassifiedasbeingseverelypoorstoodat14percentin2011.In2011,Otjombindeconstituencyhadthehighestincidenceofseverepoverty,at21percent.ItisfollowedbyOtjineneandEpukiroconstituencies.

Table 21: Omaheke Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Omaheke Region 2001 Rank 2011 Rank Change

Aminius 35.2 3 14.7 4 -20.5

Gobabis 8.4 7 7.6 7 -0.8

Kalahari 28.6 5 13.6 5 -15.0

Otjinene 35.5 2 18.0 2 -17.5

Otjombinde 37.4 1 21.2 1 -16.2

Steinhausen 21.8 6 13.0 6 -8.8

Epukiro 31.1 4 17.3 3 -13.8


3.3.11 Omusati regionOmusati region borders Angola in the north and also the Kunene, Ohangwena and Oshanaregions of Namibia. It has a total area of 26 573 km2.Withapopulationof243166,or11.5percentofthenationalpopulation,theregionhasapopulationdensityof9.2personspersquarekilometre,makingitoneofthemostdenselypopulatedregionsofNamibia.Theregioniswellknown for its Mopani trees and mopaniworms.Anestimated95percentofthepopulationlivesin rural areas and the people of the region are mainly engaged in mixed farming.


Overthe2001to2011period,theregionregistereda22percentagepointreductioninpovertyincidence,withallconstituenciesshowingsignificantpovertyreductions.Thecurrentpovertyheadcountisestimatedat28.6percent(69545people;46935fewerthanin2001).Nineofthetwelveconstituenciesregisteredpovertyreductionsofmorethan20percentagepointsandonlyEtayihadareductionoflessthan10percentagepoints.ThehighestreductionwasregisteredinOshikukuconstituency(32percentagepoints),followedbyOutapiandTsandi(28percentagepoints).

Table 22: Omusati Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Omusati Region 2001 Rank 2011 Rank ChangeOshikuku 49.3 8 17.0 12 -32.3Elim 48.1 9 23.5 11 -24.5Ogongo 47.2 10 25.5 10 -21.8Okahao 52.3 5 26.1 9 -26.2Tsandi 53.9 4 26.3 8 -27.6Outapi 55.4 2 27.2 7 -28.2Anamulenge 54.0 3 27.5 6 -26.5Ruacana 51.6 6 28.0 5 -23.5Otamanzi 46.7 11 30.3 4 -16.4Etayi 40.7 12 30.9 3 -9.8Onesi 51.0 7 34.6 2 -16.4Okalongo 59.0 1 36.0 1 -22.9Regional rate 50.9 28.6 -22.2

FromTable22,above,andasshowninMap21,below,in2011thehighestincidenceofpovertywasrecordedinOkalongo(36percent),whilethelowestpovertyincidenceofpovertywasinOshikukuconstituency(17percent).Overthe2001to2011period,thepopulationoftheregiongrewataslowaverageof0.6percentperannum,withonlyOutapi(1.6percent)andRuacana(2.8percent)constituenciesregisteringpopulationgrowthratesofmorethan1percent.

Map 21: Omusati Region Poverty Headcount Rate in 2011 (upper bound poverty line)


Omusatiregion’sliteracyrateisestimatedat88percent,whileanestimated10percentofthepopulationhasneverattainedformaleducation.Ruacana(20percent)andOnesi(15percent)have the highest proportion of people with no formal education. The economically activepopulationisestimatedat49percent.Ofthese,42percentareunemployed.About50percentof the employed population is in the agricultural sector, while manufacturing, tourism andlogisticsemploy2percenteach.Construction,andwholesaleandretailtradearealsoimportantsectorsaccountingfor4.4and3.6percentoftheregion’semployment,respectively.Thepoorerconstituencies ofOkalongo, Etayi andOntamazi have old age pensions as theirmain sourceofincome,whileinOnesi,subsistencefarmingisthemainsourceofincome.Anestimated52percentofthehouseholdshaveaccesstosafedrinkingwaterbutinOtamanzi,oneofthepoorestconstituencies,only25percentofhouseholdshavethisaccess.

Map22showsthatwiththeexceptionofEtayi(9.8percent),Otamanzi(16percent)andOnesi(16percent),allconstituenciesrecordedpovertyreductionsofmorethan20percentagepointsoverthe2001to2011period.TherecentproclamationofformervillagesinOutapi,Oshikuku,OkahaoandRuacanaconstituenciesastownsandtheattendantinvestmentinpublicservicescouldpartlyexplainthereductioninpovertynotedintheseconstituencies.

Map 22: Omusati Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Table23showsthattheincidenceofseverepovertyisestimatedat14percent,havingdeclinedby a remarkable 18 percentage points between 2001 and 2011. Okalongo and Onesi are the two constituencieswiththehighestincidenceofseverepoverty,at19percent.


Table 23: Omusati Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Omusati Region 2001 Rank 2011 Rank ChangeOkalongo 39.8 1 18.7 1 -21.1Outapi 36.1 2 13.2 7 -22.9Onesi 33.9 3 18.6 2 -15.3Elim 33.0 4 10.7 11 -22.4Okahao 32.4 5 12.4 8 -20.0Tsandi 32.3 6 12.2 9 -20.1Anamulenge 32.3 7 13.6 6 -18.7Ruacana 30.4 8 14.6 5 -15.8Oshikuku 29.3 9 7.5 12 -21.8Ogongo 27.0 10 11.7 10 -15.3Otamanzi 26.5 11 14.7 4 -11.8Etayi 23.9 12 15.3 3 -8.7Regional rate 31.6 14.1 -17.5

3.3.12 Oshana regionOshanaisoneofthethreeregionswhichdoesnothaveaninternationalboundary.ItisborderedbyOmusati,Kunene,OshikotoandOhangwenaregions.In2011,Oshanahadapopulationof176674,accountingfor8.4percentofthenationalpopulation.Intermsofgeographicsize,thisisthesmallestofthethirteenregions,coveringatotalof8653km2.Oshanaregionhasapopulationdensityof20.4personspersquarekilometre.Overthe2001to2011period,thepopulationoftheregiongrewby0.9percentperannum.ThemostlyurbanconstituenciesofOngwendiva,OndangwaandOshakatiEastregisteredgrowthratesof2.4percent,1.5percentand1.2percentperannum,respectively.TheOshakati-Ongwediva-Ondangwacomplexhasexperiencedarapidrateofurbanisationandan influxofpeople fromotherpartsof thecountry.Together thesetownsformanimportantcommercialhub,providingemploymentopportunitiesforpeopleinnorthern Namibia. Table 24: Oshana Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Oshana Region 2001 Rank 2011 Rank ChangeOngwediva 22.8 8 14.4 10 -8.4Oshakati East 19.0 9 14.9 9 -4.1Oshakati West 13.5 10 15.6 8 2.1Ondangwa 26.6 7 18.1 7 -8.5Uuvudhiya 42.6 3 24.1 6 -18.5Okatana 41.9 4 27.4 5 -14.5Ompundja 42.7 2 30.2 4 -12.5Okatyali 49.1 1 32.7 3 -16.3Okaku 38.2 5 33.2 2 -5.0Uukwiyu 36.0 6 36.0 1 0.1Regional rate 28.3 21.1 -7.1


In2011,theincidenceofpovertyintheregionwasestimatedat21percent(37278people),havingdeclinedby7percentagepointsfromthe2001figure.UukwiyuisthepoorestconstituencyinOshana,withanestimated36percentofthepeopleclassifiedaspoor.ItisfollowedcloselybyOkakuandOkatyali,wheretheincidenceofpovertyis33percent.PovertyincidenceislowestinOngwendiva,at14percent.Thegreatestreductionintheincidenceofpovertyoverthe2001to2011periodwasrecordedinUuvudhiya(19percentagepoints),Okatyali(16percentagepoints),Okatana(15percentagepoints)andOmpundja(13percentagepoints),whileOkaku(5percentagepoints)recordedtheleastprogressinreducingpoverty.PovertyincreasedinOshakatiWestby2percentage points over the same period.

From Map 23 it can be seen that poverty is highest in Ukwiyu and lowest in the mainly urban constituenciesofOngwendiva,OshakatiWestandOshakatiEast.

Map 23: Oshana Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Map24showsthatthegreatestdeclineinpovertyoccuredinUuvidhiya,OkatyaliandOkatanaconstituencies,asdiscussedabove.Theregionhasaliteracyrateof96percent,whileanestimated6percentofthepopulationhasneverattendedschool.Theeconomicallyactivepopulationisestimatedat61percentofthepopulation,ofwhich37percentareunemployed.Theagriculturalsector employs 26 percent of the employed population, while manufacturing, tourism andlogisticsemploy3.9percent,3.5percentand3.9percentof theworkforce, respectively.Theconstructionsectoremploys6.3percentoftheregion’slabourforce.Okatana,OkatyaliandOkakuhaveunemploymentratesofmorethan30percent.Salariesandwages(40percent)andoldagepensions(19percent)arethemainsourcesof incomefortheregiongenerally,whileoldagepensionsalonearetheleadingincomesourceinOkaku(41percent)andOmpundja(44percent).


Map 24: Oshana Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

In2011,theincidenceofseverelypoorindividualswasestimatedat10percent,adeclineof5percentagepointsfrom15percentin2001.Uukwiyuconstituencyhasthehighestincidenceofseverepoverty,at19percent,followedbyOkaku,OmpundjaandOkatyaliconstituencies. Table 25: Oshana Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Oshana Region 2001 Rank 2011 Rank Change

Oshakati West 6.2 10 6.8 8 0.6

Oshakati East 10.4 9 6.5 9 -3.9

Ongwediva 12.6 8 6.3 10 -6.3

Ondangwa 13.6 7 8.4 7 -5.2

Uukwiyu 18.2 6 18.8 1 0.6

Okaku 21.7 5 16.9 2 -4.8

Okatana 22.5 4 13.7 5 -8.9

Uuvudhiya 22.6 3 12.2 6 -10.4

Okatyali 26.2 2 15.9 4 -10.3

Ompundja 26.2 1 16.7 3 -9.4



3.3.13 Oshikoto region

OshikotoregionishometoEtoshaNationalPark,whichisoneofthemajortouristattractionsinNamibiaandSouthernAfrica.Theregionhasapopulationof181973,ofwhichthevastmajority(87percent)livesinruralareas.In2011,theincidenceofpovertyintheregionwas43percent(77 520 people), representing a 15 percentage point reduction from the 2001 figure of 57percent.ThepoorestconstituencyintheregionisOkankolo,with63percentofthepopulationclassifiedaspoor.ItisfollowedbyEengodi(55percent)andOnyaanya(50percent).TheleastpoorconstituencyisTsumebwhereanestimated19percentofthepopulationisclassifiedaspoor.Overthe2001to2011period,thegreatestdeclineinpoverty,of23percentagepoints,wasrecordedinOnayenaconstituency.Thiswasfollowedbya21percentagepointreductioninOmunteleconstituencyand20percentagepointsinOniipaconstituency.

Table 26: Oshikoto Region Poverty Headcount, 2001 - 2011 (upper bound poverty line)

Oshikoto Region 2001 Rank 2011 Rank Change

Tsumeb 18.8 10 18.5 10 -0.3

Olukonda 48.5 9 31.5 9 -17.0

Oniipa 52.4 8 32.6 8 -19.9

Onayena 62.1 5 39.2 7 -22.8

Guinas 54.1 7 43.9 6 -10.3

Omuthiyagwiipundi 61.2 6 44.8 5 -16.5

Omuntele 66.9 3 46.1 4 -20.9

Onyaanya 62.2 4 50.4 3 -11.8

Eengodi 69.1 2 54.7 2 -14.5

Okankolo 71.7 1 62.9 1 -8.8


Map 25: Oshikoto Region Poverty Headcount Rate in 2011 (upper bound poverty line)


AscanbeseenfromTable26andMap25,above,withtheexceptionofTsumeb,allconstituencieshave poverty levels higher than 30 percent and many are considerably higher. Between 2001 and 2011thepopulationof theregiongrewby1.2percentperannumalthoughtheurbanareasrecordedahighergrowthrateof4.6percentperannum.EengodiandTsumebconstituenciesgrewby3.6and2.9percentperannum,respectively,whileOnkankolo,thepoorestconstituency,hadapopulationgrowthrateof1.9percent.Theregionhasaliteracyrateof88percent,withabout12percentofthepopulationaged6yearsandabovehavingneverattendedschool. InGuinasconstituency,withapovertyheadcountof44percent,morethanonethird(37percent)ofpeoplehaveneverattainedformaleducation.ThisisfollowedbyEengodi(20percent)andOkankolo(18percent).

The economically active portion of the population is estimated at 57 percent. Of these, 40percentareunemployed.Theagriculturalsectoremploys49percentoftheworkingpopulation,whilemanufacturing,tourismandlogisticsaccountfor3percent,2.6percentand2.7percentofemployment,respectively.Otherimportantsectorsprovidingjobsareconstruction(4.7percent),wholesale and retail trade (5 percent), and mining (2.5 percent). Unemployment is highestinOmuthiyaqwiipundi andOnayena constituencies, at 43percentof the labour force.Accessto safedrinkingwater is estimated tobe70percent for the regionbut varies greatly acrossconstituencies.Onlyanestimatedonethird(33.3percent)ofthehouseholdsinOnkankolo,thepoorestconstituencyintheregion,haveaccesstosafedrinkingwater,withthecorrespondingfigureforOnayenabeing39percent.Inthesecondpoorestconstituency,Eengodi,morethanhalf (57percent)ofhouseholdshaveaccesstosafewater.Whileabout11and20percentofhouseholdsintheregionuseelectricityforcookingandlightningrespectively,inOkankolo,only2and3percentofhouseholds,respectively,doso.

Map26 confirms that, over the2001 to2011period, all constituencies recordeddeclines inpovertyheadcountrate,albeitinvaryingdegrees,withthegreatestreductionsbeingrecordedinOnayenaandtheleastinTsumebconstituency.

Map 26: Oshikoto Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)


Table27presentstheincidenceofseverepovertyinOshikotoregion,andchangesbetween2001and2011.Anestimated27percentofthepopulation isseverelypoor.WiththeexceptionofTsumeb,alltheconstituenciesinOshikotoregistereddeclinesintheincidenceofseverepoverty,withOnayenaandOniiparegisteringthegreatestdeclines.Despitea10percentreductionintheincidenceofseverepovertybetween2001and2011,Okankoloconstituencystillhasmorethan40percentofitspopulationlivinginextremepoverty.

Table 27: Oshikoto Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Oshikoto Region 2001 Rank 2011 Rank Change

Tsumeb 8.9 10 9.0 10 0.1

Olukonda 28.6 9 17.3 9 -11.3

Oniipa 35.3 7 18.6 8 -16.7

Onayena 41.8 6 23.0 7 -18.8

Guinas 34.4 8 27.2 6 -7.3

Omuthiyagwiipundi 41.9 5 28.1 5 -13.8

Omuntele 43.7 3 28.6 4 -15.1

Onyaanya 43.2 4 32.3 3 -10.9

Eengodi 47.9 2 36.2 2 -11.7

Okankolo 54.0 1 43.8 1 -10.2


3.3.14 Otjozondjupa regionOtjozondjupa region has a surface area of 105 185km2,accountingfor12.8percentofthelandareaofNamibia,andishometo6.8percentofthepopulation.Theregionislargelysemi-aridwithannualrainfallrangingfrom300to600mm,andlikeOmahekeregion,itischaracterisedbylargecommercialranches.Otjozondjupaisdividedintosixconstituencies-Grootfontein,Otavi,Otjiwarongo,Omatako,Okakarara,OkahandjaandTsumkwe.ThecentraltownofOtjiwarongoservesastheadministrativeheadquartersoftheregion.TheregionishometoOhorongocementfactory and B2 gold mine.

Between2001and2011,thepopulationgrewatanaveragerateof0.6percentperannum.Overthepastdecadetheregionhasexperiencedrapidurbanization,with54percentofthepopulationcurrentlylivinginurbanareas,comparedto41percentin2001.Theregionhasarelativelyyoungpopulation,withanestimated36percentof thepopulationbeingunder15yearsofage.Anestimated83percentofthepopulationisliterate.Ninetyfivepercentofhouseholdshaveaccesstosafedrinkingwaterand56percentuseelectricityforlighting.

AlthoughOtjozondjuparegionisknownforitspotentialforlargescalecommercialfarms,themainsourceofincomeissalaryandwages(60percent),withagriculture,businessandpensionsjointlyconstitutingthemainsourceof incomefor10percentofthepopulation.Seventytwopercentofthepopulationiseconomicallyactiveandofthese,37percentisunemployed.At27.5percent(39573),thepovertyheadcountrateisslightlyabovethenationalaverage,whiletherateofseverepovertyisestimatedat14.9percent.


Poverty ishighest inmostly rural constituenciescharacterisedbysubsistence farming.At the2011timepoint,Tsumkwewasthepoorestconstituency,withapovertyheadcountrateof65percent,representingamarginaldecreaseof1percentagepointfrom2001,whiletheleastpoorconstituencyisOtjiwarongowithapovertyheadcountrateof17percent.Tsumkweconstituencyalsohasthelowestliteracyrate,at58percent,withmorethanonethird(36percent)ofpeopleaged15 years and abovehavingnever attended school. The labour force is estimated at 69percentofthepopulationandmorethanhalf(52percent)ofthisgroupisunemployed. Table 28: Otjozondjupa Region Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Otjozondjupa Region 2001 Rank 2011 Rank Change

Grootfontein 23.7 5 23.8 5 0.0

Okahandja 20.8 6 18.7 6 -2.1

Okakarara 49.7 1 37.2 2 -12.5

Omatako 27.4 3 28.6 4 1.2

Otavi 25.9 4 32.1 3 6.2

Otjiwarongo 16.8 7 16.5 7 -0.3

Tsumkwe 65.7 1 64.6 1 -1.2


Map 27: Otjozondjupa Region Poverty Headcount Rate in 2011 (upper bound poverty line)

Between 2001 and 2011 the poverty headcount rate in the region declined by 3 percentage points,withOkakararaconstituencyexperiencingthehighestreductionof13percentagepoints.However,notall constituencies registereda reduction inpoverty,asOtaviandOmatakohadincreasesof6percentand1percent,respectively.Accesstowaterisestimatedat92percentofthepopulationwhileanestimated51percentofthepopulationuseselectricityforlighting.


Map 28: Otjozondjupa Region Change in Poverty Headcount Rate, 2001 - 2011 (upper bound poverty line)

Thechangesintheincidenceofseverepovertymirrorcloselythoseofpoverty,withTsumkweconstituency having the highest proportion of severe poverty (45 percent) and Okakararaconstituencyexperiencing thehighest reduction in severepoverty. The incidenceof severelypoorhouseholdsinOtjozondjuparegionisestimatedat15percent,areductionof3percentagepointsfromthe2001figure.

Table 29: Otjozondjupa Region Poverty Headcount Rate, 2001 - 2011 (lower bound poverty line)

Otjozondjupa Region 2001 Rank 2011 Rank Change

Grootfontein 13.6 4 11.6 5 -2.0

Okahandja 9.4 6 8.4 6 -1.0

Okakarara 33.5 2 21.9 2 -11.7

Omatako 15.4 3 15.2 4 -0.3

Otavi 13.1 5 17.0 3 3.9

Otjiwarongo 7.5 7 7.1 7 -0.4

Tsumkwe 45.9 1 44.9 1 -1.0



4 CONCLUSIONS AND RECOMMENDATIONS

4.1 Conclusions

PovertyinNamibiahasdeclinedoveradecadeonaggregate.Atthenationallevel,therewasadeclineof11percentagepointsinthepovertyheadcountoverthe2001to2011period,withthegreatestdeclinesbeingregisteredinthenorthernregionsofOhangwena,Omusati,KuneneandOshikoto,aswellastheeasternregionofOmaheke.However,thedeclineinpovertyheadcountwasnotuniformacross thethirteenregionsof thecountry.Whileelevenoutof thethirteenregionsreporteddeclinesinthepovertyheadcount,tworegions(ZambeziandKhomas)recordedincreases in the incidence of poverty over the same period.

At the constituency level, the biggest percentage point reduction in the poverty headcountwasregisteredinthenorthernregionsofOhangwenaandOmusati,whilethebiggestincreaseoccurredinthenorth-easternZambeziregion.Eenhana,Endola,Engela,OkongoandOngengaconstituenciesinOhangwenaregionandOshikukuconstituencyinOmusatiregionallregisteredareductioninpovertyheadcountofmorethan30percentagepoints,whileKatimaMuliloUrbanandKongolaconstituencies inZambezi regionhadan increase inpovertyheadcountofmorethan 10 percentage points over this period. Theincidenceofseverepovertyinthecountrydeclinedby9percentagepointsoverthe2001to2011period.However,whileseverepovertydeclinedinelevenoutofthethirteenregions,two regions (Zambezi and Khomas) recorded increases in the severe poverty incidence. Attheconstituencylevel,thegreatestdecline,intermsofpercentagepoints,intheincidenceofseverepovertywasrecordedinOkongoconstituencyinOhangwenaregion,followedbyEngela,OngengaandEndolaconstituencies inOhangwenaregion,aswellasSesfonteinconstituencyinKunene.Alloftheseregisteredreductionsofmorethan25percentagepoints.Ontheotherhand,thegreatestincreaseintheincidenceofseverepoverty,of7percentagepointsorhigher,wasregisteredinSibbinda,LinyantiandKongola,allinZambeziregion,aswellasinKapakoinKavango region.

Overthepastdecade,povertyinNamibiacontinuedtoexhibitanurban-ruraldivide.Thesevenpoorestregions–Kavango,Oshikoto,Zambezi,Kunene,Ohangwena,OmusatiandOtjozondjupa–hadpovertyincidencesabovethenationalaverageof26.9percent.Theseareregionswherethemajorityoftheirpopulationlivesinruralareas,whilethelesspoorregionsofKhomasandErongo,theeconomichubsofthecountrywithrelativelymoreemploymentopportunities,havelargelyurbanpopulations.

In2011,KavangowasthepoorestregioninthecountryandKhomaswastheleastpoorregion,whileEpupawas thepoorest constituency in thecountry,andWindhoekEast the leastpoorconstituency.


4.2 Policy Recommendations

Thefollowingpolicyrecommendationsaredrawnfromthefindingsofthisstudy:

• Having located where the poor are, there is a need for targeted interventions in termsresourceallocation,public/privateinvestmentandservicedelivery.

• The widely acknowledged negative relationship between education and poverty clearlyexists inNamibia. Therefore, it is important that, as a long term strategy, investment ineducationisusedtoreducepoverty.However,shorttermservicedeliveryisalsocrucialforpovertyreduction.Thiswillhavetheeffectofalsoaddressingmigrationwhichhasbeenanincreasing contributor to poverty in urban areas.

• Nationalpoliciesandanationalagendaforpovertyreductionneedtobelocalisedinorderto make a notable impact.

• Itisalsoimportanttogainabetterunderstandingofthesectors,programmesandprojects,aswellastheinstitutionalfactorsdrivingthereportedreductionsinthepovertyheadcount.Thus,furtherresearchtoidentifythefactorsdrivingpovertyreductionisrecommended.


REFERENCES

1. Dobson,A. J. (2002)An Introduction to Generalized Linear Models, SecondEdition,BocaRaton,London,NewYork,WashingtonDC:ChapmanHall.

2. CentralBureauofStatistics(2008)A Review of Poverty and Inequality in Namibia,Windhoek:CentralBureauofStatistics,NationalPlanningCommission.

3. Elbers, C., J. O. Lanjouw and P. Lanjouw (2003) ‘Micro-level estimation of poverty andinequality’,Econometrica,71:355-364.

4. National Planning Commission (2012) Fourth National Development Plan, Windhoek,Namibia.

5. NationalPlanningCommission(2004)Namibia Vision 2030,PolicyFrameworkforLong-TermNationalDevelopment,Windhoek,Namibia.

6. Namibia Statistics Agency (2012) Namibia 2011 Population and Housing Census Report, Windhoek,Namibia.

7. Namibia Statistics Agency (2012) Namibia Household Income and Expenditure Survey 2009/2010,Windhoek,Namibia.

8. Namibia StatisticsAgency (2014)Namibia 2011 Population and Housing Census Regional Profiles,Basicanalysiswithhighlights,Windhoek,Namibia.


Anne

x 1:

Det

aile

d Po

vert

y He

adco

unt u

sing

the

Upp

er P

over

ty L

ine

2001

and

201

1

Regi

on an

d Co

nstit

uenc

y20

0120

11Ch

ange

Pove

rty

Rate

Popu

latio

nPo

orPo

verty

Ra

tePo

pulat

ion

Poor

Pove

rty

Rate

Popu

latio

nPo

orNo

n-po

or

Zam

bezi

32.0

79 82

6 25

544

39.3

90 59

6 35

604

7.310

770

10 06

0 71

0 Ka

bbe

42.2

14 96

2 6 3

14

49.1

14 51

8 7 1

28

6.9-4

44

814

-1 25

8 Ka

tima M

ulilo

Rura

l38

.314

566

5 579

43

.016

399

7 052

4.7

13

1 473

-1

460

Katim

a Muli

lo Ur

ban

6.422

704

1 453

17

.228

362

4 878

10

.85 6

58

3 425

2 2

33

Kong

ola47

.44 4

19

2 095

58

.15 6

58

3 287

10

.7-

1 193

-1

193

Linya

nti41

.213

985

5 762

49

.415

477

7 646

8.2

1 492

1 8

84

-392

Si

bbind

a45

.89 1

90

4 209

55

.010

182

5 600

9.2

992

1 391

-3

99

Eron

go9.3

107 6

63

10 01

3 6.3

150 8

09

9 501

-3

.043

146

-512

43

658

Aran

dis5.4

7 590

41

0 6.3

10 09

3 63

6 0.9

2 503

22

6 2 2

77

Daur

es38

.310

289

3 941

20

.111

350

2 281

-1

8.21 0

61

-1 65

9 2 7

20

Karib

ib16

.912

084

2 042

15

.413

320

2 051

-1

.51 2

36

9 1 2

27

Omar

uru

9.67 1

56

687

11.7

8 577

1 0

04

2.11 4

21

317

1 104

Sw

akop

mund

4.426

310

1 158

3.6

44 72

5 1 6

10

-0.8

18 41

5 45

2 17

963

Walv

is Ba

y Rur

al3.4

16 29

3 55

4 3.7

26 91

6 99

6 0.3

10 62

3 44

2 10

181

Walv

is Ba

y Urb

an3.4

27 94

1 95

0 2.4

35 82

8 86

0 -1

.07 8

87

-90

7 977

Ha

rdap

20.4

68 24

9 13

923

17.2

79 50

7 13

675

-3.2

11 25

8 -2

48

11 50

6 Gi

beon

30.2

11 54

1 3 4

85

24.8

12 12

2 3 0

06

-5.4

581

-479

1 0

60

Marie

ntal R

ural

27.5

13 94

6 3 8

35

23.3

15 30

8 3 5

67

-4.2

1 362

-2

68

1 630

Ma

rienta

l Urb

an15

.413

109

2 019

15

.615

557

2 427

0.2

2 448

40

8 2 0

40

Reho

both

Rura

l27

.17 5

24

2 039

24

.27 2

88

1 764

-2

.9-2

36

-275

39

Re

hobo

th Ur

ban E

ast

13.0

12 89

1 1 6

76

13.8

18 03

5 2 4

89

0.85 1

44

813

4 331

Re

hobo

th Ur

ban W

est

9.09 2

38

831

4.011

197

448

-5.0

1 959

-3

84

2 343


Kara

s18

.069

329

12 47

9 14

.577

421

11 22

6 -3

.58 0

92

-1 25

3 9 3

45

Berse

ba38

.59 0

64

3 490

27

.210

589

2 880

-11

.31 5

25

-609

2 1

34

Kara

sbur

g21

.615

758

3 404

20

.816

470

3 426

-0

.871

2 22

69

0 Ke

etman

shoo

p Rur

al25

.86 3

99

1 651

23

.07 2

19

1 660

-2

.882

0 9

811

Keetm

ansh

oop U

rban

9.815

777

1 546

9.9

19 44

7 1 9

25

0.13 6

70

379

3 291

Lu

deritz

9.714

542

1 411

7.0

13 85

9 97

0 -2

.7-6

83

-440

-2

43

Oran

jemun

d7.6

7 789

59

2 2.9

9 837

28

5 -4

.72 0

48

-307

2 3

55

Kava

ngo

57.9

202 6

94

117 3

60

53.2

223 3

52

118 8

23

-4.7

20 65

8 1 4

63

19 19

5 Ka

heng

e60

.330

903

18 63

5 60

.629

799

18 05

8 0.3

-1 10

4 -5

76

-528

Ka

pako

55.8

26 26

3 14

655

62.6

26 98

3 16

891

6.872

0 2 2

37

-1 51

7 Ma

shar

e75

.316

007

12 05

3 60

.515

688

9 491

-1

4.8-3

19

-2 56

2 2 2

43

Mpun

gu60

.818

660

11 34

5 60

.320

787

12 53

5 -0

.52 1

27

1 189

93

8 Mu

kwe

65.2

27 25

0 17

767

58.2

27 69

0 16

116

-7.0

440

-1 65

1 2 0

91

Ndiyo

na69

.319

565

13 55

9 56

.620

633

11 67

8 -1

2.71 0

68

-1 88

0 2 9

48

Rund

u Rur

al W

est

46.8

26 62

3 12

460

45.1

38 28

1 17

265

-1.7

11 65

8 4 8

05

6 853

Ru

ndu U

rban

30.3

19 17

3 5 8

09

18.6

20 95

3 3 8

97

-11.7

1 780

-1

912

3 692

Ru

ndu R

ural

East

61.5

18 25

0 11

224

56.3

22 53

8 12

689

-5.2

4 28

8 1 4

65

2 823

Kh

omas

3.425

0 262

8 5

09

4.634

2 141

15

738

1.291

879

7 230

84

649

Tobia

s Hain

yeko

4.834

348

1 649

9.6

45 91

2 4 4

08

4.811

564

2 759

8 8

05

Katut

ura C

entra

l6.2

21 24

3 1 3

17

4.024

608

984

-2.2

3 365

-3

33

3 698

Ka

tutur

a Eas

t4.5

17 74

5 79

9 4.1

18 50

1 75

9 -0

.475

6 -4

0 79

6 Kh

omas

dal N

orth

1.627

950

447

2.443

921

1 054

0.8

15 97

1 60

7 15

364

Sowe

to2.9

13 86

5 40

2 2.1

15 12

1 31

8 -0

.81 2

56

-85

1 341

Sa

mora

Mac

hel

3.329

382

970

4.350

110

2 155

1.0

20 72

8 1 1

85

19 54

3 W

indho

ek E

ast

0.217

674

35

0.122

712

23

-0.1

5 038

-1

3 5 0

51

Wind

hoek

Rur

al11

.320

212

2 284

7.7

22 25

4 1 7

14

-3.6

2 042

-5

70

2 612

W

indho

ek W

est

0.542

201

211

0.453

438

214

-0.1

11 23

7 3

11 23

4 Mo

ses/G

aroë

b1.4

25 64

2 35

9 8.4

45 56

4 3 8

27

7.019

922

3 468

16

454


Kune

ne53

.768

735

36 91

1 38

.986

856

33 78

7 -1

4.818

121

-3 12

4 21

245

Epup

a76

.813

129

10 08

3 69

.217

696

12 24

6 -7

.64 5

67

2 163

2 4

04

Kama

njab

37.0

6 454

2 3

88

19.9

8 441

1 6

80

-17.1

1 987

-7

08

2 695

Kh

orixa

s34

.311

501

3 945

18

.812

566

2 362

-1

5.51 0

65

-1 58

2 2 6

47

Opuw

o65

.220

892

13 62

2 44

.127

272

12 02

7 -2

1.16 3

80

-1 59

5 7 9

75

Outjo

22.7

9 154

2 0

78

18.0

12 44

7 2 2

40

-4.7

3 293

16

3 3 1

30

Sesfo

ntein

69.0

7 605

5 2

47

40.0

8 434

3 3

74

-29

829

-1 87

4 2 7

03

Ohan

gwen

a62

.822

8 384

14

3 425

35

.324

5 446

86

642

-27.5

17 06

2 -5

6 783

73

845

Eenh

ana

62.6

18 69

0 11

700

31.1

21 08

9 6

559

-31.5

2 399

-5

141

7 540

En

dola

62.0

24 80

4 15

378

27.7

25 59

1 7 0

89

-34.3

787

-8 29

0 9 0

77

Enge

la59

.221

832

12 92

5 26

.024

271

6 310

-3

3.22 4

39

-6 61

4 9 0

53

Epem

be72

.214

860

10 72

9 48

.416

229

7 855

-2

3.81 3

69

-2 87

4 4 2

43

Ohan

gwen

a57

.817

887

10 33

9 29

.917

468

5 223

-2

7.9-4

19

-5 11

6 4 6

97

Okon

go73

.721

551

15 88

3 41

.025

698

10 53

6 -3

2.74 1

47

-5 34

7 9 4

94

Omun

daun

gilo

76.1

8 115

6 1

76

51.8

7 855

4 0

69

-24.3

-260

-2

107

1 847

On

dobe

51.0

22 25

3 11

349

39.8

23 95

4 9 5

34

-11.2

1 701

-1

815

3 516

On

geng

a65

.321

706

14 17

4 32

.322

075

7 130

-3

3.036

9 -7

044

7 413

Os

hikan

go58

.9 25

221

14 85

5 31

.628

635

9 049

-2

7.33 4

14

-5 80

7 9 2

21

Omulo

nga

63.1

31 46

5 19

854

40.4

32 58

1 13

163

-22.7

1 116

-6

692

7 808

Om

ahek

e41

.668

039

28 30

4 26

.271

233

18 66

3 -1

5.43 1

94

-9 64

1 12

835

Amini

us50

.912

392

6 308

28

.412

306

3 495

-2

2.5-8

6 -2

813

2 727

Go

babis

16.9

15 11

9 2 5

55

17.1

20 99

3 3 5

90

0.25 8

74

1 035

4 8

39

Kalah

ari

44.9

9 443

4 2

40

27.1

7 611

2 0

63

-17.8

-1 83

2 -2

177

345

Otjin

ene

55.1

7 790

4 2

92

33.2

7 306

2 4

26

-21.9

-484

-1

867

1 383

Ot

jombin

de57

.06 5

60

3 739

36

.36 8

51

2 487

-2

0.729

1 -1

252

1 543

St

einha

usen

35.5

9 600

3 4

08

26.1

10 06

0 2 6

26

-9.4

460

-782

1 2

42

Epuk

iro52

.57 1

35

3 746

31

.26 1

06

1 905

-2

1.3-1

029

-1 84

1 81

2


Omus

ati

50.9

228 8

42

116 4

81

28.6

243 1

66

69 54

5 -2

2.314

324

-46 9

35

61 25

9 An

amule

nge

54.0

12 61

7 6 8

13

27.5

13 41

0 3 6

88

-26.5

793

-3 12

5 3 9

18

Elim

48.1

10 85

0 5 2

19

23.5

11 40

6 2 6

80

-24.6

556

-2 53

8 3 0

94

Etay

i40

.735

130

14 29

8 30

.935

101

10 84

6 -9

.8-2

9 -3

452

3 423

Og

ongo

47.2

19 61

1 9 2

56

25.5

19 54

6 4 9

84

-21.7

-65

-4 27

2 4 2

07

Okah

ao52

.317

751

9 284

26

.117

548

4 580

-2

6.2-2

03

-4 70

4 4 5

01

Okalo

ngo

59.0

28 71

9 16

944

36.0

30 60

9 11

019

-23.0

1 890

-5

925

7 815

On

esi

51.0

12 99

5 6 6

27

34.6

13 14

9 4 5

50

-16.4

154

-2 07

8 2 2

32

Oshik

uku

49.3

8 299

4 0

91

17.0

9 093

1 5

46

-32.3

794

-2 54

6 3 3

40

Outap

i55

.431

496

17 44

9 27

.236

934

10 04

6 -2

8.25 4

38

-7 40

3 12

841

Ruac

ana

51.6

11 20

4 5 7

81

28.0

14 85

7 4 1

60

-23.6

3 653

-1

621

5 274

Ts

andi

53.9

27 04

9 14

579

26.3

28 01

8 7 3

69

-27.6

969

-7 21

1 8 1

80

Otam

anzi

46.7

13 12

1 6 1

28

30.3

13 49

5 4 0

89

-16.4

374

-2 03

9 2 4

13

Osha

na28

.316

1 916

45

822

21.1

176 6

74

37 27

8 -7

.214

758

-8 54

4 23

302

Okak

u38

.220

354

7 775

33

.219

007

6 310

-5

.0-1

347

-1 46

5 11

8 Ok

atana

41.9

15 35

2 6 4

32

27.4

14 80

1 4 0

55

-14.5

-551

-2

377

1 826

Ok

atyali

49.1

2 812

1 3

81

32.7

3 187

1 0

42

-16.4

375

-339

71

4 Om

pund

ja42

.74 4

48

1 899

30

.24 6

59

1 407

-1

2.521

1 -4

92

703

Onda

ngwa

26.6

31 69

4 8 4

31

18.1

36 84

6 6 6

69

-8.5

5 152

-1

761

6 913

On

gwed

iva22

.826

700

6 088

14

.434

065

4 905

-8

.47 3

65

-1 18

2 8 5

47

Osha

kati E

ast

19,0

24 26

9 4 6

11

14.9

27 22

7 4 0

57

-4.1

2 958

-5

54

3 512

Os

haka

ti Wes

t13

.519

862

2 681

15

.620

676

3 225

2.1

814

544

270

Uukw

iyu36

.012

047

4 337

36

.012

092

4 353

0

45

16

29

Uuvu

dhiya

42.6

4 378

1 8

65

24.1

4 114

99

1 -1

8.5-2

64

-874

61

0


Oshi

koto

57.3

161 0

07

92 25

7 42

.618

1 973

77

520

-14.7

20 96

6 -1

4 737

35

703

Eeng

odi

69.1

15 21

7 10

515

54.7

21 73

2 11

887

-14.4

6 515

1 3

72

5 143

Gu

inas

54.1

9 033

4 8

87

43.9

10 73

0 4 7

10

-10.2

1 697

-1

76

1 873

Ok

anko

lo71

.713

065

9 368

62

.915

831

9 958

-8

.82 7

66

590

2 176

Ol

ukon

da48

.59 2

26

4 475

31

.59 5

59

3 011

-1

7.033

3 -1

464

1 797

Om

untel

e66

.915

160

10 14

2 46

.116

865

7 775

-2

0.81 7

05

-2 36

7 4 0

72

Omuth

iyagw

iipun

di61

.223

674

14 48

8 44

.826

183

11 73

0 -1

6.42 5

09

-2 75

9 5 2

68

Onay

ena

62.1

15 45

9 9 6

00

39.2

15 39

2 6 0

34

-22.9

-67

-3 56

6 3 4

99

Oniip

a52

.424

730

12 95

9 32

.624

939

8 130

-1

9.820

9 -4

828

5 037

On

yaan

ya62

.220

536

12 77

3 50

.420

902

10 53

5 -11

.836

6 -2

239

2 605

Ts

umeb

18.8

14 90

7 2 8

03

18.5

19 84

0 3 6

70

-0.3

4 933

86

8 4 0

65

Otjo

zond

jupa

30.4

135 3

84

41 15

7 27

.514

3 903

39

573

-2.9

8 519

-1

583

10 10

2 Gr

ootfo

ntein

23.7

23 79

3 5 6

39

23.8

24 87

8 5 9

21

0.11 0

85

282

803

Okah

andja

20.8

18 07

1 3 7

59

18.7

24 45

1 4 5

72

-2.1

6 380

81

4 5 5

66

Okak

arar

a49

.721

820

10 84

5 37

.222

747

8 462

-1

2.592

7 -2

383

3 310

Om

atako

27.4

26 90

8 7 3

73

28.6

17 61

9 5 0

39

1.2-9

289

-2 33

4 -6

955

Otav

i25

.912

378

3 206

32

.112

488

4 009

6.2

110

803

-693

Ot

jiwar

ongo

16.8

23 41

2 3 9

33

16.5

31 81

3 5 2

49

-0.3

8 401

1 3

16

7 085

Ts

umkw

e65

.79 0

02

5 914

64

.69 9

07

6 400

-1

.190

5 48

6 41

9 Na

mib

ia37

.91 8

30 33

0 69

3 695

26

.92 1

13 07

7 56

8 418

-11

.028

2 747

-1

25 27

7 40

8 024


Annex 2: Technical Notes

Foreachperiod,therewasonedatasetavailable,theNHIES,withquiteaccurate incomeandexpenditure data, but with a limited (stratified and clustered) sample. Conversely, there isanotherdataset,theCensus,withnoincomeorexpendituredata,butwithamuchlargerandmorerepresentativesample. Following theapproachofElbersetal. (2003), theanalysis reportedhereused theNHIES topredicttherelationshipbetweenasetofobservablesandincome,andthenusedthisrelationshipto impute the likely per adult equivalent income for each household in the Census using the samesetofobservables.Oncethesevalueswereimputed,thepovertylevelsforeachhouseholdineachconstituencywereaggregated.(Notethatthetermincomeisusedthroughoutbut,asitisregardedasbettermeasuredinadevelopingcountrycontext,itwasexpenditure that was modeledandestimated.)

Throughout itwasassumedthat loggedhousehold incomecanbemodeledastheestimatedlogged household income plus an error component.

Ifitisassumedthatthisrelationshipholdsfortheentirepopulationandthatthequestionsofinterestarecomparablebetweensurveys,thenthismodelcanbeusedtopredictthe loggedconsumption in theCensus.Suppose itwasbelieved that itwouldbepossible toadequatelypredict consumption using only three predictors, the education level of household head,whetherthehousehasacar,andhouseholdsize.Thentherelationshipbetweenthesethreevariablesandhouseholdconsumption(shownintheregressionbelow)canbeusedtopredicttheexpectedlevelofconsumptionforeachhouseholdintheCensus,sinceinformationonallthreeoftheseattributesisavailableintheCensusaswell.

Coef seEducation 0.069*** 43.286Household Size -0.130*** -55.001Car 1.133*** 59.650Constant 6.724*** 371.311Adjusted R2 0.553note:***p<0.01,**p<0.05,*p<0.1

Thisexampleignorestwoimportantissues,whichwillnowbediscussedbriefly.50

Annex 2: Technical Notes

For each period, there was one dataset available, the NHIES, with quite accurate income and expenditure data, but with a limited (stratified and clustered) sample. Conversely, there is another dataset, the Census, with no income or expenditure data, but with a much larger and more representative sample. Following the approach of Elbers et al. (2003), the analysis reported here used the NHIES to predict the relationship between a set of observables and income, and then used this relationship to impute the likely per adult equivalent income for each household in the Census using the same set of observables. Once these values were imputed, the poverty levels for each household in each constituency were aggregated. (Note that the term income is used throughout but, as it is regarded as better measured in a developing country context, it was expenditure that was modeled and estimated.) Throughout it was assumed that logged household income can be modeled as the estimated logged household income plus an error component.

𝑙𝑙𝑙𝑙 𝑌𝑌!! =𝐸𝐸 𝑙𝑙𝑙𝑙𝑌𝑌!!|𝑋𝑋!! + 𝑢𝑢!!

𝑦𝑦!! = 𝐸𝐸 𝑦𝑦!!|𝑋𝑋!! + 𝑢𝑢!!

where 𝑐𝑐 is the subscript for cluster, ℎ is the subscript for the household within the cluster 𝑌𝑌!! is the per capita expenditure of household h in cluster c, 𝑋𝑋!!is the household characteristics for household h in cluster c and 𝑢𝑢!! is the error. If it is assumed that this relationship holds for the entire population and that the questions of interest are comparable between surveys, then this model can be used to predict the logged consumption in the Census. Suppose it was believed that it would be possible to adequately predict consumption using only three predictors, the education level of household head, whether the house has a car, and household size. Then the relationship between these three variables and household consumption (shown in the regression below) can be used to predict the expected level of consumption for each household in the Census, since information on all three of these attributes is available in the Census as well.

Coef se Education 0.069*** 43.286 Household Size -‐0.130*** -‐55.001 Car 1.133*** 59.650 Constant 6.724*** 371.311 Adjusted R2 0.553 note: *** p<0.01, ** p<0.05, * p<0.1

This example ignores two important issues, which will now be discussed briefly.


1. Not every household will earn exactly what they are predicted to earn

While the regressionmodelmightdowell inpredicting theunderlying relationshipbetweenobservablesandexpectedincome,itmightnotdowell inpredictingtheexpectedincomeforeach household individually. Using the predicted values of this regression unchanged assumes thateveryonewhohasthesamesetofobservableswillallhaveexactlythesameincome,whichisobviouslynotthecase.ThisisapparentwhenthedistributionofactualincomeintheNHIESiscomparedtothedistributionofthefittedvaluesinthesamedataset(seebelow).Thefittedvaluesfollowanarrowerdistributionthantheactualincomevalues.

Toaddressthisconcern,themodel incorporatedMonteCarlosimulations.Inthesecondpartofthemodeling,onethousanditerationswererun,eachtimedrawingastochasticerrortermforeveryhousehold.Theaveragepovertyrateisthenestimatedoveralltheiterations,foreachregion,constituencyandforthepopulationasawhole.Thismethodhelpstoaccountforthestochastictermthatwouldotherwisehavebeenignored.

2. Households within clusters are somewhat alike

Usually the outcomes of are considered to be independent of one another. This is unlikelytobetruewhenmeasurementsaretakenfromrelatedsubjects.InthecaseoftheNHIES,forinstance,theprimarysamplingunits(orclusters)representhouseholdsthatarefromthesameneighborhood.Oneusuallyfindsthatthesehouseholdsarefairlysimilar,notonlywithregardtoobservables,butalsowithregardtoattributesthatmaynothavebeenrecordedinthesurvey.Ignoringthiswithin-areacorrelationwouldproducemisleadingresults.

AccordingtoDobson(2002),thestandarddeviationofthemeandifferencesinincomesbetweentwoareaswillbeunderestimatediftheobservationswhicharecorrelatedareassumedtobeindependent.Forinstance,supposeitwasdesiredtocomparetheheightofwomenfromtwodifferentcountries,butwomenfromspecificcitiesonlywereselectedinbothcountries.Whilethemean differencewould still be unbiased, the standard deviation of themean differencebetween two countries would be overestimated if the observations that are correlated areassumedtobeindependent,sincethedatawouldhaverecordedtoolittleoftheactualvariationin each country.

0.1

.2.3

.4.5

4 6 8 10 12x

actual values fitted values


TheGLSmodelsmakeitpossibletocontrolforthiswithin-areacorrelation.ThisaddedflexibilitycomesatacostthoughsincetheGLSmethodismuchmorecomplexandtime-consumingthantheconventionalOLSmodel.Unfortunately,hastobeestimatedfromthedatabyaniterativeprocess since it is not known - FGLS.

52

Decomposing the Error

Unlike conventional models, it was not assumed that the errors are independent of one another. Allowance was made for the errors to be correlated within clusters. The error will consist of two components: the location effect (or cluster component) 𝜂𝜂! which will capture the intra-‐cluster correlation, and the idiosyncratic effect 𝜀𝜀!!.

𝑢𝑢!! = 𝜂𝜂! + 𝜀𝜀!!

where 𝜂𝜂! is the cluster component/location effect and 𝜀𝜀!! is the household component.

The GLS-variance covariance matrix, Ω, will look as follows:

Ω =

𝜎𝜎!!! + 𝜎𝜎!! 𝜎𝜎!!

!

𝜎𝜎!!! 𝜎𝜎!!

! + 𝜎𝜎!!0 00 0

0 00 0

𝜎𝜎!!! + 𝜎𝜎!! 𝜎𝜎!!

!

𝜎𝜎!!! 𝜎𝜎!!

! + 𝜎𝜎!!

The GLS models make it possible to control for this within-‐area correlation. This added flexibility comes at a cost though since the GLS method is much more complex and time-‐consuming than the conventional OLS model. Unfortunately, Ω has to be estimated from the data by an iterative process since it is not known -‐ FGLS.


NOTES

National Planning CommissionGovernment Office ParkLuther StreetPrivate Bag 13356WindhoekTel.: +264 61 283 4111Website: www.npc.gov.na

Republic of Namibia

ISBN: 978-99945-0-085-7

Namibia Poverty Mapping Macroeconomic Planning Department

Documents

Transcript of Namibia Poverty Mapping Macroeconomic Planning Department