ReportNo. 6934 EducationPolicies for Sub-SaharanAfrica: Adjustment,Revitalization, and Expansion

Public Disclosure Authorized september15, 1987 Educationand Ermployment Division Populationand Human Resources Department Policy,Planning and Research FOR OFFICIAL USE ONLY Public Disclosure Authorized Public Disclosure Authorized

Public Disclosure Authorized Documentof the WorldBank

Thisdocument has a restricteddistribution and may be used by recipients only in the performanceof their officialduties. Its contents may not otherwise be disclosedwithout World Bank authorization.

. K FOR OrICIALusr ONLY

This paper was preparedby Peter R. Moock, task leader, and Ralph W. Harbison, under the generaldirection of Aklilu Habte and the immediatesupervision of Dean T. Jamison. Janet Leno draftedmany of the Boxes and providedwriting and editorial as.stanc.e throughjut. Rosemary Bell.w had principal responsibility for preparing the Statistical Annex. Birger Fredriksen and John Middleton wrote the initial drafts of Chapters 4 and 7, respectively.A numberof backgroundpapers were commissioned, and they are noted in the Bibliography. Others who contributed importantlyto this effortinclude Wadi Haddad, Kenneth King (Universityof Edinburg), and Adriaan Verspoor.

Thanks to financial assistance from the Norwegian Ministryof DevelopmentCo-operation, this paper has benefited from the comments received on an earlierdraft from African policymakersat two internationalmeetings held in Ethiopia and Cote d'Ivoitein early 1987.

The judgmentsexpressed in this paper do not necessarily reflectthe views of the World Lank'smanagement or those of its board of executive directors or of the governmentsthat they represent. The maps used in this docu&ent are solely for the convenience of the readerand do not imply the expressionof any opinion whatsoever on the part of the World Bank or its affiliatesconcerning the legal statusof any country,territory, city, area, or of its authorities,or concerningthe delimitation of its boundariesor nationalaffiliation.

Thisdocument has a restricteddistribution and may be used byrecipients only in theperformance of their officialduties. Its contentsmay not otherwisebe disclosedwithout World Bank authorization. CONTENTS

EXECUTIVE SUARY X

INTRODCTION 1

PART ONE: THE POLICYCONTEXT 3

Chapter 1. The RemarcableProgress of AfricanEducation 4

1.1 Educationprior to independence 4 1.2 Advancesin educationafter 1960 6 1.3 Expenditureon education 10 (a) Publicdomestic expenditure on education 10 (b) Privatespending on education 13 (c) Foreignaid flows to African education 14

Chapter 2. Educationand the ExternalEnvironment: A Cy-le of 15 DeterioratingProspects

2 *.1 The demographicchallenge 15 2.2 Macroeconomicadjustment and fiscalausterity 20 2.3 Investmentin education:breaking the cycle of 20 deterioratingprospects? (a) Educationand labor productivity 22 (b) Loweredfertility and other nonmarket 25 benefitsof education (c) Educationand economicgrowth: 26 macroeconomicevidence

Chapter 3. Major Problems: EnrolluentStagnation and 31 Quality Decline

3.1 Enrollmentstagnation 31 (a) Schoolenrollments 32 (b) Adult literacyand training 33 (c) Expansionand equity 35 3.2 Decliningeducation quality 36 (a) The nature and importanceof quality 36 (b) Evidenceof decliningquality 38 - ii -

PART TWO: POLICYOPTIONS FOR AFRICANGOVERNMENTS 43

Chapter4. The Foundationt PrimaryEducation 46

4.1 Measuresto improveschool qcality 47 (a) Trainingand use of teachers 47 (b) Instructionalmaterials 50 (c) Physicalfacilities 53 (d) Languageof instruction 53 (e) Nutrition and health: ensuring teachable pupils 56 (f) Summary 57 4.2 Unit cost containment 58 (a) Teachers'salaries 58 (b) More intensiveuse of teachers 59 (c) Reductionin repetitionand dropout 63 (d) Appropriateconstruction standards 65 4.3 Mobilizingresources for primaryeducation 65

Chapter5. Consolidationof Competence: Secondary 71 Educationand Training

5.1 Meeting the demand for expansionby reducingunit costs 75 (a) Distanceeducation: self-studyschools 76 and extramuralprograms (b) Day schools 79 5.2 Issues of equity: increasedparticipation of females 83 5.3 Trainingfor vocationalcompetence 85 5.4 Financingsecondary education and training 89 (a) Cost-sharingin ,ublic secondaryschools 89 (b) The role of private schools 92 (c) Financingtraining 93

Chapter6. Preparationfor Responsibility:Higher Education 94

6.1 The challengeand the promise 95 6.2 Issuesin higher education 96 (a) Inappropriateoutput mix 97 (b) Low quality 103 (c) High costs 105 (d) Inequitableand inefficientfinance 109 6.3 A programfor structuraladjustment of higher education 110 (a) Improvedquality 110 (b) Increasedefficiency 111 (c) Constrainedoutput 112 (d) Expandedcost-sharing with beneficiaries 113 (e) Feasibility 113 - iii -

Chapter 7. Using Resources Wells The Mandate 115 for EducationalManagers

7.1 Improvingorganizational structure 116 (a) Schoolmanagement 116 (b) Decentralizationof policy implementation 117 (c) Structuralsimplification for policy development 120 7.2 Improvinginformation: testing, statistical, 121 and accountingsystems (a) .ducationaltesting 121 (b) Statisticaland accountingsystems 122 7.3 Strengtheninganalytic capacity 124 7.4 Managerialstaff development 125 7.5 Prioritiesand resourcerequirements 128

PART THREE: AN AGENDA FOR ACTION 130

Chapter S. Policy Packagesfor EducationalDevelopment 131

8.1 Adjustment 133 (a) Diversifyingsources of finance 133 (b) Unit cost contaiament 134 8.2 Revitalization:restoring quality 136 (a) Instructional materials 137 (b) Academic standards 138 (c) Operationand maintenanceof physical 138 plant and equipment 8.3 Selectiveexpansion 138 (a) Renewedprogress toward universal 139 primary education (b) Distanceeducation programs 140 (c) Training 140 (d) Research and postgraduateeducation 141 8.4. The task of policy designand implementation 142 - iv -

Chapter 9. InternationalAssistance for African 143 Educational Development

9.1 Aid to educationIn the recentpast 143 9.2 The comparativeadvantage of aid: past and present 147 9.3 Internationalassistance for policy design 151 and Implementation (a) Support for developmentof nationalprograms 151 (b) Organizinginternational support for 153 program implementation 9.4 Future allocationsof aid to African education 154 (a) Aid for revitalization 154 (b) Aid for selectiveexpansion 155 9.5 Call to action 159

Bibliography 160

Definitionsand Acronyms 170

Footnotes 174

STATISTICALAMNEX 179 ~~~~~- v

INSEZ1D TABLES.FIGUES, ANDBOXES Tables

1.1 Schoolenrollments andenrollment ratios, 1980 and 1983 1.2 Indicatorsof educationalprogress 2.1 Sub-SaharanAfrica prliary and secondary populatlons andenrollments, actual (1983) and projected (2000) 2.2 Thecost lopicatlors of maintalning1983's enrollment ratio throug the year 2000 3.1 Fubilc recurrentexpenditure per studentIn aLb-SaharmnAfrica, 1970-1983 4.1 Mbdimof Instructlonby former oolonlal status 4.2 Primaryanrollment ratios andprlwy education'ssharoe In the educatIonbulgt 5.1 EnrollmentcharacterIstics and education exPenditure by secondaryenrollment groups 5.2 Threeuses for distancoeducation 5.3 lherelative per studentcosts of day schools,boarding schocls, aid self-study schoolsfor secondary uation in Malawi 8.1 Ratiosof graduatestartirg salaries to Percapita income 6.2 Indexof starting salarles In pubilc servioeby educatlonlevel (*Omlevel * 100) 8.3 LUit costs of public educationat the various levels as percentageof Percapita GNP 8.4 Pubilc expenditureper studentIn tertiary educationas multiple of pubic expenditureper puPIl at lower levels 9.1 Estimatedannual xternal resc.roeflows to Africaneducation and training by dbnorgroup and use, 1981-1983average 9.2 Direct educatIonaid by recipient groupand expenditure category, 1981-1983 average 9.3 Dlrect educationaid by level ard source,1981-1983 average 9.4 Allocation of direct educationald by cperEditurecategory c.id source,1981-1983 average

Fiaures

1.1 PublIceducation epniture as a percentagpof natlonal Income,Africa andeisewhere 2.1 Sub-SaharanAfrica populationage structure, 1984 2.2 Grawthof school-agepopulations, 1965-1984 and 1984-2000 Boxes

1.1 AggregatepublIc domestlc expenditure oneducation Insub-Saharan Africa 2.1 Relativerates of returnto primary an secondaryeducation InKenya 2.2 Educationand mortallty decline: research from Nipria, Ghana, and Sudan 2.3 Edi±oatlonandeconomlc growth In sub-Sahartn Africa 3.1 Thestory of Ziibabw 3.2 Arecogiltive outcomes really Important? TheInterplay ofability, skills, and schoolIng 4.1 Olstancelearning for teachers InTanzania 4.2 InteractiveeducatIonal radio In Keny,a 4.3 Reducingteacher costs In Burkina Faso 4.4 ReducIngper pupil costs throughdouble-shifting In Senegal 4.5 Low-costschool construction In FrancoovoneAfrica - vi -

5.1 Pollcles toard secondaryedcatlon 5.2 Analternatlve route to secondaryquallfIeatlone In Malawi 5.3 Cormnityfinancing of secondaryschools In Zambla 6.1 Utilization andearnlins of university technilcalgraduates In Coted'lvolre 6.2 CrlsIs of qLallty In hlheredueatlon 7.1 Managementof reform In Ethiopia 7.2 ExaminatIonreform In Kenya 7.3 Livlingstandards surveys (LSS) - vii -

SIATISTICALANNEX

Part A. Education Indicators

A.1 Prlmaryenrollment A.2 Secondaryenrol lmnt A.3 Distribution of secondaryenrollment by type of ecatlon A.4 Tertiary enrollment A.5 Distribution of terttdry enrollmentby field of study A.6 Total enrollment A.7 Primaryenrollment ratios A.8 Secondaryenrollment ratios A.9 Tertiary enrollmentratlos A.10 Teachersand schools A.11 Student-teactwrratios A.12 Studentflow an officIency Indicators A.13 Percentageof studentsenrolled in private schools A.14 Total publIc expenditureon e ucation A.15 Distributionof total public expendItureon educatlonby component(recurrent or capItal) A.16 PublIc recurrentapendIlture on educationby level of educatlon A.17 Pubilc recurrentexpenditure per prl,ary pupiI A.18 PublIc recurrentexpenditure per secondarystudent A.19 Public recurrentexpendlture per tertiary student A.2D Distributlon of prlaary public recurrentexpWnditure bypurpose A.21 Distribution of secondarypublic recurrentexpendIture bypurpose A.22 Distributlon of tertiary public recurrentexpenditure by purpose A.23 Primaryand secondary teachers' average salaries A.24 Externalaid to African educationand traIning by donorgroup, 1981-83 average A.25 Concessionaleducation sector aid from OECDand OPEC members by recipient ocutry, 1981-83 average A.28 Externaleducatlon sector aid from OECDand OPEC donors by level andtype of educatlon A.27 Externaleduatlon aldfrom OECD and OPEC donors by purpose A.28 ExternalpubiIc debt of the soucationsector

Part B. Social and Economic Indicators

B.1 Basic Indicators B.2 Languagesof sub-SaharanAfrica B.3 Populationgrowth and projections B.4 School-agepopulation growth and projectlonm B.5 Demographyand fertillty B.6 Urbanizatlon 8.7 LaborForce B.8 Growthof production 8.9 CentralgDverrment expenditure B.10 Disbursementsof officlal developmentassistance 8.11External pubilc debt service ratios - viii -

Part C. Supplemntary and Suwmay Tables

C.1 Selectedcomwarative statistlcc forcountries with population underone-half million people C.2 EnrollmentsInsub-Saluran Africa C.3 Estimatedaverage years of educationattended by workingae population C.4 Literacy C.5 Cross-natlonalcomparisons of achlevementIn mathematics,readlg, andsclence C.6 Countriesgrouped by grossprlmary enrollment ratios Cl7 Enrollmentcharacterlstics andeducatIon expenditure by secondary enrollmentgroups C.8 Unit cost IndIcatorsof secondaryenrollment groups C.9 Percentageof natIonalsamong teachIng staff In post-prImary education - ix -

USCUTIVE SUI@(&Y

Since their independenqe,the nationsof sub-SaharanAfrica have investedheavily in education.- In many Africancountries, however, enrollmentshave stagnatedrecently, and the qualityof educationhas apparentlydeclined. These reversalshave occurredin an environment characterizedby unprecedentedpopulation growth, mounting fiscal austerity,and often tenuouspolitical and administrativeinstitutions. Each of these factorshas hurt education,and the ensuingdeterioration in educationalservices has made it more difficultto solve the region's economicand social problems,thus fuelinga cycle of erodingprospects for the people of the region. To break this cycle, there is value in identifyingpolicies that will renewprogress in Africa'seducation and training.

Any discussionof policiesand prioritiesfor a regionas vast and diverse as sub-SaharanAfrica naturallyruns the risk of overstating commonalitiesand understatingdifferences. The educationalachievements and aspirationsof individualcountries and the socio-economicand politicalconstraints within countriesvary too much for any single comprehensiveset of measuresto be completelyor uniformlyapplicable. This paper is about diverseand variablyapplicable golicies for African educationald4velopment, not about a monolithicand universallyapplicable 2olicy. In the end, nationalauthorities in each countrymust decide the prioritiesfor educationalinvestments and tailor those investmentsto the country'sspecific needs, conditions,and aspirations. This paper is intended4o providea frameworkthat may assistAfrican policymakers in this process.

Educationaldevelopment in Africa -- past progress and the currentchallenge

Africansocieties have a long and rich historyof educationand training. Indigenouseducation was practicedby all groups and remainsan importanttransmitter of culturalidentity from one generationto the next. Pre-colonia'Christianity and Islam,with their rootsextending back centuries,had a pervasiveinfluence on communitylife and perceptionsin many parts of the region.

In the colonialera, missionariesand metropolitangovernments openedup a networkof Western-typeschools in Africa. The administration of educationsystems was dominated,however, by expatriates,as was teachingbeyond the primary level. Moreover,during this period, accessto educationwas quite limited,especially in the thinlypopulated areas of FrenchWest Africa. By 1960, the gross primaryenrollment ratio in all of sub-SaharanAfrica was still only 36%. This was about half the levelsthen found in Asia and Latin America. Many countries-- includingThe Gambia, Cote d'Ivoire,and Senegalin West Africa and Tanzania and Somaliain East Africa -- had literacyrates below 10% at the time of independence.

11 Most of the discussionand all of the Africa-widestatistics in this paper refer to thirty-ninecountries south of the Sahara. Whenevertlte term "Africa"or 'African"is used, it refers only to these countriesand their residents. - x

Progresssince independence.The systemsof educationinherited by the Africannations at the time of independencewere thus quite inadequateto meet the needs of self-governanceand rapid economicgrowth. From this low startingpoint, the progressachieved in African education has been spectacular. Quantitativeexpansion has been particularly impressive. Between 1960 and 1983, the numberof studentsenrolled in African institutionsat all levelsquintupled to about 63 million students. Enrollmentsincreased about 9% annuallyduring the 1970s,double the rate in Asia and triplethat in Latin America. At the primarylevel, the gross enrollmentratio rose from 36% in 1960 to 75% in 1983. At the tertiary level, the numberof studentsenrolled in African institutionshad reached 437,000by 1983, growingfrom just 21,000 in 1960. The substantial expansionof educationsince independencehas allowedthe increased participationof some groupswho had littleor no prior access to formal education.

The consequenceof this massive educationalexpansion has been a substantialdeepening of the human capitalstock. The estimatedaverage educationalattainment of working age men and women in the median African country increasedfrom less than half a year in 1960 to more than three years in the early 1980s. The adult literacyrate in the median country rose from about 9% to 42%.

The currentchallenge. The advancessince the early 1960s are now seriouslythreatened -- in part by circumstancesoutside education. Africa'sexplosive population growth greeUly increasesthe number of childrenseeking access to schoolsand the numberof potentialilliterates. Between 1970 and 1980 Africa'spopulation grew at 2.9% a year, a full percentagepoint higher than the worldwiderate. Between 1980 and the end of the century,Africa's population is projectedto grow at 3.2% a year, its primaryand secondaryschool-age population at 3.3X. If the growthof educationalplaces is to keep pace with the growth of the school-age population,more schools,teachers, books, and other inputsare required each year. This requirementcomes at a time when the re ent economic declinehas necessitatedsignificant cutbacks in public spending. Public spendingon educationin Africa has not been spared,declining from $10 billion in 1980 to $8.9 billion in 1983. These fundamentalfacts sharply constrainthe optionsopen to policymakersand have seriousimplications for African educationpolicy.

The main educ4tionalissues in Africa today are the stagnationof enrollmentsand the erosionof quality. Althoughtotal enrollmentsin sub-SaharanAfrica grew at an averageannual rate of 6.5% during 1960-70 and 8.9% during 1970-80,the rate of increaseplummeted to 4.2% in the first three years of the 1980s. The slowingof enrollmentgrowth affected all levelsof educationbut was most evidentat the primarylevel, where the rate of growth fell from 8.4% a year during 1970-80to 2.9% in 1980-83. If the populationof primaryschool-age children increaAes at the projected averageannual rate of 3.3%, a 2.9% increasein enrollmentswill not even keep race. And so long as enrollmentsstagnate, there is little likelihood that currentinequalities in access to educationcan be eliminated. - xi -

Complicatingthe problemsof stagnantenrollments are the low levelsand recent erosionsof educationalquality. Levelsof cognitive achievementamong African studentsare low by world standards,and there is some suggestionof furtherdecline recently. Much of the evidenceis indirect,focusing on criticallylow levelsof key inputs (especiallybooks and other learningmaterials) and on recentdecline ir the use of these inputs relativeto that of teachertime and physicalfacilities. Less is known on the output side. But the performanceof African students in cross-nationalstudies of academicachievement has been sufficiently poor to be a major cause for concern.

Addressingthese issuesof stagnationand low qualitywill require additlonalresources. More important,it will requireprofound changes in educationalpolicy for many countries. Indeed,for most African countries, the first will not be obtainablewithout the second.

A Dolicy framework-- adiustment,revitalization, and selectiveexpansion

Hard decisionson educationpolicy shouldnot be postponed. In most African countriesthe cost would be continuedenrollment stagnation and qualitydecline through the 1990s. This paper urgentlyrecommends that each African nationnow embracethe task of formulatingand implementingan internallycoherent set of policiesthat reflectsthe nation'sunique history and aspirationsand one that will effectivelyaddress its own recentlyexacerbated problems in the educationand trainingsector. While the particularsof the policy packagesthat emerge from this exercisewould vary from one countryto the next, it is nonethelessclear that every country-specificpackage will need to contain,in varyingproportions, elementsof policy along three distinctdimensions. These three are adjustment,revitalization, and selectiveexpansion. Moreover,if new policiesare in fact to be implemented,management practices will need to be impro'ved.

Althoughundoubtedly painful and politicallydifficult, adjustment policieswil1 alleviatethe burdenof educationand trainingon public budgets. Measuresfor revitalizationand expansion,on the other hand, will certainlyrequire an incrementalflow of resources. Thus, in the contextof ongoingausterity in Africa, resolutemovement toward adjustment is a necessarycondition for implementingforward-looking policies on the other two dimensions.

Adiustmentto currentuemographic and fiscal realities,though it will be difficult,is essentialif the disruptiveeffects of these external factorsare to be minimizedin the years ahead. Adjustmentwill take two main forms:

o Diversifying sources of finance. This will be a necessary part of country-specificpolicy packages. This diversificationcan be achievedthrough increased cost-sharing in public educationand throughincreased official tolerance and encouragementof nongovernmentalsuppliers of educationalservices. - xii -

o Unit cost containment. The rigorouscontainment of unit costs will be just as importantin the adjustmentprocess, and in many countries,probably more .important,than policies to diversify sourcesof finance. The most promisingareas for the containmentof costs are teachers'salrice, teacherutilization, construction standards,and studentrepetition and dropoutbehavior.

Revitalizationof the educationalinfrastructure that now exists, in order to restorequality, is the second dimensionof a properly conceivededucational strategy. This involvesa renewedfocus on the fundamentalsof educationalprovision so that maximum advantageis extractedfrom the currentcapacity of educationand trainingsystems. Three kinds of measuresare necessaryfor the reitorationof quality:

o Academicstandards. There must be a renewedcommitment to standards,principally through the strengthening of examination systems.

O Instructionalmaterials. Effortsshould be made to restorean efficientmix of inputs in education. The provisionof a minimum package of textbooksand other learningmaterials is usuallythe most pressingneed in this respect.

O Operationand maintenanceof RhysicalPlant and equipment. Greater investmentmust be made in the maintenanceof physicalplant and equipment,and there is a need for greaterexpenditure on other Inputs that would resultin increasedutilization of these capital assets.

Selectiveexpansion to addressneeds for furthereducational servicesis the third dimensionof any completestrategy for educational development. Measures in this area, viable only after measuresof adjustmentand revitalizationhave begun to take hold, will concentratein four areas;success in all will dependupon a generaleffort to safeguard the qualityof instructionalstaff at all levels.

O Renewedprogress toward universalprimary education. Expanding access to primaryeducation is the new investmentthat will bring the highesteconomic and social returnsin many countries.

o Distanceeducation programs. At the secondarylevel, and later on at the tertiary,expansion of enrollmentsin selectedsubjects and streamswill be necessaryin most countriesin the years ahead. To accommodatethese increasesin post-primaryeducation, most countrieswill need to consideralternative delivery modes that shift more of the burden for learningonto the studentsthemselves; now is the time to begin planningfor such programsand developing their requisitesupport infrastructure (correspondence materials, radio programs9and examinationssystems). - xiii -

o Traininz. The amountof trainingthat occurs once individualshave enteredthe labor force must be increased;this trainingshould serve both school leaversand also those who have had no exposureto formalschooling, so that individualscan acquirethe necessary job-relatedskills and renew these skills duringtheir working lifetimein responseto changingmarket conditions.

o Researchand postgraduateeducation. Expansionof Africancapacity to produce its own intellectualtalent to fill the highest scientificand technicaljobs in educationalestablishments, in government,and in the private sectoris an importantmatter to be addressedin buildingfor Africa'sfuture.

For most Africancountries, the formulationof a comprehensiveand coherenteducation development program, derived from a balancedpackage of policiesfor adjustment,revitalization, and selectiveexpansion, will be a new experience. Each countrywill organizefor the task in its own way. In many countries,however, a fruitfulapproach to policydesign might be expectedto includethe following: establishinga nationalcommission to oversee the work; constitutinga technicalstaff to supportthe commission; for both, drawingupon the best politicaljudgment and analyticaltalent from ministriesof financeand planningas well as education,and from the nation'sinstitutions of tertiaryeducation and research;building a nationalconsensus through public debateof emergingfindings and recommendations;and taking advantageof the experienceof other African countriesin developingthe nation'sown educationdevelopment strategies. All such activitieswould, of course,entail costs, and these would need to be financed. Budgetaryresources would have to be sufficientto cover not only the personnelcosts of the nationalcommission but also its operating expensesfor travel,communications, publications, and specialized contractualservices (such as data collectionand processing,expert technicalconsultants, targeted research and analysis).

While carefulelaboration of educationdevelopment programs is urgent and essential,African capacity for implementationwill ultimately determinetheir impact. Improvementin educationmanagement is a necessary concomitantto policy reformand must be given immediateand continuing attention.

Here the most importantmeasures involvethe potentialreturns to downwardand outwarddelegation of various administrativefunctions. While some functionsmust appropriatelyremain at the centralministry level,and the performanceof these functionswill need to be improved,the toughest challengeto improvingmanagement lies closerto the classroom,at the level of individualschools and districts. Africanpolicymakers should considerhow, with adequatesafeguards against abuses, schools and the local communitiesthey serve can be given increasedauthority in the acquisitionand utilizationof the resourcesessential to eff-ective classroomteaching and learning.

In addition,central ministries must tend more seriouslyto their own managementdevelopment needs, especiallyin the areas of performance monitoringand policyplanning and analysis. Improvementsin examination systems (whichwas mentionedabove with referenceboth to academic standardsand to distanceeducation), in the nature and timelyavailability - xiv - of statisticaland financialaccounting information, and in the numbersand qualificattonsof staff engagedin full-timeanalytical work are among the necessarymeasures. Incentivesin many ministriesof educationare insufficientto attract,motivate, and retain able staff. Governments comnittednot only to formulationof an educationaldevelopment pt)gram but also to its expeditiousimplementation will have imaginativelyto address the issue of incentivesat all levelsof the educationalsystem.

Policy optionsby level of education

The mix of adjustment,revitalization, and expansioncomponents that is appropriatein light of country-specificconditions and goals can be expectedto differ. Within a country,the mix will differ also among levels of education.

At the primarylevel in most African countries,there is only limitedscope for adjustment,in the form of either loweredunit costs or increasedcost sharing. There is, however,good potentialfor improvements in the qualityof output,that is, for revitalizingprimary education, througha shift in the input mix towardtextbooks and learningmaterials. At the secondarylevel there is far more scope for containingunit costs -- partly throughmaking fulleruse of availableresources and partly throughswitching to cheaperways of providingservices. Higher education poses a set of problemsuniquely its own. Rapid expansionhas left in its wake an abundanceof institutions,programs, and graduatesthat are often of low qualityand dubiousrelevance. Modest consolidation,an adjustment measure that would lower unit costs, and increasedcost sharingare the first steps on the road to higher education'srevitalization, which should be regardedas a prerequisitefor any furtherexpansion of the sub-sector.

Primaryeducation. Duringthe 1960s and 1970s some analysts warned that sub-SaharanAfrica's preoccupation with the quantityof educationwould lead to a seriousdeterioration in quality. In the countrieswhere educationalstandards have deterioratedmost, tilechoice between expansionand qualityis no longeran either-orchoice. Without some basic revitalizinginputs, particularly textbooks and instructional materials,almost no learningcan be expectedto occur. Ensuringthe availabilityof essentialinputs is a prerequisiteboth for quality improvementand for expansion. Beyondthis minimal level,however, is the questionof findingfor each countryan appropriatebalance of qualityand enrollment-- of identifyingefficient approaches to qualitative enhancementand quantitativeexpansion, and of financingimprovements in both.

A reviewof possiblemeasures for improvingthe qualityof primary educationresults in two broad conclusions:

o First, the safestinvestment in educationalquality in most countriesis to make sure that there are enoughbooks and supplies. These materialsare effectivein raisingtest scoresand, almost invariably,have been underfundedrelative to tecachers'salaries. This is also an area where externalaid has a comparativeadvantage. Other areas that appearto have the potentialto improvequality - xv -

includeschool feedingand healthprograms, intensive use of radio, in-serviceeducation of teachersin subjectmatter, and strengthened inspectionand supervisionsystems.

o Second,some investmentsare not likelyto a have a noticeable effect on primary schoolquality despite their potentiallyhigh costs. These investmentsinclude reducing class size (withinthe range between25 and 50), providingprimary teacherswith more than a generalsecondary education, providing teachers with more than minimalexposure to pedagogicaltheory, constructing high-quality buildings,and introducingclassroom televisions or computers.

Even with no qualityimprovements (and assumingno improvementsin efficiency),the resourcesdevoted to the primarylevel would need to increasemore than 3% a year just to keep pace with populationgrowth. There is, however,little likelihood of reducingunit recurrentcosts at this level significantly,especially if countrieshope also to improvethe quality of education. There is more scope for reducingunit capitalcosts, but even so, it is unlikely that overall per pupil costs can be reduced very much at the primarylevel. Thus, the furthergrowth of primary educationwill requireadditional resources.

Secondaryeducation and training. Can adjustmentmeasures at the secondarylevel generatesufficient savings to providethe necessary nonsalaryrecurrent inputs (booksand supplies)and achievesignificant capacityexpansion? Substantialeconomies are possiblein the operationof regularschools. There can also be economiesfrom reductionsin boarding. And most significant,there can be economiesfrom the creationof distance educationsystems combining radio and correspondencetechniques that would allow expansionof reasonablequality secondary education to many more communitiesthan could be reachedin any other way for the same price.

Beyond this generalsupport for expandingaccess to secondary education,African governments should give seriousconsideration to policiesdesigned to remedyexisting inequalities in schoolparticipation. Females are the largestunderrepresented group. Small,community-based schools (whetherrelying on distanceeducation or not) will tend to attract girls more readilythan largerschools located in urban centersand at greateraverage distances from homes. Althoughsmaller conventtonal schoolsmay imply higherunit costs,this is not necessarilytrie if larger schoolatend to includeboarding facilities while smallerschooLa do not. Increasingthe number of femaleteachers may also attractmore girls. especiallyin Islamicareas.

Another set of issuesto be addressedmainly at the secondary level concernsthe relevanceof the curriculumto the needs of individuals and societies. The main questionsthat many countries face are when and how to make the transitionfrom subjectsthat have broad vocational relevance(language, mathematics, and science)to programsand subjects tnat will prepareindividuals for specificjobs or clustersof jobs. Internationalexperience shows that a strong generaleducation, which schoolsprovide efficiently,greatly enhances an individual'sfuture - xvi - trainability. It also shows that job-specific training is very important. Such training usually is most efficiently provided after initial job decisions have been made and in institutions under, or strongly influenced by, the ultimate employer. Occupation-speciZic and job-specific training need not provide individuals with degrees or credentials.

Because of the high costs and lack of vocational relevance of many school-based approaches directed to specific jobs and occupations, there is an urgent need to develop (through incentive schemes and technical assistance) local capacity to provide on-the-job training, other firm-based skills development programs, and industrial training centers. Governments interested in laying the groundwork for a more technically-oriented economy may also wish to consider placing heavy emphasis on general mathematics and scientific skills iuithe secondary and post-secondary curriculum. These ptograms tend to be relatively inexpensive and are likely to prove more conducive to economic growth than an emphasis on in-school vocational education.

Hither education. Preparing and supporting people in positions of responsibility -- in government, in business, and in the professions -- is the central and essential role of the continent's universities. In numbers at least, tne universitles have risen impressively to this challenge. Enrollments grew from 21,000 in 1960 to more than 430,000 in 1983.

Higher education's contribution to development in Africa is being threatened, however, by four Interrelated weaknesses. First, higher education is now producing relatively too many graduates of programs of dubious quality and relevance, and too little new knowledge and direct development support. Second, the quality of these outputs shows unistakable signs in many countries of having deteriorated so far as to cast doubt on their fundamental effectiveness. Third, the costs of higher education are needlessly high. Fourth, the pattern of financing higher education is socially inequitable and economically inefficient.

Wherever the foregoing diagnosis of weaknesses in higher education can be confirmed, policy reform should seek four objectivess (i) improving quality; (ii) increasing efficiency; (iii) changing the output mix, which may imply smaller enrollments in certain fields of study; and (iv) relieving the burden on public sourcer of financing, by increasing the participation of beneficiaries and their families. But quality improvements, the first objective, will cost money. Thus, implementing adjustment policies to achieve the other three objectives will, almost everywhere in Africa, be a prerequisite for freeing the resources needed to achieve the first.

The role of the International donor coumuity

This paper argues that adjustment measures are needed to alleviate the burden of education and training on public budgets in Africa. The "savings' generated from the lowering of unit costs and from increased cost sharing and increased tolerance of private provision of educational services can be used to help fund the necessary revitalization and ultimate expansion of the sector. - xvii -

Regrettably,all such savingsfrom adjustmentmeasures will not be sufficient,in most countries,to cover the very substantialresource requirements of the policies needed to revitalize and build African educationto the extentessential for futuredevelopment. International aid will remain a criticallyimportant determinant of the pace of progress of educationin the region. However,the rapid evolutionof African needs, as summarizedin the three dimensionalframework for policy reform,demands correspondingchanges in the organization,nature, and level of internationalaid for African educationaldevelopment.

The pressingrequirement is for aid in supportof policy reform. The internationaldonor communityshould offer three relatedkinds of supportfor the designof nationalpolicy.

o The first form of planningsupport is simple: seed money, quickly provided,to cover both local and foreigncosts of policy developmentand managementimprovement. The willingnessof internationaldonors to bear a part of these extraordinaryexpenses, perhapson a matchingbasis, would providean importantincentive to African governments.

o The secondkind of supportthe internationalcommunity should provide to an African countrythat embarkson a seriousreform of educationalpolicy is to ensure ready accessto the ongoing experienceof other countriesin formulatingand implementingsuch reform. Intensivecollaboration across countries, so that they share their accumulatedexperience widely, will pay high dividends as countriesgrapple with common issues.

o Third, the internationaldonor communitycould establishand finance a source of high-qualityspecialized technical expertise, without financialor politicalties to any governmentor international donor. African governmentscould call on this source for help in formulatingpolicies at the outset and in monitoring,evaluating, and correctingthem during implementation.

Appropriatemechanisms do not now exist for meeting these three interrelatedneeds for the improvementof policydevelopment. The requirementis for expeditiousaction to developthem. Any initiativeiLl this area that would take more than a year to adopt and implementwould be an inadequateresponse to the needs of African governments.

Beyond its assistancefor policy development,the international communityshould help financethe implementationof sound programs. These programswill typicallyrequire more resources,sustained over a longer period, than can be mobilizedinternally. Countriesthat have demonstrated their willingnessto addresspolicy issuesshould have accessto increased, longer-term,and more flexiblyoffered internationalaid. To the extent that a country'spolicy package involvesthoroughgoing reform, there are likely to be substantialone-time transition costs to a new and more sustainablepolicy regime. - xviii -

The internationalcommitment to the reformprogram must be seen from the beginningas continuing,something that has been missing in the past. In addition,the sum total of aid to differentlevels and expenditurecategories should reflect,at least in very rough terms, the prioritiesgiven to these levelsand categoriesin the nationalprogram. In recentyears, only 7% of internationalaid to African educationhas been used for the supportof primaryeducation, as comparedwith 16% for general secondaryeducation, 33% for vocational/technicaleducation (including teacher training),and 34% for highereducation. In terms of expenditure categories,only 11% of aid has been used to support "operationalcosts", consistingof local salaries,consumable supplies, and instructional materials.

Why meet the challenge? The expectedbenefits of educationand training

Greater investmentin educationand trainingcan, at this time in Africa'shistory, be expectedto yield broad economicbenefits. These benefitsinclude higher incomesand lower fertility. Assessmentsof labor market returnsto investmentsin educationconsistently find rates of return above 10% and sometimesabove 20% -- rates that compare favorably with those in most sectorsin Africa today. A recentstudy, undertakenin preparing for this paper, on the developmentimpact of educattonal investmentsin 31 Africancountries over time corroboratesthe microeconomicfindings of education'shigh returns.

Increasedinvestment in educationcan also be expectedto reduce fertility. In general,there is a strongnegative relationship between how much educationa woman receivesand the numberof childrenshe bears during her lifetime. More-educatedindividuals, in additionto having fewer children,tend to live healthierand longerlives. And numerousstudies have shown that parents'education, particularly that of the mother, affectschildren's survival, as well as their physicaland cognitive development.

The benefitsof educationgo far beyond those for incomesand fertility,however. The rapid transitionin Africa from colonialstatus to self-governmentto participationin the internationalarena was possible only becauseAfrican educational systems produced people to replace expatriatesat all levels. The nurturingof leaderswho can addressthe increasinglycomplex tasks of nation-buildingis a continuing responsibilityof African education. In addition,the stock of human capitalin Africawill determinewhether Africans can harnessthe universal explosionof scientificand technicalknowledge for the region'sbenefit -- or whetherAfrica will fall fartherand fartherbehind. Above all, educationis a basic right, an end in itself,an intrinsicpart of life and development.When all of the benefitsof educationare considered,the case for revitalizationand expansionof schoolingand trainingin Africa is compelling,even in this period of unusualscarcity. INTRODUCTION

Since their independen e, the nations of sub-Saharan Africa have investedheavily in education.-7 The achievementsin the sectorhave been impressiveboth absolutelyand in relationto other sectorsand other countries at other times. In many Africancountries, however, enrollments have stagnated recently,and the qualityof educationhas apparentlydeclined.

These reversalshave occurredin an environmentcharacterized by unprecedentedpopulation growth, mounting fiscal austerity,and tenuous politicaland administrativeinstitutions. Each of these factorshas hurt educationin the region,and the ensuingdeterioration in educationalservices has made it more difficultto solve the region'seconomic and socialproblems, fueling a cycle of erodingprospects for the people of the region. To break this cycle, there must certainlybe value in identifyingpolicies that will renew progressin Africa'seducation. The role of human skills in development is critical.

The returnto investmentin human skillsdepends, nevertheless, on the macroeconomicpolicy environment. The UnitedNations and iYe World Bank have spelledout the essentialingredients of this environment.- And most African governments have initiated the necessary economic reforms -- with some good results. In 1986 real incomesper capitarose in low-incomeAfrica for the first time in the 1980s. To sustainthis growth in personalincome, however, there must be a moderationof the region'sunprecedented population growth. The World Bank, in a 1986 paper on PopulationGrowth and Policiesin Sub-Saharan Africa, summarizedthe widely acceptedelements of populationpolicy for Africa and noted the very hopeful recenttrends in governmentpolicy. Clearly,the 1980s mark an importanttransition for Africa'seconomic and demographic policiesin settingthe contextfor productiveinvestment.

Any discussionof policiesand prioritiesfor a region as vast and diverseas sub-SaharanAfrica naturallyruns the risks of overstating commonalitiesand understatingdifferences. Readers of this paper will fully appreciatethe enormousdiversity that exists amongAfrican countriesalong severaldimensions -- economic,political, institutional, cultural, linguistic, educational,and (its importanceshould not be underestimated)the extentof a country's own internal diversity. The many averages cited throughout the paper,while they reflectthe 39 most populousAfrican countries,apply specificallyto none. Individualcountries, country blocs, and languagegroups will thus provideexceptions to all conclusions. In the end, national authoritiesmust decidethe policiesand prioritiesfor educationalinvestments and tailorthose investmentsto specificconditions, needs, and aspirations. For these reasons,there is no attempthere to prescribean educationpolicy for the continent-- that would be inap?ropriateand futile. Instead,the focus is on generalizations-- on addressingtrends and issuesthat, becauseof their importancein many Africantcountries, assume importance for the whole continent. This paper -- on investmentin educationand training-- is the first in a series that the World Bank is preparingto stimulatediscussion of sectoralpolicies for Africa in the 1990s and beyond. It has three main objectives. The first is to identifyand describecommon problems and issues of educationaldevelopment in Africa. The second is to provide leadersin each countrywith comparativedata and analyticaltools for developingtheir own policiesand priorities. The third is to suggestspecific policy directions for considerationby nationaleducation authorities and by donors. The key word here is "consideration."This paper will meet its objectivesif it helps initiateserious reflectionand debate in Africa on future directionsfor the sector. PART ONEs THE POLICYCONTEXT

Discussion of policies for African education in the years ahead must occur within the contextof the system'shistory and its externalenvironment. An attemptis made in Part One of the paper to addressthose aspectsof history and environmentmost pertinent to understandingthe currentproblems and prospectsof Africaneducation. Chapter 1 chroniclesthe extraordinary accomplishmentsof African nations in buildingup their systemsof education and training duringthe 25 years of the post-independenceera. Chapter2 depictscurrent demographicand economictrends in the region and assessesthe extent to which these trendsthreaten past educationalachievements. The importantpositive impact of educationalinvestment on economicgrowth and other social outcomesis also discussedin Chapter 2. Chapter3 turns to the educationsector itselfand elaboratesthe key issuesof enrollmentstagnation and quality decline. - 4 -

Chapter 1. THE REMARKABLEPROGRESS OF AFRICANEDUCATION

Around the time that most countriesof sub-SaharanAfrica gained independencefrom colonialrule, this region laggedfar behind the rest of the world on nearly every indLcatorof Western-styleeducational development. Effortssince then, especiallyduring the 1960sand 1970s,while failingto close the educationgap, have been truly dramatic. The recordof this period is a tributeto the determinationdemonstrated by African leadersand the sacrificesendured by Africanparents in their quest to providea better standardof livingfor their children'sgeneration.

1.1 Educationprior to independence

African societieshave a long and rich historyof educational traditions. Indigenouseducation was practicedby all ethnicand linguistic groups and remainsan importanttransmitter of culturalidentity from one generationto the next. It aims to Instill in childrenthe attitudesand skills appropriatefor male and femalesocial roles,emphasizing the dutiesand privilegesderived from culturalvalues. Impartedthrough language and example in the contextof the home environment,as well as formal lessonsand rituals outsidethe home, indigenouseducation responds to the concreteproblems of local communities. It preparespolitical leaders and ordinaryfermers, and it engendersa sense of citizenshipin the people of the community.

Africa'searly Christianheritage represents a second important elementof educationin the region,with roots extendingback long beforethe colonialperiod. Especiallyin NortheasternAfrica and the Nile Basin, Christianityhas thrivedin Africa for over fifteen-hundredyears. In about the year 450, the EthiopianChristian Church, a key example,established a comprehensivesystem of educationthat providedan underpinningfor Ethiopian cultural,spiritual, literary, scientific and artisticlife.

A third major antecedentto the colonialperiod is the influenceof Islam on African education. Arab cultureand languagewere adoptedin much of North Africa,while conversionto the Islamicfaith occurredalso in the Sahelianzone, along the Coast of East Africa, and in much of the Horn of Africa. Both formaland nonformalschool systems were establishedto teach the ethicsand theologyof Islam, includinga small number of elitistcenters of excellencesuch as the ones at Tombouctouin Mali and at Lamu on the East Coast. Designedto impartskills and knowledgewithin the religiousrealm, the system emphasizesreading and recitationin Arabic.

The Westerncolonial period in sub-SaharanAfrica began with the arrivalof the Portuguesein the fifteenthcentury and ended only quite recently. Of the African countriescovered in this paper,Ethiopia and Liberia alone have been sovereignstates for longer than 30 years. All of the other countriesachieved sovereign status within the recentand relativelyshort period between 1957 (Ghana)and 1980 (Zimbabwe). C3lonialprecedents are still much in evidencein most of Africa,and they sometimesconstrain the degree to which governmentsare free to initiatenew policies. - 5 -

The prlncipalsuppliers of Western-styleeducation prior to independencewere the Roman Catholicand variousProtestant churches, through their Africanmissions, and the colonialgovernments themselves. The division of administrativeand financialresponsibilities between the two differedfrom one colonialregime to the next. The British,fcr example,were generallymore tolerantof religiousand local communityautonomy than were the French.

The economicchatnges that the colonialpowers set in motion in Africa helped create a demand for Western-styleeducation that, in many areas, seemed nearly insatiable. Educationbecame the vehiclefor moving,within one generation,from peasantryand poverty to the topmostranks of society. This fact of modern-daylife escapedfew African parentslooking for ways to promote a better futurefor their children.

In their quest for convertsand literateAfrican subjects, the missionariesand colonialgovernments opened up a networkof schoolsin the region. Many of these were of a high standard. Yet the curricula,based for the most part on overseasmodels, reflectedlittle in the way of African content. The administrationof 'modern"education systems in Africawas dominatedby expatriates,as was teachingbeyond the primary level.

Access to educationwas quite limited,especially in the thinly populatedareas of FrenchWest Africa. In 1960, the gross primaryenrollment ratio in all of sub-SaharanAfrica was still only 36%.1 This was about half the levelsthen found in Asia (67%) and Latin America (73%). The enrollment racio was 38% in the Francophoneterritories (50% in the Belgiancolonies and just 31% in the French)and 40% in the Anglophone. Many countries,including The Gambia,Cote d'Ivoire,and Senegalin West Africa and Tanzaniaand Somalia in East Africa,had over 90% illiteracyat the time of independence.

There were also significantdifferences in educationalaccess and participation within colonial territories -- between urban and rural populations,males and females,and members of differentethnic or religious groups. Such patternsstemmed from a varietyof causes. DifferentAfrican peopleswere regardedand treateddifferently by colonialadministrators; the costs of providingeducation differed, certainly between urban and rural areas; some populationgroups were more responsiveto educationalopportunities than other groups;and most Africans,responding to the incentivesimposed by patrilinealcustoms, preferred education for their sons to educationfor their daughters. As a result,problems of unequaleducational perticipation frequentlytranscended colonial boundaries. Participationpatterns in northern Nigeria, for example,had less in commonwith those in the south of the same Britishterritory than with those in the north of neighboringFrench Cameroon. Such within-countrydifferences and between-countrysimilarities remain evident today.

Transitionrates from one educationallevel to the next were low in 1960, and drop-outrates were high. As a result,enrollment pyramids were typicallyvery narrow at the top. Only 6% of all sub-Saharanenrollments in 1960 were at the secondarylevel, and tertiaryeducation was virtually non-existentuntil the very end of the colonialperiod. The gross enrollment ratio at the secondarylevel in Africawas 3% in 1960, as comparedwith 14% in Latin Americaand 21% in Asia. The ratio at the tertiarylevel was 1500, about one-sixtieththose then found in Asia and Latin America. Accordingto - 6 -

UNESCO Zigures,at the time of independencethere were only 90 African universitygraduate. in all of Ghana, 72 in SierraLeone, and 29 in Malawi. When Botswanabecame independentin 1966, 96X of higher-levelposts in the countrywere filledby expatriates.

1.2 Advancesin educationafter 1960

The systemsof educationinherited by the Africannations at the time of independencewere altogetherinadequate to meet the needs of self-governance and rapid economicgrowth. From this low startingpoint, the progressachieved in African educationhas been remarkable.

Quantitativeexpansion has been particularlyiaipressive. Between 1960 and 1983, the numberof studentsenrolled in African institutionsat all levels quintupledto about 63 million students(Annex Table C.2). Enrollments increasedabout 9% annuallybetween 1970 and 1980, doublethe rate in Asia and triple that in Latin America. The substantialexpansion of educationafter independencehas allowedthe increasedparticipation of some groupswho had littleor no prior access to formaleducation.

Primary-schoolenrollments increased the most in absoluteterms, growingfrow approximately11.9 millionpupils in 1960 to 51.3 million pupils in 1983 (see Table 1.1). The gross primaryenrollment ratio rose from 36% to 75% over this period. In only three of the region's39 countries(Congo, Lesotho,and Mauritius)was the enrollmentratio higher than 802 in 1960; by 1983, 16 countrieshad achievedthis milestone(see Annex Maps 3a and 3b).

(Table 1.1]

In fact, 12 of the 39 countriesnow have gross primaryenrollment ratios equal to or greaterthan 100% (AnnexTable A.7). Enrollmentratios in excess of 100% arise from the definitionof the gross enrollmentratio, in which the denominatoris a measureof the officialschool-age population, and the numeratorcounts all pupils enrolledin school,regardless of age (see foot-ote1).

Enrollmentsat the higher level increasedthe most in relativeterms, especiallybetween 1960 and 1970. This resultedfrom the emphasisthat most Africannations were givinghigher educationin an effortto alleviatemanpower shortagesand also from the low level of base enrollments. In 1960, there were only about 21,000university students in Africa (approximatelyone in 500 of the age group) and a few thousandstudying in foreignuniversities; by 1983, 437,000 (seven per 500) were enrolled in African institutions, and a further 100,000Africans were studyingabroad. The relativeincrease in tertiary enrollmentswas particularlydramatic in French-speakingAfrica. In the group of 18 Francophonecountries, there were 40 times more studentsenrolled in higher-levelinstitutions in 1983 than in 1960. By way of contrast,in the 16 Anglophonecountries, where enrollmentswere much higher initially,enrollments had increasedby only a factorof 15. For every studentenrolled in higher educationin the Francophonecountries in 1960, therewere 4 enrolledin the Anglophonecountries; by 1983,the ratio was only about 1-to-1.5.

The buildingof schoolsand trainingof teachersundertaken to accommodatethe additionalstudents throughout the regionwere mammoth achievements.Between 1960 and 1983, the number of primaryschools in Table 1.1 School Enrollmentsand EnrollmentRatios in Sub-SaharanAfrica, 1960 and 1983

Year

Level 1960 1983

PrimaryEducation Enrollments 11.9 million 51.3 million Gross EnrollmentRatio 362 75%

SecondaryEducation Enrollments 800 thousand 11.1 million Gross EnrollmentRatio 32 202

Higher Education Enrollments 21 thousand 437 thousand Gross EnrollmentRatio 0.2S 1.42

Total Enrollments 12.7 million 62.9 million

Note: Based on Annex Table C.2. - 8 -

sub-SaharanAfrica increasedfrom about 73 thousandto roughly 162 thousand, and the npumberof primaryschool teachers, from 310 thousandto over 1.3 million.-FInterestingly, the averagepupil-teacher ratio remainedroughly the same over this period (approximately39-to-:); on the other hand, the average primary schoolsize increasedfrom 162 pupils in 1960 to 317 in 1983,almost doubling.

The numberof teachersemployed at the secondarylevel increased eightfold,going from about 46,000 in 1960 to about 373,000in 1983. This understatesthe increasein Africanteachers, since many expatriateteachers were replacedduring this period. At the tertiarylevel, the number of institutionsin sub-SaharanAfrica has more than tripledsince 1960 and today exceeds80.

The consequenceof this massive educationalexpansion, which began in some countriesin the 1950sand intensifiedeverywhere after independence,has been a substantialdeepening of the human capitalstock. Using data from countriesthat conductedcensuses around 1980, it is possibleto examinethe changingeducational attainment of successivecohorts of individualswho were of school age at differentt: 1 mes over the 30 years beginningin 1950. For the few countrieswith information,the proportionof males who reachedadulthood withouthaving attendedschool declined from 31% in the early 1950sto 22% in the late 1970s in six Anglophonecountries, and from 59% to 42% in four Francophonecountries. For the same countriesand over the same period,the proportionof males who attendedand completedprimary schoolrose from 47% to 58% in the Anglophonecountries, and from 16% to 30% in the Francophone. Estimatedmean years of schoolingfor the male population(over age 15) increasedfrom 4.6 to 5.4 years in the Anglophonecountries and from 2.0 to 3.4 years in the Francophone.

An alternativeto using census data for estimatingstocks of educationis to "add up" past enrollments. In conjunctionwith demiographic assumpcions,this allows estimatesto be made, countryby country,of the educationalattainment of the workingage population. Table 1.2 reports estimatesof educationalstocks produced in this fashion. The estimated averageeducational attainment of working age individuals(men and women) in the median African countryincreased from 1.0 years in 1970 to 3.3 years in 1983. The figuresin the table underlinethe wide disparitiesthat remain in this variable,especially between the Sahelian(low-income semi-arid) countries and the rest of Africa.

[Table1.2]

Table 1.2 also shows how literacyrates have risen since 1960. In the median African country,the percentageof adults reportedlyable to read and write has increasedby a factorof nearly five. This progressreflects both the growth of formaleducation and, in many countries,successful programs to promote literacyamong adults and young people not attendingschool. Approximatelya quarterof all African countrieshave launchedprograms of adult illiteracyeradication. - 9 -

Table 1.2 Indicatorsof EducationalProgress

Median estimated Median numberof years of literacy schoolattended by rate working-agepopulation (percent)

Latest 1965 1983 1960 available year

Economicstatus groups

Low-incomesemi-arid 0.1 0.9 2 15

Low-incomeother 0.5 2.9 10 41

Middle-incomeoil importers 1.3 4.2 19 72

Middle-incomeoil exporters 0.7 3.6 16 56

Linguisticgroups

Francophone 0.5 2.4 7 40

Anglophone 1.2 3.4 18 58

Sub-SaharanAfrica 0.5 3.3 9 42

Notet Based on Annex Tables C.3 and C.4. - 10 -

Other forms of adult and nonformaleducation have also been strengthenedin the first two decadesof independence.Departments and institutesdedicated to the productionand study of adult educationhave emerged in many universities,and nongovernmentbodies concernedwith functionalliteracy and income generationin urban and rural areas have proliferated.

In additionto the remarkablequantitative achievements, many other significantchanges have occurredin African educationover the past quarter century. Especiallyat the primary level,Africanization of the curriculumhas been a widespreadaccomplishment. In nearly every part of sub-SaharanAfrica, texts have been adaptedand new texts written so that basic skills are now taughtwith referenceto Africancustoms, the local environment,and the area's own history. Twenty-oneof the 39 countriesin the regionnow officiallybegin instructionin one or more African languages,rather than askingchildren to use a Europeanlanguage as the medium of instructionfrom their first day in primary school. These reformshave transformedthe Africanclassroom and made it far more conduciveto an African child'srapid acquisitionof the cognitive competenciesstressed in the curriculum.

1.3 Expenditureon education

To appreciatethe high degreeof commitmentof African nationsto educatior£ developmentand betterunderstand the significantadvances that have occurred in the sectorsince 1960, it sufficesto look at the flow of resourcesallocated to the sector.

(a) Public domestic expenditure on educaticn

In 1970, the 39 countries of sub-Saharan Africa allocated approximately $3.8 billion of public domestic expenditure to the education sector. This is a World Bank estimate of total, i.e., capital olus recurrent, expenditure(see Box 1.1), expressedin constant1983 U.S. dollarsa.3 By 1975, this figurehad reached$6.3 billion,up 66% over 1970. By 1980, it was nearly $10.0 billion,up another58%. Between1980 and 1983, public expenditurefell somewhat down approximately$1.1 billion,to $8.9 billionin constant1983 dollars.-/ Althoughdata are unavailablefor 1984 and 1985, expendituresmay have resumedgrowth subsequent to 1983, when the economiccrisis was at its peak.

[Box 1.1]

Given that the use of officialexchange rates may give a false picture of levelsof expenditureover time and may also distortcomparisons of expenditurebetween regions or countries,it is importantto consideran alternativemeasure of publicexpenditure on education. Expenditureas a percentageof nationalincome is an especiallymeaningful indicator of a government"s'effort' in the area of education. Figure 1.1 shows public domesticexpenditure on educationas a percentageof nationalincome in Africa and elsewherein four specificyears between 1970 and 1983. The figure shows that sub-SaharanAfriza has been allocatinga largershare of total income to educationthan have the developingcountries in general,but still a smaller share than developedcountries have managedto do.

(Figure1.1] - 11 -

BoX 1. 1 AOGRDITB PUBLIC D@SIC UD11 TRZ 0N XIDCATIONINU3-SAERhA AMlC

lh diboslon of ptilic Oxatlion sqiture In the text Is baed on Inforutll providd by UNESCOalm esrnted In kmx TablesA.11A.23. Secas. for iW glvu ywu, oditure datahte be reportedfor ally a stet of couitrles, the total of pblic drnic e#latlon oxndltsue for all 30Afrlcoan omitrles In that year Is not kn. For exmale,for 1910,the. Is oadlture Infonratlonavailable for only 29 of the 39oountries (use Amex Table A.14).

In an atteapt to bypassthis data l Imtation, the table In this boxreports fa I Lo ll of pubilc sedcatlon SqpendltureIn 1070.10S5, 1080, aid 1083for the 39¢aoutrles as a gros. Noesaagrptes wer arrived at by takingall oouIntrle for thIch actual MUt a avallablefor the partlcular yearand sstirntirg. for thl8 suet of outrles, the averp relatlwislhp btun 4dfts ad curtaln othsr mares Includi per capita Iram ad srollmit at eachof the three slation laels. fTe tlad rolatlonilip wu thn usd to IrPute hditre flga's for the rualnlg coltrles for abidh eabl value wereMo reprted. Firally, thOs stimatud valu wre aded to the sm of reorted valuesto arrlv at the aggrefgtevalues gIvn In the tale.

Public dometic zpeuditure on education, 1970-1983

1970 i9n5 1980 1983

amtrla far hlrb AClUAIdat avallable(me hM TableA.141 der of coIrles 27 28 29 28 %of Ub-SahranAfrlIcn populatlon 801 l 89% 75% Eat d aM dtira for all db-rAn African nmtrlms (1983U.S. dollars, In mlIwos) CapitaleapedltUe $459 $1,767 $1,318 $885 Raurent mewnditre $3,328 $4,518 $8.638 $8,032 Total upaiture $3.788 #6,283 $9,952 $8,897 IONdvaltM (me trah lawmv Total sndlttre 100 1e 2B3 235 Total urol lMit 100 148 235 288

280- ox Toitl Enwohvient

240 Total Expenditum

180 / RecurrentEpenditure 200 180

140-

120 -

40-

so 40

1970 1915 1960 1983 - 12 -

Figure 1.1

PublicEducational Expenditure as a Percentage of NationalIncome, Africa and Elsewhere

7.01 -

8.15 6.11 6.l5 6.01 -

5.01~~~ ~ ~ ~ ~~~~~~~~~~55

4.0_

2.0X

1970 1975 1980 1983

0 SUB-SAflARANATRICA + AIJ DEVEWPIN(GCOUNTRES O ALL DEVELOPEDCOUNTIES - 13-

Anotherway to look at public expenditureon educationis as a percentageof total public expenditure. On this indicator,the African countrieshave clearlyin the past treatededucation generously. The median African countryallocated 17.6% of the tbidgetto educationin 1970, 17.4% in 1975, 18.5% in 1980, and 15.3% in 1983.-1 At the lower end of the spectrum, Malawi,Nigeria, and Somaliaall spent below 10% of the centralgovernment budget on educationin 1983. At the upper end, education'sshare was above 20% in Burkina,Cote d'Ivoire,Mali, Niger, Rwanda,and Togo.

With regard to the distributionof expenditureacross levelsof education,the recurrentexpenditure on primaryeducation was between 43% and 49% of total recurrentexpenditure in the median African countryover the period 1970 to 1983. The share going to secondaryeducation was between25% and 31%, and the share going to tertiary,between 13% and 18%. The 'non-specified"category, which, as a residual,cannot be taken very seriously but which would includepresumably most allocationsfor centraladministrative servicesin ministriesof educationand for adult education,remained nearly constantat about 10% (AnnexTable A.16).

(b) Privatespending on education

The focus in the Annex tables on public expendituresreflects the relativelybetter data availableon this, as distinctfrom other sourcesof educationalfinance. Althoughinformation on the total of private spendingis not availablein most countries,it is clear that familiesand nongovernment organizationsbear a significantand growingportion of the financialburden of educationin much of Africa. For example,it has been estimatedthat privatE expenditureaccounted for 14% of total nationalspending on educationbetween 1975 and 1980 in the Sudan, 23% in Tanzania,31% in Zimbabwe,48% in Sierra Leone, and 53% in Ghana.

These figurestend to be largertoday than they were in the early years followingindependence. Government-collected fees (whichin official statisticsare often countedas part of "public"expenditure on education)have been imposedor increasedrecently in many Africancountries. In additionto these fees, there are significantprivate outlays that are not channeled throughgovernment. These difficult-to-quantifyflows includemost fees paid to privateeducation institutions; the privatelyborne costs of such items as trasisportation,school uniforms, textbooks, and supplies;and very important, especiallyat the primary level,family and communityoutlays, either in cash or in kind, used for the constructionor repairof school buildingsin public education. Commitmentof studenttime is, of course,another valuable input to the educationalsystem, but one that this paper does not attemptto accountfor separately.

During the colonialperiod, much of Western-styleeducation in Africa was organizedand, to a significantextent, financedby missionarygroups based overseas,often relyingheavily as well on communityself-help. In the period after independence,government assumed responsibility for most such education in nearly every Africancountry, Lesotho being a major exceptionto this rule. During this period,wherever school fees had been charged,often thesewere reducedor eliminated. More recently,however, in responseto growingfiscal and demographicpressures, African leadershave condoneda reversalof this post-independencetrend. In the late 1970s and 1980s,students, their families - 14 - and communitieshave been asked to contributeever largeramounts towards the costs of education. Often, to avoid having to collectschool "fees," governmentshave referredto these subscriptionsby anothername, such as developmentfund contributionsor activitypayments.

(c) Foreignaid flows to Africaneducation

Foreignaid also accountsfor a significantpart of total spendingon African education(Annex Tables A.24-A.27). It is estimatedthat, in the period 1981-83,the total of bilateral,multilateral, and privatevoluntary organizationaid to African education,granted through central government ministriesof education(and not throughother departmentsof gover ent, nor directlyto individuals)was in the order of $916 million per year.- Nearly a quarterof this was in the form of scholarshipsfor African studentsto study abroad;a significantportion of aid to education(442 of Westernbilateral and multilateralaid) was in the form of technicalassistance.

African governmentsreceived an estimatedadditional $394 millionof externalaid each year to financeproject-related training in sectorsother than education. In addition,when African studentsstudy abroad,indirect costs are incurredby the countriesthat host them; these have to do with the nonexplicitsubsidies that students,including foreign students, receive at most universities. A very rough estimateof the total of such subsidiesto Africanstudents who were studyingabroad between 1980 and 1982 is $245 million annually.

In conclusion,the total of all categoriesof externalaid to African educationand trainingin the early 1980s was on the order of $1.6 billion annually. This is a significantamount, nearly $4 per capita,but there is no evidenceto suggestthat externalaid has increasedin recentyears to compensatefor the decline in African governmentexpenditures. To the contrary,tfs total of all aid to Africa for all purposeswas smallerin 1983 than the 1980-82average, and smallerin 1984 than in 1983. This paper returns,in Chapter9, to furtherdiscussion of externalaid to educationin Africa. - 15 -

Chapter 2. EDUCATIONAND THE EXTERNALENVIRONMENT: A CYCLE OF DETERIORATING PROSPECTS

The impressivegains recentlywon in African educationare now seriouslythreatened by a numberof circumstancesexternal to the sector, includingmost importantly: (1) Africa'sexplosive population growth, which swells the numberof potentialilliterates on the continent,and (2) recent economicdecline, which has necessitatedsignificant cutbacks in public spending. While economicprospects may have brightenedsomewhat since 1983, and populationgrowth has begun to slow in a handfulof countries,it cannotbe denied that, for most of sub-Saharan Africa, currenteconomic and demographic realitiesundermine both the quantitativeand qualitativeeducational advances achievedin the period since independence.Retrogression in educationwill, in turn, make the solutionof these problemsmore difficult. And so the pattern is repeated, and it is destined to continue until extraordinary efforts are made to interruptthis cycle of deterioratingprospects.

2.1 The demographicchallenge

Between 1970 and 1980, while the world'spopulation was growingat an averageannual rate of 1.9%,Africa's population grew at 2.9%, one-and-a-half times the world's rate. Between 1980 and the end of the century,Africa's populationis projectedto grow even faster,at about 3.2% per annum; further declin!,in growth rates are projectedfor all other major regionsof the world.-

Africa'srapid populationgrowth creates serious problems for education. In order for the growthof educationalplaces to keep pace with the growth of school-agechildren, more schools,teachers, books and other inputs are requiredeach year. Moreover,because as populationsgrow, the numberof school-agechildren increases more rapidlythan the number of working adults, the burden of supportingan expandededucation system falls on an adult populationthat is shrinkingas a proportionof the overallpopulation. Sub-SaharanAfrica today has the youngestpopulation of any regionin the world. One in three persons is of primaryor secondaryschool age in Africa (see Figure 2.1), as comparedwith just one in five persons in Latin America and in Asia, and one in six personsin the industrializedcountries.

[Figure2.1]

It is projectedthat, by the year 2000, Africa'sprimary and secondary school-agepopulation is projectedwill have reached220 million,which is 70% or 90 million largerthan the number in 1984. This representsa massive explosionin the potentialdemand for educationalservices and one that is virtuallyinevitable over the short time horizon. This fact is of fundamental importance. It sharplyconstrains African planners' options, and its implicationscolor all of the analysiscontained in this paper.

[Figure2.2] - 16 -

Figure 2.1 SUB-SARANMAFRICA POPULATION ACE STRUCTUPRE. 1984

too-

ftpur 2.2 GROUTHOf SCHOOt-AGEPOP.TICRS, l965-I98 A101986-2000

.~~~~s l404 l I~~~~6 i 120

80

s rm - ~- - 17 -

For example,there were 51.3 millionprimary and 11.1 million secondaryschool places in Africa in 1983. These figureswould have to reach 90.7 million and 19.7 million,respectively, by the year 2000 just to maintain participationrates at their 1983 levels (see Table 2.1). The goal of universalprimary education(UPE), wh'.ch African education planners at one time hoped could be achievedeven by 1980, unfortunatelystill seems very far away from the perspectiveof the mid-1980s. For the regionto have all primary school-agechildren enrolled by the year 2000 would requirethat 131.8million primary schoolplaces be made available,a 157% increasein just 17 years. The 80.5 million additionalplaces requiredare more than all of the places that exist now aftermany decadesof educationaldevelopment.

[Table2.1]

The investmentin the constructionof new classroomsand the recurrent costs necessaryto supportsuch massiveexpansion in education,while not altogetherunreasonable if the achievementsof the 1960s and 1970sare simply extrapolatedforward, unfortunately do not seem very likelyin the contextof long-runausterity beginning in the 1970s. Table 2.1 shows that to maintain gross enrollmentratios at their 1983 levelswould requirethe additionof 39.4 million primaryand 8.6 million secondaryschool places in the 17 years between 1983 and the turn of the century. Assumingthat per pupil expenditureswill remain at their 1983 levelsin constantdollar terms (roughly$50 per primary, and $250 per secondaryschool pupil),the recurrentcosts of just the first two levelsof educationwould go up by approximately$4.0 billion,becoming $9.4 billionby the year 2000. This is alreadymore than the total amount allocated by African governmentsto all three levelsof educationand their central administrationin 1983.

In additionto meeting the swollenrecurrent costs, school buildings and equipment would need to be financed -- both to meet the needs of new students,and to replace (at an estimated2% per year) oatmodedfacilities. Assumingvery modest capitalcosts of just $150 per primaryplace and $1,250 per secondaryplace (thisbest-practice scenario involves capital costs well below actual averagecapital costs in Africa today),the requiredinvestment in facilitieswould be another $1.4 billioL.annually. The total of recurrentand capitalexpenditures for primaryand secondaryeducation alone would come to nearly $11 billionannually. This is $2.1 billionmore than the total spent on all aspectsof educationin 1983. In sum, the demographictide requires massive effortsjust to stay even in terms of enrollmentratios. Table 2.2 summarizesthe cost implicationsat the primary and secondarylevels of just stayingeven with demographicgrowth.

[Table2.23

Ironically,in additionto explosivepopulation growth, much of Africa suffersfrom anotherdemographic condition -- low populationdensity. The combinationof rapid growthand low densitymay seem strangesince, given enough time, the firstwill eventuallyeliminate the second. In the meantime, however,low populationdensity often implieshigh unit costs in education, especiallyin rural areas,where economiesof scale in the provisionof educationare precludedgiven presentmethods for deliveringeducation. At this time, Africa'spopulation density is about 19 people per squarekilometer, as comparedwith a worldwidedensity of about 36. - 18 -

Table 2.1 Sub-Saharan Africa Primary and Secondary Populatiots and Enrollments, Actual (1983) and Projected (2000)

PrimaryedicatIon Secorary Icatlon

%chr %cug 1983 2000 1983-2000 193 200 182000

School age population

Actuml 68.5 - - 55.6 - -

Projected - 120.9 76% - 98.6 77%

School enrollment

Attal 51.3 - - 11.1 - -

Projctedaswlrg:

Grossrrolint ratio saw as In 1983 - 80.7 77X - 19.7 77%

Errollhwt ratio 10w,no repetitionand no drApout - 120.9 136% - 98.6 788%

Enrollent ratio 100X, roWttiln aid dropwt sae as In 183 - 131.8 157% - 105.5 8505

Note: Pulation anderrollmnt flgres In elilrIs. Bad n AmexTables A.1, A.2, A.7,A.8, A.12,ard 8.4. - 19 -

Table 2.2 The Cost Isplications of Maintaining 1983'. Enrollment Rates through the Year 2000

1983 2000

Required Increontof Averagp Recurrent Recwrrent Recurrent ArnualCapital Level ErrolimentsExpenditures Enrollmnt8 Expenditures Expenditures Investment

Primary 51.3 2.8 90.7 4.5 1.9 .5

Secondary11.1 2.8 19.7 4.9 2.1 .9

Total 82.4 5.4 110.4 9.4 4.0 1.4

Note:Enrol lEnts are In elIjjJ.m ofstudents; uq2enditures areIn bill lnamof 1983dol lars. - 20 -

2.2 Macroeconomicadjustment and fiscalausterity

The economiccrisis, initiated with the oil price increasesof the 1970s,accelerated after 1980, leavingmost Africaneconomies in serious disarray. (Oil exportersbenefited of course,at least in terms of short-term availabilityof foreignexchange.) Agriculture,which accountsfor the largest share of goods produced(about a third of total production),was hurt by the droughtthat affectedmuch oi the region,by a marked deteriorationiT terms of trade (decliningagricultural prices combinedwith rising energyprice;), and by the continuationof nationalpolicies that discriminateagainst the sector. Moreover,the investmentrate in sub-SaharanAfrica fell from above 18% of incomein the 1970sto below 15% in 1983 and is currentlythe lowestof any developingregion, thus threateningto undermineAfrica's long-termproductive capacity.

Economicstagnation, combined with rapid populationgrowth, has meant a decline in livingstandards, a fatal declinefor many individualsin those countriesworst hit. Overall,income per capita fell nearly 4% annually between 1980 and 1984; it is lower in the regiontoday than it was 20 years ago. The externalpublic and publiclyguaranteed debt increasedeleven-fold between 1970 and 1984,and debt-servicepayments more than tripledas a percentageof GNP, from 1.2% to 4.4%. Centralgovernment expenditures increasedwhile revenuesremained about the same, and fiscal deficitson the continentrose to about 10% of GNP.

Becausecontinuation of the above trendsspelled economic disaster, African governmentsrecognized that macroeconomicpolicy reformswere an absolutenecessity. In this regard,they have made substantialprogress since 1980, along the generallines prescribedat the U.N. SpecialSession on Africa in May 1986. Thus many governmentshave reconsideredtheir exchangerate policies,instituted wage and salary reforms,and begun to eliminateprice distortionsthat penalizedfarmers. They have also reducedpublic spending, includingspending on education(Section 1.3).

It is expectedthat these structuralreforms will graduallyencourage higher investmentand lead to increasedconsumption levels. What is more, the droughtin Africa has recentlyabated, and there has been an improvementin the region'sterms of trade. Low-incomeAfrica registereda growth in per capita income in 1986, for the firsttime in the 1980s;this performanceis expected to be repeatedin 1987. In sum, there may be room again for cautiousoptimism with regardto the economicenvironment and long-termgrowth prospects of the region,but this relativeoptimism should by no means give rise to complacency.

2.3 Investmentin education:breaking the cycle of deterioratingprospects?

Extensiveexperience from Africa and elsewhereprovides strong evidencethat increasedinvestment in educationand trainingat this stage in Africa'shistory can be expectedto yield very broad economicbenefits -- including,of criticalimportance in the Africancontext, raised incomes and reducedfertility. Althoughdirect evidence for developingareas and especiallyfor sub-SaharanAfrica is less extensiveand rich than for the developedcountries, enough informationexists to concludethat the direction of the relationshipbetween education and variousindicators of economic - 21 - well-beingis the same everywhereand that the positiverelationship holds true for both formalgeneral education and for training. The evidencesuggests, however,that the magnitudeof education'simpact is somewhatlarger in developingcountries, because of education'srelative scarcity there.

Educationcannot, in and of itself,bring about economicgrowth, but the evidenceindicates it to be a vital factor. It providesthe fertileground withoutwhich other developmentinitiatives will not take root. Educationacts as an "accelerator"in the growthprocess, working as an essentialcomplement to other factors. Thus, for example,although education is associatedwith increasesin agriculturalproductivity, its impactis found to be much greater within the contextof an alreadymodernizing environment. Likewise, investmentsin the provisionof familyplanning services have been shown to have 'higherimpact if the women using the service (or potentiallyusing it) are better educated.

The economicrationale for enhancedactivity in African education, elaboratedbelow, may be complementedwith severalbroader and politically compellingexplanations of why Africancountries have investedso heavilyin educationsince independenceand with argumentsfor why they should endeavorto protectand expand their commitmentto the sector in the future. First,the essenceof sovereigntyis the controlof nationalsover the destinyof their countries. The transitionfrom colonialdependence to self-governmentand to active independentparticipation in the internationalarena was possibleonly becauseAfrican educationalsystems were able to produceindigenous personnel at all levels to replaceexpatriate rule. The qualitativenurturing and quantitativeregeneration of that leadershipelite, so that it can addressthe increasinglycomplex challenges of nationbuilding in the future,is a continuingrequirement for the effectivemaintenance of sovereigntyand, as such, a fundamentalresponsibility of Africansystems of education.

Second,the rate and extentof the growthof the human capitalstock in Africancountries, attained through the improvementand expansionof educationat all levels,will ultimatelydetermine whether the universal explosionof scientificand technicalknowledge can be harnessedfor the benefitof the region,or whetherAfrica will be left behind and deniedthe enormousbenefits accruing to technologicalchange. Finally,education is everywherea "meritgood," a basic human right,an end in itself;indeed, educationis an intrinsicelement of the developmentprocess. It is unthinkablethat African governments,and their internationalpartners, would permit a declinein the fractionof the populationhaving access to education's many benefits. And yet that is the spectrethat now menacesmany African countries.

The economicevidence indicating large payoffsto educationis of three principalsorts, each providingcorroboration for the others. The first two rely on individual-leveldata; the last consistsof macroeconomicevidence. The first one examinesthe relationbetween an individual'seducation level and his or her productivityin the labor market as a wage employeeor self-employed worker. The second,also at the individualor householdlevel, links the educationof individualsto importantoutcomes of householdbehavior such as fertilityrates or child survival. Finally,there is macroeconomicevidence, which relatesthe growth rates of nationaleconomies to prior investmentsin education,controlling for other factorspresumed to influencegrowth. The macroeconomicevidence is an important complementto the individual-leveldata, - 22 - particularlythe data on educationand earnings,in that it providesevidence to assessthe view, occasionallyadvanced, that educationimproves the lot of individuals,by reallocatingsocietal resources, but does not .ncreasethe overallflow of resources. The remainingpages of this chaptersummarize evidenceof all three types, includingan analysisconducted in conjunction with the preparationof this paper that addressesthe macroeconomicissues.

(a) Educationand laborproductivity

Economistshave assessedthe economicvalue of educationby observing differencesin the earningsof workerswith differentlevels of education, having controlledfor other differencesthat exist between the groups,and then comparingthe adjustedearnings differences with the costs of the education. The indicatorthat summ rizes the informationon costs and benefitsis the social rate of return.#2 A recentsurvey of cost-benefitstudies conducted in 16 African countriessuggests average social rates of returnto investmentin African educationof the followingmagnitudes: primary26%, secondary17%, and tertiary 13%.

Althoughthese rates of return certainlyseem attractiverelative to many other forms of investment,there are importantcaveats. These studiesmay overstatethe actual,current rates of return to educationin Africa,because the data on which many of the estimatesare based are out-of-date-- the cross-sectionalearnings data used in these studiesare, on average,10 years old, and there has been considerableeducational expansion over the decade since then. A more recentstudy from Kenya (see Box 2.1) does indicatethat rates of returnhave fallenoff somewhatand that the substantialdifference (nine percentagepoints) observed in the past betweenprimary tnd secondary educationmay have completelydisappeared, which would reflectthe declining relativescarcity value of primaryeducation. Even so, the Kenya study confirmsthat rates of returnto both levelsof education,after adjustingfor other factors,remain substantial (around 12%).

[Box 2.1]

The relativelylower rate of returnto educationat the tertiarylevel should not be interpretedto mean that high-levelskills are not importantin Africa. What the findingsreflect most of all are the extremelyhigh unit costs of tertiaryeducation. In addition,most studiesascribe all costs at this level to teachingand none to other outputsof the tertiarysystem, such as research. If the costs of tertiaryeducation were to be prorated,and the value of the researchand other outputsof the educationassessed (including some that may accrue to societyin general,and not narrowlyto thosewho receivethe education),then the rate of returnto tertiaryeducation would certainlyappear higher.

Many of the early rate of returnstudies relied on earningsdata observedfor workers employedin the formalwage sector. To the extent that the percentageof primaryschool leaversable to find employmentin the formal sector has declinedrecently (relative to the percentageemployed among uneducatedpeople), and to the extentthat earningsin the formalsector are "protected"and thus remainhigher than elsewherein the economy,the estimated rates of returnmight have been inflated. The perceivedvalidity of suc;h benefit-coststudies has been considerablyenhanced, therefore, by a numberof recentstudies of education'simpact, including some done in Africa, that are - 23 -

Box 2.1 RELATIVERATES OF RETURNTO PRIMARYAND SECONDARYEDUCATION IN KENYA

Perhaps the most Influentlalfact to emerge from twenty-fiveyears of rate of returnstudies In developingcountries concerns the relativerates of return to primaryand secondaryschooling. Most studles,Including most of those conductedIn Africa,have reported substantiallyhigher rates of return at the primary level. The Implicationgenerally drawn from this ranking Is that top priorityshould be given to primaryeducation as a form of Investment In human resources. Recent studiesfrom Kenya throw some new light on this Issue.

Returnsto education are measured by the gaps In wages, and presumably productivity,between workers with differentlevels of education. It Is often assumedthat the averagewage of labor with a given level of education measures the wage recelved by the marginal (most recently recruited)worker wlth that levelof schoollng.However, the averagemay not always Indicatethe marginalwage.

Due to rapld expansionIn the educationalsystem, the labormarket conditlonsfaced by those Just leavingschool are very different from the conditionsfaced by earliercohorts when primaryschool leaverswere In short supply. For those enteringthe labormarket a generation ago, a primary school certificatewas a passportto a white-collarjob, and typicallythose who obtained Jobs at that time remain In them today. But, due to the expansion of the educational system, today'sprimary school completerIs fortunateto get even a menial blue-collarjob, and his chance of obtaininga white-collarposition Is virtuallynil.

The process by which successivecohorts of workersat a particular educatlonallevel enter lowerskilled Jobs Is called Ofiltering down.0 A recent study took this process Into account In calculatingaverage and marginalrates of return ln Kenya. The study found that the rate of return to primary schoolingIs highlysensitive to the distinctionbetween average and marginalrates of return, whereas the rate of return to secondary schooling Is not. The averagereturn to primaryschooling, as conventionally measured,Is 17%; the marginalreturn Is 12%. The marginalreturn Is lower for two reasons. At the primary level,there Is both substantialfiltering down and large differencesIn wages by occupatlon,whereas for the uneducated there Is less scope for filterlngdown and wage differencesby occupatlonare small.

The return to secondaryeducatlon, by contrast,Is not affected by the equivalentadjustment: the averageand marginalrates of returnare both 13%. Because the degreeof filteringdown of primaryand secondarycompleters Is similar, the dlfferential In earningsbetween the two groups ts llttle affected. In Kenya,moving from the averageto the marginal concept erases the usually reporteddifference In rates of return:at the margin, the rate of returnto secondaryeducation (13%) about equals the rate of return to primaryeducation (12%). - 24 -

based on informationabout self-employedworkers. Since the self-employeddo not receive "wages'as such,many of these studieshave sought to estimate education'svalue directlyby lookingat the value of what more educated individualsproduce. Most of the researchto date has focusedon agriculture and, within agriculture,especially on crop production.

A recent reviewof 18 studiesof farmereducation and farm productivityin 13 countriesconcluded that farmerswho have completedfour years of educationproduce, on average,about 8% more farm output than farmers who have not gone to school,controlling for differencesin the use of physical inputs. Moreover,the percentageincrease in output associatedwith four years of educationis foundto be about 10% in "modernizing"environments (indicated by such factorsas the availabilityof new crop varietiesand reasonable incentives);in more traditionalenvironments, where technologyand opportunityare changingonly slowly,there is littlepayoff to education.Only one African country,Kenya, was includedamong the 13 countriesin the review, but the resultsthere were consistentwith the generalfindings. A recent study of this kind based on data from South Asia identifynumeracy and literacy among the criticalcognitive skills throughwhich education'seffect on farmer productivityis mediated.

One may concludefrom all of the availableevidence that ratesof return to investmentat all levelsof educationin Africa today compare favorablywith realizedrates in most other sectors,including some sectorsfor which the World Bank and other developmentagencies have been providing relativelylarge amountsof money in recentyears. The Bank'sown experience with project lendingreinforces what researchershave concludedregarding the value of education. The probabilitythat a developmentproject in any sector will be successfulhas been observedto increaseas a functionof the human capacityavailable to plan, implement,and benefitfrom it. The absenceof importantskills, particularly in the area of projectmanagement, is often cited as the primaryreason for the failureof particularprojects to meet their objectives. More generally,the educationallevel of the entire populationinvolved in a project -- farmerswho would be using a new packageof inputsmade availableunder a project,or women who would be visitinga new health or familyplanning clinic -- is often a key elementin determining whether or not the projecthas a significantimpact. Investmentin education, in sum, increasesthe returnto investmentsin other sectors.

There are other ways that education may enhance worker productivity. It has been shown, for example,that educationincreases the propensityof individualsto migrate. This will increaseeconomic growth to the extentthat individualsmove, as they will naturallyseek to do, from employmentin lower productivitysectors to employmentin higher ones. Moreover,employed urban migrantstypically remit a part of their earningsto their familiesback home, and this resourceflow has been shown to result in increasedinvestment in Africa'srural areas.

There is also evidenceto suggestthat educationcan, under certain conditions,be expectedto fulfillequity goals. Since it was first introduced in Africa,Western-style education has been a vehicleby which able childrenof poor familieshave managedto move to higher levelsin society'soccupational and income structure. As access increases,the additionalexpenditures on educationwill flow increasinglyto the disadvantagedelements of society,to the poor, those in rural areas, and girls. The study referredto in Box 2.1 - 25 - above with referenceto Kenya comparesthe experiencein that country,where secondaryeducation was allowedto expand rapidlyduring the 1970s as a functionof demand,with the experienceof Tanzania,where the expansionof secondaryeducation was relativelyconstrained. Althoughboth Kenya and Tanzaniaexpressed concern over the widely dispersedwage structuresthat existed,Kenya's policy resultedin substantiallygreater compression in the distributionof earningsover the period than occurredin Tanzania. In addition,the largersize of the secondarysystem has meant greateraccess for the childrenof the poor and uneducatedin Kenya than in Tanzania.

Educationyields the recipientsother importantbenefits that are not (immediately)reflected in the form of increasedearnings or a more equal distributionof earnings. These benefitshave to do with fertility,health, and individualfulfillment.

(b) Loweredfertility and other normarketbenefits of education

In general,there is a strong negativerelationship between how much educationa woman receivesand the numberof childrenshe gives birth to during her lifetime. Althoughin very low-incomeareas, fertilitymay, becauseof education'simpact on women's health and fecundity,actually increase with the first few years of schoolinggiven to females,nevertheless, research on the determinantsof familysize in Africa and elsewhereindicates convincingly that raisingthe educationalattainment of women resultsultimately in fertility reduction. This longer-termbehavioral change occurs througheducation's direct effectson family formationdecisions and throughits indirecteffects on child survivaland women'semployment opportunities.

From the householdperspective it is desirablethat, through education,women be given the power to controlthe number and spacingof the childrenthey bring into the world. Moreover,in light of the economic problemsaesociated with Africa'shistorically unprecedented population growth rate, the social benefitsof education'simpact on fertility-relatedbehavior can be expectedto be high.

In additionto having fewer children,more-educated individuals are found to live healthierand longerlives. One recentmulti-nation study demonstratedthat a one-percentdifference in the nationalliteracy rate is associatedwith a two-yeargain in life expectancy,controlling for per capita income and food energyconsumption.

idrents'education, particularly mothers', has been shown to affect the physicaland cognitivedevelopment of children. This occurs, in part, through education'seffect on familyincome and, in part, throughits effects on parents'knowledge of and use of good health and nutritionalpractices. Also, intellectualskills acquiredthrough one's own educationtend to "rub off' on one's children.Studies have shown that childrenof more educated parents are more likelyto be enrolledin school,and once enrolled,are more successfulin school and continuehigher up the educationalladder.

There is, followingfrom education'seffects on nutrition and health, a consistentand stronglypositive relationship between parental education and child survival. A reviewof researchin this area concludesthat the addition - 26 - of one year of mother'sschooling reduces child mortalityby 9 deaths per 1,000 live births. Box 2.2 summarizesinteresting new researchfindings from Africa on the impactof maternaleducation on child survival.

[Box 2.2]

(c) Educationand economicgrowth: macroeconomic evidence

Developmenteconomists concerned with the role of human resource investmentsin determiningeconomic growth have foundmicroeconomic evidence of the sort just discussedto be stronglysuggestive of education'simportance. Yet there remain importantconcerns. Perhaps education's high ratesof return are accountedfor, indirectly,by reductionsin the earningsof the less well educated. Or, conversely,to the extent that the better-educatedenhance the productivityof those aroundthem (throughtheir entrepreneurialactivities or technicalcontributions, for example),measurement of income differentialsby educationlevel may understateeducation's impact. For these reasons, macroeconomicevidence on educationand economicgrowth becomes an important complementto the extensivemicroeconomic literature.

Macroeconomicanalyses attempt to explaindifferences in growth rates (and other developmentoutcomes) across countries in terms of differences, among other things,in theirpatterns of investmentin education. The World Bank's 1980 World DevelopmentReport undertooka major analysisof this sort and concludedthe following: (i) increasesin literacycontribute both to increasedinvestment and (giventhe level of investment)to increasesin output per worker; (ii) literacy,as well as nutritionand income,affects life expectancy;and (iii)variations in life expectancy,literacy, income and the strengthof familyplanning programs explain between them most of the variation in fertilityrates acrosscountries.

While the World DevelopmentReport's findings on the development impactof educationalinvestments are quite clear,an importantcaveat, for the purposesof this paper, is that the analysiswas undertakenon a sampleof all developingcountries, and somewhatdifferent results might pertainto Africa. More recentanalyses suggest, however, that levelsof investmentin education are, if anything,more importantin explaininggrowth rate differencesamong Africancountries. Box 2.3 summarizesthe resultsof an analysis,undertaken in conjunctionwith the preparationof this paper, of the role of educationin contributingto the growth of 31 African countriesover severaldecades. The analysisconcludes, in short,that investmentsmade in educationhave contributedsignificantly to growth in GDP; indeed,perhaps 30% of GDP growth has resultedfrom educationalinvestments. The aggregateevidence thus corroboratesthe microeconomicfindings of education'shigh returns.

[Box 2.3]

Economicand demographicpressures will, clearly,constrain the capacityof societies-- and, even more, of governments-- to invest as fully as would be desirablein the educationlevels of their futurepopulationa. Over a period of time, however,governmental policy reformand individual responsesto reform,combined with developmentassistance from abroad,will create an environmentin which sustainedgrowth can resume. Indeed,economic - 27 -

Box 2.2 EDUCATIONAND MORTALITY DECLINEs RESEARCH FROM NIGERIA, GHANA AND SUDAN

A number of survey and census analyseshave reflectedan inverserelationship between a child'schances of survival and the mother'slevel of education. Data from the 1960 Census of Ghana, for example,reveal that the rate of child mortalityis almost twice as high for mothers with no educationas for motherswith elementary education,and nearly four times higher for motherswith no education than for those with secondaryschooling. The patternsare much the same for childrenin urban and rural areas.

A more comprehensivestudy was conductedin Nigeriaas part of the 1973 ChangingAfrican Family Project Survey. One componentof this study compriseda probability sample of 6,606 Yi.rubawomen, between the ages of 15 and 59, in the city of Ibadan. The second componentconsisted of a probabilitysample of 1,499 Yoruba women, over age 17, living in southwesternNigeria. Analysisof these data allowedfor an examination of rural-urban differences,which serve as a reasonable proxy for differencesin accessto modern healthservices. The study consideredchild survival in relation to medical servicesat childbirth,their practiceof birth control,and familyincome as measured by father's occupation. The analysis concluded that the singlemost importantinfluence on child survival is the level of mother'seducation. In Ibadan,the child mortality index for women with some primaryschooling was 68% of that recorded for women with no schooling,and the index for women with more than primary schooling was 39% of that for women with no schooling. In southwesternNigeria, the figureswere almostthe same, 68% and 41%, respectively.

While father's education was also found to be significant, it was less important in explaining differential rates of child mortality than mother's education. Controlling for education, family income was of little importance. Although child mortality was higher in polygamous than in monogamous homes, the effectof a mother's educationto the secondarylevel was at least a 50% reduction in mortality in both polygamous and monogamous homes. Once other factorswere controlledfor, child survivalwas found to be higher among parents who practicedbirth control,which might be explained by the greatercare accordedchildren in smallerfamilies.

Results in the Sudan from the sawm Changing African Family Project Survey confirmedthe findingson the importanceof mother's education to child survival. - 28 -

Box 2.3 EDUCATION AND ECONOMIC GROWTH IN SUB-SAHARAN AFRICA

Between 1985 and the early 19809, Africa Invested heavily In educatlon. Growth of primary school enrollments,which rose from 12 million In 1960 to over 50 million In 1983, Is one measure of this Investment. The average cost of Investment In education In sub-Saharan Africa In recent years has been between 4% and 5% of national Income, plus direct parental contributions and the potentlal earnings foregone by students while they attend school.

In order to guide future allocatlons to the educatlon sector, It Is Important to evaluate the extent to which African governments' past Investments In education have contributed to recent economic growth. Such an analysis was undertaken, using data appearing In the Statistical Annex to th!s paper and other World Bank sources. The analysis utilized a 'production functiona that relates the GDP of each country In each time period (5 time Intervals between 1965 and 1983 were used) to a number of factors potentlally Important In determining GDP. These Included: size of the working age population; land area under cultivation; available physical capital stock; and available human capital stock, measured by the number of years of secondary education attended by members of the working age population. A measure of the number of years of primary education attendance was also available, but since the two education variables are highly collnear, one was dropped from the analysis.

These data permitted use of standard econometric methods to estimate the parameters of the production function that relates GDP to the avallabilityof the determining variables. Specifically, a Cobb-Douglas production function was estimated through examination of how changn In Input availability resulted In changes In GDP. With an estimated productlon function, and knowledge about growth In the avallability of the Input factors, It Is possible to account for the growth In overall GDP by decomposing It Into components that reflect the rate of growth of each Input's availability and the relative Importance of that Input for determining output, as assessed from the production function.

The table below reports the results of this analysis. The first column of the table lists and defines the variables used In the analysis; the second Indicates their growth rate over the period 19e5-1983. For example, for the 31 countries included In the analysis, GDP grew at an average rate of 4.3% per year over this period and the size of the working age populatlon grow at a rate of 2.5% per year, so that the growth of GDP per member of the working age population was, on average, about 1.8% per year. The column showing per worker growth rates Indicates that the amount of land per worker grew more slowly than GDP per worker, but that physical capital and, particularly, human capital stocks grew much more rapidly.

The final two columns show how these Input growth rates accounted for the 1.8% growth rate of GDP per worker. Increases In the stock of human capital accounted for 0.55 of the 1.8 percentage points, or 31% of total GDP growth. This Is an Impressive fraction, Indeed, and clearly demonstrates the - 29 - lox 2.3 (entinued)

Importanceof InvestmentIn educationIn explalning GDP growth In Africa. The most nearly analogous figure from EdwardDenIson's work on eco'.OmIc growth In the United States between 1929 and 1976 is 26%. Growth In avallabilityof physical capitalaccounted for 71% of the total growth. A aresidualpof unaccounted-forgrowth was actuallynegative In this analysis, resulting In a decreaseof 0.1 percentagepoints per annum In the levelof growth that would have been expectedon the basis of growth In the measured Inputs. This residual (often Interpretedas technicaland institutional Improvement)Is positiveIn studlesundertaken wlth data from the United States and other Industrlalizedcountries and often accounts for a substantialfractlon of total growth. That It has been negative over this period In Africa may be consistentwith Impressionisticaccounts of recent sharp technologicaland Institutionaldeterloration, and Is disturbing.

Contributorsto GDP growth In 31 countriesof sub-SaharanAfrica, 1965-83

Contributlonto 1.8% growthrate of GDP per memberof working Growth rate (X per year) age population

Per member of workingage Percentage Percent Varlable Absolute populatlon points of total

GDP 4.3 1.8

Labor (size of working 2.5 - - - age populatlon)

Physicalcapital stock 7.8 5.3 1.25 69%

Land (area under 3.6 1.1 0.09 5% cultivation),

Human capital(years 18.3 15.8 0.55 31% of secondaryeduca- tion In working age population)

Residual - - -0.09 -5% - 30 -

statisticsemerging in the past 18 months have reflecteda returnto growth.To take full advantageof an improvingeconomic policy environment,sustained infrastructuralinvestments must be made, both public and private. The evidenceis powerfulthat educationis a key area for such investment. African governmentsand peopleshave successfullyimplemented massive educational investments,and evaluationof the consequencesshows the returnsto have been high. We come, then, to the first centralrecommendation of this report:

Recommendation1. On groundsof unusuallyfavorable economic returns relativeto costs, as well as implementationfeasibility, education is prominentamong the sectorsin Africa where greatereffort is essentialfor engenderinglong-term development. Most Africancountries should aim to increasetotal (publicplus private)education expenditures from their recent levels. Government expenditure, currently averaging between 4% and 5% of nationalincome in the region,should in most countriesbe graduallyincreased, particularlyin those countriesnow dedicatingsignificantly less than this proportionof public resourcesto the sector or in which attainmenton key indicatorsof educationaldevelopment is laggingfar behind regionalaverages. Such.increases in public spendingare warranted,however, only in conjunction with effortsat all levelsto enhanceeducation's internal efficiency and financialviability. The broad array of measuresappropriate for these objectiveswill typicallyinclude increased non-government contribution to, and provisionof, educationalservices, thereby raising private resourcesdedicated to the sector at least in proportionto the increasein public resources. - 31 -

Chapter3. MAJOR PROBLEMS: ENROLLMENTSTAGNATION AND QUALITYDECLINE

AlthoughAfrican nations have made enormousprogress in educationas describedin Chapter 1, much remainsto be done. Africa still lags behind other developingregions on most indicatorsof educationaldevelopment. Moreover,the externalfactors described in Chapter 2 providean inhospitable environmentfor eliminatingthe educationand traininggap.

Indeed,because of the ..nvidiouscombination of rapid population growth and economicstagnation, the gap between sub-SaharanAfrica and the rest of the world appearsto be wideningat the presenttime; certainlyit is no longerclosing as in recentdecades. Unless steps are taken to addressthe seriousproblems in education,what was a gap will in time become a gulf. Given the vital links discussedat the end of Chapter2 betweeneducation and a people'sprospects for economicgrowth and developmentin the future,this cannot be allowedto happen.

This chapterdescribes the alarmingstate of educationin Africa along the two principaldimensions of educationaloutput, quantityand quality.On the quantitativeside, school enrollmentshave fallenas a proportionof school-agepopulations in many Af-ican countries;absolute declines actually have occurredin a few. Althou adult educationfor literacy(to the extent that it providesa secondchance for those who miss the opportunityto enroll in the formalschool system)can often be justifiedon equitygrounds, it shouldbe seen primarilyas a stopgapsolution. For the longer run, the educationalgoal should remainuniversal primary enrollment,which ultimately obviatesthe need for adult educationof this kind.

With respectto the qualitativedimension of educationaloutput, althoughthere is less "hard"evidence that the performancein Africa is unsatisfactoryand that it has declinedover the recentperiod, the limited evidencepoints convincingly,nonetheless, to these conclusions.

3.1 Enrollmentstagnation

Between 1960 and 1970, total enrollmentsin sub-SaharanAfrica grew at an average annual rate of 6.5%. Between 1970 and 1980, this rate rose to 8.9%. During the first th ee years of the 1980s,however, the rate of increase plummetedto 4.2X.

Although enrollmentincreases declined at all levelsof education,the drop was most pronouncedat the first level,where the rate fell from 8.4% (approximately2.9 million additionalpupils each year) between 1970 and 1980, to 2.9% (approximately1.4 million additionalpupils each year) between 1980 and 1983. Given that the populationof primaryschool-age children is now increasingat an averageannual rate of 3.3%, a 2.9% increasein enrollments will not even keep pace. This sectionbegins with an analysisof the problem of stagnatingschool enrollments, then turns to an accountof the role of nonformalliteracy and other adult educationprograms as a partial (at best) substitutefor enrollmentexpansion, and ends with a discussionof the equity implicationsof the currentenrollment trends. - 32 -

(a) Schoolenrollments

In half (19) of the 39 countries,the presentgrowth rate of the school-agepopulation exceeds that of primaryschool enrollments. In these countries,unless efficiencyis much enhanced,gross enrollmentratios must inevitablyfall. For the region as a whole, if the growth in primary enrollmentscontinues into the future at a rate below 3%, by the year 2000 the gross enrollmentratio will have fallenback below 70%, the level of the late 1970s.

While absoluteenrollments continue to rise in most countries,this is no longer the case in all countries. Between1980 and 1983, first-level enrollmentsactually fell in four countries-- Angola,Mozambique, Somalia, and Togo. Less completeand reliabledata for the more recentperiod suggestthat primary enrollmentshave fallenin Benin, in some statesof Nigeria,and in severalof the least advantagedparts of the Saheliancountries. Post-primary enrollmentsalso declinedbetween 1980 and 1983 in severalcountries, including four (Cote d'Ivoire,Ghana, Senegal,and Swaziland)where secondaryor tertiary enrollmentsdeclined even though primaryenrollments continued to increaseat close to or better than pre-1980rates.

On the demand side, enrollmentdeclines reflect the currenteconomic situationin Africa. Childrenwho might have attendedschool in bettertimes are kept out or pulledout becausethey are needed to work at home or, in very dire situationssuch as those that accompaniedthe recentdrought, because their familieshave been forcedto migrate. Also, family incomeshave fallen at the very time that many countrieshave introducedor raisedschool-related fees. Finally,the privatebenefits of education,especially inferior quality education,may have fallenduring the recentperiod of economicstagnation; educationalqualifications for many jobs have risen as a functionof the rapid expansionof graduates. Until issuesof schoolquality are addressedand economicrecovery is well under way, the demandfor schoolingin Africawill remainweaker than in the past, when it was extraordinarilystrong.

To some extent,too, the decelerationin the expansionof educationin much of Africa is a naturaloutgrowth of past quantitativeachievements in the sector. As describedin Chapter 1, there was tremendousprogress in the early post-independenceyears, and proportionateenrollments in some African countriesare alreadyclose to world standards,especially at the primary level. The gross primaryenrollment ratio for the 39 sub-Saharancountries taken as a whole was 75X in 1983, and in 16 of the 39 countriesthe enrollment ratio exceeded85%. This level of participationwas reachedin Latin America and Asia only quite recently(in the early 1960s).

In general,given a country'spresent enrollment level, the rate of enrollmentgrowth is closelyrelated to that country'sincome per capita. Althoughthis relationshipholds in generalterms, the countrywith the fastest growingprimary educationsector in recentyears is Burundi,defined as a "low income"country under the World Bank's systemof classifyingcountry economies; Burundi'sGNP per capita in 1984 was only $220. The countrywith the most impressiverecent gains across all three educationallevels is Zimbabwe. First-levelenrollments in this countrygrew at nearly 15% annuallybetween 1980 and 1983, second-levelenrollments at 54X, and third-levelenrollments at 23%. Zimbabwe'sefforts stand out as a notablecounter-example in contrast with the generalregional trend toward enrollmentstagnation (see Box 3.1). It - 33 - would appear that, over and above the abilityto pay, commitmentto the sector is an importantfactor in determininga country'srate of educational expansion.

[Box 3.1]

(b) Adult literacyand training

Fallingenrollment ratios must resulteventually in adult populations less "schooled"and, as a consequence,even less productivethan those living and working in Africa today. Even prior to the currentfiscal problems, at the beginningof the 1970s,widespread concerns about educationalbudgets and about patternsof inequalityin the formaleducation system led to an appeal for "nonformal"education for adult literacyand other purposes. Althoughdata are not availableconcerning the numbers involvedin these programsand the extent to which they have grown or contracted,adult programsare importantin a number of countries.

By the late 19709,adult educationand trainingactivities included, in most Africancountries, not only literacyteaching but also some or all of the following: out-of-schooltra.ning for the informalsector; skill acquisitionand incomegeneration for women; the trainingof trainersfor health,nutrition, and agriculture;and a varietyof activitiesdesigned to provideequivalency to the severallevels of the formaleducation system. As nonformaleducation has developedin Africa, it is characterizedby the followingmain features: It serves predominantlyyoung adults,many of whom are of the same age as their counterpartsin the formalschools; much of the activityis organizedlocally and takes place with littleor no direct subventionand controlfrom the state beyond some minimal registrationand supervision;the young recipients(or their families)often pay for courses that are, in some sense,equivalent to those offered in the schoolsor in formal industrialtraining; even those who pay for instructionare typically drawn from the poorerelements of rural and urban society;they are often obligedto combinetheir educationor trainingwith work; many of the courses taught,especially those organizedthrough non-governmental organizations, have been negotiatedwith the participantsor their representativesand, to this extent, reflectthe priorityneeds of the community.

Many of these characteristicsof nonformaleducation programs are innovativeand likelyefficiency inducers. Some are typicallyabsent from and could,with appropriatemodification, be emulatedin the formaleducation system. The possibilitiesfor changewith respectto all elementsof the educationsystem will be consideredin the discussionof policy optionsin Part Two of the paper.

The presencear.d recent growth of nonformaleducation programs raise three major questionsfor African governments.First, do the present equivalencyprograms that offer educationand skillsoutside the formal sector point to ways that would dramaticallyopen up, to a wider public,many of the currentlyunder-utilized institutions of the ministriesof educationand labor? Second,are there ways in which the concernsof Africangovernments about the vocationalpreparation of young people for informalsector employmentcan be met throughexisting low-cost, out-of-school programs rather than by way of newly creatednational schemes for vocationalstudy within schools? And third, to what extent,if any, can the existingnonformal education structures - 34 -

Box 3.1 THE STORY OF ZIhSABWE

While primary and seeondary enrollmentrates have fallen In many sub-SaharanAfrican countries since 1980, enrollment rates have Increased dramaticallyIn Zlmbabwe. That year marked the beginningof majorityrule In Zimbabwe,15 years after a white minorityregime had unilaterallydeclared Independence(UDI) In what was then calledRhodesia.

Rhodeslan society -- IncludingIts educationalsystem -- had been structuredbasically along linesof raclal Inequality.Separate *European' and 'African3 educatlon departments led to restrictedprovision and lower qualityof education for blacks relative to whites. Public per pupil expenditures were 12 times greater In the 'Europeangsystem than In the 'Africangsystem at the prlmary level,and nearly three times greaterat the secondary school level. In 1979, however,In responseto the intensifying war of liberation,the coalition government merged the two education departments and began the processof raclal Integrationof schools.Although these reforms Increasedblack accessto education,enrollments continued to be restricted,especially at the secondaryand highereducatlon levels.

After IndependenceIn 1980, the now governmentbegan to redressthe InequalityInherIted from the colonialand UDI periods.Education reform was based on the premise that education, In additlon to contrlbutingto a nation'seconomic growth and development,was also a basic human right. Educatlonwas viewedas an effectivevehicle for promotingmany of Zlmbab'we's goals, Includingthe developmentof nonracist attitudes, a new national Identity, and loyalty to the state. The governmentgave top priorityto reopeningand reconstructingschools that had been closed during the war (nearlyone-third of all primaryand secondaryschools In rural areas)and to the expansionof educationat all levels,with particularemphasis given to secondaryeducation.

Parents and local communitieslent their supportto the expansion effort by contributing fees and their labor for the constructionand rehabilitationof schools.This privatesupport stemmed from the belief that educationIs the key to an Individual'sobtaining a Job In the modern sector of the economyand Improvingone's standardof llvIng.

Between 1979 and 1985, primaryenrollments In ZimbabweIncreased about 160%, and secondary enrollments about 650%. The gross primary enrollment ratio now exceeds100%. and given the low reportedrate of pupil repetition(1%, see Annex Table A.12), It seems safe to conclude that Zlmbabwe has achievedthe elusivegoal of tniversalprimary educaticn. The secondary enrollment ratio has shot up also, from below 10% before Independence,to above 40% today. While enrollmentIncreases have been achlevedat some cost -- In terms of Increased expendituresand consequent constralnt In macroeconomic flexlbillty-- they, nonetheless,clearly representremarkable achievements. - 35 - compensatefor slowdownsin expansionof schoolenrollments? This paper returnsto these questionsat appropriatepoints in the chaptersthat follow. The last question,however, having to do with complementaritiesbetween the formaland nonformaleducation systems deserves some brief commentat this point.

Unlike primary,general secondary, vocational, and university education,adult literacyeducation is not a regularpreoccupation of the ministryof educationin most Africancountries. Usually,it commands attentionon an ad hoc basis for the purposesof a particularcampaign or seriesof campaigns. When this happens,the tendencyhas been for the national literacycoordinating committee to make use of many of the mechanismsthat have routinelybeen associatedwith other forms of adult and nonformaleducation -- teachersreceiving token honoraria,students working duringvacations, school buildingscommandeered for use in the evenings. Part and parcel of most successfuladult literacycampaigns has been a powerfulsense of the commitment of unpaid or modestlypaid volunteers. While such relativelylow-cost modalitiesare much in evidencein the first months or years of the campaign, the process of institutionalizingthese effortscontinues to be a challenge;in the medium to long term, relianceon volunteersand a campaignspirit probably cannotbe sustained.

Moreover,the accomplishmentsof the successfulcampaigns in, say, Ethiopiaand Tanzania,could rapidlybecome dilutedif these countries'primary school systemscannot attract and serve the ever-largercohorts of children during the 19809 and 1990s. In countrieswhere the primary system fails to meet this challenge,the call for universalliteracy, if issued at all, can never be met in a single campaign. In this environment,the campaignwould have to be a perennialevent, continuedwith additionalexpense and efforttill such time as the birth rate stabilizes. Unless the problem of low school enrollmentratios in the regionis vigorouslyattacked, the pool of young adult illiterateswill be refilledas rapidlyas illiteracyeradication campaigns can deplete it.

In conclusion,adult and child literacy(and illiteracy)are thus intimatelyconnected. Adult literacycampaigns have played an important mobilizingrole for universaleducation in severalcountries, and more general programsof adult or communityeducation are vital to the maintenanceof new schoolingtraditions in parts of Africa. Singlecampaigns of adult literacyor UPE will both prove to be poor investmentsunless they are consolidttedby productiveopportunities to put new skills to use. Lifelongeducation opportunitiesare essentialfor providingindividuals with the informationand skills needed to enablethem to adapt to changingeconomic circumstances. The issuesrelating to vocationaleducation, training, and the learningof occpation-specificskills will be discussedas part of Chapter5.

(c) Expansionand equity

A criticalissue relatedto the currentenrollment stagnation in sub-SaharanAfrica is what this impliesfor equity. A great virtue of educationalinvestments, properly designed, is their potentialfor reducing inequitiesin enrollments.Conversely, the recentstagnation of enrollments seriouslyundermines the prospectsfor eliminatinginequities that now exist. In practice it is difficultto separatethe fact of unequalparticipation in educationfrom that of low overallparticipation. Policies that increase - 36 - overallparticipation will necessarilybenefit disadvantaged groups, if only in the long run; improvementsin the statusof disadvantagedgroups are unlikely to occur when educationconditions in generalare stagnant or deteriorating.

Ethnic and rural-urbandifferences in schoolparticipation, though considerablyattenuated since pre-independencedays, remain an issue in most of Africa today. Male-femaledifferences, though close to being eliminatedat the primary level, at least in overallregional terms (in 1983 girls comprised44% of Africanprimary schoolenrollments), are still much in evidenceat the secondarylevel (34% of enrollments)and especiallyat the tertiarylevel (21%). Illiteracyis still much higher among adult females (73% in 1980) than among adult males (48%).

3.2 Decliningeducation quality

The impactof rapid populationgrowth and economicstagnation on the quantityof educationalservices in sub-SaharanAfrica is relativelyeasy to document. Anotherprobable, though more difficult-to-measureeffect of the recenteconomic and demographictrends in the region is the erosionof quality in education.

(a) The nature and importance of quality

The quality of a school or education system is properly defined in terms of the performanceof its studentsand graduates. In practice,however, becauseinputs into the teachingprocess are generallyeasier to measure than education'soutputs, quality is often gauged in terms of the formerrather than the latter. Althoughinformation on input availabilityis important,caution is reqvtiredin drawingconclusions about qualityfrom informationon inputs alone. The discussionthat followsuses both directmeasures of school outcomesto provide informationon qualityas well as measuresof input availability.

When an attempt is made to measureoutputs as a direct indicatorof quality,the most commonapproach is to concentrateon cognitiveachievement test scores. This makes sense to the extent that enhancingcognitive achievementis prominentamong education'sgoals and contributescentrally to a student'sultimate productivity. This paper will follow the conventional practiceand focusprincipally on cognitiveoutcomes, because researchsuggests that these outcomesare highly important(see Box 3.2), and becausemost of the same factorsthat foster qualityin learningappear likelyto strengthenthe school'simpact in other domains.

(Box 3.2]

The goals of schoolingencompass, however, more than just academic achievement. Qualitypertains also to how well the schoolor schoolsystem preparesstudents to become responsiblecitizens of a nation and to have a set of attitudesand values relevantto modern society. Schoolsdo achievethese goals, and they are importantones.

Frequentlywhen an Africanchild entersschool, he or she is faced with a situationquite foreignto his or her life at home in a rural villageor urban slum. The languageused in the schoolis often different,the method of communicationis written ratherthan verbal,and existenceis categorizedinto - 37 -

Box 3.2 ARE COGNITIVEOUTCOMES REALLY IMPORTANT? THE INTERPLAYOF ABILITY,SKILLS, AND SCHOOLING

Earningsgo up with education,it is generallybelieved, because educationimparts cognitive skills to workers,and employerspay more for these skills. Criticsbelieve other factorsare at work.

If the criticsare right,the value of educationmay well have been exaggerated. Researcherstested the argumentsof the critics,in a study conductedin Kenya and Tanzania,by assessingthe impactof three factorson earnings: innateability, years of schooling,and acquired cognitiveskills (likenumeracy and literacy). Theirwork provides answersto the followingquestions.

Do acquiredcognitive skills and innateability matter? Workers with high cognitiveskills do earn much more than others. But the direct effectsof innateability improves earnings only slightly(although innateabilitity does have a large indirecteffect on earningsby enhancingthe acquisitionof cognitiveskills).

Do cognitiveskills affect earninzs in each educational category?There are great variationsin skill levelsamong secondary graduatesas well as primaryones. In skill tests, the top third of workers scored50% to 100% highermarks than the bottomthird in each category. This top third can expect to earn 50% more than those in the bottomthird in Kenya and 35% more than those at the bottom in Tanzania. Moreover,the top third of primary-schoolleavers will earn almost as much as the bottom third of secondarygraduates.

How large are the indirecteffects of schoolingand innate ability?In Kenya, secondaryschooling raises earningsdirectly by 21%, but raisesthem indirectlyby 252 by improvingcognitive skills. The pattern is similarin Tanzania.

The study stronglysupports the view that educationendows workerswith cognitiveskills that draw a premium in the labormarket. It shows that: (1) innatetalent is not sufficient-- it must be convertedto cognitiveskills through education; and (2) long years spent at school are not sufficient-- they must be used productivelyto acquire skills. - 38 -

subjectsthat are not referredto at home. A child'smajor challengeat this stage may not be that of learningnew skills,but simply one of adjustingto an altogetherdifferent environment. In additionto basic cognitiveskills such as literacyand numeracy,therefore, the qualityof educationhas also to do with the successfulintegration of what childrenalready know when they enter school,with what they will need to learn as they proceedthrough the education system;quality entails the accommodationof modern,market-oriented skills to traditional,home-based values and needs. Policiesfor the achievementof these qualitativeobjectives include adjusting the schoolcalendar to take accountof the child'seconomic functions at home; giving first instructionin the child'smother tongue;integrating subjects around the life of the child and his community;and involvingstudents in the applicationof theory so that learninghas utilitybeyond that of qualifyingindividuals for the next level of education. These changesencourage children to look at and react to their environmentin new and more productiveways.

When either academicor post-schoolperformance is used to measure school quality,it is necessaryto control for the effect of nonschoolfactors (e.g.,innate ability, family background, and eerly childhoodeducation). The impactof school inputson performanceis the performancegain attributableto these inputs,after controllingfor the effectsof nonschoolfactors. Although this paper focuseson school factors,the importanceof some nonschoolfactors affectingstudent performance are consideredalso, especiallythose dealing with the child'shealth and nutritionalstatus.

Finally,it shouldbe rememberedthat what constitutesan acceptable standardof school qualityis a relativematter. As with other servicessuch as housing,nutrition, and health,the appropriatelevel of qualitydiffers between differentcountries at differentlevels of developmentand for any particularcountry over time. For example,as African countriesmore than doubledtheir enrollmentratios over a comparativelybrief period, it was to be expectedthat the abilityof studentsto perform at a superiorlevel would declineon average, simply as a functionof the region'shaving moved from a small elitistsystem to a systemserving many children,including those who are disadvantagedwith respectto the out-of-schoolfactors that affect learning outcomes.

It is sometimesargued that this declinein averagequality is acceptableso long as the performanceof those individualsat the top of the achievementdistribution does not suffer. Given, however,the seriousrecent deteriorationin learningconditions in much of the region (see next section), there is reason to believethat the performanceof the highestachieving studentshas also declined. Certainly,in order to maintain a constantaverage qualityof educationin a rapidlyexpanding system, there is a need to improve the overalleffectiveness of the system. This need remainsto be satisfiedin most of sub-SaharanAfrica.

(b) Evidenceof decliningquality

Levels of cognitiveachievement on the part of African studentsare low by world standards,and the evidencepoints to a declinein recentyears. Much of this evidenceis indirect,however, and focuseson quantitiesof particularinputs (especiallybooks and other learningmateriais, management, and maintenanceof capitalassets) and on the recentdecline of these relative to other inputs (especiallyteachers). - 39 -

On the outputside, informationis limited;however, the information that is availableis compelling,and it is disturbing. The International Association for the Evaluation of Educational Achievement (IEA) based in Stockholm,Sweden, has served as the coordinatingagency for a number of large cross-nationalstudies of studentachievement since the mid-1960s. Although few developing countries participated in the earlierstudies, the latestrounds of IEA surveys do include more LDCs, including some African countries. The second IEA study of achievement in mathematics,for example,which was conducted in 1981-82, includes (in addition to 14 industrial market economies) information on three upper middle-incomecountries and three lower middle-incomecountries. Included in the last group were two African countries, Nigeria and Swaziland. The third country in this income group was Thailand.

For the mathematicsstudy, tests in five sub-fieldsof mathematics were administeredat the end of the 1981 schoolyear to nationalsamples, rangingin size from 800 to 8,000 students,all of whom had reachedthe age of 13 years by the middle of the schoolyear. Althoughthere were differences across countries,the upper middle-incomecountries as a group performedon a par with the industrializedones (for detailssee Annex Table C.4, column a). The lower middle-incomecountries, however, and particularlythe two African countries,performed much worse. Studentsin Nigeria and Swazilandanswered just over half as many items correctlyas studentsin Japan,the highest scoringcountry, and about 65% as many items correctlyas studentsin the 17 better-offcountries. These differencesare highly significant.

Earlier IEA studiesof achievementin readingcomprehension and general scienceincluded no Africancountries, but the same tests used in these studieswere subsequentlyadministered to a small sample of studentsin Malawi (AnnexTable C.4, columnsb and c). Althoughthe Malawianstudents were, on average,six years older than the IRA-surveyedstudents, their performancewas less satisfactorythan that of any of the other 16 groups in readingand lower than all but three in science. The number of items answeredcorrectly by the Malawian students was just over half the average number in the ISA-surveyed countrieson the readingtest and about 84X on the science test.

Examples of specific test items illustrate more concretelythe magnitude of the quality problem. In November 1986, an educational research institute in a FrancophoneAfrican countryadministered, to fifthgrade students, the same mathematics test that had been administered to all incoming fifth grade students in France a few months before. Three of the questions on the test are translated and substantially reproducedbelows - 40 -

(1) 1322 X 0 - ?

(a) 0 (b) 1322 (c) 13220

(2) A roast weighs 2kg, 50g, or:

(a) 2050g (b) 2500g (c) 250g

(3) A purchaseof 4 cartonscosts 3.80 Francs.

A purchaseof 8 cartonswould cost:

(a) 30.40 Francs (b) 7.60 Francs (c) 6.60 Francs

Since, for each question,students were given a choiceof only three responses,guessing would give a 33% chanceof choosingthe correctresponse. In fact, only 33% of the studentswere correcton the first item; 30% were correcton the second;and 26% were correcton the third. Each of the items tested reflectedcontent that the students'teachers felt was covered in the curriculum. On the test as a whole (40 items),the studentsperformed no better than chancewould p,edict,even though the tests had been administered in reasonablygood schoolsin the capitalcity. The conclusionis inescapable: studentsin primaryeducation in this countryare learningvirtually no mathematicst

The generalconclusion to be drawn from these studies is that the qualityof educationin sub-SaharanAfrica is well below world standards. One explanationfor this low qualityis that expenditureper student,a highly aggregatedproxy for educationalinputs, is very low by world standards. This is especiallytrue at the primarylevel.21 Another explanationfor the low qualitymay be that what little is spent on each student is poorly utilized, i.e., that there is internalinefficiency in the educationsystem.

Public recurrentexpenditure per studentat the primary level in the industrialmarket economiesexceeded expenditure in the countriesof Africa by a ratio of 30-to-Iin 1980. The industrialmarket economieswere spendingover $2,200per pupil in current 1980 dollars (medianexpenditure in 16 countries with 1980 data available),and Africa (23 countrieswith data available)only about $65. East Asia (5 countries)was spendingabout $190 per pupil on the averagein the same year, and Latin America (20 countries)about $155. Only in five countriesof South Asia was spendingper primaryschool pupil lower yet than the Africanaverage.

Per studentexpenditure in Africaneducation is not just low, it is declining(see Table 3.1). The combinationof essentiallyconstant budgets (since1980) and rapidlyexpanding enrollments has made the financingof education'srecurrent costs ever more difficultover time. Between 1970 and 1983, recurrentexpenditure per studentat the primarylevel expressedin constant1983 dollarsfell from $67 to $48, and at the secondarylevel from - 41 -

$362 to $223 in the median Africancountry (AnnexTable C.5). Only in higher education,which involvesrelatively few studentsand where unit costs have been very high traditionallyby comparisonwith lower levelsof educationin Africa, was there no obvious downward trend in expenditure per student.-1

[Table3.1)

Low expenditureper studenthas certainlyconstrained the level of learningachievement in sub-SaharanAfrica. But low qualityresults also from a misallocationof expenditure.At least until recently,there was a tendency, often encouragedby the behaviorof donor agencies,to emphasizedevelopment expenditurein education,especially expenditure involving the constructionof new facilities,and to ignore the recurrentcost implicationsof new projects as well as the recurrentinputs requiredto assurethe productivityof existing establishments.For example,the ratio of expenditureon books and suppliesto expenditureon teachers'salaries is far lower than most knowledgeable educatorswould consideroptimal. Moreover,because teachers' unions are politicallypotent and most salariesare protectedby contractualobligations, the salary-nonsalaryratio in African publiceducation has tended to increase in recentyears as budgetshave been cut.

Data for the most recentyear availableshow that teachers'salaries and benefitsin sub-SaharanAfrica accountfor about 91% of recurrenteducation expenditureat the primary level, 70% at the secondary,and 68% at the tertiary (medians-- see Annex TablesA.20-A.22). Even if all of the rest were availablefor the all-importantteaching materials, this does not leave much to spend on them, nor on the maintenanceof schoolbuildings and equipment. At the secondaryand higher levels,where transfersto studentsfor welfare costs (food and lodging)claim significantsums, the crunchon nonsalaryinputs is even tighterthan the figuressuggest.

For example,educational materials account for just 1.1% of the recurrentprimary education budget in the medianAfrican country (Annex Table A.20). This allocationamounts to less than $0.60 per pupil per year, which buys very little in the way of books, slates,wall charts,and writing implements. Even if one concedesthat some educationalmaterials are purchased out of developmentbudgets (becauseof the unwillingnessof many donor agencies to finance'recurrent" costs), and notwithstandingthe fact that some (perhaps significantamounts, in some places,of) materialsare purchasedprivately by studentsfor their own use, the pictureremains stark by comparisonwith other places. The developedcountries spend a largerpercentage of a much larger budget on insttuctionalmaterials, close to 4% at the primarylevel, which amountsto about $100 per pupil per year. With only $0.60 spent per pupil per year on educationalmaterials in Africa, even the most ingeniousteacher would find it hard to teach childrenvery much. Only by reallocatinga substantial multipleof this amount from other expenditurecategories (or by mobilizing this multipliedamount in the form or additionalresources) can teachersin African schoolsagain become pedagogicallyproductive. - 42 -

Table 301 PublicRecurrent Zzpenditura per Studentin Sub-SaharanAfrica, 1970-1983

1970 1975 1980 1983

Expenditureper primarystudent $67 $61 $51 $48 As % of GNP per capita 161 191 16X 15%

Expenditureper secondary student $362 $308 $195 $223 As Z of GNP per capita 111X 93Z 62X 62%

Expenditureper tertiarystudent $2,462 $3,090 $2,798 $2,710 As multipleof GNP per capita 11 12 7 8

Notes Medianvalues. Per studentexpenditures in constant1983 dollars. From Annex TablesA. 17-A. 19. - 43 -

PART TWO: POLICYOPTIONS FOR AFRICANGOVERNMENTS

Three of the four chaptersin Part Two will deal with the education system by level -- primary,secondary and higher. Becausethe problems encounteredat these three levelsdiffer, the substanceand structureof the chapterswill differaccordingly. Yet a number of common issuesrecur in each of the chapters,and these will be discussedin generalterms here.

Chapter 3 raised concernsabout recenttrends in both the quantitative and the qualitativedimensions of Africaneducation systems' output. Quantitativeand qualitativeoutputs of the systemcan be at undesirablylow levels for either of two reasons. One is inefficiency,and the other is a lack of resourcesor inadequacyof finance. The paper argues that both these factorsare operative,but placesparticular emphasis on the importanceof strengtheningmanagement to reduceinefficiency. Because of the key role of managementimprovements, a separatechapter is devotedto the subject.

The first set of issues,then, concernsefficiency in education. Efficiencyhas to do with how well inputsinto the educationalprocess are used in producinga schoolor a school system'sdesired educational outcomes. Efficiencyexists whenever the output obtainedfrom a given budget is at a maximum,or when the budgetused to achievea given output is at a minimum. Educationaloutput, as alreadyseen, comprisestwo main dimensions. The first is quantitativeand has to do with how many studentspass 'successfully" throughthe educationalprocess. The other is qualitativeand has to do with what happensto these individualsduring that process,i.e., with the gains that occur in the form of cognitiveachievement and other dimensionsof output. In light of these two dimensions,it is useful in discussingpolicies for enhancingeducation's efficiency to distinguishbetween policiesthat would involveadditional participants in the system,given the averageachievement gain, and others that would raise achievement,given the number of participants.

Savingsrealized via more efficientproduction methods may become availablefor redeployment.Those realizedin a given schoolor educational sub-systemcan be put back into the same unit to increasethis unit's output, or they can be shiftedelsewhere in the economy-- to anotherunit within the same sectoror to a differentsector. Althoughin any particularinstance there may be good reasonsfor doing otherwise,policymakers should, before doing anythingelse, considermaking savingsavailable to the same unit from which they were extracted. Otherwisethere is usually little incentivefor managers to reducecosts.

The secondcluster of policy issues relatesto the other factor constraining,systemoutput, adequacy of finance. Especiallywhen economicor politicalcriteria would call for an increasein the quantityof education provided (whetheran "increase'is definedin terms of additionalstudents, or higher achievingstudents), the financialquestion looms large on the policymaker'sagenda: How does a societymobilize additional resources for education? To this question,there are three possibleanswers: (1) government - 44 -

can allocateadditional public resourcesto the sector; (2) communitiesand familiescan allocateadditional local or privateresources to the sector;and (3) the donor :ommunitycan allocateadditional international resources to the sector.

Given education'simportance in the developmentprocess, additional resourceswill need to be mobilizedfor African education-- from all sources, public and private,domestic and international,although in many countries, public domesticbudgets have alreadybeen stretchedquite close to the limit. Under the conditionsof austerityexisting on the continenttoday, therefore, for most African countriesthe strategyof utilizingresource flows more efficientlyoffers the greatestpotential for amelioratingproblems of access and quality;hence the emphasisin this paper on managementand efficiency. But diversifyingsources of financebeyQnd the public budget is also important, and the forms of such cost sharingthat might be appropriateto each level of educationwill be discussedin the chaptersthat follow.

A third set of issuesto be addressedconcerns the relevanceof the curriculumto the needs of individualsand societies. The main issuesthat many countriesface in this domain concernwhen and how to make the transition from subjectsthat have broad vocationalrelevance (e.g., language and science skills)to programsand subjectsthat will prepare individualsfor specific jobs or clustersof jobs. There are severalrelevant lessons from internationalexperience. The first is that a strong generaleducation, which schoolsprovide efficiently,greatly enhances the futuretrainability of an individual. The second is that job-specifictraining is also very important, and that such trainingis usuallymost efficientlyprovided after initialjob decisionshave been made and in institutionsunder strong influenceof the ultimateemployer. Occupation-and job-specifictraining need not provide individualswith degreesor credentials.

There is no longerserious discussion about the appropriatemain emphasisof the primary (and lower secondary)curriculum: to impartthe basic skillsof languageand mathematics,as well as a basic knowledgeof the social and naturalenvironment, is what is most relevantat this levelto future productivity. The chapteron primary educationdoes not discussthis issue further. At the upper secondarylevel, however, the trade-offis more real between continuingwith generallyrelevant education on the one hand and specializationon the other. And, as specializationbecomes preferable for an increasingproportion of students,the questionarises of whether schoolsare efficientinstitutions for providingspecialized training or whether institutionsat or closerto the ultimatework-site would be better. The answermay well differfor differentoccupations and, perhaps,for different countriesat differenttimes. Chapter5 discussesthese difficultissues further,and the chapteron higher educationdeals with closely related questionssuch as, for example,the relativeimportance of producinglawyers versus engineers.

A fourth class of issues involvesequity. African governmentssince independencehave stressed,almost uniformly, the importanceof poverty reduction-- of eliminatingextremes in the distributionof incomeand achievingbroad participationin the developmentprocess. World Bank experiencehas shown that one of the most effectiveways to addressthe needs of the poor is to give them a productiveasset such as education,rather than to transferincome to them directly. - 45 -

Fortuitously,many educationpolicies serving important equity objectiveshave other important benefitsas well. Furtherextension of primary education,which is perhapsthe most importanteducational policy to address unequal income distribution,is also characterizedby high rates of return; educationthat empowerswomen also enhancestheir productivity,both as income producersand also in the managementof the home environment;policies that reducethe levelsof subsidiesgiven to relativelywell-to-do students will not only reduce inequitybut also result in improvedefficiency in the provisionof learningopportunities. A great virtueof educationalinvestments is their potential,properly designed, for reducinginequity; the paper returns frequentlyto this point in the pages that follow.

* * *

Becausethis is a sectoralpolicy paper, the discussionhere focuses inevitablyon what African educationpolicymakers must considerdoing in order to put their own houses in order. To the extentthey manage to do so, the claim that the sector can legitimatelyplace on nationalresources is strengthened. In the end, of course,the successof policy reformin education will depend very much on the institutionof appropriatecomplementary policies in other sectors,and in the overallmacroeconomic framework within which all sectorsoperate. - 46 -

Chapter 4. THE FOUNDATION: PRIMYRYEDUCATION

The growth in primaryschool enrollmentsin sub-SaharanAfrica during the 1960s and 1970s may be unparalleledat any time or place in history. Through this period,however, some analystswarned that the region's preoccupationwith quantitativegrowth would lead to a seriousdeterioration in the qualityof education. Warningsof fallingquality became even more persistentin recentyears as the financialsqueeze between ever-tighter budgets and ever-largerpupil cohortsstarved educational systems of essential operatinginputs such as textbooksand facilitymaintenance. Poor qualityin primary educationis a seriousmatter, because this is the only formal educationthat a majorityof today'sAfrican children can hope ever to receive,and becausethe qualityof primaryeducation plays a major role in determiningthe qualityof all higher levelsof education.

Even so, the extent to which any particularcountry can increasethe relativeemphasis given to educationquality in the near futuredepends on the progressalready achieved with respectto quantityin the past. In this regard,the diversitywithin the region is considerable.The median gross enrollmentratio for the ten lowestprimary enrollment countries (LPECs) is 37Z, rangingfrom 21% to 49%. The enrollmentratio for 13 middle enrollment countries(MPECs) ranges from 53% to 79%, with a median value of 67%. Finally,the enrollmentratio for the 16 highest enrollmentcountries (HPECs) ranges from 87% to well above 100%, the median being 106% (see Annex Table C.5).

Furthermore,within every countrythere are large enrollment disparitiesbetween differentpopulation groups. In 13 out of the 39 African countries,the share of girls in primaryenrollments is smallerthan 40%. Much largerthan the disparitiesbetween girls and boys at the primary level are those found between rural and urban areas. In most countries,one observes,in effecttwo educationalsystems, an urban systemand a rural system. The formerhas close to universalenrollment wnile the latter is composedof scattered(often incomplete) schools and has much lower enrollment ratios. A recentsurvey in Cote d'Ivoire,for example,showed en.ollment ratios in Abidjan and other urban centers30 percentagepoints higher on the averagethan in the countryside.Thus, regardlessof the overallenrollment ratio in a country,increased coverage may well involvethe additionof childrenwho are more difficultto attractto school (becauseof high opportunitycosts or negativeparental attitude towards school) and for whom provisionis more costly (becauseof scatteredpopulation).

At first blush, the trade-offbetween improved quality and further expansionof primaryeducation seems stark. To pose the dilemma: whereas improvementof educationquality will need to assume a more prominentplace on the agenda of educationpolicymakers in the years ahead,few if any African countriescan yet afford to take for grantedthe dimensionof education quantity. A questionremains, however, as to the most efficientmeans of raisingthe academiccompetence of pupilsby any given amount. If, by increasingthe annual flow of certaincrucial inputs (such as instructional materials),it becomespossible to bring pupils up to the same level of competenceas before in fewer years than before (or to a much higher level of - 47 - competencein the same number of years), therebysaving on the use of other, much more costly factors(notably, capital costs and teachers'salaries), then it almost certainlypays to do so. Seen in this light,there is a cost of low-qualityinstruction, and the trade-offbetween quantity and quality becomes a moot issue.

In the countrieswhere educationalstandards have deterioratedthe most, the choicebetween expansion and quality is no longerreally an either-orchoice. Withoutsome minimalpackage of basic inputs,including most importantlyan adequateamount of textbooksand other instructional materials,virtually no learningoccurs. Under these conditions,ensuring the availabilityof essentialinputs is a prerequisiteboth of qualityimprovement and of expansion. For the remainingcountries, however, it is a questionof findinga balance that reflectscurrent levels of enrollmentand quality. The remainderof this chapterdiscusses efficient approaches, first to quality enhancement,then to quantitativeexpansion. The concludingsection of the chapterdeals with the questionof how to financeimprovements in both.

4.1 Measuresto improveschool quality

This sectionreviews a number of possiblemeasures for maintaining and raisingthe qualityof primaryeducation. Five categoriesof inputsinto the productionof educationquality are identifiedand discussed:(a) teachers (classsize, teachertraining and morale); lb) instructionalmaterials (textbooksand other readingmaterials; writing implements;radio and other instructionalmedia); (c) schoolbuildings and facilities;(d) languageof instruction; and (e) nutritionand health of children. A sixth important ingredientto improvingquality is a strongexamination system. External examinationsplay an importantrole in qualityimprovement in two ways: throughperformance measurement and curriculumimprovement. Since this device is essentiallya managementtool, it is treatedin Chapter7, which discussesthe strengtheningof management.

The potentialfor adjustingany of these inputsto improvequality efficientlyvaries, and the purposeof the discussionthat followsis to summarizewhat experiencehas to say about the effectivenessas well as the financialand administrativefeasibility of the alternatives. In the final analysis,the effectivenessof any input dependson how it is used in conjunctionwith all other inputs;an effectiveschool is one that offers a coherentpackage of inputs.

(a) Trainingand use of teachers

Two dimensionsof the teacherinput may be distinguished:quantity as reflectedin class size, and qualityas reflectedby factorssuch as length and contentof trainingreceived, classroom organization, and teachermorale.

Class size. There is substantialresearch evidence in supportof the propositionthat, within broad limits (between25 and 50 pupils), changesin class size influencepupil achievementmodestly or not at all. Given that most classroomsare not designedfor more than 50 pupils,however, and given that disciplinecan be a problem,classes much largerthan 50 pupils should probably be discouraged. - 48-

In primaryschools operating under a single-shiftsystem, the pupil-teacherratio is a good proxy for class size. The median pupil-teacher ratio in all of sub-SaharanAfrica is 39-to-1. This suggestsacceptable class size on average. For classesthat are at or close to this number,only a substantialreduction in class size, too costly to be consideredby any countryat the presenttime, would be likelyto raise achievements.

However,even in countrieswhere the averageclass size is below 45 pupils, there will still be many classes,especially _n urban areas, that are much largerthan this and where overcrowdingmay detractfrom a pupil's opportunityto learn. Moreover,nine of the 39 African courtriesreport averagepupil-teacher ratios in excessof 50. In such settings,modest reductionsin class size (or largerclassrooms) could prove to be cost-effective,especially if this strategywere accompaniedby other measures to increasethe number of classesthat teachersteach on average (through,for example,double-shifting).

TeacherTraining. There is an incressingbody of evidenceon the payoff of various amountsand kinds of teachertraining. The resultstend to show that preservicetraining for primaryschool teachers consisting of more than generalsecondary education and a minimum exposureto pedagogicaltheory is not cost-effective.Long residentialcourses tend to be quite expensiveto produce,and teachers'salaries are usuallyclosely tied to the amount of such trainingreceived.

In recognitionof these two cost factorsand the difficultyof changingthem, policymakersin some countries,e.g., BurkinaFaso, have chosen to shortenthe durationof preservicetraining for primary schoolteachers. In other countries,the cost of preservicetraining has been substantially loweredin recentyears thx.ughuse of distanceteaching (i.e., correspondence educationsupplemented by raiio and tutorialsessions). Nigeria,Zimbabwe, and Tanzaniahave all successfullyapplied distance teaching methods for preserviceteacher training; Box 4.1 summarizesTanzania's experience.

(Box 4.1]

The optimalmix of trainingmodet (generaleducation, preservice classroomstudy, supervised teaching practice, learning on the job, and in-servicetraining) will dependon the relativecosts of the variousmodes and on the salary structurefor teachers. In many countries,however, an emphasison in-servicetraining has been found to be cost-effectiverelative to preserviceprovided that constructivesupervision is providedas a concomitantto the training. Some countries(e.g*, Botswana, Ghana, Kenya, Lesotho,and Malawi) are now making successfuluse of distanceteaching methods to upgrade the pertinentknowledge and skillsof servingteachers.

Perhapsmore cost-effectivethan trainingper se as a way of taising the professionalquality of the teachingforce may be to introduce,where labor market conditionspermit, more selectivecriteria for admissLonto the profession. The recentincrease in unemploymentamong universitygraduates in the liberalarts and secondaryschool leavers should help in atitractingmore able individualsto the sector,although many potentiallyable teachersmay lack formalcredentials. A facilitatingpolicy here would be to admit individualsto the professionbased on performanceon subject-mattertests, then to providethem brief initialand in-servicetraining. - 49 -

Box 4.1 DISTANCELEARNING FOR TEACHERSIN TANZANIA

In 1974 Tanzaniamounted an effortto achieveuniversal primary school enrollmentby 1977,despite serious resource constraints. It was estimatedthat 40,000teachers would be requiredto reach the goal, and that it could not be accomplishedthrough conventional teacher training methods. Further,the pool of secondaryschool leavers who might be pressed into service as primary school teacherswas small,because the governmenthad focusedon developingprimary and adult education. Thus, Tanzanianeeded a new strategyto fill its primaryteaching ranks. It chose to use primaryschool graduateswith some experiencein adult educationand to trainthem on the job.

Traineeshad to be between 17 and 28 years old, live in an area where teacherswere in short supply,and have taughtadult literacyfor at leasttwo years. The strategy consisted of providing an initial six-week residential training course, followedby supervisedprimary schoolteaching. While working in the schoolsand teaching22 periodsa week, trainees followedcorrespondence courses and listenedto related radio programs. In addition,opportunities were provided for them to meet and discuss theirwork with fellowtrainees and supervisinghead teachers.Trainees were examinedeach term, and a final, nationally organized, examination was administered at the end of the three-year course. Of the 45,534students who began the course between 1976 and 1978, 37,325 (82 percentof thosewho started)completed it, and 35,028 (77 percent)passed their final examinations,thus gaining qualified teacherstatus.

A comparison of these trainees with a control groupwho attendeda regular teacher training program found the first group performed slightly less well in academic knowledge, but betteron measuresof classroombehavior. Becausethey were recruited locally, their level of motivationmay have been higher. The combinationof the practicalclassroom apprenticeship with study at a distanceappeared to be an effectiveway to respondto the criticalprimary teacher shortage. Further,the strategyrealized important savings, since teachers are employed during their training period and the costs of residenceat trainingcollege is minimized. The costs of the distance teaching strategyin Tanzaniawere calculatedto be approximatelyone-quarter the cost of conventionalteacher training. - 50 -

In summary,in many African countries,ineffective teachers are a constraintto pupil learning,deserving of the policymaker'sattention. In most of these countries,in-service training is likelyto prove a more cost-effectivemeans of amelioratingthis problemthan alternativeprograms of preservice training.

Teachermorale. The teacher'suse of time and other classroom resources is known to be a principal determinant of pupil achievement. To the primary schoolteacher falls the major task of turninggovernment-imposed strategiesfor qualityimprovements into reality. But to be successfulin this regard requiresthat the teacherbe motivatedand dedicated. In the past the appellation'teacher' was a term of considerablerespect. The profession commandedhigh status in the community,and positionswere eagerlyfilled. Today, primaryschool teachersare often insteada beleagueredand dispirited force, their statusmuch eroded and their working conditionspoor.

The liftingof teachers'morale reprosentsa major challenge,since most countrlescannot affordadditional monetary incentivesto achievethis end. Not all measuresto regenerateteachers' professional pride and enthusiasm,however, need be costly. Increasedprovision of instructional materials,and better supportand supervisoryservices from inspectoratesand ministrieswill help improveworking conditions,particularly for the many teacherswho must work in the relativeisolation of rural areas.

(b) Instructionalmaterials

Evidencepoints stronglytoward the provisionof instructional materials,textbooks especially, as being the most cost-effectiveway of increasingthe qualityof primaryeducation. The scarcityof learning materialsin the classroomis the most seriousimpediment co educational effectivenessin Africa. It is certainlyhere that the gap in educational provisionbetween this regionand the rest of the world has grown widest.

Given that many primaryschool teachersin Africa have less formal educationand teachertraining than is usuallythe case in more developed regionsof the world, the use of teachers'guides and other materialsdesigned to assistlow-qualified teachers in the organizationof classroomactivities could prove to be especiallycost-effective in the African context. The advantageof such materialsis that they are intendedto supplementthe teacher'sown knowledgeand to promotethe proper sequencingof learning activitiesIn the classroom.

The availabilityof all such instructionalmaterials has declinedin recentyears as increasedfiscal stringency has led to severecuts in nonsalaryitems of expenditure.The problemof the scarcityof appropriate teachingmaterials, however, goes well beyond the availabilityof funds. Most African countrieshave yet to developa nationalcapacity for the development of low-costteaching materials that are pedagogicallysound and relevantto the nationalcurriculum.

With regardto writtenmaterials, an objectivecommon to all African countriesshould be to developnational skills to adapt and edit such materials. For most countriesin addition,enhanced capacity to write and publishsuch materialsin a feasibleshort-term objective. Local printing, however,should not be taken for grantedsince small countrieswith limited - 51 - educationalmarkets are particularlycostly to serve economicallywith local presses. The economiesof scale achievablethrough the combinationof sophisticated(and very expensive)high-speed multicolor presses and relativelyinexpensive but highly skilledlabor often makes it more economical for African countries -- and for many American and European publishing houses as well -- to print textbooksfor mass circulationabroad. In those African countrieswhere the local printingindustry is reasonablydeveloped and efficient,the awardingof contractsfor paper procurementand textbook printingthrough internationalcompetitive bidding will assurethat prices paid do not exceed the lowestprices obtainable elsewhere by more than the normalmargin of domesticPreference accorded to goods and servicesof nationalorigin. Cooperationamong groupsof small Africancountries affords the possibilityof economiesof scale in printing,as well as the possibility of producingmaterials in those local languagesthat cut acrossnational borders.

In additionto their developmentand production,instructional materialsneed to be storedadequately, and distributedto schoolsin a timely manner, and teachersneed to be trainedin their use. All this requires organizationand planningand, above all, funds for transport,an item in very short supply.

Exercisebooks and pencils are basic to the learningof literacyand numeracyskills. To recovercosts, some countrieshave transferredto parents the cost of these basic supplies. An advantageof this approachis that it lends protectionto the provisionof these relativelyinexpensive but pedagogicallycrucial inputsduring periods of financialstringency.

Apart from tangibleinstructional materials, radio broadcaststo the classroomhave proven in some countries(e.g., Ethiopia, Kenya, Tanzania)to be effective. For primaryeducation, the use of radio would usuallyrepresent an add-oncost (whereasin secondaryand higher education,radio can be used as a partialsubstitute for teachers-- see Chapter5). In return for its modest inflationaryeffect on unit costs, the use of radio can be expectedto yield significantlearning dividends in primaryeducation, especially in the absenceof highly skilledteachers. In Kenya, for example,there was recently a major effort to teach Englishto primaryschool studentsin regionswhere Englishwas not widely spoken;a carefulevaluation showed very substantial learninggains (Box 4.2). Elsewherein the world there has been very favorableexperience with the use of radio for teachingmathematics in primary school;the radio package (includingworkbook) substituted for textbooksat a quite low incrementover what textbooksalone would have cost.

[Box 4.2]

Although radio,when thoughtfullyused, has an importantpotential for improvingschool quality,there is no evidencethat the use of more costly technologies(e.g., television) has any advantageover radio. Indeed,the cost and complex implementationof the once massive educationaltelevision system for primaryeducation in Cote d'Ivoireled to its eventualabandonment. - 52 -

Box 4.2 INTERACTIVEEDUCATIONAL RADIO IN KENYA

Interactive radio differs dramatically from traditional educationalbroadcasting in its reliance on student participation,or interaction,with the program. Unlike the instructionaldesign of traditionaleducational radio that encourages passivity as students listen to lecture-styleinstruction, the design of interactiveprograms makes creativeuse of radio. The Kenya Radio Language Arts Project (RLAP), a good example of interactiveradio, involved primary school childrenas activeparticipants in a pedagogically sound dialogue that taught the Kenya languagecurriculum.

What might have lookedlike pandemoniumin a RLAP classroomwas actuallya well-designed,tightly-controlled lesson called "English in Action"whose key attributewas that it involvedstudents actively in the learning process. The thirty-minute daily broadcasts, which were punctuatedby music and littledramas, incorporateregular pauses for the childrento respondand receive immediate reinforcementfor answers. Responses could be sung, spoken,provided in writing or throughphysical action. Typically children were given the chance to respond over one-hundred times duringeach thirty-minuteperiod. Carefulevaluation of students'learning gains after introduction of radio found highly significant increases. The cost of extendingRLAP to additionstudents is estimatedto be well less than $1 per studentper year.

The effortto involvechildren in a conversationwith the radio demands precisiontiming and carefulobservation of how childrenrespond to radio prompts. RLAP designershave achieved this precision through trials, observations, repeated pretesting, and classroommonitoring. Teacherstended to be supportiveof the interactive radio experiment; they saw the program as a way to enhancetheir work, and not as a way to replacethem in the classroom. With assistance from teachers' guides, they work along with the radio programs,calling on individualchildren as cued by the radio,overseeing written responses,and providing closer overall supervisionthan would be possiblewithout interactiveradio.

Kenya's RLAP was based on the successfulexperience of Radio Math in Nicaragua. That program'suse of interactive learning produced consistently superior results among children in radio mathematics classes,compared with childrenin conventionalclasses.

Experience has shown that interactive radio can be used effectivelyby untrainedclassroom monitors, as well as trainedteachers, with littletraining or specialsupport. Further,once radio lessonsare developed, the annualper pupil cost is modest, since few supplementary learningmaterials are required. - 53-

(c) Physicalfacilities

Dilapidatedconstruction, missing or broken desks and chairs,and the absenceof good ventilationand sanitationfacilities are commonplaceamong African schools,especially in rural areas. Not very much is known about how constructionstandards and schoolupkeep affect the qualityof educationas indicatedby pupil achievement,nor about the effectsof the presenceand conditionof other school facilities. One effectof low-standardand poorly maintainedfacilities may be to discouragepupil attendance. For those who attend,little can be learned,surely, on a rainy day under a leaky roof.

Inadequateplant maintenanceand missing or broken furnitureare problems that have been aggravatedby the currentbudgetary crisis because, in most African countries,the responsibilityfor maintenancerests with central government. The generaltrend toward greater local financingof the capital costs of education,if extendedto includecapital maintenance, might alleviatethese problemssomewhat, and the use of more localmaterials in the constructionof schoolbuildings and classroomfurniture may make it possible to reducetheir costs. Whoeveris responsible,central government or local authorities,the failureto maintainphysical facilities not only curtailslearning, it can increaseoverall costs when prematurereplacement of the facilitiesis the result.

(d) Language of instruction

The diversityof linguisticbackgrounds in sub-SaharanAfrica greatly complicatesthe pedagogicalprocess; indeed, linguistsidentify over 1,250 languagesin use today in sub-SaharanAfrica. Only 9 of these languagesare spokenby over 10 millionpeople as a first or second language. The African educator'sresponse to this challengehas reflected,in part, attentiongiven to colonialprecedents and, in part, a healthysense of pragmatism. Although some countrieshold literqcyin one or more African languagesas an explicit goal of the educationsystem and others do not, the combinationof colorial heritageand the relativeabsence of publishedmaterials in these vernaculars has led most African countriesto adopt the languageof the formercolonial governmentas the nationallanguage and to introduceit as the medium of instructionat some level in the formaleducation system.

Differencesin languagepolicy have to do with how high on the educationalladder and how gradualthe transitionto the nationallanguage is expectedto be. The colonia'lpowers in Africa pursueddifferent policies with respectto media of instructionin schools,and African nationssince independencehave often been true to these traditions. As shown in Table 4.1, 11 of 15 formerFrench colonies and all three formerPortuguese colonies officiallybegin instructionin the nationallanguage, from the first day of primary school. On the other hand, 13 of 15 formerBritish colonies begin instructionin one or more African languages,teaching English at first as a subject;only later is Englishintroduced as the medium of instruction.

[Table4.1]

While the decisionto use or not to use an Africanlanguage for instructionalpurposes often echoes the colonialpast, the feasibilityof doing so dependsalso on the number of speakersof the languagein the nation. An examinationof currentschool language policies across African countries - 54 -

Table 4.1 Medium of Instruction by Former Colonial Status

Medium of instructionin first year of primary school

Metropolitan One or more language African only languages

FormerBelgian 1 2

FormerBritish 2 13

Former French 11 4

FormerPortugues.a 3 0

Note: Entriesin the table are the numberof countries in each category. Based on Anne" Table B.2. - 55 - suggestsconvincingly that the size factorhas influencedpolicymakers. In 21 countriesthat utilizeone or more Africanlanguages for instructingbeginning primarypupils, the number of (first-and second-language)speakers of the most widely spokenAfrican languageis more than five million people in the median case. On the other hand, in 16 countriesin which a Europeanlanguage is used as the medium of instruction,the numberof speakersof the most widely spoken languageis only about two-and-a-halfmillion (analysisbased on Annex Table B.2).

Policy toward languageof instruction-- whether and when to use the nationallanguage or an African language -- must be devised by African governmentsthemselves based on political,as well as economic,imperatives. For most African countries,a centralobjective of primary educationis that pupils emerge orally fluentand literatein the nationallanguage. Fluency in the nationallanguage may help to promotepolitical stability and build nationalunity, as well as serve economicpurposes.

On purely pedagogicalgrounds, however, the benefitsof mother-languageinstruction in the initialyears of primaryschool now seem to be establishedeven when literacyin the nationallanguage is the ultimate objective. Current researchsuggests the following:(1) second language acquisition,both verbal fluencyand literacy,is most successfulwhen there is a strong foundationin the first language;(2) conversationalskills in a second languageare learnedearlier than the abilityto use the languagefor academiclearning; and (3) academicskills learnedin schooltransfer readily from one languageto the other, so that skillstaught in the first languagein transitionalprograms do not have to be re-learnedin the second language.

In light of the above,most linguistsare in agreementthat, even in situationswhere instructionis ultimatelyto be given in a languageother than the child'smother tongue,the most effectivepolicy educationally is one of initialinstruction using the mother language(vernacular) as medium, followedby a gradualtransition to the nationallanguage as medium. Ideally, study of the first language,as a subjectat least,will continueonce the transitionis complete. The pedagogicaladvantage of this approachis more pronouncedin a transitionperiod in which the teachersthemselves are not particularlyfluent in the nationallanguage, a situationthat is fairly typicalin many Africanprimary schools today.

The effectivenessof this kind of bilingualeducation policy relative to the costs involved,however, depends on both the size of the population group being educatedand on the degreeof linguisticheterogeneity within individualclassrooms. In Uganda,for example,perhaps an unusualcase, 47% of primary school classroomscontain students who speak four or more different mother tongues;in this type of environment,instruction in Englishfrom the outsetmay be the only feasibleapproach. Experiencefrom the Rivers State of Nigeriademonstrates, however, that it is possibleto preparemother tongue textbooksfor a substantialnumber of separatelanguage groups, each one of small size, if that policy is chosen. Country-specificcircumstances here will be decisive.

In many countries,however, providing initial literacy in the African languagewill not be the principalproblem. Rather,the problemwill be that of effectivelyintroducing a nationallanguage that many of today's primary-levelteachers speak and read only poorly,if at all. Good materials - 56 - are importantto rectifying this, as are teachertraining and selection. Use of radio also has a particularcomparative advantage in languageteaching; the successof the "RadioLanguage Arts Program'in introducingEnglish into rural Kenyan primaryschools (Box 4.3 above)exemplifies radio's potential.

(e) Nutritionand health: ensuringteachable pupils

The recentdrought in much of Africa and its immediateimpact on nutritionalstatus and human survivalrates receivedwidespread, if belated, internationalattention. Unfortunately,although the absenceof rainfallwas a more or less temporaryproblem, serious nutritional deprivation predates the recent crisisand will continuein many parts of Africa even as normal harvestsare achieved. Childhoodmalnutrition and concomitantdebilitating diseasesare certainto continuewell into the foreseeablefuture as perennial problems.

The incidenceof malnutritionand diseaseis especiallyhigh among preschool-agechildren. By the time such deprivedchildren reach school-enteringage, a large proportionare physically"stunted' (of below-normalheight), which is a frequentconcomitant of impairedmental ability. Many of those who are malnourishedand sick will never attend school. Those who do enroll tend to be listlessfrom hunger and weakenedfrom their frequentbouts with diarrheaand fever; their attendanceand academic achievementobviously suffer. The high benefitspredicted to accrue from investmentin educationare never realizedin the case of these sick and malnourishedchildren.

To the extent that healthproblems continue and imbalancespersist (or worsen) between food supplyand population,it will be essentialto ascertainwhat remediesthere are for childrenand what the consequencesof failingto adopt those remedieswill be, for the educationsystem in particularand for societymore generally. Programsto promote family planningand provideprimary health care, includingprenatal care; programsof nutritioneducation for mothers;preschool education programs for children; and in-schoolfeeding programs for thosewho enroll -- these all have the potentialfor yieldingvery high returns. Becausethey do not fall neatly within any one ministry'srealm of responsibility,however, but instead overlapministerial responsibilities, programs thac addressthe complexof problemslinking health, nutrition, and intellectualdevelopment, especially as they affect those not in school,tend to fall betweenthe bureaucratic cracks. Governmentsmay wish to considergiving more attentionto the area of preschoolchild development,especially to nutritionand primaryhealth care, so as to identifyapproaches that are effeetiveand feasiblewithin particular nationalcontexts.

School feedingprograms targeted to those at greatestnutritional risk could,under some circumstances,provide the most effectivemeans for improvinga child's abilityto learn. The food requiredfor such programsis often availablein kind from externalsources such as the 4orldFood Programme. - 57 -

(f) Sumwry This review of measuresfor improving primaryschool qualityhas providedonly a limitedyield of attractiveoptions. The evidencecan be summarizedin terms of two principalconclusions:

(1) The followingkinds of investmentsare unlikelyto have ary noticeable effect on primary schoolquality despite their potentiallyhigh cost: reducingclass size, providingprimary teachers with more than a general secondaryeducation, providing teachers with more than minimal exposure to pedagogicaltheory, constructing high qualitybuildings and furniture, and introducingclassroom television or computers.

(2) The safest investmentin educationalquality in most countriesis to make sure that there are adequateoooks and supplies. These are effectivein raisingtest scores and, almost invariably,have been underinvestedin relativeto teachers. This is also an area where externalaid has comparativeadvantage. Other areas that appear to 1ave potentialinclude school feedingand health programs,intensive use of radio ("interactive radio"),in-service education of teachersin subjectmatter skills,and strengthenedinspection and supervisionsystems.

While these conclusionswere formulatedwith referenceto developing countriesgenerally, they apply in Africa. Indeed,the currentsevere shortageof teachingmaterials in Africanprimary schools -- in African schoolsat all levels,for that matter -- gives particularforce to the conclusiondealing with textbooksand materials. In terms of rectifiable inefficiency,the relativeimbalance of inputsto the detrimentof critical learningmaterials stands out as a pervasiveproblem. Hence the second recommendationof this papers

Recommendation2. A prerequisitefor both quality improvementand systemexpansion in African educationis the assuredavailability of nonsalary recurrentinputs. In countrieswhere recent fiscalconstraints and persistent pressuresto expand enrollmentshave combinedto distortthe balancebetween salary and nonsalaryexpenditures, governments should take steps to restorean efficientmix of inputsinto the educationalproduction process. The provision of a minimum packageof textbooksand instructionalmaterials is usuallyth6 most pressingneed in this respect,and this is criticallyimportant if productiveuse is to be made of the other (muchmore costly) recurrentinputs into education,namely, teacher and studenttime. The problemof inadequate suppliesof books and materialsexists at all levelsof education,but it is particularlyacute (and relativelyinexpensive to rectify)at the primary level,where an expenditureof about $5 per pupil per year shouldmeet minimum requirements. Similarlywith regardto physicalplant and equipmentin African educationalinstitutions, the balancebetween development outlays and maintenanceexpenditure should, in most countries,be adjustedto give relativelymore weight to the latterin order to ensure that full benefit is derived from existingfacilities. - 58 -

4.2 Unit cost containment

The precedingsection has reviewedoptions for qualityenhancement and arguedthat a minimumpackage of textbooksand teachingmaterials is a prerequisiteboth for qualityimprovement and for furtherenrollment expansion. To help financethis minimum package,it is essentialto identify all possibleways of containingunit costs.

To demonstratethe importanceof unit cost containment,it is worth recallingthe enrollmentdisparities noted at the beginningof this chapter and the fact, noted in Chapter2, that in 19 of the 39 countriesdealt with in this paper, primary enrollmentratios have actuallydeclined since 1980. In at least four of these countries,not only have enrollmentratios declined, but also the absolutenumbers of enrollees.

To addressthese problems,this section reviewsthe scope for unit cost reductions,stressing those measuresthat preservequality. The concludingsection then turns to a discussionof the scope for mobilizing additionalresources for primaryeducation.

(a) Teachers'salaries

The salarybill for teacherstypically accounts for between85% and 95% of primaryeducation's recurrent budget in an African country. Any discussionof strategiesfor cost savingsmust includean examinationof the scope for reducingwhat the averageteacher gets paid.

The severescarcity of teachersover much of the past three decades, the (hitherto)attractive alternative employment opportunities for persons with the educationalqualifications of teachers,and the heavy politicaland economicpower of the teachers'labor organizationswould all tend to contributeto high salariesfor teachersrelative to averageearnings in Africa. It is not surprising,therefore, that internationalcomparisons of primary teachersalaries as a multipleof per capita incomereveal that primary school teachersin Africa,especially in the Francophonecountries, earn relativelymore than their counterpartsin other regions. In the late 1970s,the averageprimary teacher's salary as a multipleof per capita income was 2.5 in OECD countries,2.4 in Latin America,and 2.6 in Asia. For 22 African countriesfor which informationwas availablearound 1983,the average primary teacher'ssalary was 5.6 times GNP per capita in the median case (Annex Table A.23). For the 10 Francophonecountries, the figurewas 8.8 times GNP per capita,and in one, Mauritania,a remarkable15 times. For the 12 Anglophonecountries, by contrast,the averageteacher salary was 3.6 times GNP per capital in Uganda the two were pract.cally on a par. Couchingthe comparisonsin terms of a technicallymore appropriate(although not readily available)indicator, such as averageteacher salary as a multipleof non-agriculturalGDP per capita,or of averagewage in the modern sector, might attenuatethe overalldifferences and reorderthe rankingof individual countries,but it would not alter the generalconclusion that teachersare expensivein the skill-shortAfrican context.

This is not to say, however,that teachersare necessarilyhoverpaid' -- i.e., compensatedmore than is necessaryto keep them in teachinggiven employmentopportunities open to them elsewherein the economy. There are reportsof seriousteacher shortages and absenteeism,especially in remote - 59 - rural primaryschools, and of teachersleaving the profession,particularly from secondaryschools and universities,to take up more remunerative opportunitiesin other sectorsof the economy. At the same time, many adequatelyeducated young people are activelyseeking employment in most Africancountries, and they providea pool of talent that (with appropriate hiring and in-servicetraining policies) might welcome reasonabledownward adjustmentsin teachers'salaries if this policy openedup employment possibilities. Even bearing in mind the problemof low teachermorale, and recognizingthe politicaldifficulties involved in cuttingnominal salaries, theremay be some scope for loweringthe overallstructure of primary teachers'hourly earnings.

However,even in labormarkets where the salaryof a teachertrained to a given level cannot be reduced,it is possibleand perhaps desirableto recruitteachers who are less well-trainedand, hence, less costly-- as has been done in BurkinaFaso (see Box 4.3). Data for six Francophonecountries show that the differencein the startingsalaries of teacherswith two and four years of preservice training (followingcompletion of juniorsecondary education)is 401 at the median. In view of the modest effectson educational quality (as measuredby pupil achievement)of the durationof professional pedagogicaltraining for teachers,recruitment of the more trainedgroup of teachersis not likelyto be cost-effectivein these countries. Indeed,use of essentiallyuntrained but educatedyoung people as primaryteachers -- A-levelor Baccalaureategraduates prior to (or Insteadof) entranceinto universities,or universitygraduates fulfilling a nationalservice obligation -- is an option that could reducethe averagecost per teacherand, particularlyif combinedwith intensiveshort-course pedagogical training, would likelynot involveany sacrificeof quality.

[Box 4.3]

Finally,a countrycould reduce the averagerate at which teachers move from one step on the salary scale to the next. At present in most countries,salary incrementsare given every two years or so, more or less automatically.As an alternative,a new norm of four or five years between promotionsmight be established. This time could be reduced,however, by two or even three years, for primaryteachers willing to teach in rural areas,for post-primaryteachers in subjectsdifficult to cover (e.g.,the sciences),or for teacherswith outstandingrecords of attendance. In general,a strong case can be made in theory for the rationalizationof salarystructures -- i.e., for rewardingwith higher pay those characteristicsof teachersthat are in short supply insteadof linkingsalary progression very rigidlyto age, seniority,and entry qualifications.Certainly a start in the right direction could be made simply by associatinga below-normtime-in-step to rural service or successfulcompletion of in-servicetraining programs.

(b) More intensive use of teachers

The salary bill requiredto sustaina given enrollmentis definedby the averagesalary per teacherand by the average ratioof pupils to teachers. Given the difficultiesinvolved in reducingteachers' salaries, using teachers more intensivelywould appearto offer greater cop9 for reducingthe salary costs per pupil. As shown in Annex Table C.5, the pupil-teacherratio differs considerablyamong Africannations. Of the 38 countriesfor which 1983 data were available,18 had ratiosof 37 or below (including7 with ratios below - 60 -

Box 4.3 REDUCING TEACHER COSTS IN BURKINA FASO

Efforts to expand provislon of primary education In Burkina through reduction of unit costs have been hampered by the relatively high cost of teachers. Although primary teacher salarles are modest by International comparlsion, they are hlgh In comparison with Burkina's Income level (more than 10 times the GNP per capita, versus 2.4 times In Latin America and 2.6 In Asia). In light of the fact that 98% of the primary education budget goes for teacher salaries, any attempt to reduce the costs to this sub-sector had to address this central Issue.

The primary teacher work force consists mainly of two categorles: lnstituteurs-adjoints (level 82 In the civil service salary scale), recrulted from graduates of the lowiersecondary schools and tralned for two years; and OInstituteurs (level 81), recruited from graduates of the upper secondary schools or through examination promotion from the instituteur-adjolntm level. Liberal standards for promotion from the B2 to BI level added to the budgetary pressure by driving up teacher salaries at a fast pace.

Efforts to reduce teacher costs through an Increase In the pupil-teacher ratio were not attempted, since the national average was already quite hlgh at 65-to-1. Instead, Burkina Is lowerlng unit costs by restructuring the primary teacher corps and Improving Internal officiency through better-focused teacher training. It has been demonstrated that by reclassifying newly-recruited instituteurs-adjolntsFfrom level B2 to Cl, and reducing tralning for this group from two years to one, unit costs can be brought down signlficantly. Implementationof these policy changes, combined with a revision of current fellowship and subsidy policies, would produce resource shifts In favor of prlmary education expansion. Burkina thereby could attain Its unofficial 60% enrollment target for the year 2000, while realizing simultaneous quality improvements through an Increase of teaching materials. If no policy changes were implemented,the gross enrollment rate would rise only In proportion to overall budgetary growth, falling far short of the 60% target.

The new uInstituteurs-adjointsm, who will comprise the standard teacher category In primary educ.lon, will receive training at the National Prlmary Teachers College In Loumbila. The course consists of pedagogical training (40%). general subjects (40%), and complementary courses such as agriculture and physical oducatlon (20%). These teachers are to receive field assistance through the provision of textbooks and teaching materials, regular podagogical support, and a gradual decrease In class size.

Further, the primary teacher promotion examination Is being upgraded through the Introductionof general subjects In order to Introduce a measure of selectivity Into tho process. This Is expected to raise promotion requirements, thus slowing the overall rate of promotion, and thereby leading to a gradual reduction In average teacher salaries. - 61 -

32). In these 18 cases at least, it should be possibleto increasethe number of pupilswithout increasingcommensurately the number of teachers. The extent to which this can be done will, of course,be greater,the greaterthe populationdensity in a particularcountry.

Even a modest increasein the pupil-teacherratio can result in quite dramaticsavings. For example,in a systemwhere the ratio is currently 35-to-1,a 202 increasewould make the ratio 42-to-1. If the systemwere now enrolling1.4 millionfirst-level students (an average-sizesystem in the region),the same numberof teachers (40,000)could now accommodatean additional280,000 students given the new ratio. Where teachers'salaries now accountfor 90% of the recurrentprimary educationbudget, if the pupil-teacherratio were not increasedbut kept at 35-to-1,adding the same 280,000students to enrollmentswould add a full 182 to the budget. This differenceis not at all trivialwhen one is talkingabout an annualprimary educationbudget in the tens or hundredsof millionsof dollars.

There are three ways that the pupil-teacherratio can be increaseds (1) by increasingthe averageteaching load of teachers-- i.e., the numberor classestaught per schoolyear; (2) by reducingthe averageattendance load of pupils -- the number of classesattended per school year; and (3) by increasingthe averageclass size -- the number of pupils in a classj/

Teachingloads can be increasedby increasingthe numberof classes taughtper week and/orby increasingthe numberof weeks in the schoolyear. The latter measure offers particular promise in a number of African countries where the schoolyear is now quite short. Of course,for these measuresto increasethe pupil-teacherratio, the number of classroomhours must be extended only for teachers and not for pupils. This implies some system of 'double-shifting"(i.e., splitting the schoolday, schoolweek, or school year),with each of the two sessionscatering to half of the total numberof pupilp enrolled. Such measuresmay be impossibleto implementwithout compensatingteachers financially, which tends to defeat the originalpurpose. Doing so can still reduceaverage per pupil cost, however,so long as the increasein the pupil-teacherratio is greaterthan the increasein teachers' salariesand providedthat a heavier teachingload does not seriously underminethe averageteacher's classroom effectiveness. Our assumptionis that, in many instances,teachers will be willingto accepta reductionin hourly earningsin order to increasetheir annualearnings (by being given the opportunityto teach additionalhours per year). In the interestof reducing unit costs and extendingenrollments, Senegal has recentlydecided to implementa poiicy of double-shifting(see Box 4.4).

[Box 4.4]

In additionto raisingteaching loads, countries should make every effort to enforcewhatever teaching loads are officiallyon the books. Absenteeismis quite widespreadin many countries. Enforcementof disciplinewith respectto duties is essentialfor efficiency;it may entail replacingslack teacherswith currentlyunemployed graduates. In addition, countriesshould minimize the extentto which those trainedand paid as teachersare assignedto other, extraneousduties; many teachersperform office and supervisorytasks that could probablybe performedless expensively and at least as satisfactorilyby somebodytrained for such work. - 62 -

Box 4.4 REDUCINGPER PUPIL COSTS THROUGHDOUBLE-SHIFTING IN SENEGAL

Primaryenrollment ratios In Senegalwere expectedto fall by 11% by the year 2000 owing to primary school populationIncreases and growing pressureto containeducational expenditures. In an effort to prevent this decline, the government of Senegal has decidedto Implementa numberof officlencymeasures to lowerunit costs In primaryeducation. Senegal ranks hlghestamong low-incomeAfrican economies In terms of primaryeducation unit costs (US$ 101 In 1983,Annex Table A.17). Teachersalaries, between seven and nine times GDP per capita,account for much of the problem. A secondkey factor responsiblefor driving up unit cost If the significantly lower student/teacherratios In rural areas as comparedto urban. Unit costs vary by as much as 110% In the extreme cases. The use of teachers In administrativepositlons also accountsfor high unit costs.

One measure to lowerunit costs, Increaseenrollments, and reduce the numberof overcrowdedurban classes(which, It Is belleved,have playeda role In declining pass rates In the prlmarycompletion exam since the late 19608)would be the Introductlonof a double-shiftingsystem. Under this scheme,one teacherwould teach two shiftsand receive25% of his or her base salaryas extra compensation.Weekly classroom hours for students would be reducedfrom 28 to about 20; to compensate,the schoolyear would be extended by 30 days. If this systemwere ImplementedIn about 20% of the overcrowded classrooms, and It were Initlatedat grade 1 In the fIrst year and extended to grades 2, 3, and 4 over a four-yearperiod, It would then be possible for 20,000 additional students (or 4% of currentenrollment) to galn accessto primaryeducation In Senegalby the end of the projectperlod.

If the double-shiftingpolicy were Implementedalone, with no other policy Interventions,It would reducethe decilne In enrollmentsfrom 11% to 9% between1986 and 2000. In order to have an even more pronouncedImpact on enrollment rates, double-shiftingcould be Institutedas part of a broader policypackage for reducingunit costs. Other elements that Senegal will Implement Include: (1) Increasingthe proportionof sinstituteursadjolnts* (who have lses training, and are thus paid at a lower level than Instituteurs); (2) redeploying400 teachersfrom administrativepositions Into e'lassroomteaching; (3) reducing the growth of expenditures for secondary education, higher education, and education administration; (4) reducingpublic fundingfor educationfellowships; and (5) limiting the Universityof Dakar'scampus servicesbudget.

It has been ostimated,If all of these pollcymeasures, Including double-shifting,were Implementedtogether as a package, that a total of $176.3 million could be reallocatedto primaryeducation between 1986 and 2000. Such a reallocationwould represent18.3% of the cumulative budgets for oducatlon In Senegalbetween now and the end of the century. Moreover, were the full range of pollcymeasures to be put Into place, the primary enrollment rate could be expectedto Increasebetween 31.2% and 65.5%,and unit costs to fall between12.3% and 19.8%. - 63 -

Reducingthe averageattendance load of pupils,either by reducing the number of classesattended per week or by reducing(for pupils)the number of weeks in the schoolyear, is the secondmajor strategyfor increasingthe pupil-teacherratio and, thereby,reducing the salary costs per pupil. The savingsgenerated by such measuresmust be weighed againstany loss in the averageachievement of pupils. For some countries,however, present financial constraintsmay necessitatethat the trade-offsat least be considered.

In particular,countries that now providethe same numberof weekly hours (e.g.,30) for all primarygrades might considerreducing the lengthof the school day for, say, the first three gradesso as to permit a second shift. The most vocal oppositionto a shorterschool day is likelyto come from parents employedin the urban wage sector,since they may have to make alternativeafter-school (or before-school)childcare arrangements. Parents in rural areas,on the other hand, may welcome a shorterschool day, as this will reducethe opportunitycost of children'sschool attendance.

The thirdway of increasingthe pupil-teacherratio is to increase the avexage class size. Even in countries where the average class size appears "aboutright" (in the 35-to-45pupil range),the averageusually masks large differencesbetween urban and rural areas. In urban areas, classesof 50, 60, and even -3 pupils are not uncommon. In such settings,smaller classesshould probably be the objective. This can be achievedby double-shifting,with pupils attendingschool only half of each day (or every other day).

In rural areas, on the other hand, the problem is usually not overcrowdingbut ratherthe reverse-- classesthat are uneconomicallysmall. In these areas,special efforts should be made to increaseclass sizes. Several approachesshould be considered. First, in areas where new schools are being built, carefulschool-location planning will help. A second approach,which to be successfulwould probablyrequire special materials and trainingfor teachers,is multi-gradeteaching. A third (morecontroversial thougheasier to implement)approach is that of admittingpupils only in alternateyears (as is done now in rural areas of Mali). In other words, the birth dates of childrenentering the first grade in a particularcommunity in a particularyear would span two years insteadof just one, and in any given year, only three years of a six-yearprimary cycle would be taught (i.e.,in Year 1, new studentswould be admittedand grades 1, 3, and 5 would be taught; in Year 2, no studentswould be admittedto first grade and grades 2, 4, and 6 would be taught).

(c) Reductionin repetitionand dropout

Even where the cost per pupil year is reasonableowing to cost-efficientutilization of teachers,the cost per completerof the primary cycle is everywhere higher than need be because of large numbers of pupils who repeatgrades and otherswho drop out before they completethe cycle. For sub-SaharanAfrica as a whole, repeatersaccount for 16% of primary enrollments(23% in Francophoneand 8% in Anglophonecountries), and because of dropout,only 61% of thosewho enter the first grade reach the final grade of primaryeducation (median values -- see Annex Table A.12). As a result,it is estimated that the cost of each completer in the median country of Africa is 50% higher than it would be in the absenceof repetitionand dropout. In general,the situationis worse, the lower the incomeof the country. In 10 - 64 - out of 24 "low-ircomeeconomies" (GNP per capitabelow $400 in 1984),the cost per completeris more than doublewhat it would otherwisebe. Althoughit is useful to note these high costs per completer,it is essentialto bear in mind that even quantitative measuresof educationaloutput have a qualitative dimension: The numbersof repeaters(and, to some extent,dropot'ts) can changewith promotionstandards. Ultimatelythe efficiencyof these and other policiescan be assessedonly in terms of their impacton learninggains.

Dropoutsoccur for a varietyof reasonswhose relativeimportance varies among countriesand individuals.Students may drop out becausethey fail a grade or otherwisehave becomediscouraged of their chances for success,because opportunity costs of continuationbecome too high, or simply because of lack of opportunityto continuein systemswith large proportions of schoolsoffering only some of the primarygrades. Althoughsome benefits undoubtedlyaccrue to pupilswho drop out before completion,it can be assumed that these benefitsare small wheneverdropout occurs prior to the attainment of basic literacyand numeracy,at about the fourth grade. Beyondthat, there is littleevidence of any inefficiencyassociated with droppingout before completionof a cycle.

The extent to which repetitionshould be regardedas waste is a controversialpoint. The proponentsof repetitionclaim that repetitionis useful in that it remediesinadequate achievement and helps pupilswho are emotionallyand intellectuallyimmature when they enter school. On the other side, the criticsof repetitionclaim that achievementdepends principally on nonschoolfactors, that valid tests cannot be developedto separatefailures from promotees,that repetitiondoes not improvethe achievementof slow learners,and that by callingattention to their poor performance,it hurts the repeaters'self-image and their prospectsfor future success. Most imporra,-tin the developingcountry context, repetition extends the duration of study and, hence, raisesthe cost per year of school finallycompleted. To eliminateriePetition, these criticspropose that pupils be promoted automatic;ly from one grade to the next. The proponentsof repetitionoppose this policy,saying that it lowersacademic standards, destroys pupils' incentiveto learn and teachers'incentive to teach, and createspedagogical problemsby increasingthe abilityrange within the classroom.

Since one reason pupils repeatgrades, especially toward the end of the cycle, is (they believe)to better their chancesof passingthe examinationsfor entry into the next level of education,a useful strategyto reducedropouts involves separating the primaryschool leavingexamination from the secondaryschool entranceexamination. Under this arrangement, pupils take the school leavingexamination at the conclusionof the cycle and, if successful,receive a diplomacertifying completion of primary school. The secondaryschool entranceexamination, however, is taken independently. Individualstudents can do what they want to improvetheir chancesof success on the entranceexamination, e.g., throughindependent study or private tuition,but not by remainingin a governmentprimary school after completion.

Aajor reductionsin dropoutand repetitionrates could be accomplished,but they would requiresizeable, and costly, improvementsin the school and classroomfactors that cause pupils to repeator drop out. Thus, at least in the short term, the areas of repetitionand dropoutprobably offer littlescope for cost savings. - 65 -

(d) Appropriate construction standards

There is scope in many Africancountries for the developmentand use of new school designs which, wVile meeting minimum standards, are much cheaper than those used typicallyto date. The cost of facilitiesis now often a substantialfraction of the economiccost of providingprimary education. Annualizedcapital costs are, in some cases, the equivalentof 80% of annual recurrentcosts. Strategiesto increaseteacher utilization would be facilitatedin the futurethrough the choice of appropriateconstruction designs. In addition,appropriate standards would emphasizethe use of local materials,and this, in additionto reducingbuilding costs, would facilitate the ongoingtransfer of responsibilityfor primaryschool constructionand maintenancefrom centralgovernment to local communities.

In many cases,greater relianceon local materialsis also a way of improvingconstruction quality. For example,in Niger, the cost of a classroomconstructed in concreteis five times that of one constructedfrom "banco"(the most commonlyused constructionmaterial in rural areas). Yet the latter is cooler in summer and warmer in winter than the former.

Box 4.5 describesa projectrecently implemented in Senegalthat demonstratesthe feasibilityof low-costschool constructiontechniques on a large scale.

(Box 4.53

4.3 Mobilizingresources for primaryeducation

While there are some policiesby which countriescan irvprovethe efficiencyof resourceuse in primaryeducation; there is, when all is said and done, little likelihoodof reducingunit recurrentcosts at this level significantly,especially if countrieshope also to improvethe Qualityof education. There is more scope for reducingunit capitalcosts, but even so, it is unlikelythat overallper capitacosts can be reducedvery much at the primary level. The principalreason for the low qualityof primaryeducation in Africa is the comparativelylow resourceuse per pupil. In 1983, the median per pupil public expenditurewas under $50. By comparison,public spendingper pupil in East Asia and Latin Americawas approximately two-and-a-halftimes this amount,and in the industrializedcountries, over 30 times. Moreover,real per pupil expendituredeclined nearly 30% in Africa between 1970 and 1983 (Anne:'Table A.17).

Thus, furthergrowth of primaryeducation will require,inevitably, that additionalresources be mobilized. Even with no qualityimprovements (and assumingalso no improvementsin efficiency),the amountof resources devotedto the primary level would need to increaseby more than 3% annually just to keep pace with populationgrowtth. In principle,these additional resourcescan be made availableeither by reallocatingwithin the education sector in favor of the primary level or by increasingthe total allocationto the educationsector. The lattercould be achievedthrough an increasein public or private spending,or throughan increasein foreignaid for education. In the recentpast, only about 8% of foreignaid to the education sector in Africawent to the primarylevel. While this is partially understandablein an aid environmentemphasizing discrete project investments, the more program-orientedpattern of aid flows that is advocatedin Chapter 9 - 66 -

Box 4.5 LOW-COSTSCHOOL CONSTRUCTION IN FRANCOPHONEAFRICA

The high cost of classroomconstruction In Senegal In past years has stemmed from reliance on expensive Imported materials, designs, and techniquesInappropriate to the local setting, on Inefficientprocurement methods, and on the rapid deteriorationof facilitiescaused by insufficlent malntenance. Despiteefforts over the past 15 years to bring construction costs down, In 1985 the Senegalese government was still facing Initial Investmentcosts of around$300 per studentplace. However,a recent pilot project undertaken by the governmenthas broughtthe cost per studentplace down to $155.

Eighty-sixprimary school classrooms In the Kolda and Tambacounda areas were constructedunder this project. Costs were held down a result of a numberof strategies. The tenderingprocedure was based on a bidding process organized as a competition among partnershipsof architectsand contractors. The outcome was a replicable, low-cost, low-maintenance construction technology that Is labor-Intensiveand maximizesthe use of localmaterials. Fullyeighty percentof total constructionmaterials are locally available. The 'orelgn exchange component of this technology represents approximately28%, compared with about 52% for classical constructionmethods.

The project relies on a generalcontractor to providetransport, technicalknow-how, and skilled labor,and to train localunsk''led labor as masons and brick-layers.The additlonalIncome generated for the community from wage recelpts,at the prevailingwage of $4 per day, Is $1,800 per classroom constructed. Throughthe tralningof unskilledlabor on the job, local artisansand craftsmenwill be created,and a new reglonal employment market will be developed.Furthermore, the design'shigh thermalmass will counteractthe uncomfortabledaytime temperaturescommon In Senegal and, thus, providea betterenvironment for learning.

Similarprojects are currentlyunderway In Mall, BurkinaFaso, Nlger and the CentralAfrican Republic. Implementationof such pilot projects has permitted governments to gain experience In managing programs for the constructionof small buildingsIn largenumbers of widely scattered sites using a new technology.At the same time, educationfinancing studies have pointedout the difficultiesthat centralgovernments are likelyto face when they assume full responsiblityfor financing capital InvestmentIn the sector,particularly with respectto the construction of primary schools. Funds generated through local taxes are likely to play an Increasingly Importantrole In this regard. Thus, the task for central governments and their ministries of education will be to manage and monitorexisting educatlonfacilities, to provide Incentivesto regionalauthorities, and to insurethat minimumstandards are met. - 67 - of this paper would providea more generousshare to the primary level. Foreignaid for primaryeducation will be dealt with in Chapter9. The remainderof this chapterdiscusses domestic resource mobilization.

In many Africancountries, a good case can be made for graduallyincreasing the proportionof public educationbudgets allocated to primary education.21 This is so particularlyin the low primaryenrollment countries (LPECs),where the median share of educationspending going to the primary sub-sectoris only 34% (AnnexTable C.5). Here, an adjustmentin favor of primaryeducation wculd be both profitableto the nationaleconomy and equitable(expansion of primaryeducation benefits less advantaged populationgroups disproportionately). As the data in Table 4.2 clearly suggest,inadequate relative allocations within the educationsector may be an importantreason for the low enrollmentlevels of the LPECs.

(Table4.2]

However,although adjustment of sharesmay be desirableon efficiency groundsand, too, in terms of equity objectives,politics could stand in the way. Obviously,if primaryeducation's share is to be increased,the share going to somethingelse must be diminished. An economicallyattractive source of public funds would be the budget for living-expensesubsidies to students at the tertiarylevel. The effecthere could be significant. A study based on 12 African countriesfound that, on average,primary enrollments could increaseby a remarkable18% with the public funds saved if university studentspaid their own living expenses.

Increasedtotal publicspending on educationwould have to be financedeither throughincreased economic growth, maintaining constant the share of resourcesdevoted to education,or by an increasein education's share of the total budget. As regardsthe former,the growthprospects for sub-SaharanAfrica over the foreseeablefuture are modest at best. In its 1986 World DevelopmentReport, the World Bank examinedscenarios of average annual rates of GDP growth rangingfrom 3.2% ("low case")to 4.0% ("high case") for the low-incomeAfrican countries over the next ten years. If the 'low case' is the more accurateprojection, this means that these countries may suffer anotherdecade of fallingreal incomesper capita. Even in the 'high case," GDP growthwill just barely keep pace with the projected populationincrease. Thus, over the next decade,the additionalresources requiredfor doing more than simply maintainingpresent educational coverage and qualitywill not be derivedfrom economicgrowth.

In 1983, the countriesof sub-SaharanAfrica allocated15.3% of total public expendituresto educationin the median case (AnnexTable A.14). Whether this figureis on the high side, on the low, or about right dependson one's assessmentof the relativeprofitability of public spendingin different areas. There are, in any case, marked differencesbetween African countries in the prioritygiven to education. Whereasone out of five countries allocatedmore than 20% of public expenditure to educationin 1983, one out of three countriesallocated less than 13%. Countriesin the lower half of the distributionmight wish to reviewwhether educationis reallygetting its appropriateshare of total expenditure,or whether a higher budgetarypriority for the sectormight not be warranted,as was recommendedin Chapter 2. - 68 -

Table 4.2 Primary Enrollment Ratios and Primary Education's Share in the Education Budget

Median Percentage Median Gross of Recurrent Primary EducationBudget CountryGroups a/ EnrollmentRatio Devotedto Primary

Low primaryenrollment (LPECs) 37S 34%

Medium primaryenrollment (MPECs) 67% 45%

High primary enrollment(HPECs) 106Z 46%

All sub-SaharanAfrican countries 77% 43%

a/ LPECs - 10 countrieswhere the gross primaryenrollment ratio is below 50S. NPECs a 13 countrieswhere the ratio is between50% and 80%. HPECs - 16 countrieswhere the ratio is above 80%.

Notes Based on Annex Table C.5. - 69 -

Additionalresources to improvethe qualityof primaryeducation can be mobilizedthrough various strategies to recoverthe costs of the education from the recipientsof the serviceprovided. Educationalcosts may be recoveredthrough school fees or other monetarylevies, or they may be recoveredin kind, e.g., in the form of free labor for school constructionand maintenance. It is estimatedthat privateexpenditures now cover about 6% of costs in East Africa and 11% in West Africa (thesepercentages are only about one-thirdas high at the universitylevel).

A distincttrend in recentyears, particularlyin the Anglophone countries,but elsewherein the regionas well, has been the shift in major responsibilityfor the constructionand maintenanceof primary school classroomsfrom centralgovernment to parentsand local communities. Driven by the inadequacyof governmentfunds to supportthe sizablecapital budget for primaryeducation, many African governmentswere quite willing to accepta much increasedrole in this regard on the part of the recipientsof the educationalservices. This policyhas enabledprimary school development programsto proceedat a more rapid pace than would have been possible otherwise,and it has tendedto ensure that new schoolsare openedwhere the demand for primaryeducation is the strongest.

For items other than classroomconstruction and maintenance,private financingplays a less significantrole in Africanprimary education. This is as it shouldbe, given the likelihoodthat fees will resultin the very poorestAfrican children being deprivedof receivingany educationat all. RecentAfrican experiencesuggests that parentsare much less willing to pay for basic tuitionthan they are to pay for institutioiualmaterials; where tuition fees have been imposedin public primaryschools, some enrollment declineshave been experienced. Very few countriesanywhere else in the world charge for publicprimary education.

As furtherjustification for subsidizationof primaryeducation costs, there is increasingevidence of large externalitiesthat accrue to investmentat this level. A recentstudy of educationand productionin Nepal showed that farm yields depends,not only on the farmer'sown educational attainment,but also on the averageeducational attainment of the communityin which the farmer lives. A recent evaluationof Bank-financedproject-related trainingsuggests that the effectivenessof such trainingis high in countries where the adult literacyrate is greaterthan 50%, and very much smallerin countrieswhere the rate is less than 50%.

Althoughcharging primary pupils for instructionshould be generally discouraged,there may be situationswhere the judicioususe of modest fees might be used for the explicitpurpose of increasingaccountability in -_qtion.For example,a purchasefee or rentalcharge for textbooksand other materialsthat are crucialto high levelsof pupil achievementwculd help to ensure that these inputsare not eliminatedfrom the budget daring times of fiscalausterity. Another efficiencyenhancing mechanism would be for parentalgroups to "top up" primary-levelteachers' salaries in proportion to how much time teachersactually spend in the classroom,an incentiveto reduce academicabsenteeism. - 70 -

Privateschool enrollmentsaccount for about 62 of total primary enrollmentsin the medianAfrican country. Privateeducation plays a somewhat largerrole in AnglophoneAfrica (22Z of enrollments)than in Francophone (3%). A "privateschool," however, is a school that is managed by nongovernmentauthorities; the financingof privateeducation can come from public,as well as from privatesources. Indeedmuch of the private education that exists in Africa is at least partiallysubsidized by government. Hence, data on enrollmentsgive an exaggeratedidea of the level of privatefinancing of primaryeducation. Nonetheless,a greaterdegree of tolerancethan many African governmentshave shown in the past for the variousprivate alternativesto public educationat the primary levelwould almostcertainly facilitateexpansion. - 71 -

Chapter5. CONSOLIDATIONOF COMPETENCEs SECONDARYEDUCATION AND TRAINING

Secondaryeducation builds on the foundationprovided by primary education. Its goal is the consolidationof competence,particularly the strengtheningof general intellectualskills relevantto many occupationsand to furthereducation, so as to prepare individualsfor adult responsibilityand for the world of work, where the majorityof job-specificskills will be learned. Quality enhancementwas identifiedin Chapter4 as the immediategoal with respectto primaryeducation for most Africancountries. The qualityof secondaryeducation is also a major concernsince, as in the case of primary, the essentialnonsalary recurrent inputs are not being providedin adequate amounts. For ne&rly all African countriesin the years to come, however,the biggest challengewith respectto secondaryeducation will be expansion-- how to satisfythe burgeoningdemand for the now-limitednumber of secondaryschool places.

As recentlyas 1960, only two of the 39 countriesof sub-Saharan Africa enrolledmore than 6% of the relevantage group in secondaryeducation. These two were Mauritiusand Ghana,where the gross enrollmentratios in 1960 were 22% and 19% respectively.For the rest of region,secondary education remaineda highly selectiveenterprise.

Today, participationin secondaryeducation varies greatlyacross countriesin Africa. At one end of the spectrumare ten "low secondary enrollmentcountries' (LSECs) with gross secondaryenrollment ratios below 10% in 1983. There are sixteen 'middle"enrollment countries (MSECs) with ratios between 10% and 20%. Finally,at the top are thirteen"high" enrollment countries(HSECs) that have surpassedthe 20% mark, an impressive accomplishmentin the light of historicalprecedents.-I Three of the HSECs -- Mauritius,Zaire, and Congo -- alreadyhave secondaryenrollment ratios above 50%.

As would be expected,gross secondaryenrollment ratios are relatedto per capita income. Nine of the 13 HSECs are classifiedby the World Bank as 'middle-income'economies on the basis of.their1984 GNP per capita ($400 or more). Conversely,all ten of the LSECs are "low-income'(GNP per capitabelow $400). The 18 Francophonecountries are equallydistributed across the three secondaryenrollment groups. The 16 Anglophonecountries, on the other hand, are skewedtowards the high end of the enrollmentspectrum, with seven such countriesclassified here as HSECs,six as MSECs, and only three as LSECs.

Table 5.1 shows how the three countryenrtl1ment groups differ witi respectto other key secondaryeducation indicators. Those with higher enrollmentratios are not only richer,they also spend a greaterproportion of nationalincome on education(column b). The greateremphasis put on secondary educationhas Lme at the expenseof primaryeducation. Among the 13 HSECs, the proportionof public expenditureon educationallocated to secondary education(column c) ranges from a low of 16% to a high of 56% (median35%), whereasamong the 10 LSECs, this figureranges from 14% to 40% (median30%). Among the HSECs,the ratio of secondaryenrollments to primaryenrollments (columnd) rangedfrom a low of 0.13 to a high of 0.57 (median0.24); among the - 72 -

LSECs, the ratio rangedfrom just 0.02 to 0.26 (median0.10). The progression rate from the last grade of primaryto the first grade of generalsecondary education(column e) was only 54% for the median HSEC, comparedto only 16% for the median LSEC.

[Table5.1]

Even in countrieswhere secondaryparticipation rates are below the Africa-wideaverage, the growthof the secondaryeducation sub-secto_ in recent years has often been quite robust. Of the three major educationlevels, secondarywould appearto have been the least affectedby the recenteconomic depression. The rate of growthof secondaryplaces on the continent which had been 12.4% between 1960 and 1980, fell only to 10.9% after 1980, about a 12% decline in the growth rate. The growth rates of post-secondaryand primary enrollmentsfell much more precipitously,39% and 59% respectively(Appendix TablesA.1-A.4). Reflectingthis shift in favor of enrollmentsat the secondarylevel, public expenditureson secondaryeducation increased relative to expenditureson primaryand tertiaryeducation in a majorityof African countriesbetween 1980 and 1983 (AppendixTable A.16).

Politicalpressures have contributedto this reallocation.The earlierrapid expansionof first-leveleducation in African countriesresulted in an ever-largerpool of primaryschool leavers,many of whom were no longer findingjobs in the modernwage sector,which is what they had come to expect. The heightenedscramble to fill the limitednumber of secondaryschool places generatedpolitical pressures for expansion. Generouspublic subsidization furtherfueled the demand for secondaryschool places.

The case for the expansionof secondaryoducation, however, is more than just political. There is a strong economicrationale for expanding nationalsystems of secondaryeducation in most Africancountries, so long as hard measuresare taken to increasethe efficiencyand equity of these systems. Evidencewas reviewedin Chapter2 indicatingthat social ratesof returnto investmentat the secondarylevel are high by comparisonwith most alternatives in the region,including many educationalalternatives. While much of the researchto date has shown even higher rates of return for primary education (owing,at least in part, to the much lower per studentcosccs at the primary level),there is recent evidence,reviewed in Box 2.1, to suggestthat the rate-of-returndifferential between primary and secondaryeducation in Africa has narrowedconsiderably over time, increasingthe relativeeconomic attractivenessof the secondarylevel.

In assessingthe economicconsequences of substantialexpansion of secondaryeducation, it is interestingto comparethe recentexperiences of Kenya and Tanzania. Kenya has pursued a relativelyexpansionary policy with respectto secondaryeducation, whereas Tanzania's approach has been much more restrictive. Box 5.1 suggeststhat an importantpart of Kenya'smnuch more rapid growth in averagewage levelscan be attributedto its investmentsin the quantityand qualityof secondaryeducation.

(Box 5.1]

Secondaryeducation, in conclusion,will and shouldcontinue to expand in sub-SaharanAfrica for sound economicand politicalreasons. Prerequisite to the furtherexpansion of secondaryeducation will be, in today'seconomic - 73 -

Table 5.1 Enrollment Characteristics and Education Expenditure By Secondary Enrollment Groups a/

Public Secondaryas Gross expenditure percentof pubilc Ratio of ProgressionFeales as secondary oneducatIon recurrent secondaryto rate from percentof enrollment as a percent expenditureon prImary prl.,ary to seconosry ratlo of GNP education erollmentsseconw ary enrollments

(a) (b) (c) (d) (e) (f)

Lowsecontary enrol lent countries(LSECs) 5 3.3 30 .10 16S 32

MediWusecondary enrollcent cesntrles(MSEWB) 15 4.0 30 .21 45X 33

Highsecondary enrollment countrIes(HSECs) 24 5.3 35 .23 541 39

Sub-SaharanAfrica 16 3.9 31 .20 40X 33

Note:Entries are percentages forthe medlan country. Based on Annw Table C.7

a, ISECS:gross secondary enrollment ratlo below 10. bSECs:ratio bDtween101 and 201. HSECs:ratlo above20X. - 74 -

Box 5.1 POLICIESTOWARD SECONDARY EDUCATION

Major differences between educational pollees In Kenya and Tanzania,particularly with respectto secondaryeducation, emerged In the late 1960s. In Tanzania, the government reduced the share of public spendingallocated to secondaryschooling and, until quite recently, also restricted the formation of private and communitysecondary schools. In contrast,the Kenyan governmentencouraged the growth of secondary education In both the publicand privatesectors. By 1980 those divergentpolicies led to secondaryschool enrollment ratlos that were six time greater In Kenya than In Tanzania.

Important qualitative differences In secondary education also emergedduring this period. Tanzaniaplaced greater stress on the teaching of agriculture and other vocational skills,thus divertingtime from the teachingof generalacademic skills. The Tanzanian system also put great emphasis on the use of Swahiliat the primary level,perhaps making It more difficultfor studentsto learn In Englishat secondaryschool. As a measure of quality, research Indicatesthat for any given combinationof Inputsof Individualability and years of secondaryschooling In the two countries, cognitive outputs (I.e., scores on academic achievmont tests) are substantlallyhigher In Kenya than In Tanzania.

Since the quantityand qualityof secondaryeducation are greater In Kenya than In Tanzania,It was to be expectedthat over timo the cognitive skill levelof the averageemployee In Kenya would rise relativeto that of the average employee In Tanzania. In accordance with this predIctlon, resultsof cognitivetests revealthat Kenyan employeesare more literateand numerate than their Tanzanian counterparts,despite the fact that average scores of Kenyans and Tanzanians on tests of reasoning ability are essentially the same. Since analysishas shown that workerswho are more literateand numerateare more productivo,one may concludethat Tanzaniahas paid a price In output foregone by restrainingthe growthof secondary educationand reducingeducational quality.

One way to quantifythis price Is to estimatehow much greater the cognitiveskill and productivityof the Tanzanlanlabor force would be If the quantityand quality of education were Increased to the Kenyan level. Results show that a simultaneous Increase In quantityand qualitywould Increasethe cognitiveskills of Tanzanians by 31% and earnings by 13%. Roughly 40% of the currentdifference In mean wages betweenthe Kenyanand Tanzanianworkers can be accountedfor by the lowercognitive skill of the Tanzanian labor force. DlfferencesIn the qualityof educationaccount for more than half the differenceIn the skill level, and thus In the mean earningsof the laborforce.

This suggeststhat the opportunitycost to Tanzaniaof constralning the quantityand qualltyof secondaryeducation has been substantial. The divergencebetween Kenya and Tanzania In educationalpollcy In the late 1960s appearsto have been a major factor In their diverging economic performance since then. - 75 - climate,the rigorouscontainment of unit costs; Section5.1 indicatesthat, in sharp contrastto the situationat the primary level,far more is possiblewith cost containmentthan is being realized. As costs are containedand expansion can occur, it will be possibleto enrollmore femalesand other members of disadvantagedgroups into secondaryeducation; this is discussedin Section 5.2. Secondaryeducation has often been turned to for the provisionof occupation-specificskills, and Section5.3 discussesthe appropriatenessof this task for secondaryeducation, under differingcountry circumstances and in the contextof the full range of alternativesfor training. Section5.4 discussesissues of financingand resourcemobilization.

5.1 Meeting the demand for expansionby reducingunit costs

Secondaryeducation is often expersivein Africa, in both absoluteand relativeterms (AnnexTable A.18; also Table A.17). Ranging from below $50 (Ghanaand Guinea-Bissau)to above $700 (Coted'Ivoire and Tanzania),public recurrentexpenditure per secondary-levelstudent was $223 in the median African countryin 1983. This equalled62% of per capitaGNP in the median Africancountry; again, the rangewas substantial,from around 20Z in several countries(Ghana, Mauritius, Central African Republic, and Kenya) to well above 3002 in one (Tanzania).What is spent on each secondary-levelstudent in the median countrycould be used to educatefour additionalprimary pupils at current levelsof expenditure.

Further,public spendingper studenttends to be highest in those countrieswhere enrollmentratios are the lowest. Annex Table C.8 compares unit costs and some of the key determinants of this variable across the three secondary enrollment groups (low,medium, and high enrollment countries)T.O In both absoluteterms and with respectto per-studentexpenditure as a percentage of GNP per capita,the low secondaryenrollment countries (LSECs) spend more on secondaryeducation than do countriesin the two higher groups. But most strikingis the fact that secondaryper-student costs relativeto primary per-studentcosts are much higher in the LSECs than in the other two groups.In the typicalLSEC, expenditureper studentin secondaryeducation is nine-to-ten times that in primaryeducation, whereas a ratio just above three characterizes the rest of the region.

Given the tight limitson real public resourcesin Africa and the enormouscompeting demands of other educationlevels, especially primary, the key to satisfyingthe high demand for secondaryeducation in Africa,especially in the LSECs but in most higher enrollmentcountries as well, lies in greater cost-sharingin~ secondary education combined with substantialreduction in unit costs. The first approachis the subjectof the final sectionof this chapter. The second,treated here, inevitablyinvolves measures that save on the input of teachersand reducecapital costs.

Chapter4, on primaryeducation, discussed Issues of capitalcosts and also issuesof teachertraining, compensation, and utilization. It addressed the questionof whetherthere existsmuch scope for economiesin these potentiallyimportant areas. It concludedthat, in many countries,there probablyis scope for improvedefficiency in teachertraining and the use of lower-costconstruction standards, for modest downwardadjustments in average teacher salaries,and for modest increasesin averageclass size and teacher utilization. Overall,however, the potentialfor cost containmentin primary educationthrough improved policies seems limited. - 76 -

With one exception,this paper'sconclusions concerning unit cost containmentat the secondarylevel parallel those for primaryeducation, and there is no need to repeatall of the analysis. As with primaryeducation, differentapproaches to teachertraining (relying much more on in-service,and less on preservicemethods) and lower-costconstruction standards are likelyto prove cost-effective.

The exceptionconcerns capital and teacherutilization, which offer much greaterscope at the secondarylevel for unit cost reductionthan at the primary. The median student-teacherratio in primaryeducation in sub-Saharan Africa is now 39, whereas it is only 23 in secondary. There is no pedagogical rationalefor a differenceof this magnitudealthough, in some cases, theremay be a logisticalone. Policiesto increasestudent-teacher ratios at the secondarylevel could substantiallyreduce unit costs. The principalmeans that policymakersmay wish to considerfor raisingthis ratio include: larger classes;heavier teaching loads, obtai.. tble through double-shifting and extensionsof the school calendar;and multi- ratherthan single-subject specializationof teachers.

In brief,there is, at the secondarylevel, substantial potential in most countriesfor reducingunit costs by improvingefficiency within the existingsystem. African leaderswill need to implementfirm policiesin this regard. While a substantialpart of the "savings"will need to be redirected to enhancingthe provisionof recurrentinputs (the inadequacyof instructional materialsand consumablesupplies is also a problemat the secondarylevel, though usuallyless of a problemthan at the primary),some resourcesmay be left for expansion. Over and above such measures,however, and unlike the situationin primaryeducation, there is an added potentialfor achieving secondarylevel expansionthrough a dramaticreduction in unit costs, to perhapsone half their currentlevels. This possibilityis the subjectof the next two subsections,on distanceeducation programs, and on the transition from a systemof boardingto a system of day schools.

(a) Distanceeducation: self-studyschools and extramral programs

Given the extentof the unsatisfieddemand for secondaryeducation, and the fact that the pool of eligibleprimary school leavertgrows inexorably larger everyyear, policiesthat incrementallyincrease student-teacher ratios and facilitiesutilization rates and reduceunit capitalcosts will never resolvecompletely the shortageof second-levelplaces and eliminatethe politicalpressure on African governmentsto expandthe capacityof the system. To achievea quantumincrease in secondaryenrollments, and to do so without sufferinga commensurateincrease in total costs or a seriousdecline in educationquality, will require radicalchange in educationalpractice in sub-SaharanAfrica. New modes of educationare requiredto minimizethe dependenceof studentson face-to-facecontact with teachersfor the learning that they acquire.

Distanceeducation has been definedas "an educationalprocess in which a significantproportion of the teachingis conductedby someone removed in space and/ortime from the learner"(Perraton, 1984, p.4). As such, the use of textbooksin conventionalclassrooms is one of the simplest,and obviously most proven examplesof distanceeducation. Properlydesigned distance educationprojects combine the strengthsof differentmedia -- 'print for permanence,broadcasting for immediacy,face-to-face learning for individuality - 77 - and feedback"(Perraton, Block, Fryer,Spain, and Young, 1986, p. 6). The essenceof distanceeducation has to do with the recording,multiplication, and distributionof packagedteaching, using a small numberof teachersto produce set coursematerials available in print, throughbroadcasts, or via some other medium other than directcontact with the originalproducers of the materials. Economiesof scale, impossiblewhere teacher-studentratios are fixed,or nearly fixed,becomes possible under such a system.

At the primary level,printed materials and broadcastingserve, for the most part, to enrich and enhancewhat is providea.. , teachersin classrooms,and, as such, distancemethods nearlyalways representan add-on cost. At the secondarylevel and above,however, including pre-service trainingand, especially,in-service upgrading of primaryschool teachers, studentshave alreadyacquired the basic study skillsand maturitythat will permit them to make use of a varietyof educationalmedia, on their own, with a minimum of individttalizedsupervision. Beginning at this level,the potential of distanceeducation to reducecosts by substitutingfor the input of highly skilledteacher time makes this a very attractivealternative to conventional instructionfor low-incomecountries, including especially those of sub-Saharan Africa. Table 5.2 describestwo modes of distanceeducation that can be used to save on teachercosts -- "self-studyschools" and "extramuralprograms." It also contraststhese to the "enrichment"and "qualityenhancement" modes, which do not result in reducedcosts typically.

[Table5.2]

Distanceeducation io alreadyplaying a role in wideningthe opportunitiesfor secondaryschooling in a numberof African countries. SeveralAnglophone African countries,Zambia being a good example,have establishedgovernment or university-basedextramural programs. In several Francophonecountries, too, distance-teachingprograms have been set up, modelledon the French CentreNational de Tele-Enseignement.One motivation for the establishmentof such collegeshas been the desireto reducethe amount of foreignexchange spent on correspondencecourses supplied from Europe.

Often initially,the ilatendedtarget group for the distanceeducation programsconsisted of peoplewith primaryeducation, and perhaps incomplete secondaryeducation, employed in the formalwage sector and needing qualificationsfor promotion. Very soon, however,the programsbegan to attract,in additionto these establishedindividuals, very recentprimary school leaverswho had not succeededin gainingadmission to secondaryschool and were not yet gainfullyemployed. Becauseof the selectionprocess at work, the studentsenrolled in the distanceeducation programs were, on average,less academicallyable than those enrolledin the regularschools. Accordingly,in order to provide increasedsupport for such studentsand raise their chancesof passingnational examinations, some countriesmoved to supplementthe correspondencemethods by settingup study centersin which studentscould receive instructionby radio and work on their correspondencematerials, during scheduledtimes, under the guidanceof supervisors. In this fashion,a number of extramuralprograms, relying exclusively on correspondenceand broadcastingmethods, have grown into self-studyschools over time.

The classroomsupervisors in self-studysecondary schools are not usuallybecondary school terchers,though some may have qualifiedas primary school teachers. Typically,they lack the knowledgeto teach the students. - 78 -

Table 5.2 Three Modes of DistanceEducation

Wae Oascription Costsand Effects

1. Enrlduentand Radloor televislonbroadcasts are Theenrlduent use of radlo Is quite unurrunt provieldIn on or twosubJects for Inepnive but hasrelatively little between15 lrnutesard 2 1/2hors per lmpacton learning.More substantial week.Many coiztries, Including use,as in Interactiveradlo, Ethloplaand Kenya, use limited Increasescosts, but producesstrog amontsof timefor rldmnt. positiveeffects on mathemtics and Alternatively,more atstantlal use nationallanguage learning. sosetiescarries the sIn burdenof Enrldmentan eslancinentare InstructionIn a stbJsctto erane particularlysuitable at theprimary quallty (soeBox 4.2 on *Interactive level. radlo').

2. Self-study Studentsattend sdool but classesare Self-studyschools reduce cost by 20S soouls led byan older student or comnity to 30S(or ore If theytake advantap memer;radio or televisloncarries the of classroamsPace from regular m&inburden of classrocoInstructlon In schoolsafter hours). Theycan extend all suaJects.and students are expected accessby allowingfor verysnall to rely heavilyon their texbodks.The schoolsIn indIvidualcommitles, world'slargest eduatioal InstitutIon hIlchIs particularlyleportant for - theChlnese Televislon University - Increasingfemale access and for Is basedessentlally on this model, reducingreliance on boarding schools. emphsizingPost-secondary Instruction. Thecosts of televisionare hlgher So.too, Is thelilawl Correspondne thanthose assocIated with radio,and Col'eloat the secondarylevel (see Bao 5.2). the gainsIn learningoutcomes are mininal.

S. Extrarail Extraeuralprogras rely prlncipallyon ThePotential for oostsavinrs with Programs prlnt to conveyInstruction but, to the extrawal programsIs evengreater, extentfeasible, they fr entlysiPPiemnt naturally,than for self-study prInt with radioor TVbroadcasts and schools.The cost per student per tutorial meetings(which are particularly courseWay be expected to be In the IMwtantmwe postalsystems hinder rane of 10%to 25Xof conventinal exchangeof assiuents). TheBritIsh Instruction.Extramral programs have OpenUniversity Is perhaPsthe best- the addedadvantage of alloirng knrmextramural program today; the studentsto pursuetheir studIes hille U,niversityof Zablaoperates a small- employed.Their disadvantage is the scale(but lonestablIshed) extramural demndthey make on student motivation program,and Tanzanla had arkedaress (althoLghthlb Is argiblya god wIthcorrespondeence education for in- criterion for studentselection). serviceteacher tralning (see Bo 4.1). Extramuralprograms have their maln roles In higherand In-service eduatlon,but theyare alsoan option at the secondarylevel. - 79 -

They are, however,able to help studentswith the organizationof their studies and to advise them on how and where to seek furtherassistance. The Malawi CorrespondenceCollege provides an exampleof a distance-teachingprogram that was launchedto meet the needsof out-of-schoolindividuals and has sincebeen adapted to create an in-schoolalternative. Studentsin this programcan choose to study part-time_r full-time,relying heavily on radiobroadcasts and self study printedmaterials, but enjoyingthe benefitsof a classroom in which to study and a supervisorto overseetheir work (see Box 5.2).

[Box 5.2]

The existenceof extramuralprograms like the one in Zambiaand self-studyschools like the Malawi CorrespondenceCollege can be seen as ways of promotingequity in aducation,in that they providea second chance for those who would not otherwisehave accessto secondaryeducation. But the evidencesuggests, unsurprisingly, that this is a more difficultroute to secondaryqualifications than the conventionalone. Thosewho do well in their primary-schoolleaving examination now go&to regularschools. Thosewho do less well must settle for a method of study intrinsicallymore difficultto follow. If a countrywished to promotefull equitywhile at the same time xpanding secondaryeducation by takingadvantage of the low costs of distance education,then one might rather envisagea system in which all second-level studentsdid some of their study in a distanceeducation mode (preferablyin self-studyschools making full use of printedtextbooks, workbooks, and a marking;service), and all did some in regularschools, face-to-face with qualifiedteachers. Alternatively,students 'n regularschools might be charged some fraction-- 50%, say -- of the additionalcost of regularsc01ols asacompared with the cost of self-studyschools or extramuralprograms. The most likely patternof use, however,is for self-studyschools to serve smaller communitiesduring schoolhours and, in urban areas,to be used to allow eveningor weekend study (in existingschool facilities) for individuals alreadyemployed.

(b) Day schools

Transitionfrom a systemof boardingto a system of day secondary schoolsis anotherapproach to reducingboth the capitaland recurrentcosts of secondaryeducation. Althoughthe savingswould be less dramaticthan those achievedby moving to a full-fledgeddistance education system, they can neverthelessbe substantialin countrieswhere boardingschools are now prevalent. The advantagesin favorof small day schoolsthat cater for individualcomunities over largerboarding schools that draw studentsfrom greater distancesare strongerat the junior-secondarythan at the senior-secondarylevel. Especiallywhen Implementedin conjunctionwith distance education,which reducesthe need for large boardingcomplexes at any level,a systemof smallday schoolscan significantlyreduce the unit costs of secondaryeducation.

Boardingschools are sometimesjustified on nation-buildinggrounds in that they bring togetherstudents from differontregional and ethnic backgrounds. In addition,in terms of costs, they may offer economiesof scale as regardsteacher uuilisation, and the existenceof such schoolsclearly obviatesthe daily commutingcosts associatedwith day schools. Despitethese arguableadvantages, however, the housingand feedingexpenses in boarding schoolsare, in some Africancountries, as high as all of the instructional - 80 -

Box 5.2 AN ALTERNATIVEROUTE TO SECONDARYQUALIFICATIONS IN MAIAWI

In 1985the governmentof Malawl established an alternative to the formal secondaryeduNatlon system- the Malaw Correspnd Collego (MCC). It wascreated In responwto demandfor secondary school!ngthat couldnot bemAt by the forml system,where access was limited to 9 percentof primary scool graduates.MCC evolved gradually over a twenty-yearperiod as the goverrmentrecognized It as a reUatlvelylnexensive way to respondto the growingdemand for secndaryeducation. The recurent cost per student In a MCCstudy center Is lessthan oe-fifth of the recurrentcost of a studentIn a Governmentsecondary school, and the costper graduate Is slightly less than the cost per secondary schoolgraduate.

The systemIs basedon radIo, correspondence, andthe useof under-gIalI f led teachersworking with studentsIn speclal c-etnrgand in regularW=ry Schos at n1d,t. The only entrance requirementsare possesslonof a primaryschool-leaving certificate and paymentof a fee for correspondencematerials. More than 80% of MCCstudents enroll at the juniorsecondary level, and the rest at the senior seondarylevel. AlthoughMCC centers were originally deslgnedto serveworking youthson a part-timebasis, over 70% of all studentsare now studying In classroomsfor overfive hours a day.By 1985 MCC was enrolIlng more than 10,000 new students per year and providing 15 hours of radio programsper week,correspondence materials, and a markingservice for about19,000 active students.

MCCcenters are generally located In simplebul dings, often constructed by thkcommunity next to a primaryschool, and are trequentlyaccompanied bysimple housing facillties for studentsand teachers.In somcamse, canters make use of prlmarys,hool buildings that are availableduring the afternoonand eveninghours. Althoughteachers at thecenters may have only the primary teaching certificate,they have been selected by localsupervisors onthe basis of thelr sklIls and Interests. Centerteachers, who are paid throughlocal educatlonauthorities, are responsiblefor overall supervlilonof theclasss. MCCalso provides classes, In regularsecondary schools after hours. These classeare often taught by regular secondary school teachers, who receive supplementary payment through MCC. Overallpass rates for uCCcandidates on theNational Junior Certificate Examination have variedbetween 10% and 22%over the past few years. These pass rates are lowcompared with regular secondaryschools, but they are satisfactory In thelight of MCC'smuch lower admissions standards. Theprogram's attractiveness to students was enhanced when the Government madeavallable places at regularsenlor secondary schools for allMCC students passing a full JuniorCertificate In one sitting, andpraces at theUniverslty for all thosewho passed a full MalawiCertificate of EducatlonIn onesitting. Thispollcy has now changed somewhat, andMCC graduates have to competeon an equal footingwlth graduates of the regular secondary schools. - 81 - costs. Since in many African countriesthese extra costs are fully subsidized, boardingschools in fact shift to the public educationbudget the basic welfare costs of childrenthat familieshave an obligationto bear. The result is that governmentexpenditures per studenttend to be much higher in boardingschools than in day schools,by as much as three and a half times in a countrylike Somalia; with a giver,budget governmentcan ofler many fewerplaces. Table 5.3 shows a recentestimate of the relativecosts of day and boardingschools in Malsl.;and comparesthem both to the costs of the Malawi Correspondence College. Also, in additionto the monetarycosts, there can be serious social costs associatedwith young students,especially female students, living away from home.

[Table5.3]

From an equity perspective, there is, of course, the worry that exclusivereliance on day schoolsreduces the access^f rural families(and, conversely,increases the accessof urban families)to secondaryschooling. Th.s, the transitionfrom a systemof boardingto a systemof day schools becomes increasinglyfeasible in any countryover time as the networkof secondaryschools grows larger. Widespreadintroduction of small radio correspondenceschools would, of course, facilitateth4s transition. In places wb.re boardingfacilities are still regardedas necessaryfor reasonsof equity,the most equitabledecision of all is to chargeparents the full costs of boardingand feeding,with specialprovisions being made to reducethe fees impossdon academicallyable studentsfrom poor homes, since this policy generatesrevenue and makes possiblea largertotal numberof studentsin the system.

This sectionhas addressedthe questionof whether efficiencyat the secondarylevel can be increasedsufficiently both to make possiblethe necessarylevels of non-salaryrecurrent inputs (booksand supplies)and to achievesignificant capacity expansion. The conclusionis that substantial economiesare possible-- economiesin the operationof regularschools, economiesfrom reductionsin boarding,and most significant,economies from the creationof distanceeducation systems combining radio and correspondence techniquesthat would allow expansionof reasonablequality secondary education - 82 -

Table 5.3 Per StudentCosts of Day Schools,Boarding Schools, and Self-Study Schools for SecondaryEducation in Malawi

(a) (b) (c) (d=b+c)

Capital Annualized Annual Total annualcost cost capitalcost recurrentcost (K) (K/year) (K/year) (K/year) Index

Boarding schools 5,800 530 750 1,280 7.5

Day schools 3,880 360 204 564 3.3

Self-study s4hoolsa/ 1,180 114 56 170 1.0

Note: Cost figuresare in 1984 Kwacha. Calculationsdo not includethe opportunitycost of studenttime, but the boardingschool costs do includefood. Capitalcosts includeconstruction (depreciated over 30 years)and furnitureand equipment(7 years)and are annualizedusing a 7% discountrate.

a/ The Malawi CorrespondenceCollege. - 83 -

to many more communitiesthan can be reachedin anv other way for the saLle price. This leads to:

Recommendation3. Significantexpansion in secondaryand tertiary education,especially so as to reach far more femalesand more individualsfrom remote rural areas and disadvantagedsocial classes, will not be possiblein most African countrieswithout policies that substantiallyreduce unit costs. Fortunately,there existsat these levelssubstantial scope for reducingcosts. The key is to identifyand implementnew instructionalmodes that rely much more heavilythan in the past on the input of studenttime and motivationand that economize -.n expensivecapital inputs and teachertime. Increased availabilityand use of self-studyschools and extramut:alprograms, relying heavilyon varying combinationsof radio broadcastsan6 correspondence materials,can dramaticallyreduce costs by lowerivgthe requirementfor skilled teachersand by allowingthe extensionof educationalservices even to very small communities,thereby obviating the need for expensiveboarding facilities. Increasedprivatization can also be expectedto lower unit costs and enhance efficiencyin educationto the extentthat individualsare more cost-consciouswhen spendingtheir own funds than when making use of a "free" public service.

5.2 Issuesof equity: increasedparticipation of females

Beyondthis generalsupport for expandingaccess to secondary education,African nations should give seriousconsiderationi to policies designedto remedyexisting inequalities in s^hoolparticipation. Females comprisethe largestunderrepresented group.-

Althoughfemale enrollments have grown steadilyin proportionto totals in most African countriessince independence,this is not true in every country -- the female-maleratio has actuallyfallen in the three former Portuguesecolonies, and it has been virtuallystagnant in four other countries (Burundi,Rwanda, Togo, and Benin). Even where substantialprogreqs has been made, femaleenrollments continue to lag behindmale enrollments.4/

In regionalterms, the lag is no longerparticul-.i-y -ronounced at the first level -- girls comprise44% of primaryenrollments in Africa today (as comparedwith 43% in Asia and 48% in Latin America). The differencebetween male and femaleenrollments remains high, however,in post-primaryeducation. At the secondarylevel, femalescomprise 34% of enrollments(as comparedwith 39% in Asia and 50% in Latin America),and at the tertiary,only 21% (as comparedwith 33% and 45%). Moreover,for those femaleswho do manage to enroll,repetition and dropoutrates are somewhathigher for them in rmny African countriesthan for males, which is contraryto the situationin most other parts of the world.

The absenceof equal ..cess to educationalaervices is not just inequitable. It is also inefficientif those less able to benefitfrom the opportunitygain accessto educationahead of those more able. The fact that women in Africa spend, on the average,less of their lives in wage employment than do men may have convincedsome observersthat the return on investmentin girls'education is smallerthan in boys'. - 84 -

This conclusionignores, however, the major role of Africanwomen in self-employmentactivities, and it ignoresthe now proven links between mothers'education and the health and educabilityof their children,and associatedwith these, the connectionbetween education and reducednumbers of cildren (see Section 2.3). Many studieshaze shown femalesecondary education,and the final years of primaryeducation, to have the greatest impact on these householdproductivity variables. Moreover,a recentstudy that estimatesmonetary rates of returnto investmentin educationin Botswana reportsbigher rates for the educationof femalesthan for the educationof males, except in the most traditionalparts of the countrywhere women seldom enter the paid labor force.

Given these findingson the importantreturns to girls'education, specialefforts should be made, particularlyat the secondarylevel but extendingto the tertiary,to raise femaleenrollments. It may not be obvious, however,how policymakerscan influencethe relativeparticipation of females vis-a-vismales. Rarely in Africawould it be correctto attributethe lag in femaleenrollments explicitly to governmentpolicy, a notableexception being the common policy of expellingpregnant girls. Schoolplaces are, for the most part, open to both sexes.

To identifythe principalcauses of differencesin male and female participationrates, one needs to look primarilyat demand-sidefactors. Sometimeswhen the distancefrom home to school is very far, as in many rural areas,parents are afraid for their daughtersto walk alone, and a girl may be kept out of schoolunless there is someone,eepecially an older brother,to accompanyher. For this reason,too, the opportunitiesthat distanceeducation providesfor geographicaldispersion of secondaryeducational opportunity becomesan importantconsideration.

Moreover,when a choicemust be made betweensending a son or a daughterto school,African parents are like parentsin many other parts of the world -- most often they will send the son. Unfortunately,the important socialben fits derived from femaleeducation are not likely to have much impacton a family'sprivate investmentdecisions. In patrilinealsocieties, the opportunityccsts of a daughter'stime when she is attendingschool, and any other costs such as school fees, are borne by her parents,whereas the benefitsof her educationare thoughtto accrueprincipally lo her future husband'sfamily. Recently,in some areaswhere 'brideprice"has been paid traditionallyto the familyof a marriedwomen by the familyof her husbat:d. studieshave shown the emergenceof a positiverelationship between the bridepricecommanded and a woman's educationalattainment. Here is an excellentexample of how familieswill try to "internalize"the returnsto their educationalinvestments.

A key issue from the policyperspective is how to bring the private decisionsregarding the educationof girls in line with what is best for society as a whole. Nothing is likely to change unless explicit attention is given to this goal, and there are, indeed,policy measures that will increase the willingnessof familiesto allow daughtersto attend school.

Small, community-basedschools (whether relying on distanceeducation or not) will tend to attractgirls more readilythan larger schoolslocated in urban centersand at greateraverage distances from homes; althoughsmaller conventionalschools may imply higherunit costs,this is not necessarilythe - 85 - case, at least from a fiscalperspective, especially if largerschools tend to includeboarding facilities whereas smallerschools do not. Increasingthe number of female teachers may also attractmore girls, especiallyin Islamic areas. Governmentscan reducethe private costs of girls'education relative to boys' by, for example,providing girls with free books and other instructionalmaterials, charging them lower tuitionfees, or recoveringless of the cost of boardingand welfare servicesfrom girls' familiesthan from boys'. Finally,governments can providea transferpayment incentive in the form of a schoolmeal program. Such programshave been shown to be successful in increasingenrollments, especially among girls.

5.3 Trainingfor vocationalcompetence

Of centralimportance to a country'stotal factorproductivity and economicgrowth is the stock of human capitalembodied in the country'sworking population. Investmentin human capitalincludes, in additionto the learning of generalcognitive skills and maintenanceof good health,the acquisitionof specific,joS-related skills and the developmentof reliablework habits and positiveattitudes towards work of all kinds. All of these attributesare importantto an individual'ssuccessful integration into the labor market and lifetimeperformance at work.

There is little,if ar.y,disagreement with the propositionthat the teachingof generalcognitive skills such as reading,writing, mathematics, and scientificunderstanding is accomplishedmost efficientlywithin a formal school setting. Virtuallyevery societyon earth has developeda schoolsystem for the purposeof transferringthese generalskills, which are important,not only in their own right for individualself-fulfillment, bat also as the foundationupon which all occupation-specificattributes are subsequently built. Althoughthere is no disagreement,either, about the importanceof occupation-specificskills, positive attitudes to the world of work, and good work habits, there remainsdisagreement as to the most efficienttiming, methods, and venue to be used for impartingthese, more narrowlyvocational attributes. The salientquestions that policymakersmay ask includethe following: Is it most cost-effectivefor the individualto acquirethe necessaryjob-related skills when in school;or to do so after completionof schoolbut before taking a job; or to wait and learn the necessaryskills after securinga job? For individualsalready employed, should training be industry-based,i.e., in regionaltraining centers serving large numbersof workers,or firm-based? What is the optimalmix of theoreticalinstruction and practicalwork experience?

Differentcountries have successfullyadopted different strategies for transferringoccupation-specific skills to membersof the work force. The cost-effectivenessof differentskills-acquisition modes dependson the existence(or nonexistence)of many complementaryfactors; what works well in one countrymay be poorly suitedto another. Moreover,specific programs may requirea period of time in which to evolveand mature before they serve well the goals for which they were intended;in some instances,it may be more cost-effectiveto modify existinginstitutions rather than to inventnew one-

Finally,and impcrtantly,whereas certain broad-based occupational skills (e.g.,bookkeeping, typing, accounting) might be acquired cost-et'fectivelywithin schoolsbecause these skills are widely applicable throughoutthe economyand becausethe cost of teachingthem is comparatively - 86 - modest, the acquisitionof other,more specializedand technicalskills might be providedmore cheaplyand effectivelywithin firms,where the equipmentand expertiseare alreadypresent and where the identificationof needed skills is drivenby productionneeds. Betweenthese two extremes,there are examplesin many African countriesof industriesconsisting of many small firms none of which is sufficientlylarge or developedto provideeffective skills training; here the answer is likelyto be the establishmentof industrialtrainirng centers,typically operating under governmentcontrol but ideallyfinanced througha systemof taxing the beneficiariesin proportionto the benefits received. As a rule-of-thumb,schools should be asked to providegeneral skills,applicable in a wide range of jobs and householdsituations; training centerscan providemore specificskills, applicable within particular occupations;and firms are best at providingjob-specific skills.

From the early days of colonialismin Africa, there has been pressure to increasethe vocationalspecificity of the school curriculum. The rationale for vocationalizationat the secondarylevel rests on the observationthat, for most secondaryschool students,this will be their last exposureto formal education;it should therefore,according to the argument,provide the recipientswith specificvocational attributes demanded in the labormarket, therebypresenting them with betterpaying and more immediateemployment prospectsupon schoolcompletion. This rationaleis put forwardmost strongly in countrieswhere the industrialbase is still very small and where employers have abandonedtraditional apprenticeship practices -- i.e.,where the oCa-of-schoolpaths to the acquisitionof technicalskills seem few. In such circumscances,the formaleducation system, it 4s reasoned,has an obligation to impr.rtattitudes and specificjob skillsthat preparestudents for the workirg life they will soon confront. In the post-colonialera, many African governmentshave acceptedthis reasoning.

One approachto the vocationalizationof secondaryeducation has been to sft up a salf-containedstream of technical,ag-icultural, and commercial schoolsthat parallelsthe generalstream; this might be calledthe "specializedvocational school approach." A second approach("the diversified secondaryschool approach")has been to insertinto the core curriculuma range of vocationalor (as called in some descriptionsof diversifiedsecondary education)"pre-vocational" courses. Both approacheshave been tried extensivelyin Africa and have receivedsubstantial support from international donors,including the World Bank.

Rigorousevaluation of both modes of vocationaleducation has not been common in Africa. There is a clear need to examinehow differenttypes of generaland vocationallyspecific education have performedin relationto their costs, and to examinehow they have been sustainedand their functions continuedin recentyears under increasinglyaustere financial conditions. Such studies need to be conductedseveral years after the institutionshave been establishedand after they have ceasedbeing dependent(if they were to begin with) upon the supportof foreigndonors. Also, becauseof the time it takes for graduatesto find regularwork, the assessmentthrough tracer studies of the labormarket experienceof graduatesshould, ideally,not be conducted until some years after coursecompletion.

Studiesalong these lines are now being undertaken,some under the auspicesof the World Bank. Indeed,the Bank is planningto producea major policy paper on the subjectof vocationaleducationi and trainingbefore the end - 87 - of the 1980a. In the meantime,there are some few findingsthat can be xeportednow on the subject. These will be reportedin what follows,not with any intentto show that in-schoolprograms for teachingvocationally specific s;dlls should be avoidedby all countriesand under all circumstances,but simply to raisewarning flags that policymakersmay prudentlybear in mind when the introductionor expansionof such programsis being considered.

First, there is an apparenttendency, particularly in developing countries,to overstatethe need for pre-employmentoc_upation-specific skills acquisition. Most entry-leveljobs requirerelatively little in the way of specificskills, and even in tLose cases where the reverseis true, adequate trainingcan usuallybe provieedby the employerthrough on-the-job training, alone or in combinationwith classroominstruction in a varietyof ways. There is, of course,the problemof how, in very poor economies,to provideskills for an industrythat is only emergentor, as yet, nonexistent. This, however, is a classic "chicken and egg" problem. Just wishing for the existence of a modern, "high-tech"industry will not make it happen,nor unfortunatelywill the teachingof "high-tech"skills, in the absenceof many (other) pre-conditionsnecessary for the industry'sdevelopment.

Second,there is a tendencyto overstatethe power of schoolsto shape attitudesand behaviorin ways that would facilitatethe transl.tionfrom school to work. It is certainlytrue that certainbehavioral characteristics, such as punctuality,persistence, and the willi.agnessto acceptinstruction, are emphasizedin schools,and that the very same characteristicsare rewardedin the workplace.If schoolsdo not actuallyimpart these desiredbehaviors, at least they select for them. There is, however,scant evidenceto suggestthat training in preparationfor a particularoccupation will, in and of itself, predisposean individualto seek employmentin that occupation;the occupation must offer other (economic)incentives as well that attractthe individualto it. Indeed,the historyof Africaneducation is repletewith examplesof vocationalschool programs whose principalattraction can be shown to have been the "secondchance" they offeredfor studentsto re-enterthe academicstream and, thereby,qualify for high-payingjobs in occupationsother than those for which the vocationalschools were actuallydesigned to providetrainirg. Moreover,the attitudesthat are actuallylearned in school-basedvocational programsare not alwaysappropriate ones. For example,the credentialed graduatesof such programsmay possessunrealistic expectations concerning pay and status for entry-levelpersonnel.

Third, there is mountingevidence to show that, for many occupation-specificskills categories, the trainingprovided in schoolsis generally"less useful" (i.e.,the market demand for it is less strong)than trainingprovided on the job or in specializedtraining centers. When this is the case, the reasonsfor it are usuallyeasy to comprehend. Schoolshave difficultyin recruitingand retainingcompetent instructors, especially in those fieldswhere market demand is strongest;school administrators are constrainedgenerally from offeringdifferential compensation to instructorsin differentfields. Being separatefrom the industriesfor which they train, vocationalschools generally find it difficultto provide studentswith realisticwork experience. For the same reason,instructors find their own skills fallingquickly out-of-date. Equipment, too, tends to become technicallyobsolete, especially in times of fiscal austerity,which puts budgetarypressure on schoolsthat might not be felt in industry. Unlikemuch industry,schools are slow to adjustto changingmarket conditionsand change - 88 - the mix of skillsprovided. Even when good labormarket informationis transmittedto thosewho run trainingestablishments, curriculum changes are inherentlydifficult to effectwithin the formaleducation system.

Althoughbureaucratic inertia seems to characterizemost technicaland vocationalschool systemsin most developingcountries, this is not to say that it could not be overcome,given a differentset of incentivesfor school administrators.West Germany and Japan,both industrializedeconomies and extraordinarilysuccessful ones, rely on distinctlydifferent systems for the transferralof technicaland vocationalskills. Japan'sis a systemof quite broad generaleducation followed by heavy private (enterprise)investment in in-servicetraining, whereas Germanyputs considerablymore emphasison pre-employmenttraining within the formalschool system. Both systemsresult in trainedpeople who bring a high level of technicalcompetence to the lob. But both systemsare capablealso of adaptingswiftly to change,and herein probablylies the secret of their success.

Finally,policymakers should rememberthat most attemptsto vocationalizethe educationcurriculum, in Africa at least,have proven to be expensiverelative to the cost of providinggeneral education. Some of the differencein unit costs is attributableto the fact that many vocational educationinstitutions and programsin Africa have not been expanded sufficientlyto benefit from eventualscale economies. Beyondthis, however, vocationaleducation is (in most industrialfields and, to a much lesser extant,commercial fields as well) inherentlymore expensivethan general educationbecause of its greaterreliance on specializedequipment, parts, and consumablesupplies and its requirementfor smallerclasses.

At the secondarylevel, the cost of specializedvocational schools has been at leasttwice as much per studentin many Africancountries as the cost per student in generaleducation, and more than 15 times as much in some countries. A recentevaluation of the World Bank'sexperience with the alternativeapproach to vocationalization-- diversifiedr.secondary schools, which combinevocational and academicsubjects -- finds that unit costs,while not as high as in specializedvocational schools, are still higherunder a diversifiedcurriculum than un4er a generalcurriculum. In spite of the higher costs,however, the eoaploymentand earningsexperience of diversifiedsecondary school graduatesseems much the same as for thosewho graduatefrom the general curriculumstream, although the recentstudy did not follow studentsfor many years after graduation. If furtherresearch corroborates these findings,then on the face of it, it would seem that diversifiedsecondary schools are not worth their higher costs. This is a sad lessonfor the many Africancountries that investedin such programsand for the technicalexperts (oftenfrom the internationaldonor community)who advisedthem, but a lesson nonetheless,and one now to be heeded as policymakersconsider directions for the future.

Because of the high costs and spparentlack of vocationalrelevance of both these school-basedapproaches to skills acquisition,there is an urgent need to develop (throughincentive schemes and technicalassistance) industry's capacityto provideon-the-job training and other firm-basedskills development programs. Experiencewith these shouldbe evaluatedwith referenceto all of the alternatives,including (post-primary and post-secondary)industrial trainingcenters. - 89 -

While additionalinformation is being gatheredon the relative labor-marketoutcomes of differenttypes of educationand training, policymakersshould at leastbe awareof past differencesin per student costs between generaland vocationalsecondary education and recognizethat a policy of increasedvocationalization of the curriculumimplies, inevitably, that fewer studentscan be given access to education. Governmentsinterested in laying the groundworkfor a more technicallyoriented economy in the futuremay also wish to considerplacing heavy emphasison generalmathematics and scientificskills in the secondaryand post-secondarycurriculum. These tend to be relativelyinexpensive to provideand are likelyto provemore conducive to economicgrowth than an emphasison in-schoolvocational education.

At the same time, policymakersshould be exploringall of the macroeconomicpolicy instrumentsthat may help to raise the volume of training providedthroughout the economy. Out-of-schooltraining, directed at individualsalready in the labor force,has two very importantadvantages as comparedwith in-schoolvocational education. First, it has the potentialof reachingall workers -- those who have never attendedsecondary sch.ools as well as the minority (in most African countries)of workerswho have. Second,it has the potentialof continuingthroughout an individual'sentire working life, therebyallowing each individualto renew some skillsas these become "rusty", and to replace others as these become obsolete. The macroeconomic policy instrumentswhereby governmentcan encouragetraining for in-serviceworkers are many. Investmentcodes, wage structureregulations, tax write-offs,and apprenticeshipguidelines are examplesof the instrumentsthrough which governmentcan enhancethe incentivesfor firms to offer training,and for individdalsto seek it.

5.4 Financingsecondary education and training

Given the large unmet demand for secondaryschool places,a natural starti.ngpoint for promotingthe furthergrowth of the sub-sectoris to raiae student fees. Another, indirectway for governmentsto expand secondary enrollmentswithout increasingthe burden on the public budget is to encourage the developmentof privateschools, i.e., schoolsthat are managed by nongovernmentauthorities. This sectiondiscusses these two approachesto increasedcost sharingin secondaryeducation and concludeswith a discussion of financingarrangements for training.

(a) Cost-sharingin public secondaryschools

Currentlevels of subsidizationin secondaryeducation are high in sub-SaharanAfrica. This is especiallyso in the Francophonecountries where, in most cases, fees were never charged,but also in AnglophoneAfrica, where fees have declinedtypically as a percentageof full educationalcosts in the years since independence.

Obviously,the introductionor raisingof fees can be used to relieve the public financialburden of providingsecondary education. As a by-product, this policy reducesthe difficultadministrative chore that educational authoritiesnow face of having to rationa limitednumber of schoolplaces among a much largernumber of academicallyqualified applicants. Evidencefrom - 90 - countriesthat have recentlyraised studentfees indicatesthat the dropoutof those alreadyenrolled is smallerthan might be anticipated,and that the impacton overallenrollment tends to be nil, rs those who drop out are replacedby otherswishing to enroll.

It was concludedfrom a recentstudy in Malawi that the 1982 fee of Kwacha 101 (K30 for tuition and K71 for boarding),which representedcost recoveryof about 40% at the time, could be increasedsubstantially before this policywould eliminateall excess demand for places. The additionalplaces that could be financedby this change alone would be somewhere(depending on the exact price elasticityof demand assumed)between 5,000 and 11,000 students,or between20% and 50% of the secondaryenrollment at the time.

To the extentthat higher fees would discouragestudents with lesser academicability, and hence, lower probabilityof success,from enrollingin secondaryeducation, the policy of increasingfees would increaseefficiency within the sub-sector. There may be concern,however, about the policy's impacton equity,since the higher fees will tend to increasethe ratio of high-incometo low-incomeapplicants of ant given abilitylevel. This would simplyexacerbate the situationthat existstoday, wherein the substantial across-the-boardsubsidies that go to iecondaryeducation benefit high- and middle-incomegrouips disproportionately, since enrollmentsare alreadyskewed towardsthese groups.

The adverseequity effectsof ,he fees chargednow, and of any future increasein fees, can be offset by the provisionof scholarshipsfor talented low-incomestudents. It Is not inconsistentto endorsea generalpolicy that devolvesa higherproportion of secondaryeducation costs onto users,while advocatingthat the fees chargedlow-income students remain the same as before or be reduced.

Even in countrieswhere politicianshave taken a stand againstthe impositionof user fees in public educationas a matter of principle,it is still possiblefor governmentto pass on more of the c-sts currentlyshouldered by the miaistryof education. First, if care is taken to distinguishbetween the educationalservices and other (e.g.,welfare, boarding, recreational) servicesthat secondaryschools provide clients, it may be possibleto charge for the latterwithout violatingthe government'sprinciple of 'free"public education. Moreover,education authorities can rely, more so than in the past, on communityparticipation in the constructionand maintenanceof school buildings. That communitiesshould assume a major responsibilityfor the capitalcosts of primaryeducation is alreadywidely acceptedin Africa, and the processof transferringthis burden from centralgovernment to local authoritiesis alreadyquite far along in many countriesas was discussedin Chapter4. A parallelprocess at the secondaryeducation level is now under way in a number of countries(see Box 5.3 for an exampleof this in Zambia).

(Box 5.3]

Another cost-recoverymechanism that may deservemore use than given at present is to open secondaryschools in more populatedareas to evening students. Althoughthis systemwould serve fewer additionalstudents than a full secondshift in the afternoons,it has the merit of providingsecondary school accessto young people alreadyworking, whose fees can be used to give an incentiveto regular,day-time teachers to return in the evenings. - 91 -

lox 5.3 CUOIUNITY FINNUCING OF SECONDARI SCHOOLS IN ZNMBIA

Since IndrpendenceIn 1984, the governrnt of Zambia has worked to achleve baslc educatlonfor all chl dren.In stPort of this polIcy, dirIng the 19608the governmentphased out all school fees, and providedfroe board and IodgIngto all secondaryboarding schooI students. Revenue from mineral cvort salese thesecosts. Howeer,by the old-1970sZaebla faced a dramaticeconmnlc downturn, which ledto severe redctlms In public expenditures foreducatlon.

As pibIicallocatIons tosdcatIan declined over the past 10years or so, a rinber of actIons weretaken to pIck up the slack. In order to IncreaseAccess to secondary school places, the government in 1975began to enonuraga Indlvidais andorganizations toowrate private fee-paying schools. By the old-1980sthe governrentalso lnstItuted a policy that requiredall nmn-Zablansto pay tuition end boardIng fe8s toward the cot of their children's educatlon in the country,and armolmced thet Parent-TeacdAssociations muld be allowed to levy a boarding swplet on pwl Is to Improve chiIdren's dIets.

A key elent In aeoilizing nonl-publc resrces to sLpprtedcation was official encouWaeetfor the establ idhunt of Self-Help e01os drlrng the late 1970s. Cocalty reactlon- has bon positlve: bete 1981an 1984,82 Solf-lelp secondaryschels werecreated. As a reasit of the cash. terlal, ard Isbor contrIbutlon provided by coAl auntles. accessto secondary educatlom has thu beengreatly stpmnded. Moreovwr,private contribitlom havebeen ads In a variety of otlhr ws. For maple, gos comsnitles offer acdatlon for weeklyboarders In residents' huts tolle school dorsltorleo are beIngcostructed by the comsity. Otes cnstruct hWes for teachers, and In on casethe PTAhas paid rent to accodate tsacer. Further, pupi is In thse schools are levied to help f Innu project ad prorm.

In govrrut andgrant-ied schools, osmeducatlonal recurent sndlltures from the pib Ic budgthavwunoffIclallybeentraneferredto pWents or otlhr reciplents of licational services. Parentsare nw eting the costsof tebos, wolse boks, rulers, eathutlcal sets, erasersad otheruterials, nd schooluiform. For boardingschooi pup Is, parentspay for bedding,soap and other tolletries. Part alsoontrib to conr theoost of sJhitmemas entertalruentan sarts, chooldevelopmt ad sinteiwc, andprOdctlQn unlt activities.

hiIe suchprivate oontrlbutlomhave been essntlal to the contiriad provisIonof e catlon for Zoblan children,the gwvenuint'sreoponse has beonalvalent. Noguildellnes have been Issued regarding hlchItem schoolsnay charg parents for, andthe determintlon of levels of contributlons to variousfund haabEen left largply to schoolauthoritIes ard their PTAs. So2afear that the abse of governs mitewing eliht leadto fIn-clal olsrtent by unwrupulou8school heads and PTAs and argA that cl-out off lolal policy Is deoanded.In Ilghtof Zambla'scontinuing esonolc diffIcultIes andthe gremifhof the schl-ap populatlon,hewver, governmentreaponsibi Ilty for educationIs likely to decline ftrt~6r and private suWort to take anan inreasIrngly Importantrole In the financgngof edwatlon. - 92 -

(b) The role of privateschools

"Private"schools include those that are sponsoredby religiousor other special-interestgroups. schools run by labor organizationsor business enterprises,private proprietary schools, and fully or semi-autonomous communityschools. African governmentstook over the runningof m'ny private schoolsin the years soon after independence.Nevertheless, private-school studentsstill comprisedmore than 25% of all second-levelenrollments in 1970. By 1983 this numberhad decreasedto fewer than 15%. The usual practicein Africa has been to subjecteducation to strictcontrols, and these have tended to stifleprivate provision and also to prevent individualproviders from respondingto the changingneeds of their constituentpopulations.

The nature of governmentcontrol over privateeducation ranges from outrightprohibition in a few countries,to a long list of regulationscovering such things as the level of fees, qualificationof teachersand contentof curriculain most countries. On the one hand, it is arguedthat such regulationsare necessaryto protectfamilies from choosingunwittingly an inferior-qualityprivate-sector alternative to public schools,and on the other hand, that they are needed to preventthe developmentof a high-priced, superior-qualityeducational alternative, which only the rich can afford.

Whateverthe merits of these opposingarguments, many countriesseem willing today, in the light of the financialsituation that exists,to recons'dertheir policiesvis-a-vis private education. An easingof the restrictionsthat have been imposedwill encouragenew providersto enter the educaticnalmarketplace. The harambee(or self-help)secondary school movement in Kenya contributedimportantly to the rapid growthof secondaryenrollments in this countryduring the latterhalf of the 1960s and early 1970s (over 15% annual growth). By way of comparison,in neighboringTanzania the expansionof secondaryeducation was deliberatelycurtailed, and a recentstudy attributes at least some of the differencesin prasenteconomic vitality between these two countriesto the earlierdifferences in their policieswith regardto secondary education(Box 5.1, above).

It is often assumedthat privatizationin the African contextshould involvesome type of accreditationsystem imposedby government. Such systems, however,are difficultto implementand, indeed,are unreliablein that they rely on externalsurrogates for educationalquality such as teachers' qualifications,student-teacher ratios, and schoolbuilding stanfards. National or regionalexaminations can be used as powerfulvehicles for shapingcurricula and for promoting(and not simplymonitoring) quality in education.Results on examinationsconstitute a signal zo consumersthat they can use as a basis for choosingamong educationalalternatives. When alternativesare available, schoolscannot take studentsfor granted. Competitionfor studentswill tend to encourageexperimentation in publicand privateschools alike, and the result shouldbe increasedefficiency in the educationsystem.

To minimizedifferences in the privateeducational costs encountered by householdssending children to differentkinds of schools,governments may choose to subsidizepartially private education to the same (presumably reduced)extent that it does public education. In many African countries today, there exists a wide range of financialarrangements, from the public schools,where subsidiesare usuallythe highest,to partiallyaided private schools,to completely "unaided" private schools. - 93 -

(c) Financingtraining

In general,consideration should be given to structuresfor governance (ownership)of vocationaltraining that parallelsthe patternsused to finance it. Where trainingis supportedthrough taxes and other levieson employing enterprises,these enterprisesshould have a major role in the managementof expenditures. Frequently,this will mean ownershipand managementof training institutionsby individuallarge enterprisesor by associationsof employers with common skills needs. A similarmodel can be appliedto public sector institutionssufficiently large to justifyownership and managementof specializedtraining centers and schools.

Relaxed regulationof proprietaryschools and centersfor teaching vocationallyspecific skills is a second option. Again,while proprietary schoolshave functionedwell in other contexts,it is not clear how well they might functionin many African situations. In the servicessector, where capitalizationcosts are relativelylow, such policiesmay be more productive than in trainingfor industry. In agriculture,where consumerincomes are relativelylow, subsidizedtraining managed by public agenciesmay continueto be required.

Where proprietaryschools generate revenues through the marketingof trainingservices, management of effectiveinstitutions is heavily influenced by consumerpreferences as informedby governmentaccreditation. Institutions that fail to meet qualitystandards or cannotprovide trainingat acceptable priceswill not long survive. Such structuresincrease the responsivenessof trainingcurricula and costs to employmentdemand, while structuresthat financeand govern independentof demand for skillsare generallyless responsive,thereby loweringtraining efficiency.

Public financingand managementof vocationaltraining institutions may continueto be t.,emost feasibleoption in economiesor skill areaswhere privateenterprises are relativelysmall and lackingin managementand training capacity. In such cases ecoz.omiesof scale may be realizedthrough training institutionsserving a numberof firms,particularly when schoolsand training centers are locatedin close proximityto employers. Institutional arrangementsthat give small employersa voice in the curriculaand management of these traininginstitutions in returnfor even modest financial contributionsshould be considered,both to increaseimmediate responsiveness of trainingto demandand to provide a startingpoint for eventuallystronger roles for enterprises.

Policiesthat alter the structureof trainingcoordination require investmentsin implementation.Transferring ownership of specializedsecondary vocationalschools from the ministryof educationto anotherorganization, for example,involves transition costs for the establishmentof new coordinating proceduresand units, and for the implementationtraining required both in schools and in the new structuralhome. Some of these costs representa shift of costs from one organizationto another;others are start-upcosts that will disappearbeyond the initialperiod. Cost savingsmay developeventually as the new systembecomes increasinglyefficient. - 94 -

Chapter 6. PREPARATIONFOR RESPONSIBILITY:HIGHER EDUCATION

Higher educationis of paramountimportance in the processof African development. High-levelmanpower must be trained and qualityresearch carried out if developmentpolicies are to be correctlyformulated, programs Appropriatelyplanned, and projectseffectively implemented. Preparationof and support for those in positionsof responsibility-- in government,in business,and in the professions-- is the centraland essentialrole of the continent'suniversities.

The leadershipof the region'sinsti'.utions of higher learninghave spoken fo.rcefullyand eloquentlyat two importantconferences -- in Mbabane in 1985 and Harare in 1987 -- about the urgent need for African universitiesto produce graduateswho can tacklethe complexproblems that confrontthe continent. Pronouncementsat thesemeetings were consistentwith the findings of the Fifth Conferenceof Ministersof Educationand Those Responsiblefor EconomicPlanning in AfricanMember Statas (Harare,1982) and are in line with the aspirationsand African-articulatedpolicy frameworkenunciated in the Lagos Plan of Action,elaborated in Africa'sPriority Programme for Economic Recovery (1986-1990),and reiteratedin the U.N. Programmeof Action for African EconomicRecovery and Development(1986-1990).

Universityprograms of researchand teachingthat supportthe rehabilitationand furtherdevelopment of the agriculturalsector are of great importance. Many appliedareas, such as soil and water conservationand drought and desertificationcontrol, and the naturalscience and engineering disciplinesupon which progressin all such appliedareas dependsare central to the task of strengtheningAfrica's agricultural base, and they demand immediateattention. Of equal Importanceis the contributionof institutions of higher learningto public administration,the governments'capacity to plan and directthe processof development. For this, they must producehigh qualitymanpower in the social scienceand managementdisciplines, conduct timely rRsearch,and provideadvisory services in such fieldsas economic planning,finance (including debt management),and public administration.

Despitethe centralityof these tasks to Africa's future,they cannot be successfullyaccomplished unless fundamentalchanges are made in higher education-- changesthat will dramaticallyimprove quality. If African universitiesare to provideworld-class graduates, research, and other essentialservices in responseto the demandsof their modernizingsocieties -- if they are to producethe researchscientists and =iversity teacherswho can interpretthe latestscience and technologyand harness it for the purposesof Africa's survivaland futuredevelopment -- then resourcesfor higher education must be increased,and they must be used more productively.

This paper contendsthat these goals (expansionof resourcesand increasedefficiency in the utilizationof resources,so as to enhancethe quality of higher education)can be met despitelimited economic prospects and unremittingneed for public austerity. It also suggeststhat suxecessful implementationof policiesto improvehigher educationmay eventuallyhelp eaae - 95 - the resource constraintto educationaldevelopment at the lower levels. Adjustmentof higher educationin Africa to better serve the changing imperativesfor developmentneed not be at the expen:-eof priorityobjectives in other subsectorsbut can be in supportof them.

6.1 The challenge and the promise

The developmentchallenge posed in tertiaryeducation is in one importantrespect more dauntingthan for lower levelsof the system. Real growthof public resourcesfor the educationsector as a whole In most countriesis unlikelyto keep pace with growth of school age populatioz,s, implyinga decreasein real public per capita educationexpenditure. Even with vigorousapplication of recommendedreforms at primaryand secondarylevels, these lower levelsof the educationsystem will requireadditional resources just to keep enrollmentrates and qualitystandards from deterioratingfrom their currentmodest levels. But if public resourcesfor educationare at best constantin real terms,where will additionalresources for primary and secondaryschooling be found?

A portionof those resourcescan appropriatelybe suppliedby the privatebeneficiaries of educationand their families,not the public purse. It is probably inevitablethat parents'contributions to the costs of primary education,and particularlysecondary education, will increase,despite very real concernsabout the impactof this on overallequity and efficiency. But, especiallyin the low and medium primaryenrollment rate countries,where expansionof access to lower levels educationremains an urgentpriority, all of this may fail to ensurethe increasedstocks of human capitalrequired for continueddevelopment.

The conclusionis both harsh and inescapable:To meet minimally acceptabletargets for coverageand qualityof lower levelsof educationin most countries,as a generalrule the tertiarysub-sector's share of stagnant real public educationexpenditures cannot expand further,and in some cases may have to contract. Some combinationof efficiencyimprovements, increased privatecontribution to costs,and constrainedgrowth of (in some countriesand fields,outright cutbacks in) productionof graduatesmust be sought. Even more daunting,total savingsfrom such measures,plus resources mobilizedfrom the internationalcommunity, must generallybe sufficientto financethe indispensablequality improvementsneeded in the tertiarysvbsector in the shert-run,as well as its expansionin the long run. In some cases,such as the Saheliancountries where the crisis in coverageat the lwger leveis is most acute,these savingsmay need to be substantialenough to free iw some funds for these levels.

It will take years of dogged determinationand profoundchanges in the way higher educationis organized,managed, and financedto achievethe neededsavings. In particular,unless tertiarysector managers are presented with compellingincentives to Improveefficiency, impose quantitative limits on enrollments,and mobilizeresources from the privatesector, the required savingswill not materialize. The most powerfulincentive is to ensure that savingsrealized in any specificinstitution through wise and courageous leadershipare in significantmeasure availablefor redeploymentto quality improvementsin the same institution. Conversely,policies that expropriate savingsobtained by aacrificein individualinstitutions and transferthem to a centralauthority for redistributionwill be self-defeatingbecause they - 96 -

destroythe incentiveto look for savings. Until those savingsare obtained, the limitson what can be achievedin terms of -ualityat the tertiarylevel and of coverageand qualityat the primary and secondarylevels will be tighter,and the long-termprospects for developmentin sub-SaharanAfrica more constrained,than need be.

The challenge,then, is to design and implementwith persistence, over many years and in circumstancesof severe austerity,policies that generatethe resourcesneeded for the revitalizationof the subsector. If the challengecan be met, these policiespromise to enhancethe severalcritical contributionsof higher educationto nationaldevelopment.

In broad terms, tertiaryeducation makes three such contributions. First, tertiaryinstitutions prepare the versonnelneeded to fill high-level scientific,technical, professional, and managerialjobs -- i.e., they educate the elite leadershipof a nation'sdevelopment effort. Of special importance here is the preparationof teachers,scholars, and managersfor the education sector itself,espacially for its most advancedteaching and research functions. These people are the indispensablecore of nationalcapacity for sustainedproduction of sound scientific,professional and technicalmanpower and for settingstandards, maintaining quality, and adjustingthe education system to changingcircumstances. They are the leadersin Africa'sbattle againstintellectual colonization with which the worldwideexplosion of knowledgethreatens the region. So it is that, second,African countries look to higher educationinstitutions to generatethe knowledgeand innovation needed for development,tarough indigenousscientific research and as agents for the acquisition,adaptation, and disseminationof scientificand technical knowledgedeveloped elsewhere. Third,African universities as institutionsand their facultiesas individualscan providenecessary services needed for developmentir. both the public and privatesectors.

Beyond these major responsibilities,tertiary education in Africa as elsewhereis also a sourceof analyticalperspective on societalproblems and their possiblesolutions that is independentof and, often, a usefully pluralisticcounterpoint to, politicaland religiousauthorities. Higher educationinstitutions also provide a mechanismfor indigenousself-expression, help to conserveand adapt local traditionsand values,and constitute importantsymbols of nationalprestige and attainment.

6.2 Issues in higher education

At leas,.with respectto the preparationof high-levelpersonnel, tertiaryeducation's contribution to sub-SaharanAfrican developmentsince independencehas been remarkable. The rapid growthof universitiesand enrollmentswas sketchedearlier (Chapter1). In 1960, tertiaryinstitutions in Africa graduatedabout 1.2 thousanddegree and non-degreeholders, equivalentto one person trainedat that level for each 168,000Inhabitants; the 70.6 thousandgraduates in 1983 representeda ratio of one per 5,800 inhabitants.

With this notableeasing of the high-levelmanpower constraint, however,the relativeimportance of tertiaryeducation's other functionshas increased. Today, higher education'scontinuing contribution to developmentis - 97 - threatenedby four complexlyinterrelated weaknesses. First, the mix of outputsof the higher educationsystem is no longerwell suitedto the requirementsfor development. Second,the qualityof those outputsshows signs in many instancesof having deterioratedso far as to cast doubt upon their fundamentaleffectiveness. Third, their costs of productionare needlessly high (wherecost is measured as the output forgonedue to not applying elsewherein the educational,or economic,system the resourcescurrently going to higher education). Finally,the patternof finaacinztheir productionis sociallyinequitable and economicallyinefficient.'-

(a) Inappropriateoutput mix

High level manpower. Of the three major outputsof tertiary education-- high level manpower (includingscholars and teachersfor the educgtionsector itself),knowledge and innovation(research), and development advisoryservices -- sub-SaharanAfrica today generallyproduces relatively too much of the first,and not enough of the second and third.

The proportionof tertiarygraduates in African populations24 years and older (now about 0.4%) is still small by comparisonwith other developing regions (on averageperhaps 6%). The proportionof scientific,technical, professionaland, especially,education sector positions filled by expatriates remainssubstantial in many countries. Over the long run, Africa can survive and prosperonly by maximum exploitationof its major resource,people, includingespecially those with skills acquiredin tertiaryeducation.

Nevertheless,short- to medium-termoverproduction of high-level manpower-- at least of the requisitequality -- is suggestedby the growing problemof graduateun- and under-employment.Since reliablefigures are not generallyavailable, the assertionis largelybased on the overwhelmingweight of recenttestimony from officialsof Africanministries and universities. Other considerations(like physicalcapital stock) aside, the incidenceand size of surplusesof manpower trainedat the tertiarylevel are likely to be greater in countriessuch as Guinea,Lesotho, Swaziland, Congo or Gabon where about one of every 2,000 inhabitantswas awardeda degree from tertiary institutionsin 1983 than in Mozambique,Tanzania or Burundiwhere the rate of productionof tertiarygraduates in recentyears was only about one per 17,000 inhabitants.

Data problemsnotwithstanding, long and lengtheningperiods of job searchhave been reportedamong graduatesin Zaire and Nigeria. In Kenya, a 1985 tracer study of Universityof Nairobigraduates between 1970 and 1983 documentedperiods of job search from one to three years and in additionfound a rise in temporaryemployment and a decreasein utilizationof graduates' formaltraining on the job. Manpowerforecasts for Lesotho show an excess of graduates,particularly outside the teachingprofession. In Somalia demandfor tertiarygraduates around 1990 is estimatedat 150 annuallywhile anticipated output of Somalianinstitutions, net of studentsreturning from abroad,is more than five times that amount. Fewer than 15% of the approximately1100 graduatesof tertiaryinstitutions in Mali in 1986 can expect to find employmentin the public sector,and privatesector opportunities are certainly no greater;the conclusionis that at most 30% of the graduateswill find work appropriateto their level of training. If the base for these figureswere to includeMalians returningfrom trainingabroad, the picturewould be even - 98 - darker. In Guinea,a 40% reductionin higher educationenrollments was deemed necessaryto bring annualproduction of graduatesin line with absorptive capacityof the public service,essentially the only employerof tertiary manpower.

What accountsfor this situation,now an essentiallyuniversal phenomenonon the continent? Severalhypotheses can be offered. Each is worthy of rigoroustesting in the light of country-specificcircumstances.

To some extent,the apparentsurpluses represent not so much overproductionas a general failureby Africancountries to nurtureconditions in which individualswith tertiaryeducation can be productive. Political philosophiesand developmentmodels that place primary relianceon the reasonablyunfettered interplay of market forces,internally and internationally,generate by their very nature incentivestructures that afford a premium to skillsacquired in tertiaryeducation. As Africaneconomies succeed in their efforts at liberalizationthrough structural adjustment, their abilityto get the most out of tertiarylevel manpowerwill be enhanced. However inevitableit may ultimatelybe, this increasein the capacityof African economiesto utilize tertiarymanpower productively is unlikelyto be noticeablebefore the end of the century. If recent experienceis any guide (see Box 6.1 for a case in point),structural adjustment in the short-run exacerbatesrather than amelioratesthe problemof graduateunemployment, especiallysince aspirationsand expectatioansof tertiarystudents with respect to occupationalstatus and wages are much slower to respondto new economic parametersthan the demand for the high level skills they embody.

[Box 6.1]

The more immediateexplanation for the surplus is that, on the one hand, the demand for graduatescontracted suddenly in both the public and privatesectors. The economicdownturn of the 1980s and the pressureto containpublic employmentarrived just as African civil services,almost everywherethe largestemployer of tertiarylevel manpower,had largely completedthe processof replacingexpatriates with newly qualifiednationals and were consequentlyreducing their hiring rates. Beyond reducingpublic sector demand for graduates,other aspectsof adjustmentto the downturn --includingthe retreatfrom importsubstitution industrialization, the shift away from nontradables(construction, banking, insurance), and sharperscrutiny of staffinglevels following privatization of parastatalsmay also have slowed privatesector demand for high level manpower,which in any event was often much less significantthan demandfrom the public sector.

On the other hand, the economicdownturn coincided with cominginto productionof the remarkablecapacity expansion program for tertiaryeducation which was undertakenin the 1970s. Between1980 and 1983, enrollmentsin African institutionsof higher educationincreased 30% (from 337 thousandto 437 thousand,not includingthe additional100,000 African students enrolled in foreignuniversities) and graduatesincreased 70% (from 41.5 thousandto 70.6 thousand). In short, the supplyof graduatemanpower mushroomed.

In addition,the emergingimbalance between supply and demandwas exacerbatedby structuralrigidities in the educationsystem. Indirect evidencesuggests excess supplyof graduatesin arts, humanities,and soft socialsciences is not an especiallynew problem. In 1960, reflectingthe need - 99 -

Box 6.1 UTILIZATION AND EARNINGS OF UNIVERSITY AND TECHNICAL GRADUATESIN COTED'IVOIRE

Between 1982 and 1984,Cote d'Ivoireexperienced a severe recession,reducing the GNP per capitaby almost50X. As a result, many jobs were lost, especiallyin the modernprivate sector. The higher educationsystem was not able to adjust sufficientlyto accommodate the changingdemand for its graduates. A 1985 survey in Cote d'Ivoirerevealed that 37.52 of peoplewith university diplomas in Abidjan and more than half of such people in other citieswere economicallyinactive or unemployed. More than half of them had been in that situationcontinuously for the full year prior to the survey. Among people with vocationaland technical diplomas, the situation was only marginally better: 33X of them were withoutjobs in Abidjan,and 262 in other cities. The reducedability of the formal sector -- the traditionaloutlet for university and technical graduates-- to absorbsuch graduates, especially in Abidjan, has indtucedabout 101 of them to take up work in the informalsector, mainly as self-employedentrepreneurs.

For those graduateswho do find Jobs in the formal sector, the private returns to their educationare still quite high. Each year of post-secondaryschooling adds an average of 171 to an employee's earnings. This figureis higher in Abidjan (191) than in other cities (15%). For technical and vocational education, the increase in earnings is about 9X per year of schooling,with no differencebetween Abidjan and other cities. Such elevated returns to education in combination with high unemployment among university and technicalgraduates are signs of an imperfectlyfunctioning labor market. In particular,it appearsthat as a resultof the recession, the number of jobs has been reducedwhile the salarylevels of those workers still employedhave been maintained.

As for graduateswho become self-employedinformal sector workers,their earningsare sometimeseven higherthan the earnings of their wage-employed peers. For example, in Abidjan, a self-employedworker with a vocational or technical diploma has earnings almost 50S higher than a governmentworker with the same diplomaon the average. Also, those earnings are more than three times the earnings of a self-employedperson with only a primary school diploma.

These findingssuggest the importancein the short-run of adjusting university output to levels that the labor market can absorb. At the same time, they point to adjustment mechanisms that will eventually allow full (or fuller) utilization of university graduatesas the economyrecovers and as labor market performance improves. The high productivityof graduatesworking in the informal sector,in particular,indicates that (even thoughnew graduates may have to adjustoccupational expectations downwards) college education is of broad vocationalvalue. - 100 - for indigenous administrative personnel and the less onerous requirements in terms of costly facilitiesof establishingprograms in the liberalarts than in the sciencesand engineering,602 of enrollmentsin African institutionsof higher educationwere in the arts and humanitiesand 40% in the sciencesand engineering. Today,Africanization of public servicepositions and (at least in some places)of managementpositions in the privatesector is quite advanced,so demand for arts and humanitiesgraduates is less robust. By contrast,demand for high levelmarpower in scientificand engineering professionsis, in most places,much more substantial. Yet the 60:40 ratio is the same. Enrollmentpatterns have not respondedto shiftinglabor market demands,a rigiditywhose effectis exacerbatedby lower proportionsof studentsactually graduating in sciencesand engineeringthan in the arts and humanities.

The recentappearance in a few countriesof significantun- and under-employmentamong graduatesof technicalprograms in engineering, agriculturalsciences, physical and life suiencessuggests that beyond structuralrigidities in enrollmentpatterns, the overallsize of the system in some countriesmay have exceededwhat can currentlybe justifiedby requirementsof economicgrowth. The continuingreliance on expatriatesto staff scJtentificand technicalpositions, especially in the educationsector itself,suggests that inadequatequality may also impedeabsorption into the labor force of graduatesof Africanuniversities.

Finally,public subsidizationof highereducation has so far kept its direct costs very close to zero for the individualstudent. Yet, as Tables 6.1 and 6.2 make clear,wage differentialsto tertiarytrained labor remain substantial. As a consequence,the private rate-of-returnto tertiary educationis on the order of 302, higherthan for any other regionof the world, and shows no sign yet of decline. Growth in the numberof secondary school leaverswith expectationsof acquiringhigher education,and the extraordinaryprivate returnsit has thus far conferred,has not slowed. The overall result is a paradoxof more graduateathan the economycan usefully absorb in many fieldscombined with undiminishedpressure for expansionof higher educationopportunities despite the reduced,and falling,probability of obtaininga well paying job once a diplomais in hand.

(Tables6.1 and 6.2]

Educationsector leadership.A number of specificscientific and engineeringcareers manifest excess demand for graduates,and may thus require focusedcapacity expansion. But the major exceptionto the generalrule of overproductionof graduatesis to be found in the educationsector itself. For 18 countriesfor which data are available,expatriates constituted on average 352 of secondaryschool teachers at the end of the 1970s (see Annex Table C.9); the proportionis much higher in some fields (e.g.,science, mathematics and technicaleducation). Despitesizeable staff developmentefforts, including programsof foreigngraduate training, many tertiaryinstitutions are still criticallyshort of African staff,especially in the senior academicpositions of teachingand researchleadership. In Nigeria,further ahead than most countriesin the Africanizationof staff, about 202 of teachingpositions were still filledwith expatriatecin the early 1980s. Data for 10 other countries at the end of the 1970s suggestthat expatriateson averagefilled 502 of - 101 -

Table 6.1 Ratios of GraduateStarting Salaries to Per Capita Income

Graduatestarting salary as Country multipleof per capita income

Gambia 11.1 Ghana 3.6 Kenya 14.6 Liberia 11.1 Sierra Leone 5.1 Somalia 8.3 Tanzania 14.2 Zambia 12.0

Mean 10.0 Median 11.1

Note: Circa 1979 (InternationalLabor Office, 1982).

Table 6.2 Index of StartingSalaries in PublicService by EducationLevel ("0" level - 100)

Country "Al level Degree level

Ghana 120 169 Kenya 127 269 Liberia . 339 Sierra Leone 229 342 Somalia 128 171 Tanzania 148 323 Zambia 116 169

Mean 145 254 Median 128 269

Sources Derivedfrom [InternationalLabor Office,1982]. - 102 - tertiaryteaching posts (see Annex Table C.9). And postgraduateprograms, especially those involving rigorous research training, remain in their infancy, in part because requisitestaff with doctoralqualifications and adequate experienceare not available.

Indeed,the least recognizedbut perhaps,over the long run, the most devastatingdamage to Africanhigher educationsince 1980 was to the promising start of the 1970. in buildingan indigenousAfrican capacityto produce tertiaryteachers, research scholars, and top flight analyticpersonnel for the public services. Illustrativeof the situationis Nigeria,where erosionof progresswill mean that only one-thirdof 3,000 new academicstaff requiredby universities before the end of the decade is expected to be forthcoming from existing postgraduate and staff development programs. The ahortfall is greater, of course, when doctoral level staff requirements in non-University researchand professionaloccupations are considered.

Research. Althoughsubstantiation depends more on anecdotalthan empiricalevidence, African university staff uniformlyreport that researchin their universitieswithered in the 1980s. As the financialcrisis of tertiary educationdeepened, research budgets were typicallysubject to early and severe cuts. Since a significant part of postgraduate (especially doctoral) training involves student participation as apprentices in faculty research and ultimately the solo undertaking of a dissertationproject, the feasibilityof offering graduate education of reasonable standard suffereda concomitant decline.

Stagnationor outrightdecline in African researchoutput, and in African capacityto producefuture researchers,places in jeopardythe continent'slong-run ability to take advantageof the worldwideadvance in science and technology. Africa need not be consistentlyin the forefrontof all scientificand technologicaladvance. And well into the next century,a sizeablefraction of Africa'sPh.D.'s will still undoubtedlybe earned from foreigninstitutions. But the continentnevertheless already needs substantial capacityfor absorptionand utilizationof new knowledge,and that capacityis in large measuredeveloped through the operationof indigenouspostgraduate teachingand researchprograms. For example,frontier developments in genetic engineeringand other areas of biotechnologyare potentiallyapplicable to problem.of plant and animalhealth in Africa;the impacton productionof food crops and on exportagriculture could be sizeable.Extension of the argumentto human health and even industrialprocesses -- and to such frontierareas as microelectronicsand materialsscience (ceramicsand metallurgy)-- is not hard to contemplate.

The centralpoint is that withoutAfrican mastery of the underlying scienceof such developmentsand an abilityto adapt'themto local problemsand conditions,the potentialbenefits to Africa of these advanceswill likelybe lost in large measure and certainlywill be late in arriving. World class university-basedprograms of both basic and appliedresearch and of postgraduateeducation are the breedinggrounds for masteryof scienceand technology. They are the key to sophisticatedconsumption of mankind's explodingstock of knowledgeand technology. They are a necessarycondition for African escape from intellectualdependency, In a very real paradoxical sense, no Africannation can affordto have such programsin the short run, and none can affordnot to in the long term. - 103 -

Consultacv and Advice. Even in their currentdepressed and fragile state, uaiversitiesin Africa representa nation'slargest reservoir of expert knowledgeand cosmopolitanexperience. And here is anotherparadox. The continuing demand for expatriate technical assistance from both public and privatesector organizationssuggests there is no lack of tasks requiringthe highest levelsof academictraining and professionalexperience. Yet, with only numberedexceptions, African universities as institutionsand their facultiesas individualsdo not allocatemuch time and effort to direct service activitiesabove and beyond the mountingof specialtraining programs. Much appliedwork currentlyundertaken by foreignerscould as well be done by Africanuniversity staff, who in the processwould enrich their own research and teachingby more thoroughgrounding in present-dayAfrican realities.

(b) Low quality

Direct and hard evidenceon the level and trends in qualityof the outputsof Africanhigher educationis not available. That African graduates may not be as knowledgeableas their peers elsewherein the developingworld may be suggestedby resultsof the GraduateRecord Examination, where scores on the verbal, quantitative,and analyticsections are uniformlylower than for Latin Americ&n,Asian, or Middle Easternstudents; but the real significanceof this finding is subjectto varying interpretations.Beyond this, the low (and possiblydeclining) standard in African higher education is so pervasively bemoanedby teacher,student, employer, and governmentofficial alike (see Box 6.2) as to be everywhereaccepted as conventionalwisdom, which we do not dispute.

[Box 6.2]

Nor could it be otherwise,since indirectevidence of a crisis of qualityis overwhelming.A tragic consequenceof the economicdownturn -- and of the concomitantconstriction in publicbudgets and reducedaccess to foreign exchange-- has been the virtual disappearancefrom higher education institutionsin many Africancountries of exactlythose inputsthat make physicalplant and highly trainedacademic staff educationallyproductive.

These inputsinclude: spare parts for equipmentmaintenance and repair;routine replacement and upgradingof equipment;reagents and other consumablesupplies; multiple copy textbookand monographacquisitions for libraries;subscriptions to scientificjournale; functioning computation facilities;maintenance, gas, and oil for vehiclesused in facultyresearch and studentfield trips; other researchsupport to faculty,including travel; even routineupkeep of physicalplant (e.g., sanitaryfacilities and telephone exchanges)and publicutilities (e.g., sewage, water, electricity). The level of deprivationof these non-salaryoperating expenses varies from place to place on the continent,but Annex Table A.22 suggeststhat, on average,less than 2Z of total tertiarylevel recurrentpublic expendituresis availablefor these crucialpurposes. Campusvisits make clear that, in a reversalof the situationprevalent into the 1970s, scarcityof non-salaryrecurrent inputs, not the numberor level of trainingof academicstaff, is today the governing constrainton qualityin African higher educationin nearly every country. - 104 -

Box 6.2 CRISIS OF QUALIT.-IN HIGHER EDUCATION

The scarcity of funding for capital Investment and nonsalary operating expenses has seriously undermined the quality of educatlon In African universitles. The situatlon at Nigeria's University of Ibadan Illustrates the problem.

For several months now we have been expected to run a physlcs laboratory without electricity, perform biology and zoology experiments without water and get accurate readings from microscopes blinded by use and age. Chemicals are unimaginably short. The result of all this Is a chemistry laboratory that cannot produce distilled water and hundreds of 'science graduates' lacking the benefits of practical demonstratlons. CWest Africa, 12 September 1983.]

Ghana provides other examples. At the University of Science and Technology, equlpment for the electrical engineering department has not been purchased since 1962, and most equipment In the civll engineering department dates from the 19508. Such old equipment, Including the university's computer, Is rarely In working order; It requires regular routine malntenance, Including replacement parts, for which funds are not available. The same university has been characterized as ... grossly short of... books, Paper and food' (t West Africa," 18 July 1983]. Slmilarly, reports from the University of Ghana's faculty of science tell of shorteges of chemicals and other consumables necessary for laboratory classes, and note that the scarcity of foreign exchange precludes the supply of essential materials from abroad. Lack of funds also means that some universities are operating without thl. vehicles necessary for student fleld trips and data collection, and that others do not have the resources to repair broken telephone systems.

In many countries throughol4tthe sub-Saharan region, the lack of capital funds has led to unfinished construction work on classrooms, laboratories,workshops, libraries and residence halls. A 1981 report from the Nigerian Commission on Salary and Conditions of Service of University Staff states: *The Commission was horrified to witness the disgraceful spectacle of students In the corridors and outside lecture theatres struggling to comprehend the proceedings Inside.' Reports from the University of Ibadan are also discouraging: 'Everything In the University today points to an agonizing decline. Students swarm from their hostels where there are six In a room designed for two, Into a dingy lecture room where a teacher shout- his notes across a hall of five hundred listeners* rawest Africa*, 12 September 1983]. - 105 -

The most immediateconsequences of the dryingup of non-salaryinputs to higher educationare that researchceases and instructionis reducedto littlemore than rote learningof theory from professoriallectures and chalked notes on blackboards. Chemistswho have not done a titration;biologists who have not done a dissection;physicists who have never measuredan electrical current;secondary science teachers who have never witnessed,let alone themselvesactually conducted, the demonstrationscentral to the curriculathey teach; agronomistswho have never conducteda field trial of any sort; engineerswho have never disassembledthe machinerythey are calledupon to operate;social scientistsof all persuasionswho have never collected,or conductedan analysisof, their own empiricaldata; specialistsfor whom the programmingand use of computersis essentialwho have never sat before or tested a programon a functioningmachine; lawyers who do not have accessto recentjudicial opinions; medical doctorswhose only knowledgeof laboratory test proceduresis from hearingthem describedin a lecturehall -- qualitativelydeprived graduates such as these are now appearingin countries that have been hardesthit by the scarcityof non-salaryinputs. But the threat to all applieddisciplines is nowhere far removed,even in the relativelystable university environments. Complicating the situationfurther in many institutionsare studentnumbers in excessof the carryingcapacity of crucialcomponents of the physicalplant, livingaccommodation and library study space in particular.

As a consequence,the skillsmost relevantto development,those acquiredwhen theory is confrontedwith the exigenciesof the real world, are exactlythe ones that do not get learned. The end result is that, in its stock of high-levelskills and in its abilityto generateknowledge and innovation, sub-SaharanAfrica is fallingfurther behind despitethe increasingnumbers of higher educationgraduates. Tertiaryeducation discharges ever less effecti-.lyits principalresponsibilities.

(c) High costs

Not only does higher educationnow producetoo many graduatesin many fields,too few high-levelpersonnel for postgraduateeducation and research, and not enough researchand developmentadvisory and consultancyservices. Not only are all those outputsof increasinglydubious quality. But the costs per araduate-- some of whom are not needed and many of whom are of poor standard -- are exorbitant. As a percentageof GDP per capita,which is a reasonable proxy for affordability,unit costs (costsper studentyear) of publichigher educationare six-to-seventimes higher in sub-SaharanAfrica than they are in Asia and nine times higher than in Latin America (see Table 6.3). As a multipleof the cost of a studentin primaryschool, a plausiblemeasure of opportunitycost, the unit cost of publiclysupported students in higher educationis abtut 60 times that of primarystudents (see Table 6.4),while the same statisticfor Asia and Latin Americais between 10 and 15 times.

[Tables6.3 and 6.4]

There are two parts to this problemof internalefficiency, both reflectingfundamental performance shortfalls in Africanhigher education. First, wastage -- and thus cost per graduate -- is high. Data for seven countriessuggest that betweenone-third and two-thirdsof the initialentrants to tertiaryeducation complete their studiesbehind scheduleor not at all. Thus, a sizeableportion of studentplaces are occupiedby repeatersand/or - 106 -

Table 6.3 Unit Costs of PublicEducation at the VariousLevels as Percentageof Per Capita GNP

Region/CountryGroup Primary Secondary Higher

Sub-SaharanAfrica 15 62 800

Francophone 23 86 1,000 Anglophone 12 51 600

Asia

South East Asia and Pacific 11 20 118 South Asia 8 18 119

Latin America 9 26 88

All developingcountries 14 41 370

Advancedcountries 22 24 49

Sourest Annex TablesA.17-A.19 and (Mingatand Psacharopoulos, 1985]. - 107 -

Table 6.4 Public Ependiture per Student in Tertiary Education as Multiple of Public Espenditure per Student at lower levels

Tertlaryas multipleof priury Tertiaryas multipleof secondary

1970 1975 1980 1983 1970 1975 1980 1983

Low-incaseconsoles 68 57 61 80 12 12 18 14 MIddle-lnombeconuies 55 50 43 50 7 10 10 7

Franolftm cozItrles 40 55 44 81 9 12 18 13 Mgliptne cantries 79 88 50 52 11 11 11 12

U-SalaranAfrica 55 55 50 59 11 12 14 13

Note: All flgweu aresedias. Bed onAnsne Tables A.17-A.19. - 108 - future dropouts,with the resultthat many more years of studentservices are requiredto produceeach graduatethan the lengthof the cycle would optimally riquire. Among the factorscontributing to studentrepetition and dropoutare: irsufficientlydeveloped selection mechanisms for studentsaspiring to enter higher education(in some countriesit remainsthe case that all who graduate from secondaryeducation are assureda place in highereducation); the dearth of non-salaryquality-enhancing inputs (notedearlier) which makes learning difficult,unstimulating, and unrewarding;and the absenceof sanctionsfor poor performanceof studentsand teachers,so there is no incentiveto strive hard for results.

Second,resource costs per student-yearof higher educationservices providedare needlesslyinflated. Althoughnot uniformlypresent in every country,several factors typically contribute to the unnecessarilyhigh costs of tertiaryinstitutions borne by the educationbudget of Africannations.

Within a sampleof 50 Africanuniversities at the end of the 1970s, 12 had enrollmentsunder 1,000 students,while only 13 had enrollmentsover 5,000 (and severalof thase had studentsdispersed over more than one campus). This proliferationof small institutionsprecludes exploitation of the well documentedeconomies of scale in higher educationprovision (and, in addition, means foregoingthe academicadvantages of criticalmass concentrationsof highly specializedstaff). Explicitunderstandings on an appropriatedivision of 3-'bors-- in large countriesinternally among the tert!aryinstitutions, and in 6sfallcountries across internationalboundaries -- could greatlyreduce this cause of high costs.

In many countries,universities were eitablishedinitially as self-containedcommunities on vast tractsof land outsidemajor urban areas, involvingheavily subsidized provision to all staff membersof housing, standardmunicipal services, and even welfare and social services(clinics, schools,clubs). The huge sunk costs of this infrastructureprohibit starting afresh,and the ongoingcosts of this legacymust somehowbe met. With few exceptions,the publicbudget picks up the tab.

Judgingfrom the approvedpersonnel establishments, the tatio of studentsto academicstaff is 13:1 in a sampleof FrancophoneAfrican universitiesand about 7:1 in AnglophoneAfrica, where it rangesfrom 3:1 on the low side to 12:1. By way of comparison,the ratio of studentsto academic staff in Britishand French universities,in some sense the models for their counterpartsin formercolonial possessions, is 13:1 and 25:1, respectively. In some countries,where erosionof salariesand deteriorationof working conditionshave resultedin an exodusof the most highly qualifiedacademics from the universities,the actual ratio of studentsto (in place) acsdemic staff may be somewhatabove the median. Still,staffing provision is generous by comparisonwith developedcountries where the student-staffratio is more typicallytwice that in Africa. One reason for this is the propensityto offer a very wide range of programsand courses in each institution,resulting in wastefulduplication. Courseenrollment of 15 studentsis not unmsual. This problem,too, could be alleviatedin the large countriesby cooperationamong institutionsand in the small countriesby internationalcooperation. - 109 -

Africanuniversities also usually employ large numbersof non- academicstaff, especiallyto operatemunicipal and studentwelfare services and to care for the campus. For example,in 1980-81,Nigerian universities employed52 thousandstaff for a studentpopulation of 77 thousand. In a Britishuniversity reasonably analogous to the three Ghanaianuniversities togetherin size, age, and locationon the peripheryof a large city, there are over six studentsfor each member of the non-teachingstaff and the ratio of non-teachingto teachingstaff is 1-to-6. Illustrativeof the generalnature if not necessarilythe exact quantitativedimensions of the wider problem in Africa, non-teachingstaff in the three universitiesin Ghana outnumber studentsand the ratio of non-teachingto teachingstaff is 14-to-1. Most of the differenceis in serviceareas. Guards,ground crew, and maintenance personneltotalled 902 in one of the Ghanaianuniversities and 166 in the Britishuniversity.

Of 24 Africancountries for which recentdata are available,none has a generalpolicy of chargingtuition fees which are not coveredby a government subsidy in some form. The essentiallyuniversal policy of chargingno fees for higher educationmeans iPso facto that the publiclyborne unit costs are higher than they would be if a significantpart of the burdenwere borne by students and their families. But the no-fee policy contributesindirectly to high unit costs as well, since studentsand their parentshave no incentiveto contain costs. If insteadof fee-freehigher education,students and their parentshad to shouldera significantproportion of the total cost, pressurewould surely develop to containunessential expenditure and, in general,to increase efficiency.

In a similarvein, the longstandingpractice of providingeither free room and board on campus or an allowanceto all students,thereby shifting to the public "educational'budget the considerablesums requiredto cover living expensesof students,greatly inflates publicly borne unit costs of higher educationin Africa. While severalcountries have begun to phase out support for livingcosts of students,fellowships to studentsstill constitutehalf the public expendituresin higher educationin a numberof African countries. The scope for reducingsuch public expendituresis particularlybroad in many Francophonecountries.

Finally,unit costs are inflatedby the failureto make maximumuse of expensiveteaching personnel and physicalfacilities, which is manifest in light teachingloads, in limitationof class and laboratoryhours to a restrictedportion of the day, and in lettingthe entireplant and staff lie idle at least 20 weeks per year. The scope for eveningand vacationstudents to get access to universitieshas almost nowherebeen exploited,the University of Zambiabeing an importantexception. With universitieseffectively closed to all but the fortunatefew full-timestudents, there is a substantial populationof studentstaking courses at their own expenseoverseas.

(d) Inequitableand inefficientfinance

Unlikethe situationin some other parts of the world, tertiary educationin sub-SaharanAfrica is overwhelminglypublic in ownershipand operationalcontrol. Private institutionsof higher educationare often explicitlyproscribed in fundamentallegislation and administrativeregulation. While it need not necessarilybe the case, public ownershipand controlof higher educationin Africa has meant for all practicalpurposes that tertiary - 110 - education-- includingthe livingcosts of its students,which are not properly an educationexpense -- is entirelyfinanced by the public budget. With few exceptions,students, their families,and their future employersare spared any contributionto the costs of higher educationbeyond the generalincidence of the tax system and the incomeforegone while studying. The extentto which the private return to higher educationexceeds the social return is a useful index of public subsidizationof education,since most of the differencebetween the two is due to includingthe state'scontribution to costs in the social rate calculationsand excludingthem in the private rate calculation. The available evidencesuggests that in Africa privatereturns to higher educationare, conservatively,150% greaterthan social returns,a multiplemore than 3 times higher than in Latin Americaor in developedcountries generally and more than 15 times higher than in Asia.

There are two undesirableconsequences of this insulationof the beneficiariesof higher educationfrom its heavy costs. Incomeinequalities are increasedthrough the sharplyregressive impact of higher education expenditures;the systemassures that the rich get richerand the poor get poorer.And perhapsmore important,within the educationsector as a whole, and within higher educationas a subsector,resources are allocatedinefficiently.

6.3 A program for structuraladjustment of Africanhigher education

If the above overalldiagnosis of weaknessesin higher educationin sub-SaharanAfrica were to be confirmedthrough careful analysis of the situationat the individualcountry level,the four objectivesto be sought for the short-termthrough reformed policies would be clear enough: (a) improve quality; (b) increaseefficiency; (c) constrainoutput, especially in those fieldsthat do not directlysupport economic development; and (d) relievethe burden on public sourcesof financingby increasingthe participationof beneficiariesand their families. But qualityenhancement as the first objectivewill cost money. Thus, implementationof policiesto achievethe last three objectiveswill, essentiallyeverywhere in sub-SaharanAfrica, be a prerequisitefor freeingthe resourcesneeded to achievethe first. The four objectiveswould, of course,be pursuedwith differingmixes of policies,as appropriateto particularcountry circumstances. The sequenceof policy implementationwould also depend upon the needs of each country;gradual phasing in of new policies,in stagesover some years,would have to be the rule not the exception.

(a) Improvedquality

A quantumincrease in the qualityof tertiaryteaching and research is the first objective. Long-termdevelopment goals cannotbe wet without it. However,quality improvement is unavoidablyexpensive. So, given the requirementsof other parts of the educationsystem and the environmentof constrainedpublic resources,no advancehere is possiblewithout significant progresstowards the other three resourceconservation objectives.

Quality improvementcould be achievedthrough a varietyof measures. Most immediatelynecessary is establishmentand gradualimplementation of standardsof provisionfor the full range of vital non-salaryinputs to teachingand research. Supply to the librariesof multiplecopies of basic textbooksas well as supplementarybooks and periodicalsis the highest priority,closely followed by supply to laboratoriesand workshopsof - ill -

consumablesand equipmentmaintenance and repair. Resuscitationof lo-ng-term effortsto upgradethe academicqualifications of staff is also essential. Formal postgraduatetraining in masters and doctoralprograms, for the time being mostly outsideof Africa, is an essentialpart of this. But the need extendsmuch further,to postdoctoralfellowships, facultyexchanges, collaborativeresearch, and other professionallinks with foreignuniversities whereunderAfrican academicsare exposedto new developmentsin researchand curriculumin their fields. Twinningarrangements, between a departmentin an African universityand the same departmentin one or more foreignuniversities, are an especiallyattractive device for staff developmentat these sophisticatedlevels, and can profitablyalso involveservice of foreign universitystaff in Africaninstitutions. For maximum productivity,such arrangementsneed to be sustainedover many years.

Qualityin tertiaryeducation is also enhancedby rigoroustesting programsadministered independently of the universitiesand involvingexternal examiners. This is an essentialingredient in settingstandards of performance to which individualsand institutionscan respond. Independentperformance measurementalso createsan ambient in which highly efficientindependent study programscan develop.

Finally,in the longer term, qualityimprovements will be realized and sustainedthrough the establishmentof centers-- and, in some cases, of multi-institutionalprograms -- of excellencefor postgraduateeducation and researchin which criticalmass concentrationsof staff and resourcescan be achieved. In this regard,participants at the Mbabane and Harare conferences on higher educationhave voiced their supportfor an appraisalof the advanced teachingand researchcapabilities and capacitiesof African universities,as a first step toward identifyingpotential sites for researchand postgraduate teachingin priorityareas. By establishingsuch specialized,high quality institutionsand programs,African govcrnmentswould provide able African studentswith an attractivealternative to (more costly)foreign study, create incentivesfor universityresearchers to pursue their work on the continent, and in so doing addresstwo aspoctsof the seriousproblem of "braindrain."

(b) Increasedefficiency

Reductionin unit costs (per studentyear of educationalservices provided)is the second challenge. This would be achievedwithin existing institutionsthrough such measuresas: establishmentand gradualphasing in of higher numbersof pupils for each member of academicstaff; minimum standards for class or coursesize; reducednumbers of nonacademicsupport staff for each academicstaff member (perhapsin part throughusing studentlabor in some campusJobs); increasinghours per week and weeks per year that academicstaff and physical facilitiesare utilized;expanded access for part-time,fee-paying students;introduction of self-studymethods (e.g.,radio and correspondence) to teach low-enrollmentcourses; gradually assigning the full cost for housing, food, and other welfare servicesprovided to staff and studentsto non-public sources;where feasiblewithin countries-- and prospectivelyeven among countries-- rationalizationof programsand faculties,and consolidationof institutions, so that diseconomies of small scale are mitigated. No country is likely to be able to implementall of these measures,but all can potentially reduceunit costs of tertiaryeducation by developinga coherentpolicy package from carefulselection of a few. The potentialrole of international collaborationshould obviouslybe pursuedas well. - 112 -

In additionto reducingunit costs within existinginstitutions, more widespreadutilization of extramuraldegree programshas demor.stratedthe capacityfor reducingper studentcosts to a fractionof the levelstypically found in on-campusprograms. Extramuraldegree programs rely on books, correspondencematerials, occasionally radio (or audio-cassettes),occasional on-campusmeetings, and most lmportantly,student time. Essentialto the successof such programsis the existenceof an independentexamination or accreditationsystem (furtherdiscussed below) to measure and certifyquality. Greater relianceon extramuraldegree programs frees governmentsfrom providing boardingfacilities and coveringthe cost of transport,allows fulleruse of existingclassroom and laboratoryfacilities, and allowsstudents to work while studying. Avoiding largeconcentrations of full time students,typically in capitalcities, may in some instancesentail the furthervirtue of loweringthe potentialfor politicalturbulence.

(c) Constrainedoutput

Reductionin the annualnumber of (publiclysupported) tertiary educationgraduates in some fieldstogether with decelerationin the growthof graduatesin most others is the secondchallenge to be addressed. Therewill of coursebe exceptionsto this generalprescription by countryand field,the clearestof which may be in teachertraining and graduateeducation programs in the sciences,engineering and socialsciences where outputsmay need selectivelyto be increased. This quantitativeoutput objectivecould be achievedthrough a more than proportionatereduction of intaketo (publicly supportedinstitutions of) tertiaryeducation. Limitationof the intaketo tertiaryinstitutions could be accomplishedby a mix of such measuresas tighteningup of selectioncriteria (instigation of stiff entrance examinations),establishing a systemof differentialavailability of scholarships,or even outrightimposition of quotas,by region and field;and initiatingcost sharingmore aggressivelyin the disciplineswith highest graduateunemployment. Smallerentry cohortscomposed of studentsof higher qualitywould help to decreasewastage duringthe tertiarycycle and reduce cost per graduate. Congestionof physicalfacilities, now a problem in some countries,would also be relieved.In additionto limitingthe intaketo tertiaryinstitutions, the numbersof graduatescan be furtherreduced, their overallquality increased, and importantresources freed for investmentin quality-enhancinginputs for those who remain,by conditioningthe continuation of scholarshipson fully satisfactoryacademic performance. Failure in a course could entail a "repetitionfeel and failurein too many coursescould resultin denialof the privilegeto returnthe followingyear under any circumstances.

In pursuingthis objective,policymakers should realizethat in individualtertiary institutions that have yet to expandto the thresholdof diminishingaverage cost, restrictionof outputwould imply higherunit costs. In such situations,consolidation of small institutionsinto largerones should be aggressivelyexplored as a way of constrainingoutput without sacrificing the unit cost advantagesof large institutions.The potentialpayoff to internationalcooperation in consolidatingtertiary education is high. - 113 -

Of course,in purelyeconomic terms, recourseto such directmeasures to controlthe numberof graduatesshould not be necessary. If markets generally,and the higher educationmarket in particular,functioned properly -- i.e., if prices (especiallywages for labor)were unregulated,if informationwas comprehensiveand readilyavailable, and if entry and exit of firms (includinghigher educationinstitutions) to the market were uncomplicatedand expeditious-- supply and demand for higher education graduateswould equatethemselves nicely without public intervention.As structuraladjustment of Africaneconomies is achievedand liberalizationis institutionalized,markets will work better and the need to rely on public interventionto controlenrollments in highereducation will diminish. For the moment, however,the necessaryconditions do not obtain. In their absence,and despitethe imperfectwisdom of the authoritiesin designatingthe fieldsfor expansion(or contraction)and in implementingappropriate measures (or incentives),public interventionis justifiable. During a potentiallylengthy and painfultransition to a more market-orientedsystem of higher education, direct actionmay offer the only practicalshort-term possibility to limit outputsof the system.

(d) Expandedcost-sharing on the part of beneficiaries

Recoveryof some greaterportion of the real costs of providing tertiaryeducation is the fourth imperative. This would be achievedthrough such measuresas: allowingthe establishmentof privatelyowned and financed institutionsof higher educationoperating under state-setquality standards; introducingfees in public establishments,initially for non-instructional serviceslike food and lodgingand then for instructionaltuition; imposing nationalservice obligations -- e.g., to teach school,staff distanceeducation centers.participate in adult literacycampaigns -- prior to, during,and/or after enrollmentin higher education;promoting an educationalcredit market; imposinga specialtax on earningsof tertiarylevel graduatesduring a transitionperiod to implementationof effectivegraduated income tax systems.

Expansionof cost-sharingto includethe beneficiariesof tertiary educationand their familiesdoes not mean that governmentsshould lessentheir financialsupport to the subsector. Rather,promotion of broader financial participationshould be seen as one way in which governmentscan help ensure the increasein financialflows to higher educationwhich are necessaryfor its revitalizationand ultimateexpansion.

(e) Feasibility

Such a program of structuraladjustment of higher educationin Africa will not be easy to designor implement. First, fixingupon the educationally and economicallycorrect mix, and then the sequenceand phasing in, of measures in light of specificcountry circumstances requires analytical and planning capabilitieswhich are nowhereplentiful, in some countriesexceedingly scarce, and everywhereuntested on a task of this complexity.

Second,political considerations will inevitablylimit the feasibilityof some desirableelements and sequencesof measures,many of which will in the short term be perceivedas threatsto deeply ingrainedinterests of powerfulgroups in society (civilservants, professors, and students). Determinedand very high level leadershipwill be neededto overcome resistance. - 114 -

Third, actual implementationwill demanda level of managerial competencethat is uncommon. The programwill have to incorporateactions to supportmanagement, especially through developing capabilities to monitor and evaluate the impact of new policies and to adjust and fine tune their implementation.

Fourth,very littlecan be expectedto happen quickly,for institutional change of the magnitude required is always excruciatingly slow. Great persistencewill be required.

rinally,and perhapsmost importantly,a fundamentaldilemma must be overcome. Rapidlyapplied, highly visible, and quick-yieldingmeasures to reversethe precipitatedecline in qualityof African higher educationand to assure its long-termupgrading are likelya necessary(although not sufficient) conditionfor acquiringsocietal acceptance of the painfulmeasures that will have to be taken to achievethe three resourceconservation objectives. But it is only after those measureshave been put in place and taken effect that resourceswill be availablefor sustainedmaintenance and improvementof quality. Donors could addressways to resolvethis dilemmaof transitionas a first step in restructuringinternational collaboration on highereducation in Africa.

* * *

Academics, university administrators, public sector managers and even students share an acute recognition of the intolerable situation of African higher education. In most countries,measures to increaseefficiency, constrainoutput, and diversifyfinance are alreadyunder considerationand, in some places, beginningto be implemented.As such measurestake hold, the possibilitiesfor introducingurgently needed quality enhancementwill grow. In summary:

Recommendation4. Recent deteriorationseriously threatens the abilityof most of Africa'sinstitutions of higher educationto make their several,absolutely vital contributionsto the region'sdevelopment. The qualityof these institutionsmust first be restoredand then furtherenhanced in order for the regionto extractmaximum advantage, for its own purposes, from the accelerAtingworldwide advance of scienceand technology. In most countries,however, a period of adjustmentto changedeconomic circumstances is a short-termprerequisite for improvinghigher education. Some tertiary institutions,individual campuses, academic departments, and teachingprograms need to be amalgamatedinto largerunits of economicallyviable size. Personnel reductions,especially of non-teachingstaff, are indicatedin some countries but should be sought in such a way as to increasethe averagelevel of relevant trainingand experienceof those who remain,particularly those in academic positions. The numberof studentsat most institutionsneeds to be stabilized, throughtighter admission and pJrZormancestandards, and throughthe eliminationof livingallowances and free room and board; the adverseequity Impact of this final measure can be mitigated by the provision of scholarships based on need. Such consolidationin higher educationwill permit reestablishmentof an economicallyand pedagogicallyviable base of requisite qualityfrom which to bolsterfuture expansion of graduates,research, and communityservice. - 115 -

Chapter 7. USING RESOURCESWELL: THE MANDATEFOR EDUCATIONALMANAGERS

Although enrollmentstagnation and low qualitycan ultimatelybe traced to demographicand economicadversity, they are exacerbated(and their resolutionhampered) by inefficientuse of availableresources. Reflecting this inefficiencyare widespreadunderutilization of facilities,high levelsof absenteeismof teachersand students,and a generallack of order and disciplinein the operationof educationalsystems. Administrativeand logisticalinfrastructures originally created for systemsof quite limitedsize are incapableof copingwith the vastly expandedstructures of today. This constrainsthe abilityof governmentsto plan, implement,and monitorpolicy changes that would addressthe obstaclesto improvededucational quality and coverage. While appropriatepolicy changesare necessaryto improveeducation in sub-SaharanAfrica, they alonewill not suffice. They must be coupledwith strengthenedmanagement, initially of the capacityto delivertraditional educationalservices, if the benefitsof changeand innovationare to be realized.

Despitemore than two decadesof investmentin education,management capacityremains strained and insufficientlydeveloped. This has been due to a combinationof reasons: relativelylow investmentsin this area; fragmentation in effortsof governmentsand donor agenciesthat inhibitssustained institutionaldevelopment; multiple and sometimesconflicting donor policies and procedures,which often consumea disproportionateshare of managementtime and attention;and difficultiesin adaptingmodern forms of organizationto the values and patternsof allegiancecharacteristic of many African cultures. Central,therefore, to the mandateof educationalmanagers will be the strengtheningof their own capacityto use availableresources well.

The discussionthat followsis concernedwith three levelsof management. At the bottom is school-levelmanagement; in recentyears educationalpolicy analysts have developeda strong concensusaround th 9 key importance of the school-level manager -- the principal or headteacher-/ -- and the communityenvironment in which he or she operates. Linkingthe headteacher with the policymakeris a structurefor policy implementation,which involves incentivesand local politics,budgetary choice and projectdevelopment, teachertraining and teachersupervision. At the top is managementof policy developmentfor the nationalsystem.

Shapingmanagerial performance at each of these three levelsare four main factors: (i) the managementand organizationalstructure itself; (ii) the studenttesting, general statistical, and accountingsystems that provide informationto managers;(iii) the analyticalcapacity that generatesand evaluatesoptions for managersat all levels,but particularlyfor managersat the policy level;and, (iv) the qualityand trainingof managerialstaff. These four factorsare subjectto direct improvement:the policy leversfor enhancingmanagerial capacity thus lie in these domains. The remainderof this chapterdiscusses the importanceof each of these four factorsand what is now known about potentialfor their improvement. - 116 -

7.1 Improvingorganizational structure

Organizationalstructures establish predictable relationships between people and tasks, and thus channelthe processesof gettingthings done. They are intimatelyconnected with the distributionof power and authority,and they have considerableimpact on decision-makingand resourceallocation. The need for structuresappropriate to the managementof Africaneducation is acute. Althoughthe issuesand their resolutionare complexand inter-related,three areas of policy concernstand out: schoolmanagement, decentralization, and structuralsimplification.

(a) Schoolmanagement

Schoolmanagement is a crucialcomponent of effectiveteaching and learning. Effectiveschools display common characteristics.First, they displayan orderlyenvironment. Teachersand studentsattend regularly, recordsare kept, and buildingsand groundsare clean and adequ-tely maintained. Second,they emphasizeacademic achievement. Progressis made againstthe curriculum,available materials are used, tests are given and resultsused by teachersand students. Third, teachersand principalsexpect studentsto perform at high levelsof achievement. Teachersprovide students with regularperformance feedback and remedialassistance. Studentscome to believethat work and effortare more importantthan luck in helpingthem to get ahead. Fourth,headteachers pursue an activistpolicy for effectiveness. They have high performanceexpectations for staff and students. They take an interestin classroomactivity and are able to provideprofessional advice to teachers, They take initiativein acquiringresources for the school. They interacteffectively with higher level authoritiesand with the community.

The structureof educationorganizations can help supporteffective schoolmanagement by grantingto schoolsthe authorityto generateand, importantly,to use local resources. Schoolsthat are able to invest locally generatedresources in school improvementare able to show parentsa returnon their financialsacrifice and, thereby,to assurecontinued parental support. School supervisionprovides technical and administrativeassistance for these activities. The role of headteacherscan be broadenedto includethese tasks, as well as an expandedresponsibility for the qualityof classroominstruction. Where the small size of the schoolprecludes the appointmentof an adequately qualifiedheadteacher, a first-linesupervisor might be given these managerial responsibilitiesand become, in effect,the ambulantheadteacher for a number of schools. Alternatively,a headteachermight be designatedand either releasedpart-time from classroomresponsibilities or providedadditional remunerationas compensationfor schoolmanagement activities.

The successfulimplementation of a school improvementpolicy will demand the establishmentof strongand permanentstructures for schoolsupport and supervision. These shouldbe designedto allow for regularsupervision of schools,permanently available opportunities for in-serviceteacher training, and frequentupgrading programs for first-linesupervisors designed in conjunctionwith nationalpolicy and researchinstitutions. - 117 -

(b) Decentralizationof policy implementation

With some notable exceptions, e.g., Nigeria, a pattern of highly centralized education systems has emerged in most of sub-Saharan Africa, and this tendency has become more pronounced, if anything, in the years since independence. The rapid expansion of schools, combined with the increased importance of central funding and expenditure control, has led to an increasingly centralized system of education management. Resources are controlled at the center, and lower level managers typically elevate all decisions to higher levels. Centralization in education is not, in and of itself, necessarily something to be avoided, and indeed, a convincing case is often made for it, especially in newly independent countries where a strong sense of nationhood has not yet been developed. Also, a centralized system of education may be shown to be more efficient for some purposes than a very decentralized one, to the extent that it results in less duplication of effort.

Within the African context, however, there are good reasons for believing that education systems could be made more efficient if certain functions and responsibilitieswere devolved away from central ministries of education and manpower develupment. The reasons favoring greater decentralization have to do with particular characteristics that typify most African countries: large geographic distances between individual schools and the center; great ethnic and linguistic diversity between communities; and relatively poorly developed systems of communications -- inferior transportation (the absence of all-weather roads, of functioning vehicles, and an inadequate central government budget to keep the vehicles on the road); an incomplete and frequently nonworking telephone system; slow and unreliable postal services. Under such conditions, resource and information flows that are supposed to occur between the central ministry and individual institutions frequently do not occur, and an increased reliance on local initiative might obviate the need for such flows and alleviate the consequences of their not occurring.

Decentralization is a policy that can be pursued in two different ways. The first and more obvious kind of decentralization involves changing the structure of authority and responsibility in the public education ministry or ministry that deals with training so as to increase the autonomy of units at lower levels. As implied above, rigid centralization in Africa has tended to block information and decision flows and alienate schools from their local environments, limiting their ability to respond to local needs and resource opportunities. Decentralization can, in this context, contribute significantly to better school management by supporting school autonomy, increasing the responsiveness of the school to the local community -- and of the community to the school.

Second, governments can achieve decentralizationby relaxing restrictions on the activities of private schools and training institutions. These include schools and institutions run by church groups and other voluntary organizations as well as many types of non-formal fee-paying education and training schemes operated by private concerns. This policy expands the range of educational services, and it shifts more of the costs of education to the beneficiaries. Government subsidies to such institutions, perhaps through need-based scholarships to attendees, can help assure equity and quality at relatively low unit costs. The efficiency of such schools is determined through the marketplace, as reflected in the willingness of students or their - 118 - parentsto pay the price demandedfor the servicesoffered. Public investments to supportdecentralizat:ion of this kind includethe publicationof examination resultsand pass rates,the importanceof which is furtherdiscussed later in this chapter. The effectivenessof such schoolscan be 'managed'through examinations,the settingof atandards,and inspectorateservices. Such mechanismshelp to create and enforcestandards, and to disseminateinformation to the public on the statusof individualinstitutions.

Decentralizationthrough deregulation of privateinitiative is likely to be most effectivein countrieswhere privateinstitutions exist but fail to flourishbecause of restrictiveregulations. It will work better in places and at levelsof educationwhere family incomepermits increased expenditure on education. In short,it is likelyto be more effectivein urban areas and at the secondaryand post-secondarylevels.

Planningfor decentralizationin a given educationorganization is a complextask. The purpose of a given form of decentralizationshould be clear. Decentralizationis not an end in itself,but a means to more effective administrationof some specificactivity or set of activities. Many things can be decentralized,to differentdegrees and in differentways. Work, decisions, responsibility,authority can all be assignedto various levels. In bureaucraticdecentralization, what to assign,when, and for what purposesare all key issues. Units that are having difficultycarrying out currenttasks (schools,district offices) may find the decentralizationof certaintypes of authorityhelpful -- e.g., the authorityto procuremany items locallywithout cumbersomecentral approval. On the other hand, mandatingthat schoolsmust be responsiblefor major new tasks (such as generatingand allocatingfunds to provide teachersalary supplementsand schoolmaintenance) is less likelyto succeed.

In sum, effectivedecentralization requires explicit attention to the definitionof the roles of units at variouslevels and to the establishmentof effectiveinformation flows betweenthem. The successfulmanagement of educationalreform in Ethiopia (Box 7.1) illustratescareful delineation of appropriatetasks to decentralizeand othersto maintainat the center. In general,central ministries should retainpolicy, planning, and monitoring functions. In most countries,curriculum policy and development,together with materialsproduction, will continueto be centralized,unless and until the capacityfor curriculumdevelopment is establishedat lower levels,and even then if the purposesare to maintainnational standards and accomplish nation-buildinggoals. Administrativesupport functions (such as payroll, procurement,and school construction),teacher training and certification,and statisticalservices may or may not be fully centralized,depending on the degreeto which public administrationis generally(and effectively) decentralizedto states or provinces. Here the size of the country,the nature of the politicalsystem, and the effectivenessof communicationsinfrastructure are determiningfactors, with the advantagesof decentralizationlooming larger with greatergeographic dispersion, more pluralisticinstitutions, and weaker linkinginfrastructure.

[Box 7.13 - 119 -

Box 7.1 MANAGEMENT OF REFORM IN ETHIOPIA

Implemontation of educatlon reform In Ethlopla Illustrates an organizatlonal development strategy affected both at the central and local levels of the educatlonal system. The central level -- the Ministry of Education and the Office of the President -- took responsibilityfor establishing the overall framework for change. Subsequently, a number of Important tasks were delegated effectively to the local level. It was this combination of strong central control backed up by local Involvement and Initlative that created the environment In which successfui implementationof the reform program took place.

The goals of the reform were the expansion and Improvement of education In rural areas through the construction of schools and the implementationof a completely revised curriculum. A series of education projects, begun In 1966 and backed up by a 1972 Education Sector Review, were Implemented In support of a long-term strategy for meeting the country's socio-economic needs. The institutlonalcapacity for educational planning, project management, curriculum design, and developing and distributing educational materials was gradually developed over a ten-year period. These system resources served as the foundation for further reform.

With the arrival of the now government In 1974, officlal coimitment to the reform of Ethiopia's education system was strengthened. In the context of fundamental changes In the soclal, economic, and political development of the country, the government viewed educatlon as a vehicle for social change and made educational development a high priority. Implementatlon of the reform at the central level relled on three well-established institutions: the Ethiopian Building Construction Authority, which was responsible for the physical Implementationof three eariler projects supported by the World Bank and the efficient lmplementation of this one; the National Curriculum Development Center (NCDC), which was well-equipped from prlor experience to handle the testing and finalizationof the curriculum materials under the proJect; and the Educational Materials Production and Dlstrlbution Agency.

Although the reform was carefully directed by the central government, Implementationwas highly decentralized. At the local level a capacity for teacher training, support, and supervision was estabilshed through the Awraja (district) Pedagogical Centers, which provided a permanent base for these activities. The management strategy also provided for the Incorporationof local Input. Considerable effort was made to Include technical feedback from administratorsand teachers on the new curriculum and materials, which proved to be keys to the project's success. Curriculum and textbooks were tested In 70 experimental schools prior to adoptlon nationwide. Teachers and administratorsprovided feedback to NCDC staff, who then carried out a sumiative evaluatlon. Mor¢'1over, community support and participatlon played an Important part In project Implementatlon. Student commitment was fostered by out-of-school community work that taught additional practical skills and gave students a chance to apply what they had learned In the classroom. This work was performed under the supervision of Peasant Associatlons, School Managemnt Committees, and local-level polltIcal units known as Keboles. - 120 -

Schoolsare responsiblefor the qualityof instruction,and hcadteachersplay a key role in meeting that responsibility.The headteacher should,within the frameworkof careful role definitionfor various levelsof the system,be grantedthe authoritynecessary to achievequality in the individualschool. The headteachershould have a genuinevoice, if not the final say, in all of the following: appointment,disciplining, and dismissal of teachers;adaptation of curriculumand classroomschedules to meet local circumstances;establishment of effectiverelationships with community organizations;generation of local resources;and importantly,use of locally generatedrevenues within centrallyprovided guidelines and a systemof accountability.

The functionsof intermediatelevels of administrationare also crucialto effectivedecentralization. Supervision of schoolsshould be expandedto includetechnical advice and trainingsupport for new functionsin schools,such as resourcegeneration, and not be seen narrowlyas visits from the center to collectstatistics. To the extentthat non-formaland private educationexpand under ministryaegis, supervision systems must expandto cover new types of institutionsand tasks. The linkingfunction, of moving informationboth ways between schoolsand higher authorities,increases in importance. For vocationaland technicaleducation, staff shouldbe responsiblefor establishingand strengtheninglinkages between schools and employers. This also is a specializedrole for which centrallyprovided trainingand resourcesupport may be needed.

(c) Structuralsimplification for policy development

Structuralsimplification can supportdecentralization and school mfanagementif specificmeasures are taken as part of ai broad strategyfor structuralreform. As certainadministrative functions are delegatedto lower levels,central ministry functions may be simplified. It becomespossible for centralministries to devote greaterattention to their principlefunctions, which are those of broad policy planning;design of policy implementation strategies;monitoring the consequencesof policy implementationthrough observation,testing and evaluation;and adaptingpolicy in the light of its evaluatedimpact. Delegationof power to appointand dismirsteachers to school-levelor intermediate-levelofficials greatly simplifies the personnel managementresponsibilities of the centralministry. Gatheringof statistical data is a second case in point: as the capabilityto gatherand aggregatedata is decentralized,the tasks of the centralunit are simplifiedand reoriented towardquality control,analysis, and dissemination.

In supportof this more focusedrole of centralgovernment, two sorts of structuralsimplification will often be in order. First, those countries with more than one ministrydealing with education(or ministry-levelentities such as Ethiopia'sHigher EducationCommission), and certainlythose few African countrieswith three or more such ministries,will wish to consider structuralsimplification through consolidation into a singleministry. This will encouragea broaderperspective in the developmentof educationpolicy.

The second sort of structuralsimplification will be developmentof an entity responsiblefor maintaininga system-wideview of trainingwithin the economy. It would track how trainingactivities relate to educational activitiesand how they relateto employers,and it would developappropriate - 121 -

policiestoward training. While such a trainingpolicy unit might be housed in a ministryof education(or an expandedministry of educationand labor),it might also be locatedin a nationalplanning ministry. In either case its work would both facilitateand be facilitatedby consolidationand streamliningof the ministryof education.

7.2 Improvinginformation: testing, statistical, and accountingsystems

Each managementlevel -- from the policy formulationlevel to that of the school -- requiresinformation for its own use and as a commoditythat it providesto the largersocial system. Broadlyspeaking there are two main types of informationwith which the educationsector needs to be concerned.The first is achievement test information on the quality of student performance; the second is descriptivedata concerningthe numbersand types of institutions,personnel and studentsin the system,their geographical distribution,and the financialflows that affectthem. These data are used at all levels for planningand resourceallocation; they are a necessaryinput into the analyticalfunction to be discussedin Section7.3; and test data play a key role in certificationand selectionof individualstudents.

(a) Educationaltesting

There is a long historyof educationaltesting in Africa,beginning, in the colonialera, with extensiveuse of examinationsadministered from Europe. After independence,such multinationalgroupings as the East and West African ExaminationsCouncils played criticalroles in helpingdevelop a cadre of experiencedAfrican testing professionals and psychometricians.External examinations(i.e., examinations external to the individualschool) play importantroles today in a significantnumber of Africancountries.

The role of externalexaminations should be fourfold:

o Firdt, if properlydesigned, examinations provide performance measurementof the systemas a whole and of individualdistricts and schools;this can allow trackingof performanceover time, internationaland inter-regionalcomparisons of performance,and school-levelaccountability for performance. This accountability functionis best servedby making public appropriateaggregations of scores in a timelyway. These performancemeasures can also be used as an input to the system'sanalytic efforts -- allowing,for example,for carefulevaluation of new projectsor reformefforts.

O Second,well-designed examinations play a curriculumimprovement role. They encourageclassroom level teaching (and learning)of the designatedcurricula, since teachersteach to (and studentsstudy for) the tests,however they are designed;if the tests are well-designed,teachers teach (and studentslearn) what they are supposedto. Both this functionand the performancemeasurement functionare centralto improvingeducational quality.

O Third,examinations allow objectiveselection of studentsinto the next cycle of educationor into appropriatetraining opportunities. - 122 -

o Fourth,external examinations allow objectivecertification of studentcompletion of an educationalcycle; such a certification procedureis essentialto the widespreadimplementation of independentstudy programsthat were identifiedin Chapter5 as highly cost-effective.

The performancemeasurement function can be well performedby testing only samplesof studentsat selectedtimes. To a lesserextent, this may be true of the curriculumenhancement function as well. Studentselection and certification,on the other hand, will usuallyrequire testing all studentsat appropriatepoints in their progressionthrough school; different tests may be desirable,however, for the selectionfunction than those used for certification.Most externalexamining performed in Africa today,with the importantexception of Kenya, emphasizesselection, and althoughthis provides a base upon which to build, furtherdevelopment of educationtesting systems should ensure that testingsarves all of its importantfunctions.

To do this successfully,and particularlyto serve the curriculum improvementrole, examimtions must cover the full rangeof cognitive competenciesspecified to be coveredin the curriculum. These include,in additionto those competenciesintended for the minorityof pupils who proceed to the next level,all those neededby the majorityof pupils who do not. If the examinationsfail to test for these lattercompetencies, the schoolshave little or no incentiveto teach them, nor studentsto study for them. This resultsin primary educationbeing treatednarrowly as a preparationfor secondaryeducation, and not as a providerof worthwhileterminal skills.

Recent experiencein Kenya shows that it is possibleto shift from an examinationsystem that merely screenspupils to one that stressesthe full range of skills deemedappropriate for primary schoolpupils to be taught (see Box 7.2). Experiencethere also shows that examinationscan be used as an importanttool for monitoringand, ultimately,through constructive feed-back mechanismsincluding newsletters and regularsupervision, attenuating differencesin achievementbetween schools and districts.

[Box 7.2]

(b) Statistical and accounting systems

Investmentsin testingshould be accompaniedby the creationof strong monitoringand informationsystems that provide limitedbut strategically useful informationon schools,classrooms, teachers, students, materials, and financesto managersat all levels. These systemsshould emphasizesimplicity and practicalityof data gatheringand use, and they shouldbe based on careful analysisof informationneeds at differentlevels. They shouldbe integral parts of the organizationalstructure, and the tasks inherentin their operationbuilt into positiondescriptions and supportedwith training. In many countriesthe managementof these functionsshould be closelytied to the educationpolicy and planningunit.

The educationinformation systems should operate in two directions: informationshould flow both up and down the administrativehierarchy. The relevanceof statisticsfor improvingthe qualityof educationat the classroom level must be made clear to school-levelpersonnel, responsible for collecting - 123 -

Box 7.2 EXAMINATIONREFORM IN KENYA

To be admitted Into formal,government-funded secondary educatlon, secondarytechnical schools, and some trainingschools, Kenyan students are requiredto pass the Certificateof PrlmaryEducation (CPE) examination.The test consistsof several parts, Includlng: mathematics; English; and a general sectlon that Includeshistory, geography and sclence (naturestudy, agricultureand health). During the 1970s efforts to reformthe CPE grew out of concern that the test had llmitedrelevance for those studentswho would not go on to secondaryeducatlon, and that It rewarded memorization rather than reasoning abilIty. Critics charged that the CPE thus discouraged educational Improvementthrough curriculum reform and more Imaginative teachingmethods.

In responseto these problemsand to the demand for equal accessto secondaryeducation, the governmentIntroduced a new exam system. It has attemptedto make examinationquestions more relevantto the life experlences of Kenyan studentsand to test for thinkingability. In addition, the new system disseminates examinatlon performance data to schoolsto help them preparepupils to meet the new Intellectualdemands being made of them. Guidance InformatlonIs providedthrough the use of a newsletter,distrlbuted annually to all schools, field officers, professlonaleducators, the teachers' union and the press. Its purpose is to explainthe changesbeing made In the exam, Identifykey topicsand skills that have caused candidates particulardifficulty, and suggestways In which teachersmight help students Improvetheir performance.Related changes are graduallybeing Incorporated Into textbooksand teachingmaterials as well. Further,district and school merit order lists are publicized,making CPE peformancean important public Issue.

Thus the new examinationgives teachersan Incentiveto develop In their pupils relevantskills and knowledge,thereby Improving the quality of basic education. The new CPE Is Intendedto promotethe developmentof a set of competenciesthat will be useful both to those enterlng secondary school as well as to those for whom baslc educationIs terminal. In particular,the no.;system Is designedto provideencouragement and assistance to the less suecessful schools In an effort to reducequallty differences among schools and geographicalareas. And by basingquestions on materialrelevant to the experlence of most studentsand testingfor reasoningability, the new exam Is also Intendedto be fairerto underprIvilegedgroups. - 124 -

the raw informationand enteringit into the statisticalpipeline. School-levelpersonnel should be the beneficiariesof timely analysisof data for their units and of informationabout the largereducational system.

Testingand monitoringsystems may requiresignificant investments in facilities,equipment, and staffing. These investmentsmust addresscurrent problemsof data reliabilityand timeliness. Now that the capacityto analyze data is, with the recentdevelopment of microchiptechnology, no longer,even in very poor societies,a major constraint,the collectionof informationon a timely basis assumesrelatively even greaterimportance than in the past.

7.3 Strengtheninganalytic capacity

Analysisand planningare centralto efficientresource allocation and, therefore,to the achievementof qualityeducation under conditionsof austerity. Educationalleaders must increasinglybe able to assessthe performanceof their systemsand gauge the effectsof their policies. This takes high-orderanalytical skills organized in strong,well-staffed central policy and planningoffices. (In some largercountries, some of these functionsmay be decentralizedto statesor districts.)

The work of the policy and planningstaff must be well-integratedwith the policy-makingprocess. In many countriesthe lack of staffing, information,and resourceshas reducedsuch units to the status of statistics offices,concerned primarily with meetingthe informationneeds of donors. Such units are not able to use infermationto generatea range of policy optionsfor review,or to monitor -- and thus learn from -- the implementation of policy decisions.

Considerableeffort, centered mostly aroundtraining, has been made to developplanning offices and researchcapacity. The record,however, is not encouraging. To attractand retaingood planners,education organizations need to provideadequate compensation and career opportunities.Planning must be supportedby effectiveinformation systems, as discussedabove, and adequate operationalresources. The planningunit shouldhave clear structuralaccess to the highestlevels of the policy-makingprocess, and not be locatedin a temporaryproject management unit. In short,an effectiveanalysis and planningunit must be planned as an integralpart of overallorganizational design. Ad hoc arrangementsthat train a few plannersunder project funding and that fail to providecontinuing infrastructure support will not work. Here again improvementrequires sustained (although not large)investment accompaniedby structuralchanges necessary to make the investmentproductive. Providingtraining and incentivesfor enoughplanners, as well as the informationand resourcesnecessary for their work, will cost money.

Another elementof strengtheninganalytic capacity, and a crucially importantone, is that of developingthe appliedresearch base on which policy analysiscan rest. Three lines of researchstand out as particularly important. African educationresearch institutions should, of course,play a key role in helping to ensure that researchis timely,relevant, and of high quality. Since such institutionsare only just beginningto appe&r in much of Africa, investmentto strengthentheir capacityand willingnessto conduct researchin these areas shouldbe high on the agendafor action. - 125 -

First,more descriptiveoperations-related research is urgentlyneeded on the cost, finance,and runningof educationand trainingsystems. How much do differeuttypes and levelsof educationand trainingcost? How are they financed? Who providesand who uses them? What materialsare availableto studentsand who providesthese? How much time do teachersand students actuallyspend in school,and how do they spend that time? Are schools orderly and disciplined? Simple thoughsuch questionsare, answersare frequentlyunknown; either quantitativeor qualitativeanswers would clearlybe relevantto improvingoperations. A great deal can be learnedfrom selective improvementsin routinelygathered performance and expenditureinformation coupledwith more intensiveanalytical use of these data. When these data are combinedwith performancedata from the examinationssystems, vital questions of cost-effectivenesscan be addressed.

The second type of researchinvolves use of surveymethods to assess the internalefficiency of education. The InternationalEvaluation of Achievement(IEA) program constitutesan importanttool for analysisof internalefficiency; its tests providea discriminatinginstrument with which to measure the qualityof schoolsin terms of their successin raising achievementin such core areas as mathematicsand science. Levels,trends, and determinantsof academicperformance can now be trackedon an (almost)routine basis. Such objectivefeedback on performanceis the key to accountabilityand quality improvementin education. Wider and periodicapplication of IEA-type instrumentsoffers great potentialfor improveddecisions on resource allocation. Researchhere is dependenton strengtheningof the testingsystem, as discussedabove.

A third relativelyroutinized line of researchdeals with external efficiency,i.e., with assessmentof the labormarket performanceof graduates of differentlevels and types of education. The LivingStandards Survey (LSS) approach-- pioneeredwith World Bank assistancein Cote d'Ivoireand Peru -- providesan inattumentfor assessingperformance on severaldimensions (see Box 7.3). Plans are now being developedfor 10 otherAfrican countriesto begin use of LSS techniques. LSS serves a very broad range of purposes,and it will be importantto ensurethat its potentialfor illuminatingeducational policy is in fact realized.

(Box 7.3]

7.4 Managerialstaff development

Lack of skilledmanagers and low morale are fundamentalconstraints on African educationat all levels. Past effortsto addressthese constraints have not been successful. Traininghas generallynot been tied to clear longer-termstrategies for organizationaldevelopment. As a result,training has been general in natureand not well-linkedwith managementdevelopment needs. Trainingopportunities have principallybeen providedfor higher level staff (oftentrained abroad) and not focusedon schools. And this training has been conceivednarrowly as a single input.

A broaderpolicy approachto the developmentof managementskills is needed in most countries. Fundamentally,this will requirethe developmentof strongmanagement institutions, a relativelylong-term solution. In the short-term,governments should consider a number of immediateoptions. In view - 126 -

Box 7.3 LIVING STANDARDSSURVEYS (LSS)

WuIlenational data that measuremacro-economic growth werformance areavailable in most countries, the quantitativebasis to assess the distributlon ofthe frults of economic growth and the shareaccruing tothe poor Is much weaker In the majorityof countries. In response, the World Bank launchedthe Living StandardsMeasurement Study (LSS) In 1980.The goal was to set upa datasystem that wouldallow the monitoringof the well-belngof householdsand IndilvidualsIn different soclo-economicgrowPs of the societyIn the curseof development. Unilike standard household surveys, fromwhichdataare typically avallableonlyyears after collectlon,the LSMSsystem was to produce Informationquickly and In a formreadlly suitabie for analysisthat coulddirectly Informeconomic and socialpollcies.

Thethree key features of a properLiving Standards Survey (LSS) are the coavrehensivecoverage of the marydlmensIons of weIl-beIng,the sephasIsonqualIty control, 1nd speedydata processing. LivingStandards Surveys collect informationon the demographiccharacteristics of houssholdmembers, their educatlonalachievement, health condition,migration history, useof time In the homeand the labormarket, and earnings obtained as a result. Thehousehold's housing conditlon Is surveyed,as are Its owwershipof assets and enterprlses (and the earningsthey generate), Its expenditures,borrowings, andsavings.

Tobe useful for policy, survey Information mUStbe current and of high quality. LSS qualIty assurancetecnlmques Included extensive training and refresher pgrading of eruwators,a h10w than ursalswWervisor-to-enusrator ratio, useof personalcoAuters (PCs) In the field for contimuousdata etry and Instantaneousrange and validity checks,and verificatlon of Identified errors and inconsistencmisby the enmeratorIn a seconddata collection ro.ud. ExperlenceIn the Cote d'lvoire and Peru, vwerethe LSSsystem has been tested on a natlonwidesoale, has shon that tablescan be prodluedwithin 3-4 monthsafter datacollection. Analysis;n the LSSsystem can thustake place In parallel with data collectlon. Aseach new wave of datareaches headquarters, tables can qulkly be xpdated.

TheLSS generated data base is usefulto studymany aspects of edc.atlon. In addition to ekicatlonal attainment,the educationmodule covers self-reported lIteracy and rneracy, formal and informalvocational and tectrical training, andthe schoolIngof chiIdreneho have left the household. All this ls usefulto studythe total returnsof ttshousehold's investment Ineducation. For wurrently enrolledstudents, inforeation is gatheredon the typeof school,distanc andtlme of travel to school, out-of-pocket ,an scholarships.Opportunity costs can be derived from the epilymentdata. This permitsa stuty of privatereturns to differenttypes of education.The avallability of Informationon other aepeotsof well-beingcan throw i ghton the lInksbetween educatln and pollcies In otherareas suchas health,labor markets, and migration. Ho dograduates of different educational curricula fare In obtainingobs? i eObetter edicatlonImprove one's use of health faciIltles? AMdconvrsely, does better healthleprove educatlonal attalnment? How do parental characterlsticsand migrationhistory affect acoessto educatIon?9hU1d fees be carged for schIoolIng? Howmich can Poor peopIe afford to payfor prilary schoolIng? - 127 -

of the difficultyof attractinghighly qualifiedindividuals to the education service,priority should be given to developingand advancingtalented persons within the educationsector. This will requirethat managementdevelopment be seen as a system-wideactivity, one that builds from good schoolmanagement through intermediatelevels towards the center. Trainingshould be seen as one of severalinputs. The others includeresources necessary to make staff effective,incentives to hold them in the system,definition of careerpaths, and systems for performanceassessment. A teachershould enter servicewith a clear idea of opportunitiesfor advancement,and of the kinds of performance and trainingthat will be necessaryto move ahead.

Managementdevelopment in the centralministry should be done in parallelwith developmentat lower levels. Trainingacross levels (for example,of headteachersand districtofficers together) should be commonplace, to strengthenrole relationshipsand coordination.

Such effortsare heavily constrainedby the lack of skilledmanagers and managementtrainers. One policyoption for considerationis enhancingthe incentiveand career opportunitystructures for highly skillededucation staff in order to attracttalented nationals. Such a policy is difficultto implement,of course,for a singlesector and is more likelyto succeedwhen appliedacross sectorswith carefulcontrols. Governmentsand donorsalike might well review currentpractice and fundingfor expatriateassistance to see if ways can be found to divert resourcesfrom expatriatetechnical assistance, with its very high unit costs, towardssuch a policy.

Relianceon expatriateassistance, however, is likelyto continuein many countriesover the short term. Experiencehas demonstratedsome measures that can be taken to ensurethat such assistanceis effectivelyused. First, expatriateassistance should be incorporatedwithin a longer-termplan for developmentof human resourcesin the sector (as is currentlythe practicein a number of African countries). Such planningcan give the governmentand its officialsa clear sense of controlover the use of high-costexpatriates and of the phases throughwhich their roles will be assumedby nationalstaff.

Second,expatriate staff shouldbe assignedto line operatingunits where their expertisecan be acquiredby nationalcounterparts -- not isolated in project implementationunits without counterparts.These assignmentsshould be carefullyplanned by governmentand donor alike to includeas principal responsibilitiesboth formaland informaltraining; expatriate performance shouldbe evaluatedon this basis. In some cases a trainingspecialist should be includedon expatriateadvisory teams to ensure that expertiseis effectivelyturned into trainingand trainingmaterials.

Overall,a combinationof policiesthat (a) increasethe recruitment, developmentand retentionof skillednational staff; (b) base expatriateuse on broad human resourcedevelopment planning; and (c) give priorityto the trainingrole of advisorsoffers considerablepotential for improvementin the short-term. Longer-termsolutions require the developmentof permanent managementdevelopment capacity and, ultimately,raising the abilityof African institutionsto providethe full range of managementdevelopment services, includingfirst-rate research. - 128 -

7.5 Prioritiesand resourcerequirements

Substantialresources will be requiredto createa management environmentin which Africaneducation can be improved. The policies recommendedabove envisagesignificant investments in more attractiveincentive and career paths for managers-- and investmentsin training,analysis, and the creationof what, for many governments,will be the establishmentof new or much strongerunits for testing,accrediting, monitoring, planning, and training. Managementdevelopment institutions will requiresubstantial financialsupport. Most measureswill incur significantrecurrent costs to support the implementationof eithermanagement improvement.

Such investmentspose difficulttrade-off questions in the current financialand economicclimate, in which resourcesfor educationsystem managementhave stagnatedor declinedin many countries. However,substantial investmentin managementcapacity is important to the successof other reforms, includingfinancial reforms considered necessary for establishingappropriate prioritieswithin education. Some of these costs may be financedout of savings generatedthrough policy reform. Othersmay have to come from difficultchoices in other importantareas.

A large share of these resourceswill have to come from the donor community. The patternof donor support for educationmanagement, however, will have to change significantlyif these resourcesare to be effectively applied. A much higher level of donor coordinationwill be required;this in turn will dependon the developmentof longer-termmanagement development plans for a given countryor region. Episodicproject-related investments will probablyneed to be replacedby long-terminvestments in managementcapacity per se, includinginvestments in well-designedmanagement development institutions.

The scarcityof resourcesin the face of large needs arguesstrongly for a strategicapproach to educationmanagement development that recognizes the validityand urgencyof short-termneeds but does so in the contextof longer-termgoals for institutionaldevelopment. In such circumstancesa strategyfor managementdevelopment must be developedout of the range of options-- for organizationalstructure, testing and informationsystems, analyticcapacity building and staff development-- that were discussedin this chapter. At the outset,however, as a first priority,this strategyshould identifythe investmentsand activitiesnecessary to achieveacceptable levels of schoolmanagement, with all of the supportingsystems necessary. This leads to: - 129 -

Recomendation5. The design of a policy frameworkto enhancethe efficiencyof resourceuse in educationgenerally occurs at the nationallevel, but successfulimplementation of policieswill usually depend,in the final analysis,on the strengthof managerialcapacity throughout the education system, and especiallyat the level of the individualschool. This capacity includes,among its essentialingredients, the presenceof school leaderswho possessboth strategicanalytic skills and the freedomto act on behalf of their school clients,and also the timely availabilityof relevantinformation on which to base decisions. Accordingly,the first prioritiesfor education managementshould be, in mos. Africancountries: (i) the improvementof programsfor selecting,training, and supervisingschool headteachers and principals,combined with enhancedinstitutional autonomy; and (ii) the developmentand implementationof achievementtesting systems that provide feedbackon institutionalperformance to individualschools, to their supervisors,and to the communitiesthat they serve. - 130 -

PART THREEs AN AGEND FOR ACTION

An extensivemenu of policy options for Africaneducational developmenthas been presentedin the chaptersabove. The adoptionand implementationby any one African countryof all of the discussedpolicies is probablyneither desirable nor feasible. The educationalachievements and aspirationsof individualcountries, and the socio-economicand political constraints,vary too much for any single comprehensiveset of measuresto be completelyor uniformlyapplicable. This paper is about diverseand variably applicablepolicies for African educationaldevelopment, not about a monolithic and universallyvalid policy.

Nonetheless,the analysisof the predicamentof Africaneducation and trainingsystems in Part One of this report and the discussionof policy options for differentlevels of educationand for sectormanagement in Part Two do imply a generallyapplicable three-dimensional framework for the design,at the country level,of educationalpolicies. for the region. Althoughspecific detailswill differfrom one African countryto the next, elementsof all three dimensionsare certainto be found 'n any carefullyconceived policy package tailoredto the particularcircumstances of any country in the regiontoday. The three dimensions,to be elaboratedupon in Chapter8, are labelled adtustment,revitalization, and selectiveexpansion.

While undoubtedlypainful and politicallydifficult, adjustment policieswill alleviatethe burden of educationand trainingon public budgets. Measuresfor revitalizationand expansion,on the other hand, will certainlyrequire incremental resources. Thus, in the contextof ongoing austerityin Africa, resolutemovement toward adjustmentis a necessary conditionfor implementingforward-looking policies on the other two dimensions. "Savings"generated fron adjustmentwill be needed to help fund educational improvement and expansion.

Regrettably,all such savingsfrom adjustmentmeasures will not be sufficientto cover the very substantialresource requirements of the policies needed to revitalizeand build African educationto the extentessential for futuredevelopment. Internationalaid will remain a criticallyimportant determinantof the pace of progressof educationin the region. However,the rapid evolutionof Africanneeds, as saumarizedin the three dimensional framework for policy reform, demands corresponding changes in the organization, nature, and level of international aid for African educational development. Chapter9 reviewsthe patternof public Internationaldevelopment assistance in the early 1980s,and it sketchesthe modificationsnecessary to make future assistancefully responsiveto and supportiveof Africa'sdetermination to make educationwork better for development. - 131 -

Chapter 8. POLICY PACKAGESFOR EDUCATIONALDEVELOPMENT

Hard decisions on education policy can be postponed -- but only at the cost in most African countries of continued enrollment stagnation and quality decline through the 1990s. Part Two of this paper has reviewed a series of policy measures -- some of them admittedly difficult -- that hold the promise of restoringquality and resumingan orderlyexpansion of enrollments. This chaptersuggests that each African countryshould now embracethe task of formulatingand implementingan internallycoherent set of policiesthat, while reflectingthe country'sunique historyand aspirations,will effectively address its own recentlyexacerbated problems in the educationand training sector. Althoughthe particularsof the policypackages that emerge from this exercisewould vary from one countryto the next, it is nonethelessclear that every country-specificpackage will need to contain,in varyingproportions, elementsof policy along three distinctdimensions. These three are adjustment, revitalization, and selective expansion. Moreover, if new policies are in fact to be implemented,management practicesneed to be improved.

The three dimensionsof policy are definedbelow and further elaboratedin the courseof the chapter. The closingsection of the chapter then discussesorganizational issues associated with next steps in policy developmentand implementation.

Adiustmentto currentdemographic and fiscal realities,though it will be difficult,is essentialif the disruptiveeffects of these externalfactors are to be minimizedin the years ahead. Adjustmentwill take two main forms:

o The diversification of educationalfinance will be a necessarypart of country-specificpolicy packages. This diversificationcan be achieved throughincreased cost-sharing in public educationand throughincreased officialtolerance and encouragementof nongovernmentalsuppliers of educationalservices.

o The rigorouscontainment of unit costswill be just as importantin the adjustmentprocess, and in many countries,probably more important,than diversifyingfinance.

Revitalizationof the educationalinfrastructure that now exists in order to restorequality is the seconddimension of a properly conceived educationalstrategy. This involvesa renewedfocus on the fundamentalsof educationalprovision so that maximum advantageis extractedfrom the current capacityof educationand trainingsystems. Three kinds of measures are necessaryfor the restorationof quality:

o Textbooksand learningmaterials must once again become generally availablein African classrooms.

o There must be a renewedcommitment to academicstandards, principally throughstrengthening examination systems.

o Greater investmentmust be made in maintenanceof physicalplant and equipmentand in operationalexpenditures. - 132 -

Selectiveexpansion to addressneeds for furthereducational services is the third dimensionof any completestrategy for educationaldevelopment. Measures in this area, viable only after measuresor adjustmentand revitalizationhave begun to take hold, will concentratein four areas;success in all will dependupon a generaleffort to safeguardthe qualityof instructionalstaff at all levels.

o Renewalof progress towarduniversal primary education is the new investmentthat will bring the highesteconomic and socialreturne in many countries.

o At the secondarylevel, and later on at the tertiary,expansion of enrollmentsin selectedsubjects and streamswill be necessaryin most countriesin the years ahead. To accommodatethese increasesin post-primaryeducation, most countrieswill need to consideralternative deliverymodes that shift more of the burden for learningonto the studentsthemselves; now is the time to begin planningfor such programs and developingtheir requisitesupport infrastructure (correspondence materials,radio programs,and examinationssystems).

o The amountof trainingthat occurs once individualshave enteredthe labor force must be increased;this trainingshould serve both school leavers and also those who have had no exposureto formal schooling,so that individualscan acquirethe necessaryjob-related skills and renew these skills duringtheir working lifetimein responseto changingmarket conditions.

o Expansionof African capacityto produce its own postgraduateintellectual talentto fill the highestscientific and technicaljobs in educational establishments,in government,and in the privatesector is an important matter to be addressedin buildingfor Africa'sfuture.

The message here is:

Recommendation6. To maximizeeducation's contribution to economic growth in the years ahead,African governmentsshould design and begin expeditiouslyto implementlong-term education sector development programs. Each governmentwill need to select,from among the many policy options available(including, but not restrictedto, those discussedin this paper),a consistentpackage of policiesthat effectivelyaddresses country-specific problemsand serves country-specificgoals. While the optimalpolicy package will differ from one countryto the next, no countrycan afford, in light of education'shigh costs and its crucialrole in the developmentprocess, to neglectthis task. The policypackage that emergeswill, in every country, containelements of three strategicdimensions -- adjustment,revitalization, and selectiveexpansion. Given the pressureof populationgrowth and fiscal constraints,most countries will need first to implementpainful adjustment measuresso as to generatethe necessaryresources for quality-enhancing revitalization;and the combinationof adjustmentand revitalizationshould be regardedas prerequisitesfor the longer-termreturn to selectiveexpansion of the educationalsystem. - 133 -

8.1 Adjustment

The two elementsof an adjustmentprogram for educationand training are diversifyingthe sourcesof financeand containingunit costs.

(a) Diversifyingsources of finance

The imperativehere is to acquirefrom the beneficiariesof education and traininga much larger share of the real costs of providingthese services. As regardstraining, an increasein the degree of cost-sharingis a normal concomitantof moving the locus of trainingcloser to the workplace,a move justifiedon other groundselsewhere in the paper. As regardseducation, the rationalefor increasingthe degreeof cost-sharing,relative to the situation common at present,is strongestat the tertiarylevel but may, in some countries,be pertinentat lower levels as well.

Increasedcost-sharing can be achievedthrough a varietyof measures. First, countriesmight considerencouraging the establishmentof, and relaxing regulationsthat currentlyconstrain the operationof, privatelyowned and privatelyfinanced institutions of secondaryand higher education. Encouraging and facilitatinglocal or NGO effortsat construction,finance, and school operation,especially at the primaryand secondarylevels, can be expectedto enhancethe opportunityfor resourcemobilization and improvedmanagement in responseto local needs.

Second,there is a clear rationalein secondaryand tertiaryeducation in many African countriesfor introducingor raisingthe level of family responsibilitywith respectto the costs of food, lodging,and other items of studentmaintenance unrelated to instruction.These are costs that must be borne whether or not individualsare enrolledin school. There is litt.leif any justificationfor financingthese costs out of the publicbudget, especiallyin poor countrieswhere only a minority (and, nearly always,a relativelywell-to-do minority) is enrolledat secondaryand tertiarylevels. Beyond full recoveryfrom studentsand their familiesof student welfare-relatedcosts, including especially costs of staff servicesunrelated to instruction,students themselves could be requiredto perform a varietyof instruction-relatedsupport tasks now assignedto non-teachingstaff (custodial care of teachingfacilities, upkeep of the grounds,clerical and secretarial assistance,and other administrativesupport). Studentprovision of such servicesin kind, a widespreadpractice in many developingcountries, is a potentiallyimportant approach to alleviatingthe public financialburden of secondaryand tertiaryeducation.

Finally,many countrieswill need to introduceor raise tuitionfees in public establishmentsto cover a significant part of the costs of instruction. Especiallyin higher education,however, full cost-sharingcould result in many potentialstudents finding themselves excluded from education owing to their families'inability to pay. To the extentpossible, ability to pay should not be a factorthat determineswho receiveseducation and who does not. Therefore,alternative modes of financingthe increasedprivate costs of educationwould need to be introduced. Possibilitiesinclude: opening opportunitiesfor nationalservice obligations (e.g., to teach school)prior to, during,or after a student'senrollment; promoting student loans and other - 134 - educationalcredit markets; and, during a period of transitionto an effective system of graduatedincome taxation, imposing a specialtax on the earningsof tertiary-levelgraduates.

For most policiesaimed at decreasingthe government'sshare of full educationalcosts, implementationwould need to be phased in graduallyover a period of some years. And legitimateconcerns on the policies'equity impact would need to be addressed.

(b) Unit cost containment

Unit cost containment,the more importantcomponent of adjustmentin much of Africa, shouldbe aggressivelypursued at all levelsof the education system on both the capital and recurrent accounts. The goal here is to reduce the economic costs per student, or per completed cycle, at each level (and not necessarily to reduce the total aggregate expenditure at any level).

Reducingconstruction costs and raisingphysical facility utilization rates offer considerablepotential for reducingunit capitalcosts. Since annualizedcapital costs can accountfor 40% or more of the direct economic costs of educationin Africa,careful attention to capitalcosts is a high priority. Too many existingeducational facilities (particularly those financedin part throughforeign assistance) cost a sizeablemultiple of what best practicehas shown to be possible. Chapter2 projectedthe probable magnitudeof capitalexpenditure in educationbudgets in the comingyears under the assumptionthat best practicewould prevail. In order to contain constructioncosts to the levelsthere indicated,it will be necessaryto minimizeexpenditure on pedagogicallyredundant civil works (such as boarding facilities,auditoriums, cafeterias, and sportscomplexes) and to utilize low-costconstruction methods and localmaterials. In addition,particularly at the secondaryand tertiarylevel., there is scope for more intensive utilizationof existingfacilities through extension of the teachingcalendar. There is little or no justificationin most cases for the currentpractice of closingdown facilitiesin the eveningsand duringthe long vacationperiods. Absent compellingextenuating circumstances, new constructionshould be deferreduntil full utilizationof existingcapacity, as well as of facilities under constructionbut still to be completed,is reached.

Since personnelremuneration is the largestsingle item of education costs at all levels,modification in how teachersand non-teachingstaff are paid and utilizedis a potentiallyimportant element of any strategyto contain unit recurrentcosts.

Recent experiencehas demonstratedthat in some countries,there is a large payoff to purgingfrom the educationalpayrolls so-called "ghost teachers"-- those who are not actuallyassigned to, or not fulfillingtheir assignedobligations to, classroomduties. There likely remainsa significant potentialto be exploitedin this sensitivearea, at least within the primary and junior secondarysubsectors.

Reductionin the typicallyhigh numbersof non-teachingstaff in African secondaryand, especially,tertiary educational institutions should everywherebe pursued. Transferto studentsof significantresponsibility for performanceof essentialinstruction-related custodial and administrative supportservices was includedearlier among the financediversification - 135 - measures. But beyond this, the levelsof provisionof such categoriesof personnelas messengers,drivers, watchmen, stewr-ds, and sweeperscan usually be reducedwithout sacrificeof educationalquality. But becausesalaries at this level are low, even very sharp reductionswill have onl.y a limited, althoughnonetheless positive, impact on unit costs.

Reductionof teachers'rates of pay is also a policy optionthat shouldbe consideredin countrieswhere it can be demonstratedthat a sufficientsupply of teachersof comparablequality would still be forthcoming. It is true that inflationhas underminedteachers' salaries substantially during the past two decades,as it has all public sectorwages in the region. Even if it were politicallyfeasible, further reduction in real ratesof pay for teachersalone (in isolationfrom other civil servants)could resultin loweringthe qualityof educationin some Africancountries. Certainly outritghtcuts in nominalwages would appear to be undesirableeverywhere becauseof the adverseImpact on teachers'professional motivation and commitment. All this notwithstanding,the growingnumbers of unemployed secondaryschool leaversand of universitygraduates and dropoutsin Africa suggestsa potentialpool of laborwith adequatemotivation and intellectual skills for teaching;they ultimatelymay be prepatedto acceptwages somewhat lower than currentlyprevail in the teachingprofession. Some reductionin teacherwages, probablyachieved by allowinginflation to erode the real earningsof all public servantsincluding teachers, may thus be possibleover time. This would be facilitatedif certificationrequirements for the teaching professionare relaxed,a point closelyrelated to entry qualifications,to which we now turn.

Althoughstill difficultand controversial,another policy option is to reduce,especially at the primaryand lower secondarylevels, the minimum entry qualificationrequirements for teachersin terms of their years of preserviceeducation and training-- and thus reducethe salariesthat they must be paid under the existingcompensation structure. The cost (in terms of childrenwho, thereby,are denied access to schooling)of trying to match in the world's least developedcountries the preserviceeducational requirements typicalof the world'smost developedcountries is unjustifiedfor two reasons. First,marginally lower entry qualifications-- e.g.,one or two fewer years of schoolingand preservicetraining for those teachingthe early primary grades -- may not impactnegatively on what pupils learn in the classroom,especially in Africa todaywhere the governingconstraints on learningare the lack of instructionalmaterials and lack of effectivetime on task. Second,given the relativecosts of providingpreservice and in-serviceteacher training, and given the salary structuresthat prevailin most places, in-servicetraining is usually the more cost-effectivemeans of raisingthe qualityof classroom instruction. For primaryschool teachers especially, many African countries should consider imposing further limitations on the quantity of preservice education and training, and coupling this with a policy of frequent in-service coursesfor the upgradingand refreshmentof teachers'skills, with particular attentiongiven to subject-mattercompetency and the proper use of instructionalmaterials.

More intensiveuse of teachingstaff is potentiallythe most fruitful approachfor reducingunit recurrentcosts. This will result in unit cost reductionsso long as the percentageincrease in teachingtime is greaterthan the percentageincrease in salariesneeded to motivate and compensatethe teachers for their greatereffort -- i.e., so long as teacherscan be induced - 136 - to accept some reductionin hourlywages in the contextof an increasein their total earnings. Even if averageteachers' wages do not decline,the more intensiveuse of teachersas teachers(as distinctfrom implicitlycondoning their performanceof other labor,such as tendingto the family farm or small business)may allow unit costs to be reduced.

As indicatedin Chapters4-6, the optionsto be exploredfor improved utilizationof teachersare numerous. They include: lengthening,through a varietyof schemes,the teachers'(but not the students')school day; holding classessix days a week in primary schoolswhere this is not alreadydone; increasingthe teachinghours per week in secondaryand tertiaryinstitutions to levelsmore closelyapproaching norms outsideAfrica; reducingthe vacation periods so that teachers (and facilities)are employedmuch more than the 36 weeks per year now common;increasing the minimumnumber of studentsin a class, especiallyin secondaryand tertiaryinstitutions where coursesare often thinly subscribed. In additionto all such measures,each of which would result in some modest reductionin unit costs,most developingcountries, if they aspireto bring about nonincrementalchanges in educationalaccess over the next severaldecades, will need to considersome fundamentallydifferent alternativesfor the deliveryof educationalservices (see, below, discussion on extramuralstudy programs).

An analysisof tertiaryeducation in Africa leads to a special conclusionabout teacherutilization at this level. For reasonsdiscussed in Chapter6, it will be desirablein most African countriesto stabilize,or even reduce the number of universitystudents in the short-term(i.e., the next five-to-tenyears). This adjustmentwould be achievedby contracting relativelylow priorityfaculties (e.g., arts and law) and by consolidatingthe number of institutionsand academicprograms within a countryor across several smallercountries. This contractionand consolidationwould allow for a substantialimprovement in facilitiesutilization and increasesin class sizes in core programson the consolidatedcampuses, thereby increasing the productivityof the teachingpersonnel employed there. A painful -- but absolutelynecessary -- concomitantof this adjustmentmeasure in higher educationwill be reductionsin facultynumbers beyond what would occur through normal attrition. Substitutionof expatriateprofessors x7ith (somewhat fewer) African academicswill in some fieldsmake it possibleto avoid dismissing currentlyemployed teaching staff. However,in selecteddisciplines in arts and humanities,where overproductionof personneltrained at the tertiarylevel is most severe,the redundancyof African teachingstaff is inescapableand must be squarelyfaced.

8.2 Revitalization:restoring quality

The second dimensionof an educationalstrategy for Africa involves revitalizingthe infrastructurenow availablefor educationand trainingin order to restoreand enhancequality. This amountsto a renewedfocus on the fundamentalsof educationalprovision: instructionalmaterials, academic standards,the maintenanceof equipmentand physicalplant. Studentsand teacherswaste their time (at great cost) for lack of textbooksand other learningmaterials and for lack of effectiveexamination systems to set and maintainstandards. Schoolsupervision systems, vital to educational performance,come to a halt when there is no money to operatevehicles or pay for telephoneand other communicationservices. In addition,buildings and equipmentdeteriorate for lack of maintenance,and expensivelaboratories are - 137 - not used for lack of reagentsand spare parcs. These items of non-salary recurrentexpenditure have been highlyvulnerable when budgetshad to be cut, even thoughthey amountto only a small fractionof total education expenditures. Appropriatebalance in the input mix must be restored immediately.Measures aimed at systematicand sustainedrevitalization of the educationenterprise, through renewed commitment to academicstandards and non-salaryrecurrent expenditures, are essentialsupplements to adjustmentpolicies if efficiencyand sustainabilityare to be achieved.

(a) Instructionalmaterials

The top educationalpriority in Africa today is to ensure that every child in every classroomhas accessto the pedagogicallynecessary minimum of instructionalmaterials. What that means in concreteand practicalterms is of course differentfor the several levels,and within levelswill vary by grade, subject,and curriculumcontent. Methodsof supplyand financewill also vary from countryto country. But only those countriesthat accord central importanceto provisionof instructionalmaterials can be judged to have put in place adequatestrategies for educationaldevelopment.

Difficultissues will have to be confronted: what pedagogical materialto developlocally and what to purchasefrom abroad;the trade-offs betweenhigher-cost local printingand least-costprinting elsewhere in the region or, more usually,outside Africa; the use of nontraditionalmedia such as radio. In its own recentand thoughtfulpolicy paper on education,the African DevelopmentBank spoke eloquentlyto these issuesin placinggreat emphasison the importanceof addressingthe instructionalmaterials problem within the overallcontext of restoringnon-salary recurrent expenditures to adequatelevels:

The supply of appropriate teaching materials is particularly inadequatein large parts of Africa. While this is to some extent a questionof finance,the issue of producingand distributingadequate teachingmaterials for African schoolsgoes much beyond the question of funds. As there is an urgent need not just for any teaching materialsand textbooks,but for materialsthat are more closely in tune with the realitiesand needs of Africansocieties, a major field of lendingactivity opens up here. Bank Group loans will support, not just some of the technicalassistance needed in modifyingand adapting existing textbooks and materials and preparing new materials, but also the production and distributionof these materialsin Africa. EducationalResource Centers in areas where there is a particularlyserious shortage of instructionalmaterials could be anotherexample of this generalthrust. In this area of quality and internal efficiency,as the majorityof the non-salary inputshave a directeffect on the qualitativeaspects of education, the Bank Group will give priority to assistingregional member countriesidentify and maintain minimum standards for non-salary inputs. (African Development Bank EducationSector Policy Paper, 1986, pp. 15-16.] - 138 -

(b) Academicstandards

Beyondputting instructionalmaterials in clasrrooms,the restoration and clarificationof standardsof academicperformance in educationare key elementsof qualityimprovement. This is importantat all levels. Academic expectationsfor studentsand schoolsshould be high, and they shouldbe clear. By providingsignals on performanceto which teachers,students, and parents can respond,the examinationsystem is a powerfulmeasurement-cum-incentive devicethat shouldbe used explicitlyfor raisingacademic standards. But to perform this function,most Africanexamination systems need modificationso that the broader range of cognitivecompetencies sought by a nation from a given level of the educationalsystem is sampledon the tests, ratherthan their concentrating,as is now often the case, narrowlyon those skillsmost needed for successat the next level of the system. Only if examinationsare so structuredwill the curriculumbe made pertinentfor the majorityof studentsfor whom any level of educationis now terminal.

(c) Operationand maintenanceof physical plant and equipment

Preventivemaintenance and repairof physicalplant and equipment, another item of non-salaryrecurrent expenditure, is an essentialingredient to revitalisingAfrican educationsystems. Qualityeducation is just not possible in laboratoriesand workshopsthat have no electricityor water due to deteriorationin wiring, fuses,and plumbing,and where equipmentdoes not operate for lack of spare parts and consumablesupplies. Maintainingdoor and cabinet locks in fullyworkable condition, replacing broken windows, repairing leaky roofs, changirlthe oil and filtersin heavy use fieldvehicles -- these may be simple things to do, but they are not gettingdone. Failureto do them means that vital equipmentis not availableand functioningwhen needed,and unit costs are inflatedwhen prematurereplacement is the consequence.

Once plant and equipmentare restoredto fully functionalstatus and their maintenanceattended to on a routinebasis, money is also needed so that these resourcesdo not sit idle. In this respect,for example,reestablishing adequatebudgetary provision for such simple items as petrol,postage, and telephoneservice is essentialif school supervisionsystems at the primaryand secondarylevels and fieldwork at the tertiarylevel are to make their intendedcontribution to educationalproductivity.

8.3 Selectiveewpansion

The third dimensionof an overallstrategy for educationand training in Africa entailsthe consideredand deliberateexpansion of selected educationalservices. Public supportfor the sectorhas been threatened recently,and will continueto be threatened,by the fiscalausterity that grips the region. Wise leaders,however, will do what they can to protectfrom debilitatingauts those long-termeducation investments that promise the most for their nations'future.

Althoughthe relativeweights given to differentelements may differ from one African countryto the next, most forward-lookingeducation sector programs wi'Ll need to put some emphasis on each of the following: renewed progress toward universal primary education; new programs of extramural study - 139 - at secondaryand tertiarylevels; development of a broad-basedsystem of training;and enhancementof graduateeducation and strengtheningof research capacity.

In addressingthe needs for selectiveexpansion, safeguarding the qualityof instructionalstaff will be essentialat all levels and will require specialattention. Motivated,knowledgeable, and pedagogicallycompetent teachersare a vitally importantIngredient of educationalquality. Revitalizationof African educationinvolves eliminating the currentconstraint on classroomlearning through ensuring that the instructionalmaterials that teachersneed to be effectiveare everywhereavailable. But beyond this, maintainingquality in African educationas it again becomes economically feasibleto encourageselective expansion of the system,will requireconstant professionalrenewal of the teachingforce, a matter of particularimportance to the extent the standardsfor teachers'years of preserviceeducation are relaxedin order to containcosts.

In order to maintainthe qualityof instructionalstaff, the design and implementationof cost-effectivesystems of continuousin-service training for primary and secondaryschool teachersshould be part of every country's educationpolicy package. This trainingshould focus on upgradingand updating subject-matterknowledge and on masteryof improvedpedagogical methods embodiedin widely availableinstructional materials. Distanceeducation is likelyto be the most attractivedelivery mode for much of this activity. Further,teap)bers* progression up throughthe steps of the salaryscale should be conditionedupon successfulcompletion of such regularin-service courses, therebyensuring that pay is more closelyrelated to potentialclassroom productivitythan it is now. For the tertiarylevel, beyond reestablishingthe flow of standardtextbooks, new monographs,and journalsupon which much on-the-jobprofessional improvement depends, upgrading of teacher qualitywill entail increasedinvestments in formalpostgraduate education.

(a) Renewed progress toward universal primary education

For many countries, the most important long-term investment -- in terms of its economic and social returns -- will be to renew,after adjustmentand revitalizationmeasures have begun to take hold, national progresstoward universal primary education(UPE). Renewingprogress toward UPE will, inevitably,require the mobilizationof substantialnew resources including-- but not limitedto -- increasedcommitments of public resources. Chapters2 and 3 showedhow costly and difficultis the task of merelykeeping up with populationgrowth. Yet, two lines of evidencesuggest that, in many countries,very high priorityshould be accordedto the goal of continued progresstoward UPE, while in a few countries,universal provision of nine years of educationmay alreadybe an appropriategoal. First, relativeto projectsand investmentsin other sectors,the implementationand sustainabilityrecord of edrcationinvestments Is good; absorptivecapacity has been demonstrated. Second,there is strong (and mounting)economic evidence indicatinghigh returnsto investmentin education,particularly primary education.

Those returnswill be greatlyattenuated or never materializeat all to the extent that the incidenceof diseaseand malnutritionamong young childrengoes unchecked. Policypackages for educationaldevelopment must give attentionto how to ensure teachablepupils, a point of increasedimportance - 140 - since seriousnutritional deprivation in many parts of Africa has become a perennialproblem. However,the linkagesbetween child health,nutrition, preschoolintellectual development, and attendanceat and performancein school are particularlycomplex and only imperfectlyunderstood. Regionaland internationalcollaboration in the search for more completeanswers offer the best hope of developingpractical policies to remedy the situation.

(b) Distanceeducation programs

Perhapsthe only viableway of addressingthe massiveproblems of access in Africaneducation at the secondaryand post-secondarylevels and of continuingeducation for the teacherforce involvesthe use of new extramural study programs. Accreditationexaminations allow certificationof an outside student's"equivalency" to having completeda conventionalprogram. Students can thus acquirediplomas and degreesby independentstudy, typicallyguided by correspondencematerials supplemented by radio broadcasts. The replicability of high-qualityteaching materials allows high performancestandards to be set and maintainedin the equivalencysystem. Moreover,by their very nature, equivalencyprograms tend to favor the most persistentand motivatedstudents. Often "extramural"programs actually involve the use of existingcampus facilitiesduring eveningsor vacationmonths for tutorialsessions and laboratorywork. Sometimesequivalency education takes place in specially createdinstitutions. The unit instructionalcosts of extramuralprograms are typicallyonly 20% to 40% of the unit costs of conventionalinstruction. In addition,there are often substantialsavings in studenttransportation and (publicbudget) savingson livingcosts. During the presentperiod of "adjustmentto austerity,"a country'srationale for incurringthe high constructionand incrementalcosts of the "bricksand mortar" approachto educationalexpansion deserves to be scrutinized.

(c) Training

Occupation-specifictraining is essentialfor African development. The questionsarise not about whetherto train,but ratherwith respectto when, where, and how to do so in the most cost-effectiveway. In this, experiencesuggests that Africanpolicymakers, like their counterpartsin other parts of the world, may be temptedby the undeniableurgency of the requirementsfor occupation-specificskills to adopt questionablepolicies. As summarizedin Chapter5, there are numerouspitfalls to be avoided,but only a few bedrockprinciples upon which wise policymakerswill with confidencerely for guidance.

It is particularlyimportant to formulateeducation development strategiesaround an understandingof what decidedlyis not true. Formal schoolsare, generally,neither the only nor the best place in which to train for most highly specificvocational skills. With the possibleexceptions in some circumstancesof commercialskills (such as typing and accounting)with wide applicabilityand low requirementsfor expensiveequipment and facilities, vocationaltraining is best providedafter initialemployment in venues closer to the workplaceand more directlyunder the controlof employersthan are formal schools. The emphasisin Africa shouldbe to design for each countrya systemof vocationaltraining founded on this principle. This difficulttask will be greatlyfacilitated by ensuringthat there is a single place in the government-- and not necessarily,or even desirably,in the Ministryof Education-- that is chargedwith formulation,monitoring, and evaluationof - 141 - trainingpolicy. In this process,macroeconomic policies that stimulate provisionof job-specifictraining -- in such areas as wage regulations, investmentcodes, tax incentives,and apprenticeshiprules -- have a crucially importantpart to play, and shouldbe prominentlyincluded in an overallpolicy package for improvementof educationand training.

Meanwhile,the educationalsystems of most Africancountries already includeschools whose mission it is to provideoccupation-specific training for agriculture,industry, and the services. In the courseof developingfully coherenttraining systemsp it is imperativethat current establishmentsoperate as efficientlyas possible. Here there is typicallya great deal to be accomplishedby way of assuringthat: expensivefacilities and equipmentare fully utilized,curricula are maximallyresponsive to the (oftenchanging) skill requirementsof employers,and teachersare well endowedwith real world experiencein the jobs for which they train.

(d) Researchand postgraduateeducation

The adjustmentand revitalizationmeasures for tertiaryeducation proposedin this paper will go far when implementedtoward addressingproblems related to the efficiency and quality of undergraduate programs. By itself this will not bring about an increasedmeasure of scientificand technological self-reliancein the next century. Ultimately,Africa will developonly to the extentthat it can access and take advantageof the worldwideexplosion of knowledge,and itself generateknowledge of particularpertinence to African problems. This is the functionof Africa'stop level intellectualtalent: the peoplewith masters and doctoraldegrees whose careersare in university teachingand researchand in the most sophisticatedknowledge-intensive scientificand technicalpositions in governmentand the privatesector. In fieldsmost centralto African development,such as agriculture,health, engineering,management, and the range of basic natural and socialscience disciplinesthat underlieapplied work in those areas,the continentmust intensifyefforts to developits own capacitynot only to producethese people but also to sustainprofessional environments in which such highly specializedtalent can be productive.

Thus the questionarises of whether,when, and how Africa can develop, in a few institutionsof higher learning,programs of postgraduateeducation and researchtraining comparable in qualityto the best availableoutside the continent. Can African postgraduateeducation be expandedand upgradedto the pointwhere a substantialfraction of the continent'stop professionaltalent, and more researchof the highestinternational standards, can be producedon the continent? This is a secondarea (the problem of childhoodnutrition and health5eing the first)where the dimensionsof the problemare far more clearlydefined than the solutionfor addressingit. The answer to this questionis of criticallong-run importance to all countries,but none can reasonablyaddress it alone. Given the continent'slimited resources and the unavoidablyconsiderable sums requiredto ensurehigh qualitycritical mass efforts,international cooperation -- among African countries,and betweenthem and their partnersin other regionsof the world -- to establisha coherentset of programsof excellencein nationalinstitutions will have to be centralto any solution. Country-specificpolicy packages for educationdevelopment - 142 - should includepractical mechgnisms for encouragingand effectively collaboratingwith regionaland internationalinitiatives to addresssuch transna&ionalissues.

8.4 The task of policy design and plementation

For most African countries, the formulation of a comprehensive and coherenteducation development program, derived from a balancedpackage of policiesfor adjustment,revitalization, and selectiveexpansion, will be a new experience. Each countrywill organizefor the task in its own way. In many countries,however, a fruitfulapproach to policy designmight be expectedto includethe following: establishinga nationalcommission with politicalclout to overseethe work; cotutitutinga technicalstaff to supportthe commission; drawing for both upon the best politicaljudgment and analyticaltalent of financeand planningas well as educationministries and of the nation's institutionsof tertiaryeducation and research;building a nationalconsensus throughprovision of ample opportunityfor public debateof emergingfindings and recommendationsand of their rationale;taking advantageof the experience of other African countriesin developingtheir own educationdevelopment strategies. All such activitieswould, of course,entail costs, and these would need to be financed. Budgetaryresources would have to be sufficientto cover not only the personnelcosts of such a nationalcommission and its staff but also their operatingexpenses for travel, communications,publications, and specializedcontractual services (such as data collectionand processing, expert technicalconsultants, targeted research or analysis).

While carefulelaboration of educationdevelopment programs is urgent and essential, African capacity for implementation will ultimately determine their impact. Improvement in educationmanajement is a necessary concomitant to policy reform and must be given immediateand continuingattention.

Here the most importantmeasures involve the poteustialreturns to downwardand outwarddelegation of various administrativefunctions. While some functions must appropriately remain at the central ministry level, and the performance of these functions will need to be improved, the toughestchallenge to improvingmanagement lies closer to the classroom,at the level of individualschools and districts. Africanpolicymakers should consider how, with adequatesafeguards against abuses, schools and the local communities they serve can be given increasedauthority in the acquisitionand utilizationof the resourcesessential to effectiveclassroom teaching and learning.

In addition,central ministries must tend more seriouslyto their own managementdevelopment needs, especially in the areas of performancemonitoring and policy planning and analysis. Improvements in examination systems (which was mentioned above with reference both to academic standardsand to distance education), in the nature and timely availability of statisticaland financial accounting information, and in the numbers and qualificationsof staff engaged in full-time analytical work are among the necessary measures. Incentives in many ministries of education are insufficientto attract,motivate, and retain able staff. Governments committed not only to formulation of an educational development program but also to its expeditiousimplementation will have imaginatively to address the issue of incentives at all levels of the educational system. - 143 -

Chapter 9. INTENIATIONALASSISTANCE FOR AFRICANEDUCATIONAL DEVELOPMNT

Chapter8 recapitulatedthe detaileddiscussion of policy options for African educationand training,recasting the variousoptions within a new, three-partframwork to be used by African government decisionmakers for developingimproved policy gackages on a country-by-countrybasis. Chapter9 analyzesthe volume and distributionin recenttimes of public international developmentassistance for Africaneducation and training,and it offers suggestionsfor enhancingthe impactof such assistancein the future. It advocatesan enlargedand better coordinatedinternational effort, responsive to and supportiveof the effortsof Africangovernments as they design and implementnew policiesfor adjustment,revitalization, and selectiveexpansion of educationsystems.

9.1 Aid to educationLu the recent past

Relativeto its population,sub-Saharan Africa has recentlycommanded a sizeableshare of public international4evelopment assistance -- of what is commonlycounted as "internationalaid.' With only about 112 of the populationof the world's less developedcountries (LDCs), sub-Saharan Africa in the early 1980swas receivingabout 222 of internationalaid. Africa's annual allotmentof aid is equivalentto about $19 per inhabitant,as compared with $8 per inhabitantfor the other LDCs. As a proportionof this largerper capita amount,aid for educationin Africa is about the same as it is elsewheresroughly 102 overall (althoughin a third of the Africancountries it is above 152).

Analysisof public internationaldevelopment assistance to education in Africa in the early 19809yields a number of interestingresults, the most salientof which will be summarizedhere. The detaileddata are containedin Annex TablesA.24-A.27. Becausecomparable information on aid flows is extremelydifficult to develop,and in any event not availableexcept with substantiallags, two generalcaveats are important. First,despite the care taken in preparingAnnex TablesA.24-A.27, the reporteddistribution of aid by use, level,and expenditurecategory may be distortedby difficultiesinherent, for severalmajor bilateraldonors, in accountingaccurately for resourceflows to educationwhich are integralparts of bilateralbudget support arrangements. Second,the data say nothingabout trends since the 1981-1983 period. Since then, in responseto the deteriorationof African education, several importantproviders of internationalassistance for the sector have introducedfundamental changes In aid distributionpolicies which are not yet being reflectedin the availabledata. So cautionis appropriatewhen consideringthe followingpoints.

The first point is that these externalresources are critically importantto African educationaldevelopment. In the early 19809,public internationaldevelopment assistance to educationand trainingin Africa -- "educationaid" as definedhere -- averaged$1.3 billionper year (Table 9.1, row I.A.). This was the eqifvalentof about 152 of Africandomestic public expenditureon education.A' - 144 -

[Table9.1)

Of this amount,$367 million (28%)was used to financetraining in sectorsother than education. Another $190 million (14%)was in the form of hidden subsidiesincurred by donor countriesin hostingthe approximately 100,000African studentsenrolled in educationalinstitutions abroad. The remaining$757 million (59% of the $1.3 billion in total "aid"to African educationand training)is the subjectof the followinganalysis. This is what was actuallychannelled through education ministries and expendeddirectly by the educationsector itself. It will be referredto henceforthas direct educationaid (Table9.1, row L.A., column (a)).

The sourcesof direct educationaid to Africa have been several. Three formercolonial powers (France,Belgium, and Britain)account for 40% of it; France alone contributesmore than a quarter. In the early 1980s,the dollar amountsof direct educationaid providedby the United Kingdomand the United Stateswere roughlythe same, but each was less than 20% of France's aid. All other bilateralsources together supplied 27%, and multilateral sourcesprovided 332.

Directeducation aid for the purchaseof recurrentitems was equivalentto roughly 17% of domesticrecurrent expenditure on educationin the median African country,and aid for the purchaseof capitalitems, to about 30% of domesticcapital expenditure (Table 9.2). Both proportionsdecrease steadilyas countriesbecome more developed. For the six low-incomesemi-arid African countries,direct aid is 302 of recurrentand 55% of capital expenditureson average,while for the five middle-incomeoil exporting countries,the correspondingfigures are only 4% and 3%.

(Table 9.2]

But direct educationaid per capita is less rationallydistributed. The 25 low-incomeAfrican countries receive $2.90 per capitaon averagewhile the nine middle-incomeoil importersreceive $4.16. Statisticalanalysis of the country-leveldata in Annex Table A.25 shows that neitherlevel of development(GDP per capita)nor magnitudeof educationalneed (enrollment rates for the three levels)is a significantdeterminant of educationaid per capita. The same analysisalso demonstratesthat the levelof a recipient country'spolitical support of donor governmentsin the arenasof international debate is not relatedto the amount of aid received. Indeed,among the variablesincluded in the analysis,only the size of the country (population) was a significantpredictor of per capitaeducation aid, with smallercountries gettingmore.

The distributionof direct aid by levelof educationis heavilyskewed toward higher levels (Table9.3). Only about 7% ($56 millionper year in the early 1980s)of all direct aid to African educationis used to financeprimary education,while 34% ($259million) goes to the tertiarysubsector. On a per-studentbasis, the resultsare staggeriag,-- directeducation aid to primary educationamounts to $1.10per student,aid to secondaryapproximately $11, aid to teachertraining $78, and aid to secondarytechnical education $182; aid to highereducation is $575 per student,which is well over 500 times the amountper primarypupil. In round numbers,direct educationaid to primary educationcovers only 2Z of the cost of sendingan African primary pupil to school;aid to generalsecondary education and teachertraining covers - 145 -

Table9.1 Estlated nualExternal Raeowce Flow to AfricanEducatlon and Training by DonoGrouP and Use, 1981-1983Averap

(a) (b) (c) (d)

Costof hosting Project African Educatlon related students sectorA/ trainingh/ abroadcg Total

I. OECDand OPEC donors 785 394 180 1,370

A. Conoe%linalflows (uald M) 757g/ 387 190 1,314

(1)Bilateral 507 247 190 944

France 206 - 73 279 Belgi1u 58 - 7 66 UnitedKlngdom 40 - 24 64 Allother bilaterals 203 - 85 288 Notdistributed - 247 - 247

(2)Multllateral 250 120 0 370

InternatlonalDeveopnt Association 128 54 0 183 AfricanDevelopment Fund 34 15 0 48 All othermultilaterals 88 51 0 139

B. N -concsionalflows a/ 28 27 0 56

11. EastEuropean non-warket economies and OUba 40 - 55 95

111. No-GovermentalOrgnIzations 90 - - 90

IV.Total 915 394 245 1,555

Note: MilIInsm of current U.S. dol lars. Based on Akmr Tabie A.24. Columnsmay not adddue to rourdlng.

/ Disbursementsto central ministrles of education. W Aidfor tralning In sectorsother than ecfk.ation. C Hostcountry si,sidizatlon ofAfrican students studying abroad, over ad aboveany fellowships. d/ *DirecteducatIon ald,Q the termuned In this paerto describeconcesslonal flows to theeducation sectore se. a/ Loansfrom the WorIdBark ndthe AfricanDevelwmnt Brk. - 146 -

Wble9.2 DirectEducation Aid ty RecipientSOri d EmdirtowCateo, 191-1983Arwap

(a) (b) All directOdUCtlIn aid

Aidfor rfcrent Aidfor capital (c) (d) (e) rqdlturs mmitt a8 prcontap aspercentag Asa of dmstic of do estic Total percontaep rewrnt capItal dollars Per of external ¢atlon educatian (milisla capita ald In all bkdgt bidget pr yaw) (8p yw) sectors

1. EconolcGreuP

LomInme seal-arld 29.59 54.91 $99 $3.42 9.91 LowIncm other 17.9 47.71 $477 $2.58 13.1X MIddleIrno ol Ilworters 11.1X 28.0% 8118 $4.18 10.7X

Middleicm ol prters 3.91 2.9X $ 64 $2.83 14.01

I I. LlngulsticGroups

Francophomcountrles 18.7X 49.9 a333 $3.13 13.91

Anclophonecountrles 11.91 24.29 80 $3.38 11.3u

Other 16.46 18.41 $ es $1.03 9.6X

SulkSaharanAfrica 16.71 30.5K $75? $3.38 13.1X

Note: Basedan MmnxTable A.2S. - 147 - about 42 of the cost of each student,and aid to secondarytechnical and tertiaryeducation about 502. With the bilateraldonors, the lack of balance is even more dramatict less than 41 of their directaid goes for primary educationand 422 for tertiary. Swedishaid (SIDA)with its focusedprogram of support for primaryeducation in selectedEastern African countries provides a notableexception to this generalpattern of concentrationon the tertiary subsector.

[Table9.3]

The distributionof directeducation aid by expenditurecategorv is notablefor the small proportiondedicated to financing"operating costs," which comprisesalaries of nationals of the country,consumable supplies includingpublic utilities,and all instructionalmaterials (Table9.4). This allocationwas especiallyremarkable during a period in Africawhen the absence of non-salaryrecurrent inputs to educationhad becomethe gqverningconstraint limitingeducational performance at all levels (Tabla9.4).21 Only 11 of direct educationaid is allocatedto operatingcosts. By contrast,442 of aid supportstechnical assistance in its narrowestsense -- i.e., the provisionof foreignexperts.

[Table9.4]

The predilectionin favor of technicalassistance is most marked in the case of the bilateraldonors. None of the four major bilateraldonors dedicatesless than 55X of its direct educationaid to technicalassistance, which in essenceprovides employment in Africa for its own professionals; because it suppliesso many teachersin secondaryand tertlry institutions, Franceuses an astounding832 of its aid for this purpose.- Only 132 of bilateralaid is used to financeoperating costs. Bilateraldonors also allocatea smallerpercentage of their directeducation aid than do the multilateralsto systemmanagement and to capacitybuilding activities. These activities,unallocated by level, are crucialto long term educational development.

Whereasthe bilateralsdevote a disproportionatelylarge share of their direct educationaid to technicalassistance, the multilateralsdevote a disproportionatelysmall share of theirs to non-capitalexpenditure items of all kinds. Only 35Z of directmultilateral aid to educationfinances recurrent expenditures(as comparedwith 92X in the case of the bilateraldonors). Only about 7X of directmultilateral aid is used to financeoperating costs. UNICEF,with its focus on non-salaryrecurrent inputs, is a notableexception to the generalpractice.

9.2 The comparative advantage of aidt: past and present

Caution is appropriatein drawingnegative conclusions from the above six facts on the effectivenessof directeducation aid to Africa. The apparentinconsistencies between Africa's current needs and the recent distributionof aid do not, by themselves,constitute evidence that aid in the past has been eithermisdirected or unproductive.The evolutionof the context for aid and the comparativeadvantuge of nationalgovernments and donorsas sourcesof financefor differentpurposes must enter into the calculusbefore an evaluativeqssessment is possible. - 148 -

Table9.3 DirectEducation Aid by Leveland Source,1981-1983 Average

Percentof Percentof Percentof bilateralaid multilateralaid all directaid ($507million) ($250million) ($757million)

Primary 3.4 15.7 7.4

Secondary General 20.9 6.1 16.0 Teacher Training 3.0 12.7 6.2 Vocational/Technical 14.9 20.9 16.9

Tertiary 42.4 17.5 34.2 Other 15.4 27.1 19.3

Notes Basedon AnnexTable A.26.

Table 9.4 Allocationof DirectEducation Aid by ExpenditureCategory and Source, 1981-1983Average

Percentof Percentof Percentof bilateralaid multilateralaid all directaid ($507million) ($250million) ($757million)

Capital 7.4 64.9 26.4

Recurrent TechnicalAssistance a/ 57.5 17.4 44.3 Fellowships 20.5 9.6 16.9 OperationalCosts bI 12.7 7.2 10.9

Other 1.9 0.9 1.5

Notes Basedon AnnexTable A.27.

a/ Includesprovision of teachersas well as otherforeign experts. bi Includessalary support for nationals,utilities and supplies,and instructionalmaterials. - 149 -

In an environmentin which essentiallyall direct educationaid to Africa was for discreteinvestment projects, the concentrationof aid on small countries,on tertisryinstitutions, and on the provisionof foreignexperts and fellowshipsfor study abroadwas understandableand may even have been rational. As a generalmatter, donors enjoy a comparativeadvantage over nationalgovernments as sourcesof financefor investmentprojects that are capital-irtensive;foreign-exchange-intensive; limited in the number,scope, and geographicdispersion of their componentsso as to minimizethe burden of project implementationon scarcemanagerial resources; and heavily dependent upon the expertiseof donor countryprofessionals (including teachers) and on study abroad for recipientcountry nationals The distributionof recent direct educationaid by expenditurecategory (Table 9.4) was broadlyconsistent with this assessmentof donors'comparative advantage.

So too was the distributionof directeducation aid by level (Table 9.3). Intrinsiclack of donor -- and sometimesrecipient -- interestin the primary subsectorcompared to the more visible interventionsthat can be made in tertiaryeducation, and failureon both sides to appreciatethe singular importanceof primary educationin African developmentcontribute to the distributionof aid by educationallevel. But to some extent,the relative neglectof primaryeducation in educationaid can be explainedalso by the differencesbetween primary and tertiaryeducation in the ways resourcesfor these two subsectorsare generatedand utilizedand by the implicationsof these differencesfor the comparativeadvantage of nationalversus donor financialsupport.

School constructionand teachersalaries typically account for more thar.95% of expendituresin the primarysubsector. But primaryschools are highly dispersed,the cost of an individualschool is minisculefrom the perspectiveof an internationaldonor, and no single schoolenjoys high visibilityor separatelyidentifiable impact. Moreover,the clear trend is to rely ever more heavilyon the use of local materialsin primaryschool constructionand on local communitiesto financethose materialsand to provide the necessarylabor. The potentialcontribution of Africangovernments, and since they work throughthose governments,the donor agenciesas well, to primary schoolconstruction is therebylimited, and this is even truer today than it was in previousdecades. AlthoughAfrican central governments do contributerelatively more to teachers'salaries than to constructionat the primary level, the internationaldonor communityhas been generallyhesitant to financelocal salariesin the contextof conventionalinvestment projects.

For tertiaryand secondarytechnical institutions, the situationis quite the opposite. There are few establishments.Their size and location make them highly visible. Each has a discernibleimpact on the educational landscape. Typically,these subsectorsare directlyand wholly dependentupon resourcesfrom the centralgovernment. A significantportion of the capital costs of constructionand equipmentinvo'lve foreign exchange. And supporting staff developmentthrough financing graduate study abroad and supplying expatriateprofessors -- eitherproviding them outrightor supplementing institutionalbudgets to facilitateAfrican direct hire -- are items easily providedby many donors.

While the distributionof direct educationaid by level and expenditurecategory may have been a rationalreflection of comparative advantagein a periodcharacterized entirely by project-focusedaid, the nature - 150 - of the currentchallenges facing educational development in Africa suggests that aid patternsshould now change in a numberof significantways. The immediaterequirement is policy reformto supporteducation sector (i) adjustmeAtto harsh new fiscaland demographicconstraints, (ii) revitalization to achieveenhanced quality and efficiencyof the currentsystem, and (iii) carefullytargeted and cost-effectiveexpansion in the future. This leads to:

Recommendation7. Externalaid flows to Africaneducation have been substantialin their aggregateamount -- the equivalent(depending on what is counted)of between82 and 17% of what African governmentsthemselves spend on education. Aid to educacionin most countries,however, has tended to focus on discreteinvestment projects, with littleor no effectivecoordination among donors in supportof a coherentnational strategy for the sector.Moreoever, comparedto currentAfrican needs, disproportionatefractions of what has been given in the past have been allocatedto higher educationrather than primary; in the case of the multilateraldonors, to capitalexpenditures rather than recurrent;and in the case of the bilateraldonors, to technicalassistance and overseasfellowships rather than other forms of recurrentexpenditures. The organizationalforms and substantivecontent of externalaid to African educationshould be broughtinto conformitywith the new imperativefor African governmentsto design and implementcountry-specific education secto. developmentprograms supporting policy reform. As the policycontext develops, the overallvolume of aid to educationshould increasein real terms.

This is not to say that supportfor discreteinvestment projects in the educationsector has no place in any African countrynow or anytimein the foreseeablefuture. It does mean, however,that programmaticassistance, in supportof the implementationof policiesdesigned to put the educationsector on a viable long-termfooting (combinedwith carefullyselected investment projectswhose overridingpurpose is to serve as instrumentsof policy reform) shouldpredominate in most countriesfor the remainderof the century. After measuresof adjustment,revitalization, and preparationfor cost-effectiveexpansion have been aggressivelyand persistentlyimplemented over a period of years, the policy environmentwill again be congenialfor a concentrationon conventionalproject assistance for educationalexpansion. Conversely,and with few exceptions,African countriesunwilling to embarkupon sectoraldevelopment programs are unlikelyto be attractivecandidates for conventionalproject assistance.

The designand implementationof an educationaldevelopment strategy along the three dimensionsof adjustment,revitalization, and selective expansionrepresents an unprecedentedchallenge to Africanpolicymakers. How can the internationaldonor communitybest assistAfrican countries to meet the challenge? The answerhas two parts. First, new structuresare neededto facilitateinternational cooperation for African educationaldevelopment. Second,adjustments are needed in the establishedamounts and substantive - 151 - contentof aid to Africaneducation. A new era in internationalassistance for African educationand trainingmust commencebefore the end of the decade. The final sectionsof this chaptersuggest how this can be achieved.

9.3 International assistance for policy design and implementation

When an African government decides to address the issuesof adjustment,revitalization, and selectiveexpansion, it will have to confront two imediate tasks: developingthe policybasis and implementationmechanisms of nationalprograms, and mobilizingresources to pay for them. In accomplishingboth of these tasks, new forms of internationalsupport will be helpful.There are a varietyof ways in whiichinternational support for these two purposesmight be organized. No one of them is perfect in every respect, but change is essentialif donor assistanceis to be made more effective. As productive starting points for more systematic discussion, two sets of ideas on this question are outlined below.

(a) Support for development of national programs

The design of a sound country-specific education development program to supportpolicy reformis difficult. It entailsthe carefulselection, phasing,and linkingof specificpolicies from among the full array of policy options for cost containment, for diversifyingsources of finance,for supplyingvital non-salary inputs to learning,for assuringadequate standards of performance,for operationand maintenanceof physicalplant and equipment, for primary schoolexpansion, for developingdistance education systems for secondaryand tertiaryeducation, for expandingtraining opportunities, for strengtheningpostgraduate education and researchcapacity, and for improving managementincluding in particularthe incentivestructure for managersat all levels and the knowledgebase for decisionmaking. As noted at the end of Chapter 8, there are a variety of steps which Africangovernments can take to addressthe task of policy design and implementation.In many countries, however,this complextask may possiblyrequire more experiencea.d analytical talent, in more professionaldisciplines, and appliedmuch more intensively, than local resourcesalone can supply,at least in the short run.

The internationaldonor communityshould offer three relatedkinds of support for the nationallevel processof policydesign and implementation sketchedin Chapter8. The need is for expeditiousaction. Any initiativesin these areas that would take more than a year to be adoptedand implemented cannot be judged an adequateresponse to the needs of African governments.

The first is simples seed money, quicklyprovided, to cover both local and foreigncosts of policy developmentand managementimprovement activities. Willingnessof internationaldonors to bear part of these extraordinarycharges, perhaps on a matchingbasis, would provide important incentiveto African governments.

In situationswnere budgetaryresources do not begin to cover minimal operationalneeds of educationalsystems, finding funds for nationaltask forces,technical staffs, and consultancyservices can be a lengthyand ultimatelyunproductive process, especially if part of the requirementis for foreignexchange and if budgetsbeyond that of the educationministry are involved. - 152 -

Centralto the improvementof managementof educationsystems is the strengtheningof examinationsystems and the buildingof nationalcapacity to conductresearch on education. Over the long term, Improvementsin testingso as better to serve its four purposesof improvingcurricula, measuring performance,certification, and selectionmay be expectedto pay for themselves in increasededucational efficiency. Over the short and medium term, however, significantsums will be requiredfor additionalpersonnel, for training,for equipment,and for technicalsupport. Early investmentsare needed to improve examinationssystems so that they may play properlysupportive roles in implementationof new policies.

Applied researchon education,and trainingto do it, also costs money which, in nationalcontexts of extremepublic sector austerity,is not readily available. The payoff to such expendituresis not usually immediate,or even preciselyattributable to specificinterventions. And yet, the qualityof nationaleducation development programs depends vitally on the availabilityof first-rateoperations-oriented research and on expandedknowledge of internal and externalefficiency. This kind of analyticalcapacity is generatedas much by actuallydoing such work as by formaltraining to do it. Internationalaid is a crucialsource of supportfor both.

Second, ready accessto the ongoingexperience of other countriesin educationpolicy formulationand implementationshould be providedby the internationalcommunity to African countriesthat embarkon seriouseducation policy reformefforts. Intensivecollaboration across countries, so that accumulatingexperience is widely shared,will pay high dividends,as countries grapplewith common issues -- e.g., equivalencyprograms in secondary education,consolidation of academicprograms in higher education, establishmentof systemsfor providingtextbooks in primaryschools. Serving as a catalystfor such learningfrom sharedexperience is a functionnicely suitedto donors.

Third, the internationaldonor communitycould establishand financea sourceof high-qualitynon-political specialized technical expertise, beholden to no internationaldonor or governmentin particular,which African governmentscould call upon for help in formulatingpolicies at the outset and in monitoringand evaluatingthem for correctionsduring implementation. Such help could be suppliedin the form of small teams that would, in periodic intensivevisits over a periodof months or even years, supportbut not supplantlocally constituted task forcesor commissions.The initialproduct of interactionof the nationalcommission with the outsidetechnical expertise would be a packageof policiestogether with a detailedplan for their implementation-- i.e., an educationsector developmentprogram -- ready to be consideredfor financingby the governmentand appraisedby the international donor community. The interactionbetween national commission and outside technicalexpertise might continueas programsattract funding and enter implementation.This could take the form of collaborativeefforts in on-going policy evaluationand mutual formulationof recommendationsfor policy refinementson the basis of accumulatingexperience. Importantly,if this technicalsupport is providedin sensibleways, African nationalcapacity to engage in this sort of continuouspolicy planning and analysisshould be significantlyenhanced as a resultof the process. - 153 -

(b) Organizinginternational support for program implementation

The internationalcommunity should help financethe implementationof sound nationaleducation sector development programs, which will typically requiremore resources,sustained over a longerperiod of time, than can be mobilizedinternally. Countriesthat have demonstratedtheir willingness to addresspolicy issues shouldhave access to increased,lonaer-term, and more flexiblyproffered international aid. To the extentthat a country'spolicy package involvesthorough-going reform, there are likelyto be substantial one-timetransition costs to a new and more sustainablepolicy regime. In countrieswith severe foreignexchange constraints, revitalization measures are likely to involveincreased claims on foreignexchange for importsof essential non-salaryinputs to learning. Further,in many cases the necessarypolicy changesconstitute a sharp break with the past; their implementationmust be carefullyphased, and provisionmust be made for correctionsas implementation proceeds. The internationalcommitment to the developmentprogram must be seen from the beginningas long-termand sustained,with disbursementstranched againstsequenced achievement of agreed-uponpolicy objectives, as verifiedin periodicjoint reviewsof the program. Aid must have continuityover time, somethingit has too often lackedin the past.

A promisingapproach would be to make greateruse of country-specific donor consortia. A number of Africangovernments already have successful experiencein submittingtheir macroeconomicstructural adjustment and developmentprograms to meetingsof prospectiveinternational supporters, for their review and appraisal. The ConsultativeGroup mechanism,supported by the World Bank, has involvedperiodic meetings of donorsto appraisethe macroeconomicpolicy frameworkof a country'sadjustment program, while the UNDP Round Table mechanismhas been used to coordinatedonor support for a country'sproject list within an investmentprogram. The educationsector needs its own mechanismthat buildsupon the strengthsof both these devices.

Once an Africangovernment has preparedits own strategyfor educationaldevelopment -- i.e., decidedupon the necessary policy reformsand identifiedthe operationalmeans and resourcerequirements for implementing them -- the internationalcommunity should be invitedto a forum in which the strategycan be presentedin some detail,thoroughly reviewed and discussed, and broad agreementsreached with respectto externalsupport. Two useful precedentsfor such meetingsto considermajor reformsin educationat the country level were establishedin 1986. Ghana'sambitious education sector adjustmentprogram was the subjectof a donor group meeting in Vienna, in this instanceas an outgrowthof the World Bank ConsultativeGroup set up to considerGhana's overall adjustment program. The UNDP organizeda Round Table on Educationfor Burundi,chaired by the Ministerof Educationand held in Bujumbura.

Some governmentsmay wish to invltethe World Bank to arrangesimilar specialConsultative Group meetingsdedicated to reviewingtheir education sectordevelopment program; other governmentsmay wish to ask the UNDP to arrangea specialRound Table for this purpose. In still other instances, governmentsmay wish to rely upon the good officesof the ad hoc technical supportteams, mentionedabove, to organizea meeting. Whateverthe mechanism, the importantpoint is that each Africancountry would be affordedan opportunityto meet periodicallywith all prospectivedonors, to reviewthe policy frameworkand operationalprograms of its long-termeducational - 154 - developmentstrategy, monitor progressin its implementation,and agree collectivelyon resourcerequirements and sources. Much enhancedaid coordinationin supportof the priorityneeds of the recipientcountry would be the result.

This brings us to the questionof how donorsshould spend their money rather than how they shouldorganize themselves to spend it.

9.4 Future allocations of aid to Africw education

Whether or not new organizationalforms take root, the earlier analysis of educational development challenges in Africa today and of the recent patte:ns of international assistance for African education suggest that modificationin the substantivefocus of aid is urgently needed. This will of coursebe easier to achieveto the extentthat African governmentshave embracedthe need for policy reform and the donor communityhas improvedthe organizationof its activities.

Of the three dimensionsof educationreform, the first (adjustment policiesthat seek to diversifyfinance and reduceunit costs, thereby economizingpublic resources)will usuallynot depend for its implementation upon a fresh injectionof financialresources. The estahlishmentof student loan schemes,and the prospectiveneed in some countriesto acceleratethe reductionor redeploymentof excesspersonnel, by offeringone-time compensationto the involvedindividuals, are possibleexceptions. But as a generalmatter, the implementationof adjustmentpolicies rarely entails purchaseof goods or services.and thus presentsfew opportunitiesfor increasedinternational financial aid. Contingentupon African countries' vigorousimplementation of educationsector adjustment policies, however, the internationaldonor commnuityhas a vital role to play in enablingAfrican governmentsto implementpolicies on the other two dimensionsof change-- revitalizationand resumingselected expansion of African educationsystems. In both these areas, there is an immediateand continuingrequirement for more resources. This paper concludes,therefore, with a discussionof aid for these two purposes.

(a) Aid for revitalization

Instructional materials. The revitalization of African educational systemsdepends first and foremostupon correctingthe Imbalancebetween salary and non-salaryrecurrent expenditures. Nothing short of a massive resurgence of the flow of non-salaryinputs to learningis required. Instructional materialsare centralin this effo:t. For primaryschools, textbooks and writing materialsare the essentialelements. For generalsecondary schools, the need is for textbooksand consumablesupplies for laboratory demonstrations.In addition to these items, secondary technical schools requiresubstantial doses of equipmentmaintenance and repair,including spare parts, to make the workshopsfunctional again. In tertiaryinstitutions, the prioritiesare libraries(including multiple copies of standardtextbooks), equipmentmaintenance and repair,and consumablesfor laboratoriesand workshops. At all three levels,but particularlyfor secondaryand tertiary education,maintenance of the physicalplant is the secondpriot :y for non-salaryrecurrent finance. - 155 -

To get a rough sense of the volume of resourcesrequired, assume that by 1988 there will be roughly63 millionAfrican childrenin primaryeducation, 16 million in secondary,and 0.5 million in tertiary,and further,that the coat per student of providingan initialstock of materials is, respectively, $9, $18, and $90. The start-upcosts, incurredover three years,would approximate4900 million. Thereafter,roughly one-third of that amountwould be requiredannually to replacetexts and ensure continuedflows of consumables;provision would also have to be made for expandingenrollments (for exampleat 4%, 8%, and 0.52 annuallyfor primary,secondary, tertiary education,respectively). A reasonabletarget phased in over ten years (1991 to 2000) would be to have 60% recoveryof replacementcosts of materialsfrom primary studentsand their families,and 95% cost recoveryfor secondaryand tertiarymaterials. On these assumptions,the differencebetween replacement requirementsand amountsrecovered would declineover the decadefrom $322 million to $134 million;the total non-recoveredoperational cost for that decadewould approximate$2.4 billion. BY the beginningof the next century, then, the instructionalmaterials shortfall in African classroomscan be resolvedfor a total of about $3.3 billion. This _otal would equal 33% of direct educationaid under the pessimistic(and, indeed,unacceptable) assumptionthat there will be no increasein aid flows over the thirteenyear period (1988-2000)from the 1981-1983average annual amount. This level of responsewould not seriouslyburden the capacityof the international developmentassistance community.

Repair and maintenanceof equipmentand physicalplant. The contractionof public resourcesfor educationoccasioned by the economiccrisis in Africa was typicallyfirst felt in the disappearanceof budgetedfunds for the maintenanceof capitalassets -- physicalplant and major items of equipment. Maintenancewas deferredso long that many countriesnow face an immenseburden of costly repairs. The situationis particularlyserious with respectto equipmentfor technicalsubjects in secondaryand tertiary institutions,but extendsas well to such fundamentalitems as electricaland water supply systems,sewage, lightingand ventilationsystems, and even rocfs and windows needed to providebasic protectionfrom the elements. Precise calculationsof the financialrequirements are beyond the scope of this paper. A rough estimate (basedon presumedcosts of $500 per tertiary,$100 per secondarytechnical, and $20 per generalsecondary student) suggests an order of magnitudeof $600 million. Contingentupon agreementsin the contextof an overalleducation development program to ensureand protecta budget share adequateto cover preventativemaintenance in the future,international donors should be willing to financethis backlogof necessaryrehabilitation. In some cases it will be more economicalto replaceobsolete equipment with state-of-the-artitems than to repairand then maintainthe old.

(b) Aid for selectiveexpansion

Progress towards universal primary education. Two aspects of resuming the advance towards universal primary education are especially appropriate subjects of international aid: gettingthe classroomsbuilt and assuringthat studentswho will fill them are intellectuallyand physicse'ly ready to learn.

In the contextof rapid growthof the school age populationand continuingpublic austerity,sensible education development strategies will need to encouragelocal communitiesto assume greaterresponsibility for - 156 - financing primary schoolconstruction. Nevertheless, African governments can, with the help of internationalaid, be vitally importantsupporting actors in this drama. Leading candidatesfor (donor-financed)government assistance include: applied researchon low cost constructiontechniques; establishing reasonablestandards for use of space and materials;the trainingof school constructionforemen; supplying certain standard materials that are either unavailableor pricedmuch higher at the communitylevel, and even paying the salariesof some workers,such as constructionforemen.

But much more importantlythan this, in situationsin which an African governmentis vigorouslyimplementing a comprehensiveeducation policy reform packagethat includescapacity expansion for primaryeducation, international donors shouldbe preparedto disbursea reasonableportion of the funds allocatedin supportof the overallprogram upon receiptof proof of phased completionof classroomconstruction. More bluntly,the expansionof primary educationis so centralto long-termAfrican developmentthat, even though aid funds may not be the optimalsource of financefor primary schoolconstruction, the disbursementof those funds (to be used for the purchaseof other inputsto the educationsystem, or even for generalbalance-of-payments support) should be in some measure contingentupon successfulperformance in the programof primary schoolconstruction.

In additionto classrooms,teachers, and instructionalmaterials, primarypupils need to have benefitedfrom and continueto enjoy minimum standardsof health and nutritionin order for schoolingto result in learning. In view of the soberingrecord of childhooddisease and malnutritionin Africa, it is timely to delineatethe parametersof the problem,assess the likely implicationsfor educationalsystems, and identifypractical solutions. Organizationand supportfor such activitydeserves international financial assistance.

Staff developmentfor the educationsystem. The qualityof staff is obviouslyan essentialingredient to the qualityof learning. For primary education,more teacherswill be requiredvery soon if enrollmentrates are to be maintainedand then grow. At the secondarylevel, significant expansion will probablydepend upon widespreadutilization of new distancelearning technologies,for which a new style "teacher'will be needed. Cost-effective systemsof continuousin-service pedagogical support and trainingof teachers at both levelstake professionaland financialresources to develop,and are appropriateobjects of aid.

For secondaryand tertiaryinstitutions in Africa, less concernedin the short-runwith expansion,new staff must be trainedto replaceexpatriates; and, in most countries,current staff with inadequateacademic credentials and insufficientprofessional experience need opportunitiesto upgrade their knowledgeand skills. Ultimatelymost of this activityshould take place in Africanuniversities. As a practicalmatter, however, much of the required staff developmentfor the tertiarysystem cannotbe conductedin Africa for the rest of this century. The requirementsin this area extendfar beyond matriculationof African staff (presentand prospective)in formalprograms of postgraduate(masters- and doctoral-level)study in donor universities, althougha sizeableexpansion of opportunitiesin this regardis extremely important. Whateverthe form of staff development,international aid must help to financeit. - 157 -

Distanceeducation for secondaryand tertiaryeducation. Two ways of using distanceeducation -- radio correspondenceschools and extramural programs-- have been advancedas highly efficientmechanisms for expansionof secondaryand higher education. For successfulimplementation, both forms depend on examinationsystems to assure qualityand provide accreditation. Radio correspondenceschools bring studentstogether in simpleclassrooms, supervisedby a communitymember, to follow lessonson radio and to work from texts and exercisebooks; they are highly cost-effectivefor lower secondary and secondaryeducation. They enjoy the added attractionof making it possible to provide learningopportunities in small communities,a powerfuldevice both for extendingaccess to femalesand for reducingcostly enrollmentsin boarding institutions.Extramural programs differ from radio correspondenceschools principallyin that they do not bring studentstogether in supervised classrooms;unit costs are therebyfurther reduced and the opportunitycreated for individualsto continuetheir studieswhile employed. Extramuralprograms typicallyrely heavilyon print, but may also utilizeradio and occasional tutorialand laboratoryinstruction at a campus duringevenings, weekends, or vacationperiods. Extramuralprograms are particularlysuitable for in-service and post-secondaryeducation.

The internationalcommunity can assistwith the developmentof distanceeducation in three importantways. First, as individualcountries invest in radio correspondenceschools and extramuralprograms, aid can appropriatelyfinance the capitalcosts and developmentof sound course materialsand the necessaryexamination systems. Second,a rich varietyof correspondencematerial is now availablefrom universitiesthroughout the world. Most of this is for universitylevel courses,but basic secondary courses in mathematicsand sciencealso exist. As a partialsubstitute for sendingfirst degreestudents for study abroad,donor agenciescould provide the foreignexchange for individualsto enroll in extramuralcorrespondence courseswithout leavingtheir country. A good exampleof such a program, tailoredparticularly for the needs of developingcountries, is an M.Sc. course in agriculturaleconomics offered by an eminentBritish university.

Third, and most important,a number of donorscould collaborateto help financean African distanceteaching center. Such a centerwould serve three main purposes: (i) it would provide a repositoryof experienceand expertiseavailable to assistindividual countries in establishingor improving their own distanceteaching capacity; (ii) it would preparebasic coursesthat could be adaptedat minimal expenseby nationalauthorities for their own use (including,for specializedcourses, for use in their conventionalsecondary schoolsand universities);and (iii) it would offer extramuralprograms directly,to at least a limitednumber of students,thereby strengthening its own course developmentwhile also providinga service,especially to small countrieswhich are not in a positionto investheavily in this area themselves. Among the possibleorganizational alternatives for such an African distanceteaching center would be to create it as a semi-autonomousentity, rather like the InternationalInstitute of EducationalPlanning (IIEP), under the UNESCO umbrella.

Training. Internationalaid to trainingis now mostly of two quite differentsorts. First, in the contexto- projectsthat donorshelp to finance in all sectors,assistance is often providedto supportthe trainingof those responsiblefor building,operating, or maintainingthe project. Experience with such 'project-relatedtraining" has been generallyfavorable, although its - 158 - focus on discreteprojects usually has precludedthe longer-termattention to buildingthe capacityof local traininginstitutions. Second, international aid for traininghas supportedthe constructionof vocationallyoriented secondaryschools, usually for the purposeof providinggraduates with specific job skills for the agriculture,commercial, or industrialsector. While evaluationof the economicimpact of such schoolsis in its infancy,it is known that industrial-sectortechnical schools (and to a lesserextent, agriculturalones) are enormouslyexpensive relative to more generaleducation in secondary-levelmathematics and scienceand that the labor market outcomes for students of those technical schools -- in terms of occupational placement and earnings-- are often less successfulthan what was envisioned.

In this situation,there are three broad directionsthat international aid should take to strengthentraining. First, for projectsoutside the educationsector, increasesin the proportionof resourcesdevoted to project-relatedtraining will have high payoff. At the same time, however, attentionmust be given to the developmentof local trainingcapacity in key economicsectors; sector-related training investments, designed to build institutionalcapacity, merit donor support. Second,aid for job-specific trainingin secondaryschools should concentrate on maximizingthe utilization of existing schools and minimizing their unit costs; all suc;1assistance should provide for rigorousevaluation to determinewhether labor market returnsare in fact being realized. Third, and most important,international donors should expand their supportfor developmentand utilizationof traininginstitutions that serve employersand that are partiallycontrolled and financedby them. These institutionswould providetraining programs mostly for existing employeesor new hires whose initialjob assignmenthas alreadybeen decided; completionwould be recognizedwith a certificaterather than an academic degree;and the menu of course offeringswould evolvewith labor market needs. Assisting in the developmentand financingof such employment-relatedtraining institutions,in the contextof an overalleducation sector development program, is a high priorityfor internationalaid.

g 2 pandedcaoacity for graduateeducation and research. Beyond rationalizingthe provisionof first degreetertiary education, Africa must intensifyefforts to develop its own capacityto conductresearch and to providepostgraduate education and researchtraining of world class standards in fieldscentral to African development.Leading candidates for attentionin this respectare agricultureand health (includingthe naturalscience disciplineswhich underlietlem), management (including the underlyingsocial sciences),and engineering.

Determiningwhich institutionswill be participantsin postgraduate programsof excellence,in which fieldsand at what levelswill not be easy -- for African governmentsor for their internationalpartners. Questionsof internationaland regionalcomparative advantage must be squarelyfaced, since resourceswill never be forthcomingfor every countryin Africa or even for the continentas a whole to developcapacity at the highest level in all fieldsand subfields. Ultimatelydecisions must be made on a discipline-by-discipline oasis and in light of an assessmentof regionalrequirements. Sometimes an institutionserving as the site for a programof excellencein one field may also need to be designatedas the site for programsin closelyrelated fields with which professionalinteraction is crucialand for which facilitiesand - 159 - equipmentmay be shared. Scale and sustainabilityare crucialmatters. Fragmentationin spacc and time will defeatthe purpose. Some internationalizationof managementcontrol and financemay provideprudent insuranceagainst such risks.

Here there is a strong case for Africangovernments and donor agencies to establishjointly some overallcoordinating mechanism to ensure reasonable efficiencyin developingthe programsof excellenceconcept. A useful first step in this directionwould be the appointmentof a small but very high level study group to considerthe evidenceand rendera judgmenton the feasibility, desirability,organization, priority fields, and possiblesources of finance for a sustainedeffort to transformthe qualityof Africanpostgraduate educationand research.

9.5 Call to action

The new era in internationalassistance for African education and trainingshould commence without delay. Donors and African governmentsneed now to come togetherto determinewhat concretesteps shouldbe taken to supportadjustment, revitalization, and selectiveexpansion of African education. This paper will have servedits purposeif it stimulatessAfrican governmentsto rethinkpolicies for educationaldevelopment, international &*^enciesto reflecton how their aid can be improvedand enlarged,and all parties to enter into a new partnershipto provideAfrica with the stock of human skills indispensiblefor developmentin the next century. - 160 -

SELECTEDBIBLIOGRAPHY

Note: In the courseof preparingthis policy paper,background papers were commissionedon variouspertinent themes. These reviewsand the extensivebibliographies that they containare the major sources for this paper. Ia the bibliographythat follows,the speciallycommissioned backgroundpapers are indicatedby an asteriskenclosed in parentheses(*) after the entry.

AfricanDevelopment Bank. 1986. Education ctor StrategyPaper. Abidjan, Cote d'Ivoire.

Aklilu Habte, G. Psacharopoulos,and S. Heyneman. 1985. Educationand DevelopmentsViews from The World Bank. Washington,D.C.: The World Bank.

Armitage,J. and R. Sabot. Forthcoming. "Efficiencyand Equity Implications of Subsidiesof SecondaryEducation in Kenya." In David Newberryand NicholasStern, eds., The Theory of_Taxationfor DevelopingCountries. New York: OxfordUniversity Press.

Auerhan,J., S. Ramakrishnan,R. Romain,G. Stoikov,L. Tiburcio,and P. Torres. 1985. InstitutionalDevelopment in Educationand Trainingin Sub-SaharanCountries. DiscussionPaper SeriesNo. EDT22. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Avalos, B. and W. Haddad. 1981. Review of TeacherEffectiveness Research in Africa,Middle East, Malayuia,Philippines & Thailand:A Synthesisof Results. Ottawa:International Development Research Centre.

Behrman,J. and N. Birdsall. 1983. 'The Qualityof Schooling:Quantity Alone is Misleading.' American EconomicReview 73 (Dec.):926-46.

Benson,C. 1987. "Taxonomiesof Skill Development:A Search for Criteriato Predictthe RelativeEfficiency of AlternativePrograms of Occupational Training." DiscussionPaper Series (forthcoming),Education and Training Department,The World Bank, Washington,D.C.

Bertrand,T. and R. Griffin. 1984. 'FinancingEducation in Kenya." DiscussionPaper. CountryPolicy Department,The World Bank, Washington, D.C.

Birdsall,N. 1983. "Demandfor PrimarySchooling in Rural Mali: ShouldUser Fees Be Increased?" DiscussionPaper No. 1983-8. Country Policy Department,The World Bank, Washington,D.C.

_ 1985. "Cost Recovery in Health and Education: Bank Policy and Operations." Population,Health, and NutritionDepartment, The World Bank, Washington,D.C. - 161 -

Bowcock,D. 1985. "Languagesof Sub-SaharanAfrica: A Listingby Country."' Educationand TrainingDepartment, The World Bank, Washington,D.C. (*)

Bowman,M.J. and R.H. Sabot. 1982. "HumanResources in Africa:A Continentin Rapid Change." Backgroundpaper presentedat the 1982 Conferenceof AfricanGovernmental Experts on TechnicalCooperation Among African Countrieson Human ResourcesDevelopment and Utilization,April.

Bray, M. 1986. New Resourcesfor Education. London:Commonwealth Secretariat.

Caldwell,J.C. 1979. "Educationas a Factor in MortalityDecline: An Examinationof NigerianData." PopulationStudies 33:395-413.

and P. Caldwell. 1985 "Education and Literacy as Factorsin Health." In Scott B. Halstead,Julia A. Walsh, and Kenneth S. Warren, eds., Good Health at Low Cost. New York: The RockefellerFoundation.

Cochrane,S.H. 1979. Fertilityand Education:What Do We Really Know? Baltimore,Maryland: Johns HopkinsUniversity Press.

_ 1980. Educational Differentials in Mortality of Children. Paper presentedat the AnnualMeeting of the PopulationAssociation of America. April.

Cochrane,S.H., D.J. 0'Hara,and J. Leslie. 1980. The Effectsof Educationon Health. World Bank Staff WorkingPaper No. 405. The World Bank, Washington,D.C.

_ 1982. "Parental Education and Child Health: Intracountry Evidence." Health Policyand Education(2):213-50.

Colelough,C. 1980. PrimarySchooling and EconomicDevelopment; A Reviewof the Evidence. World Bank StaffWorking Paper No. 399. The World Bank, Washington,D.C.

Court,D. aud K. Kinyanjui. 1986. "AfricanEducation: Problems in a High-GrowthSector." In Robert J. Berg and JenniferSeymour Whitaker, eds., Strategiesfor AfricanDevelopment. Berkeley,Californiat Universityof CaliforniaPress.

Denison,E. 1979. Accountingfor SlowerEconomic Growth. Washington,D.C.: The BrookingsInstitution.

Dorsey,B.J. 1986. 'Developmentand Reform in Education- Zimbabwe:A Case Study." Paper presentedat the 1986 Comparativeand International EducationSociety Conference, Tor¢nto, Canada.

Dougherty,C. 1987. "Cost-Effectivenessof TrainingDelivery Modes: A Review." Educationand TrainingDepartment, The World Bank,Washington, D.C.

Dutcher,N. 1982. The Use of First and SecondLanguages in PrimaryEducation: SelectedCase Studies. World Bank Staff WorkingPaper No. 504. The World Bank, Washington,D.C. - 162 -

Eicher, J.C. 1985. EducationalCostinA and FinancingIn DevelopintCountries with SpecialReference to Sub-SaharanAfrica. World Bank Staff Working Paper No. 655. The World Bank, Washington,D.C.

Eisemon,T.O. 1986. "Benefittingfrom Basic Educationin Developing Countries:A Reviewof Researchon the EducationalAntecedents of School Effects." Centre for Cognitiveand EthnographicStudies, McGill University,Toronto.

Foster, P. 1965. "The VocationalSchool Fallacy in DevelopmentPlanning." In C.A. Andersonand M.J. Bowman,eds., Educationand EconomicDevelopment. Chicago:Aldine Press.

_____. 1982. "The EducationalPolicies of PostcolonialStates." In L. Andersonand D. Windham,eds., Educationand Development:Issues in the Analysis and Plannintof Post-ColonialSocieties. Lexington, Massachusetts:Lexington Books.

_ 1985. "Educationin Sub-Saharan Africa: Some PreliminaryIssues." Educationand TrainingDepartment, The World Bank, Washington,D.C. (*)

Fryer, M. 1986. 'Femalesas Beneficiariesof Bank Operationsin Africa." Educationand TrainingDepartment, The World Bank, Washington,D.C. (*)

Fuller,B. 1985. RaisingSchool Qualityin DevelopinSCountries: What InvestmentsBoost Learning?. DiscussionPaper SeriesNo. EDT7. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Haddad,W. 1978. EducationalEffects of Class Size. World Bank Staff Working Paper No. 280. The World Bank, Washington,D.C.

_ 1979. Educationaland EconomicEffects of Promotionand Repetition Iractices. World Bank Staff WorkingPaper No. 319. The World Bank, Washington,D.C.

_ 1985. TeacherTrainints A Reviewof World BankExperience. Discussion Paper Series No. EDT21. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Hakuta,K. 1985. 'Generalizationsfrom Researchin SecondLanguage Acquisitionand Bilingualism."Paper deliveredto the House Educationand Labor Committee. Washington,D.C.

Haveman,R. and B. Wolfe. 1984. 'Schoolingand EconomicWell-Being: The Role of Non-MarketEffects." Journalof Human ResourcesXIX (summer):408-28.

Haves,H. and T. Coombe,eds. 1986. EducationPriorities and Aid Responsesin Sub-SaharanAfrica. Instituteof Education,University of London,London.

Hevneman,S. 1980a. "Evaluationof Human Capital in Aalawi." World Bank Staff Working Paper No. 420. The World Bank, Washington,D.C.

_ 1980b. 'Instructionin the Mother Tongue:The Question of Logistics." Canadianand InternationalEducation, Vol. 9, Number 2. - 163 -

, 1983. "EducationDuring a Periodof Austerity:Uganda, 1971-1981." ComparativeEducation Review 27 (October)s403-13.

and D. Jamison. 1980. 'TextbookAvailability and Other Determinantsof StudentLearning in Uganda." ComparativeEducation Review 24 (June)s206-220.

and W. Loxley. 1983. "TheEffect of Primary-SchoolQuality on Achievementacross Twenty-Nine High- and Low-IncomeCountries." American Journalof Sociology88 (June):1162-94.

Hicks, N. 1980. EconomicGrowth and Human Resources. World Bank Staff Working Paper No. 408. The World Bank, Washington,D.C.

Hinchliffe,K. 1985. IssuesRelated to Higher Educationin Sub-Saharan Africa. World Bank Staff Working Paper No. 780. The World Bank, Washington,D.C. (*)

. 1986. The MonetarXand Non-MonetaryReturns to Educationin Africa. DiscussionPaper Series No. EDT46. Educationaud Training Department,The World Bank, Washington,D.C. (*)

Horn, R. and A.M. Arriagada. 1986. The EducationalAttainment of the World's Population:Three Decadesof Progress. DiscussionPaper SeriesNo. EDT37. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Houlev C. 1973. The ExternalDegree. San Francisco,Washington, and Londons Jossey-BassPublishers.

Huffman,W. and J. Klock. Forthcoming. 'Assistancefor Education." In Anne 0. Krueger,C. Michalopoulos,and VernonW. Ruttan,eds., The Impactof DevelopmentAssistance to LDCs. Baltimore,Maryland: Johns Hopkins UniversityPress.

Imhoof,M. and P. Christensen,eds. 1986. TeachingEnglish by Radio: InteractiveRadio in Kenya. Washington,D.C.: Academy for Educational Development,Inc.

Inkeles,A. and D.H. Smith. 1974. BecomingModern: IndividualChange in Six DevelopingCountries. Cambridge,Massachusetts: Harvard University Press.

Jamison,D. 1982. "ReducedClass Size and Other Alternativesfor Improving Schools:An Economist'sView." In Gene V. GlasstLeonard S. Cahen,Mary Lee Smith, and NikolaN. Filby, eds., SchoolClass Size. BeverlyHills and London: Sage Publications.

and L. Lau. Forthcoming. "Educationand EconomicGrowth in Sub-Saharan Africa." Educationand TrainingDepartment, The World Baak, Washington, D.C. (*)

and P. Moock. 1984. "FarmerEducation and Farm Efficiencyin Nepal: The Role of Schooling,Extension Services, and CognitiveSkills." World Development,12 (January)t67-86. - 164 -

Jimenez,E. 1986. "PublicSubsidization of Educationand Health in Developing Countries:A Reviewof Efficiencyand Equity." World Bank Research Observer 1 (Jan.)s111-30.

_ Forthcoming. 'Structureof EducationalCosts: Multiproduct Cost Functionsfor Primaryand SecondarySchools in Latin America." Economics of EducationReview.

Kaluba,L.H. and P.P.W.Achola. 1985. "CommunityFinancing of Schools in CommonwealthSADCC Countries,A Non-GovernmentView from Zambia." Paper presentedat the CommonwealthRegional Workshop with SpecialReference to SouthernAfrica, Gaborone,Botswana. CommonwealthSecretariat, London.

King, K. 1986. "Manpower,Technology, and Employmentin Africa: Internaland ExternalPolicy Agendas." In Robert J. Berg and JenniferSeymour Whittaker,eds., Strategiesfor AfricanDevelopment. Berkeley:University of CaliforniaPress.

Knight,J.B. and R.H. Sabot. 1986. Overviewof EducationalExpansion, Productivityand Inequality:A ComparativeAnalysis of the East African Natural Experiment. DiscussionPaper SeriesNo. EDT48. Educationand TrainingDepartment, The World Bank, Washington,D.C. (*)

Lee, K. Forthcoming. UniversalPrimary Education: An AfricanDilemma. DiscussionPaper Series. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Lindauer.D.L. 1984. 'PublicSector Pay in Africa:An AnalyticalFramework." CountryPolicy Department,The World Bank, Washington,D.C.

_____, O.A.Meesook, and P. Suebsaeng. 1986. GovernmentWage Policy in Africa:Summary of Findingsand Policy Issues. DiscussionPaper Series No. 1986-24. CountryPolicy Department,The World Bank, Washington,D.C.

Livingstone,I. 1985. "Perceptionof the Intendedand ImplementedMathematics Curriculum."Mimeographed. University of Illinois.

Lockheed,M., D. Jamison,and L. Lau. 1980. "FarmerEducation and Farm Efficiency:A Survey." EconomicDevelopment and CulturalChange 29 (October):37-76.

Lucas,R.E.B. and 0. Stark. 1985. "Motivationsto Remit:Evidence from Botswana." Journalof PoliticalEconomy 93 (October):901-18.

MacMillan,D.L. 1981. LanguagePolicies for African PrimaryEducation: Summaryof the AnglophoneResearch Literature. DiscussionPaper No. 28. Populationand .iumanResources Division, The World Bank, Washington,D.C.

Malawi, Governmentof. 1984. The Impactof the Increasein SchoolFees on PrimarySchool Enrollments In 1983. Lilongwe:Ministry of Educationand Culture.

Mass, J. and G. Criel. 1982. PrimarySchool Participation and Its Internal Distributionin EasternAfrica. World Bank Staff WorkingPaper No. 511. The World Bank, Washington,D.C. - 165 -

Mbanefoh,G.F. 1980. "Sharingthe Cost and Benefitsof UniversityEducation in Nigeria." InternationalJournal of EducationalDevelopment l(July)231-43. McMahon,W. Forthcoming. "The Relationof Educationand R&D to Productivity Growth in the DevelopingCountries of Africa." Economicsof Education Review.

Michael,R. 1982. "1MeasuringNon-Monetary Benefits of Education:A Survey." In W. McMahon and T. Geske, eds., FinancingEducation. Urbana:University of IllinoisPress.

Middleton,J., HabteselassieWoldemariam, and C. Mayo-Brown. 1986. Management in World Bank EducationProlects: Analysis of Experience. Discussion Paper SeriesNo. EDT42. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Millot, B., F. Orivel,and J. Rasera. 1987. L'aide exterieurea l'education en Afrigue sub-saharienne.Discussion Paper SeriesNo. EDT65. Education and TrainingDepartment, The World Bank, Washington,D.C. (*)

Mingat,A. 1985. "La Diversificationdes Sourcesde Financementde l'EnseignementSuperieur en AfriqueFrancophone." Course material preparedfor a seminarorganized by the EconomicDevelopment Institute, World Bank, Washington,D.C.

I and G. Psachar,poulos.1985. "FinancingEducation in Sub-Saharan Africa: Issuesof Equity and Efficiencyof Investment- Some Policy Alternatives."Finance and Development22 (March):35-38.

______, and J. Tan. 1985. Improvingthe Quality-QuantityMix in Education:A Stimulationof Policy Tradeoffs. DiscussionPaper SeriesNo. EDT15. Educationand TrainingDepartment, The World Bank, Washington,D.C.

_ 1985b. "Subsidizationof HigherEducation versus Expansionof Primary Enrollments:What Can a Shift of ResourcesAchieve in Sub-SaharanAfrica?" InternationalJournal of EducationalDevelopment 5:259-68.

_ 1986. "ExpandingEducation through User Charges:What Can Be Achieved in Malawi and Other LDCs?" Economicsof EducationReview, No. 3.

_____. 1987. "The EconomicReturns to Investmentin Project-Related Training:Some Evidencefrom World Bank Projects." Educationand Training Department,The World Bank, Washington,D.C.

Moock, P. 1981. "Educationand TechnicalEfficiency in Small-Farm Production." EconomicDevelopment and CulturalChange 29 (July):723-39.

Mundangepfupfu,R.M. 1986. Economiesin the Provisionof Facilitiesfor TeachingS*condary School Science. EconomicDevelopment Institute, Educationand TrainingDesign Division, The World Bank, Washington,D.C.

Neumann,P. 1980. Publishingfor Schools:Textbooks and the Less Developed Countries. World Bank Staff WorkingPaper No. 398. The World Bank, Washington,D_7C. - 166 -

Noor, A. 1985. 'StrengtheningEducational Management: A Review of World Bank Assistance,1963-83." Mimeographed. Educationand TrainingDepartment, The World Bank, Washington,D.C.

Odtntan,S.O. 1975. "The Health of NigerianChildren of SchoolAge." Brazzaville:Regional Office, World HealthOrganization.

Organizationof AfricanUnity. 1981. Latos Plan of Actiot.for the Economic Developmentof Africa, 1986 - 2000. InternationalInstitute for Labour Studies,Geneva.

_ 1985. Africa'sPriority Programme for EconomicRecovers 1986 - 1990. Twenty-firstSession of the Assemblyof Heads of State and Governmentof the OAU. Addis Ababa, Ethiopia.

Paul, S. 1982. ManagingDevelopment Programs: The Lessonof Success. Boulder,Coloradot Westview Press.

Perraton,H., ed. 1982. AlternativeRoutes to Formal Education:Distance Teachingfor School Ec-ivalency.World Bank ResearchPublication. Baltimoreand London:The Johns HopkinsUniversity Press.

_ ___ ,C. Block,M.L. Fryer, P.L. Spain,and M. Young. 1986. Distance Education:An Economicand EducationalAssessment of Its Potenti4lfor Africa. DiscussionPaper SeriesNo. EDT43. Educationand Training Department,The World Bank,Washington, D.C. (*)

Psacharopoulos,G. 1980. Higher Educationin DevelopingCountries: A Cost-BenefitAnalysis. World Bank StaffWorking Paper No. 440. The World Bank, Washington,D.C.

. 1984. 'The Contributionof Educationto EconomicGrowth: InternationalComparisons." In John Kendrik,ed., International ProductivityComparisons and the Causesof Slowdown. Cambridge, Massachusetts:Ballinger.

_ . 1985. 'Returnsto Education:A FurtherInternational Update and Implications."Journal of Human Resources XX (Fall):583-604.

__ 1987. "To Vocationalizeor Not to Vocationalise?That is the CurriculumQuestion." InternationalReview of Education,No. 2.

and A.M. Arriagada. 1986. 'The EducationalAttainment of the Labor Force:An InternationalComparison or Moving BeyondEnrollment Ratios for AssessingPriorities in EducationalInvestment." Mimeographed. EducAtion and TrainingDepartment, The World Bank,Washington, D.C.

and W. Loxley. 1985. DiversifiedSecondary Education and Development: Evidencefrom Colombiaand Tanzania.Baltimore, Maryland: Johns Hopkins UniversityPress.

and M. Woo1hall. 1985. Educationfor Development:An Analysis of InvestmentChoices. New York: OxfordUniversity Press. - 167 -

Rogers,D.C. 1972. "StudentLoan Programsand the Returnsto Investmentin Higher Levels of Educationin Kenyt." EconomicDeveloiment and Cultural Chante 2 (January)t243-59.

Romain,R. 1985. LendinR in PrimaryEducation: Bank PerformanceReview, 1962-1983. DiscussionPaper SeriesNo. EDT20. Educationand Training Department,The World Bank, WashingtontD.C.

Schultz,T.P. 1975. "The Value of the Abilityto Deal with Disequilibria.' Journalof EconomicLiterature 13 (September):827-46.

_ 1985. "SchoolExpenditures and Enrollments,1960-1980: The Effectof Income,Prices and PopulationGrowth." DiscussionPaper No. 487. EconomicGrowth Center,Yale University,New Haven, Connecticut.

Smock, A.C. 1981. Women'sEducation in DevelopingCountriess Opportunities and Outcomes. New York: Praeger.

Somerset,H.C.A. 1987. "ExaminationReforms: The Kenya Experience." DiscussionPaper Series No. EDT64. Educationand TrainingDepartment, The World Bank, Washington,D.C.

__ 1985. "TheQuality of ElementaryEducation in Africa: Some Key Issues." Educationand TrainingDepartment, The World Bank, Washington, D.C. (*)

Stoikov, V.L. 1975. "The Economicsof RecurrentEducation and Training." InternationalLabour Organization,Geneva.

Stromquist,N.P. 1986. "EmpoweringWomen ThroughKnowledge: Policies and Practicesin InternationalCooperation in Basic Education." School of Education,Stanford University, Palo Alto, California.

Tan, J. Forthcoming. "PrivateDirect Cost of SecondarySchooling in Tanzania." InternationalJournal of DevelopmentEconomics.

_, K.H. Lee, and A. Mingat. 1984. User Charges for Education:The Ability and Willingnessto Pay in Malawi. World Bank Staff WorkingPaper No. 661. The World Bank, Washington,D.C.

Thevenin,T. 1981. PedagogicalImplications of LanguatePolicy in African Schools:A Review of the FrancophoneLiterature. DiscussionPaper No. 81-29. Populationand Human ResourcesDivision, The World Bank, Washington,D.C.

Thobani,M. 1983. ChargingUser Fees for Social Services:The Case of Malawi. World Bank Staff Working Paper No. 572. The World Bank; Washington,D.C.

UnitedNations. 1966-1985. Yearbookof NationalAccounts Stat'.stics.New York.

______1986. Programmeof Action for African EconomicRecovery and Development1986 - 1990. (A/RES/S-13/2) - 168 -

UnitedNations EconomicCommission for Africa. 1984. Human Resourcesin Africa. MonographNo. 13. New York.

__ 1985. Report of the Second Conferenceof Vice-Chancellors/Presidents/ Rectorsof Institutionsof Higher Learningin Africa. Mbabane,Swaziland, February1985. (EJECA/CM.11/47)

_ 1986. Africa'sDevelopment Priorities and the Role of Institutions of HigherLearnint: The Next Five Years. ECA/AAUThird Conferenceof Vice-Chancellors,Presidents and Rectorsof the Institutionsof Higher Learningin Africa. Harare,Zimbabwe, January 1987. (E/ECA/AAUIED/86/3)

Unesco. 1961. Final Report. Conferenceof African Stateson the Development of Educationin Africa,Addis Ababa, Ethiopia,May.

* 1983. AfricanLanguages as InstructionalMedia. Dakar:Unesco RegionalOffice for Educationin Africa.

_ 1985. AfricanCommunity Languages and Thei' Use in Literacyand Education. Dakar: Unesco RegionalOffice for Educationin Africa.

Unicef. 1985. Within Human Reach:A Future for Africa'sChildren. Unicef Report. New York.

Wheeler,D. 1980. Human ResourceDevelopment and EconomicGrowth in DevelopingCountries: A SimultaneousModel. World Bank Staff Working Paper No. 407. The World Bank, Washington,D.C.

Windham,D.M. 1987. InternalEfficiency and the African School.Discussion Paper Series No. EDT47. Educationand TrainingDepartment, The World Bank, Washington,D.C. (*)

Wolff, L. 1984. Controllingthe Costs of Educationin EasternAfrica: A Review of Data, Issues,and Policies. World Bank Staff Working Paper No. 702. The World Bank, Washington,D.C.

Woodhall,M. 1983. StudentLoans as a Means of FinancintHigher Education: Lessons from InternationalExperience. World Bank Staff Working Paper No. 599. The World Bank, Washington,D.C.

World Bank. 1980a. EducationSector Policy Paper. Washington,D.C.: The World Bank.

__ 1980b. World DevelopmentReport. 1980. New York: Oxford University Press.

______1981. AcceleratedDevelopment in Sub-SaharanAfrica: An Agenda for Action. Washington,D.C.s The World Bank.

_ 1984. Toward SustainedDevelopment in Sub-SaharanAfrica: A Joint Proum of Action. Washington,D.C.: The World Bank.

______1986a. FinancingAdiustment With Growth in Sub-SaharanAfrica, 1986-90. Report No. 6082. Washington,D.C.: The World Bank. - 169 -

_ 1986b. PopulationGrowth and Policiesin Sub-SaharanAfrica. Washikigton,D.C.: The World Bank.

_ 1986c. FinancinxEducation in DevelopingCountries. Washington, D.C.s The World Bank.

/Unesco. 1984. Joint Study of TechnicalAssistance in Bank-Financed EducationProjects. Reportof the SteeringCommittee to the Presidentof the World Bank and the DirectorGeneral of Unesco.

Young, M., H. Perraton,J. Jenkinsand T. Dodds. 1980. DistanceTeaching for the Third World: The Lion and the ClockworkMouse. London,Boston and Henley: Routledgeand Kegan Paul. - 170 -

DEFINITIONSAND ACRONYMS

International Organizations

World Bank The term 'World Bank' as used in this paper reftrs to an internationallending institution comprised of the InternationalBank for Reconstructionand Development(IBRD) and its affiliate,the International DevelopmentAssociation (IDA). The commonobjective of these institutionsis to help raise standardsof living in developingcountries by channeling rinancialresources from developedcountries to the developingworld. Lending is only done for productivepurposes that stimulateeconomic growth in its Borrowercountries, traditionally through development projects in capital infrastructure.The World Bank'sdevelopment strategy also places an emphasis on investmentsthat can directlyaffect the well-beingof the poor in developingcountries by making them more productiveand integratingthem as activepartners in the developmentprocess.

IBRD The InternationalBank for Reconstructionand Development(IBRD) is the World Bank affiliatethat provideslending for projectsin developing countriesthat are at the more-advancedstages of economicand social growth.

IDA The InternationalDevelopment Association (IDA) was establishedas a World Bank affiliateto provideassistance for the same purposesas the IBRD, but priwarilyin the poorer developingcountries -- thosewith an annualper capita gross nationalproduct of less than $791 (in 1983 dollars),on terms that bear less heavilyon their balanceof paymentsthan IBRD loans.

UNESCO The United NationsEducational, Cultural and ScientificOrganization (UNESCO)is a specializedagency of the United Nations responsiblefor promotingcollaboration among UN Member Statesthrough education, science and culture. Its activitiesin educationinclude promoting information exchanges between countries,convening meetings of ministersof education,and prov3ding advisoryservices or carryingout, as part of its operationalwork, specific practicalprojects in its Member States.

ILO The InternationalLabour Organisation(ILO) -Isa specialized intergovernmentalagency of the UnitedNations that seeks to improvelabor conditions,raise livingstandards and promoteproduetive employment. Its activitiesinclude research, advisory services and the provisionof technical assistanceto UN Member States.

ADB The African DevelopmentBank Group (ADB) is an internationallending institutionthat encourageseconomic development in its AfricanMember States by financingprojects in the agriculture,transport, public utilities, industry,education and health sectors. - 171-

ECK The United NationsEconomic Commission for Africa (ECA) is a regional intergovernmentalorganization established in 1958 to fosterthe economic developmentand integrationof its Member Statesthrough concerted intergovernmentalcooperation to promote strategiesfor regionaldevelopment.

Levels of Schooling

PrlmaryEducation, Basic Education The first levelof education,in which studentsfollow a commoncurriculum. In "primary'education students receive instructionin primaryor elementaryschools that are part of the formal educationsystem. These schoolsspan grades 1-4, 1-5, 1-7 or 1-8, with curriculaof communication,mathematics and science."Basic" education generallyrefers to instructionin literacyand numeracyskills for out-of-schoolyouths and adults.

SecondaryEducation Secondaryeducation requires at least four years of primarypreparation for entry. Curriculaat this level are eitherof a generalacademic nature or orientedto the teachingof job-relatedskills. Through a "general'curriculum students follow an academicprogram that typicallyleads to admissionto a post-secondaryinstitution; under both 'technical/vocational'and 'agricultural'curricula students are preparedfor direct entry into a trade or occupation. This level may span any subsetof Grades 5-12.

Post-Secondary.Tertiary. Hixher Level Education Educationat this level requires,as a minimumcondition of entry, the successfulcompletion of educationat the secondarylevel or proof of equivalentknowledge or experience. Instructionis given in varied types of institutionssuch as universities,vocational/technical training institutes, and teachertraining institutes.

GeneralTens

CorresgondenceEducation A type of distanceeducation through which students receivetextbooks for individualstudy in their homes and supplementarymass media broadcastson the subjectmatter. Each textbookincludes exercises to be completedand mailed to a postaltutor for marking and providingindividual feedbackto the student.

Curriculum A set of coursesin a field of study,often constitutingan area of specializationat the higher levelsof the educationcycle.

DistanceEducation. Distance Teaching An educationdelivery system that uses a varietyof media and a systemof feedbackto provideeducation to peoplewho are unable to attendtraditional schools. Distanceeducation usually involves a combinationof media broadeasts,use of printedmaterials by the student, and some kind of face-to-facestudy. Distanceteaching programs can range from in-schoolprograms in which broadcastsare used to supplementlearning activitiesin the classroom,to out-of-schoolprograms such as "correspondence lessons"in which studentsmay never meet their tutors and may have littleor no contactwith the regulareducational system.

DiversifiedSecondary Education Educationat the secondarylevel adaptedto the expectedroles of school leaversthrough the introductionof practicalor occupationalsubjects into an otherwisecompletely academic program. Two models are prevalent. The first introducespractical subjects (industrial - 172 - arts, home economics,agriculture) at the lower secondarylevel to provide prevocationalorientation and to developa positiveattitude toward work. The second includesa generalacademic stream plus one or more specialized occupationalstreams, usually at the upper secondarylevel, depending on the economiclife of the surroundingarea.

EnrollmentRatio School enrollment,both public and private,as a percentage of a given populationage group. A gross enrollmet.tratio is the total number of studentsenrolled at a given educationallevci dividedby the populationof the age group for that level. This ratiomay includeunder-age and repeater students. A net enrollmentratio is calculatedby using only that part of the total number of studentsenrolled that correspondsto the specificage groups defined for that level.

EquivalencyProgram A teachingprogram that providesopportunities for educationto studentswho would otherwisebe unable to attendformal schools and that emphasizesthe acquisitionof knowledgerather than the place where the knowledgeis acquired. Equivalencyprograms exist for all levelsof schooling.

ExtramuralDegree (Certificate),External Degree (Certificate) A degree (certificate)awarded by a universityto studentsparticipating in equivalency programs iA which studentsstudy the curriculumon their own without attendanceat formal classes.

ExternalExamination An examinationset by an organizationoutside of the individualschool, and independentof it, that is administeredto a large number of studentsfrom differentschools, to allow comparisonof results across schools.

IncompleteSchool A school that does not have the full number of grades for the levelof educationprovided.

NonformalEducation Educationand trainingfor out-of-schoolyouths and adults in classes,courses, or activitiesintended to promotelearning but not constitutingpart of the formalschool systemand not leadingto formal qualificationssuch as diplomasor specifictrade standards. Nonformal educationtypically concentrates on short programsof a few months duration.

Non-GovernmentalOrganization (NGO) A non-profitorganization with private membershipthat providesdevelopment assistance. Also calledPrivate VoluntaryOrganizations (PVOs), the NGOs includefoundations, lay and religiousaid associations,and non-governmentalcooperatives.

TechnicalEducation Trainingin specialistskills for the higher-level skilledworker, typicallyin pre-employmenttraining institutions such as polytechnics. Administrationof technicaleducation is commonlythe responsibilityof ministriesof education,labor or employment,or comes under the authorityof specificindustries. Coursesof study under technical educationusually includea generaleducation component.

Training Instructionin job-relatedskills to prepare studentsfor direct entry into a trade or occupation. Such instructioncan take place in training centers,through apprenticeships, and/or at the place of employment ("on-the-Job"training). - 173 -

VocationalEducation Training in craft or trade skills for the semi-skilled worker that is typicallyschool-based and under the directionof ministries of education. - 174 -

FOOTNOTES

Introduction

1/ Most of the discussionand all of the Africa-widestatistics in this paper refer to just 39 countriessouth of the Sahara. Wheneverthe term "Africa"or 'African"is used, it refersonly to these countriesand their residents. The StatisticalAnnex to this paper lists and providesfull comparativeinformation on these 39 countries.Excluded from the Annex tables are statisticson Namibia and SouthAfrica; on the North Africancountries; and on six emall countrieswith populationsof less than one half millionpeople in 1983 (Cape Verde, Comoros,Djibouti, Equatorial Guinea, Sao Tome and Principe, and Seychelles)tfor which only the most Lasic informationis provided(in Annex Table C.1). For the other countriessocio-economic statistics are given through 1984, but internationallycomparable educational statistics were availableonly through 1983. This is an unfortunateend-year since, in many respects,1983 was the worst year of the economiccrisis, and it is possible that a numberof educationalindicators have subsequentlybegun to improve.

2 For broaderanalysis than could be includedin this paper on educationstrategies, several important documents are available. The 'Lagos Plan of Action,' adoptedby AfricanHeads of State under the auspicesof the Organizationof African Unity (OAU) in April 1980, analyzesAfrica's problems and puts forwarda comprehensiveplan for addressingthem (OAU,Lagos Plan of Action for the EconomicDevelopment of Africa, 1981]. In accordancewith the policiesset forth in this plan, the OAU met again and agreed to a set of activitiesand priorities;these were spelledout in a second document (OAU, Africa'sPriority Programme for EconomicRecovery, 1985]. Subsequently,the U.N. GeneralAssembly adopted a set of resolutionspertaining to African -..velopment[United Nations, Programme of Action for African EconomicRecovery and Development,1986]. Concurrentlywith these efforts,the World Bank has publishedits own series of specialreports on macroeconomicissues and developmentrequirements for Africa. The first in this series,Accelerated Developmentin Sub-SaharanAfrica, was releasedin 1981; the fourth and most recent,Financina Adjustment With Growth in Sub-SaharanAfrica, appearedin 1986.

Chapter1

V1 The gross enrollmentratio dividesthe enrollmentby the total, male and female,population of school age. The officialage range differsfrom countryto countrydepending on the structureof the educationsystem. The gross enrollment'ratio is an accuratemeasure of the schoolsystem's ca-acitv relativeto the school age population. By includingstudents who are outside the officialage range,however, it overstatesthe percentageof this populationthat is actuallyenrolled. The net ratio,which excludesoverage and underagechildren, is conceptuallysuperior, but generallythis information is not available. Problemswith gross enrollmentstatistics are further discussedin the TechnicalNotes pertainingto Annex TablesA.7-A.9. - 175 -

2V The numbersQf schoolsand teachersin this paragraphand the next, while based on the figuresin Annex Table A.10, reflectreasonable assumptions about missing data in the table.

V UNESCO local-currencyexpenditure figureswere inflatedto reflect 1983 prices using country-specificGDP deflatorsand then convertedto U.S. dollarsusing official 1983 exchange rates.

41 This probablyoverstates somewhat the declinein "real"education expendituresince, in most Afric;ncountries in recent years,teachers' salaries (whichcomprise approximately 902 and 702 of recurrentexpenditure in primaryand secondaryeducation, respectively, and above 502 in tertiary)have not gone up as rapidlyas prices in generalas reflectedin the GDP deflator.

Al In terms of weightedmeans, as distinctfrom medians,education's share went from above 162 throughoutthe 1970.to below 121 in 1983. This more precipitousdrop in education'sshare of expenditureafter 1980 reflects,to a very large extent,what was happeningin a single large country-- Nigeria. See Annex Table A.14.

6 This figureincludes a small part -- about 32 -- from the InternationalBank for Reconstructionand Development(as distinctfrom the InternationalDevelopment Association) of the World Bank Group and from the AfricanDevelopment Bank (as distinctfrom the AfricanDevelopment Fund) that was given in the form of-. arket-rateloans. All of tho remaining97% can be countedas concessionalaid to Africaneducation.

Chapter 2

11 Africa'shigh populationgrowth rate reflectsthe fact that mortalityrates have declinedvery rapidlyin recentyears while fertility rates remain,for the most part, at the same high levelsthat prevailed historically. In 1960 the crude death rate in sub-SaharanAfrica was 25 per thousandand the crude death rate 49 per thousand. By 1983,owing to significantimprovements in healthcare, the death rate had fallento 17 per thousand,but the birth rate remainedpractically unchanged. Birth rates actuallyincreesed over this period in a handfulof countries.

V In technicalterms, the rate of returnis the discountrate that equatesthe presentvalue of the economiccosts and benefitsof an investment. Private rates of returnto educationare calculatedusing after-taxearnings differentialsand only those educationalcosts that are actuallyborne by studentsand their families. Social rates of return,which are more useful for public policy,are baosd on before-taxearnings differentials and on educationa'sfull resourcecosts. The full costs of educationequal the sum of all private costs plus any subsidiesgiven.

Chapter 3

11 Althoughcomparative education data more recent than 1983 were unavailablefor the purposesof this report,there is evidenceof some economic recoveryin the periodsince 1983,and this may have led to some recoveryin enrollments. - 176 -

2 This becomesless true as one moves up the educationalladder. At the tertiarylevel, in fact, per studentexpenditure is on a par with that in the industrializedcountries -- see Sect.ion6.2(c). V Expenditureper highereducation student in constantdollars for the median countrywas $2,462in 1970, $3,090 in 1975, $2,798 in 1980, and $2,710 in 1983. The weightedmean of expenditureper studentdid fall sharply, from $6,461 in 1975 (for 23 countrieswhere informationwas available)to $2,365 in 1983 (for 28 countries),but most of this declinewas accountedfor by Nigeria,a large countrywhere centralgovernment expenditure fell from $14,621to $2,181.

Chapter4

11 A fourthway to reducepupil-teaciier ratios, theoretically, is to reducethe numberof teachersin each class. It is assumed,however, that each primaryschool class is taughtby only one teacherand that this parameter, therefore,cannot be furtherreduced.

21 An increaseof the proportionof the public educationbudget allocatedto the primary subsectordoes not mean that publicexpenditures on secondaryand higher educationmust necessarilydecline in absoluteterms; they can remain the same or even go up while primaryeducation's share is going up, so long as the total public budget is growing. As arguedin Chapter 6, however,most of the increasein total resourcesfor higher education,which is urgentlyrequired in most countries,will have to come, not from growthof the public budget,but rather from increasedprivate financialcontributions to the subsector.

Chapter5

-/ It is worth rememberingthat virtuallynone of the now developed nationsof Europe and North America had secondaryenrollment ratios above 10% at the beginningof this century,and for the most part, ratios above 20% were achievedonly in the post-WorldWar II era.

2V The reader shouldbe remindedthat comparisonsof educationalcosts across countriesare fraughtwith problems. The problemshere are over and above those associatedmore generallywith the use of exchangerates to convert local currencyfigures into U.S. dollarsfor purposesof making international comparisons. In education,the collectionof statisticshas evolved largelyas a by-productof educationadministration and not as a consequenceof a policy concernto analyzeand controlcosts. This explainswhy data on students, teachers,and physicalfacilities, while certainlynot adequate (cf. Box 1.1), are neverthelessmuch more plentifulthan data on educationalcosts and financing. Great cautionmust be exercised,therefore, when using the cost data now availableas a basis for formulatingpolicy.

31 Other seriouslyunderrepresented groups includethe childrenof familiesliving in very remoterural areas, especiallynomads, and the children of politicaland economicrefugees. This last group is, unfortunately,a rapidlygrowing and often neglectedelement in today'sAfrica. - 177 -

41 The only real exceptions,where femalescomprise above 45% of secondaryenrollments, are Lesotho,Botswana, and Swaziland(in these threet virtual parity had alreadyteen achievedby 1960),Mauritius and Madagascar.In five more countries-- the Congo, Gabon,Kenya, Sudan, and Zimbabwe-- females now comprisebetween 40% and 45% of enrollments.

Chapter 6

11 The commentsin this chapteron the costs and financingof higher educationa-e based on robustempirical evidence. Findingson highereducation outputs,however, must be offeredmore in the nature of hypothesesworthy of intensiveefforts in individualcountries to gather and analyzethe data necessaryfor validationor rejection,than as incontrovertiblydemonstrated facts. Except for simplehead countsof graduates,talere is presentlya dearth of reliabledata on the levelsof outputs from the higher educationsystem and especiallyon changesin those levelsover time. African governmentsand their internationalpartners in developmentshould launch the necessarydata generationand analyticactivities, and be preparedto act promptlyon their results.

Chapter 7

11 The "headteacher"is used throughoutthis chapterto refer to the headmasteror headmistressof a primary school.

Chapter9

13 UInternational aid' as defined in this paper comprisesall flows of resources,financial and in-kind,originating from "Western"(OECD and OPEC) bilateraland multilateralsources and containinga 'concessional"element of at least 25%. It includes,therefore, in additionto grants,all loans carryingrepayment terms that are sufficientlybelow what would be attainable commerciallyfor the grant componentto exceed 25%. Escludedunder this definitionare non-concessionalloans from Western governmentsand the multilateralbanks (the World Bank and the AfricanDevelopment Bank) and, of course, all loans from private commercialbanks. Also excludedfrom this definitionare all flows from East Bloc countries(which do not routinely divulge the pertinentfigures) and flows from private,non-governmental organizations. Non-concessionaland East Bloc aid accountedfor only about 7% of flows to the educationsector in 1981-83,so the definitionof aid used here encompassesvirtually all flows. However,data on aid flows are not notable for their accuracyor consistency,and the figuresgiven here shouldbe interpretedas giving rough approximationsrather than exact magnitudes.

21 If one adds to this amount the aid flows from non-Western governments(the East Bloc, plus Egypt and India),those from private, non-governmentalorganizations, and the non-concessionalflows from the, multilateralbanks ("hard'loans from the World Bank and AfricanDevelopment Bank), the grand total of externalflows to African educationand training comes to nearly $1.6 billionannually. This is the equivalentof about 17% of African domesticpublic expenditureon education. - 178 -

V The inconsistency betweenTables 9.2 and 9.4 in the breakdownof direct educationaid betweencapital and recurrentexpenditures is due to differencesin the sourcesfor the two tablesand in the way tiese sources treat technical assistanc*l some technicalassistance, used for capacity-building purposes, is classified by some sources as an invostment item rather than as recurrent expenditure.

41 For reasons noted in the secondparagraph of section9.1 above, it is likelythat these figuresdo not give a completelyaccurate picture of current reality,althouph the overallconclusion would probablynot change even with Aore completedata. - 179 -

STATISTICALAMU - 180 -

STATISTICALANNEX

Table of Contents

AlphabeticalList of Countriesand Key to Tables...... 183

CountriesClassified by Linguisticand EconomicStatus ...... 184

List of Maps...... 185

List of Tables in Part ...... *9 186

List of Tables in Part B...... 190

Listof Tables in Part C ...... 192

Part A: Education Indicators...... ,...... 196

Part B: Economicand Social Indicators ...... 225

Part Ct Supplementaryand SummaryTables...... 243

Technical Notes...... 255

Bibliographyof Data Sources...... 275

Maps, - 181 -

INTRODUCTION

This three-partstatistical annex providesinformation on principalfeatures of educational,social and economicdevelopment for 39 countriesin sub-SaharanAfrica. Excludedare Namibia and South Africa, all North African countries,and (with the exceptionof one table) six countrieswith populationsof less than one-halfmillion people in 1984 (Cape Verde, CoGoros,Djibouti, Equatorial Guinea, Sao Tome and Principe, and the Seychelles).

The educationindicators in Part A were caiculatedfrom data suppliedby UNESCO and supplementedwith informationfrom World Bank data files and countrydocumentation. The social and economicindicators in Part B are collectedannually by the World Bank on its developingmember countries,and the tables in that annex were preparedin collaborationwith the EconomicAnalysis and ProjectionsDepartment and the Population, Health and NutritionDepartment of the World Bank. The data presentedin Part C supplementor summarizethose displayedin Parts A and B.

Althoughconsiderable effort has been made to standardizethe data, nevertheless,statistical methods, coverage, practices and defini- tions differ from countryto country. Moreover,weaknesses in developing countries'statistical systems limit the availabilityand reliabilityof the data. The indicatorsshould, therefore, be used to characterizethe trends and major differencesbetween countries and countrygroups rather than to show precisequantative measures of those trends and differences.

The formatof Parts A and B followsthat used in the Bank'sWorld DevelopmentReport. Countriesare classifiedinto two major economic groups -- low-incomeand middle-incomeeconomies. The 25 countrieswith per capita incomesof less than $400 are classifiedas low-income economies. The 14 middle-incomeeconomies are those with per capita incomesof greaterthan $400. The economiesare furtherclassified to distinguishlow-income semi-arid from other low-incomeeconomies and middle-incomeoil exportersfrom middle-incomeoil importers.

Within each economicgroup, countriesare listed in ascending order of incomeper capitawith the exceptionsof Chad, Mozambiqueand Angola for which per capita incomecould not be calculated. The alphabeticallist in the key on the next page shows the table reference number of each country.

Summarymeasures of all indicatorsappear in the bands wit...ln each table. The letterw followinga s=mmarymeasure indicatesa weighted mean; the letterm, a median;and the lettert, a total. Readersshould exercisecaution in comparingthe swtmarymeasures for different indicators,groups, and years or periods. Data coverageis not uniform for all indicators,and variationsaround central tendencies can be large. The technicalnotes, which outlinethe methods,concepts, definitions and data sources,should be referredto beforeusing any of the data. - 182 -

The tables also providesummary measures for countriesgrouped by languase. Those that could not be classifiedas eitherAnglophone or Francophoneare includedin the 'Other'category.

The Annex also containsmaps which have been preparedfor the convenienceof the reader. The denominationsused, and the boundaries shown do not imply any judgmenton the legal statusof any territoryor any endorsementor acceptanceof such boundarieson the part of the World Bank and its affiliates. - 183 -

almi3T C&LLIST OF COUNTRIESAND KEY TO TABLES

------__------In each annex table, countries are listed in their economic group in ascendingorder of GNP per capitaexcept for Chad, Mozambiqueand Angola for which no GNP per capita could be calculated. The reference numbers in the alphabeticallist below reflect the order in the tables. ------__------Angola 39 Malawi 9 Benin 17 Mali 1 Botswana 33 Mauritania 26 BurkinaFaso 2 Mauritius 34 Burundi 12 Mozambique 25 Cameroon 36 Niger 3 CentralAfrican Rep. 15 Nigeria 35 Chad 6 Rwanda 18 Congo 37 Senegal 24 Cote d'Ivoire 30 Sierra Leone 20 Ethiopia 7 Somalia 5 Gabon 38 Sudan 23 Gambia 4 Swaziland 32 Ghana 22 Tanzania 11 Guinea 21 Togo 14 Ouinea-Bissau 10 Uganda 13 Kenya 19 Zaire 8 Liberia 27 Zambia 28 Lesotho 29 Zimbabwe 31 Madagascar 16 ------___------

In the tables,the followingsymbols are used:

Not available. (. ) Less than half the unit shown. .1. Data includedin anothercategory. * Data are for a year other than that specified. + Estimatedor provisionaldata.

w Weightedmean. m Median. t Total.

Due to rounding,the totals and sub-totalsshown in the tables may not correspondto the sums of their components. ------__------184 -

COUNTRIESCLASSIFIED BY LINGUISTICAND ECONOKICSTATUS

------__------Francophone Antlo)hone Other (nm18) (n-16) (n=5) ------__---- Low-incomesemi-arid (n=6) ------__------2 BurkinaFaso 4 Gambia,The 5 Somalia 6 Chad 1 Mali 3 Niger ------__--- Low-incomeother (n-19) ------.------_------__------17 Benin 22 Ghana 7 Ethiopia 12 Burundi 19 Kenya 10 Guinea-Bissau 15 CentralAfrican Republic 9 Malawi 25 Mozambique 21 Guinea 20 SierraLeone 16 Madagascar 23 Sudan 18 Rwanda 11 Tanzania 24 Senegal 13 Uganda 14 Togo 8 Zaire ------__--- Mid4le-income oil importers (n=9) --. _------__------. 30 Cote d'Ivoire 33 Botswana 26 Mauritania 27 Liberia 29 Lesotho 34 Mauritiu,. 32 Swaziland 28 Zambia 31 Zimbabwe ------__------Middle-income oil exporters (n5) ------__------36 Cameroon 35 Nigeria 39 Angola 37 Congo, The 38 Gabon ------__---- - 185 -

LIST OF MAPS

Total population, 1984

Population density, 1984

Language groups, 1984

Income groupings, 1984

Enrollment in primary schools, 1960

Enrollment in primary schools, 1983 - 186 -

LIST OP TABLES IN PART As EDUCATIONINDICATORS"

Table 1. Primary enrollment

Total primary enrollment [ J average annual growth rate t 3 percent female

Table 2. Secondary enrollment

Total secondary enrollment t 1 averageannual growth rate t 1 percent female

Table 3. Distributionof secondaryenrollment by type of education

Percentof studentsenrolled in gezeralsecondary I1 in teachertraining - ] in vocationalitechnicalschools [ I Percentof femalesecondary studentsenrolled in generalsecondary t 3 in teachertraining [ ] in vocational/technicalschools

Table 4. Tertiaryenrollment

Total tertiazy enrollment [ ] averageannual growthrate t ] percent female

Table 5. Distribution of tertiary enrollment by field of study

Percent of tertiary students enrolled in Arts t ] in Sciences f ] Percent of female tertiary students enrolled in Arts [ ] in Sciences

Table 6. Total enrollment

Total enrollment ( ] average annual growth rate [ ] percent female

Table 7. Primary enrollment ratios

Number of primarystudents enrolled as a percentof age group t 1 total t I male ]I female

Table 8. Secondary enrollment ratios

Number of secondary students enrolled as a percent of age group t ] total t 1 male t 1 female - 187 -

Table 9. Tertisryenrollment ratios

Number of tertiarystudents enrolled as a percentof age group 1 ] total 1 1 male t ] female

Table 10. Teachers and schools

Number of teachers I1 in primaryschools t 3 in secondaryschools t ] Numberof primaryschools t ] Numberof primarystudents per school

Table 11. Student-teacherratios

Number of studentsper teacher I ] in primaryschools [ ] in secoiidary schools I I in tertiaryeducation

Table 12. Studentflow and efficiencyindicators

Repeatersas a percent of total enrollmentt I in primary schools t ] in secondaryschools I I Survivalto the final grade of primaryschool t I Cost per primarycompleter C ] Progressionfrom primaryto secondary school

Table 13. Percentageof studentsenrolled in privateschools

Percentageof studentsenrolled in privateschools t I for primary education I for secondaryeducation

Table 14. Total public expenditureon education

Total public expenditureon educationt ] averageannual growth rate t as a percentof total governmentexpenditure C I as a percentof GNP

Table 15. Distributionof total public expenditureon educationby component(recurrent or capital)

Recurrentexpenditure on educationas a percentof total public expenditureon education1 I averageannual growth rate t I Capital expenditureon educationas a percentof total public expenditureon education[ I averageannual growth rate

Table 16. Public recurrent ezpenditure on education by level of education

Percent of public recurrent expenditure allocated to primary education t I to secondaryeducation I3 to tertiary education - 188 -

Table 17. Public recurrent expenditure per primary pupil

Expenditure per pupil C I as a percentof GNP per capita £ I Expenditure per pupil on teaching materials

Table 18. Public recurrentexpenditure per secondarystudent

Expenditureper student ( ] as a percentof GNP per capita t ] Expenditure per studenton teachingmaterials

Table 19. Public recurrentexpenditure per tertiarystudent

Expenditureper student [ ] as a percentof GNP per capita

Table 20. Distributionof primarypublic recurrentexpenditure by purpose

Percentprimary recurrentexpenditure allocated to administration ( 3 teachers'emoluments [ 3 teachingmaterials ( ] scholarships t ] welfare services

Table 21. Distributionof secondarypublic recurrentexpenditure by purpose

Percentsecondary recurrent expenditure allocated to administration [ ] teachers'emoluments [ 3 teachingmaterials t 3 scholarships 1 ] welfare services

Table 22. Distributionof tertiarypublic rec.u.rrentexpenditure by purpose

Percenttertiary recurrent expenditure allocated to administration [ ] teachers'emoluments t ] teachingmaterials t ] scholarships ] welfare services

Table 23. Primaryand secondaryteachers' average salaries

Teachers'average salary [ ] in primaryschools [ ] in secondaryschools [ ] as a multipleof incomeper capita

Table 24. Externalaid to Africaneducation and trainingby donor group, 1981-83average

Amount to educationsector £ ] to projectrelated training [ 3 cost of hostingAfrican studntsabroad [ ] Total educationaid [ ] as a percentage of all externaleducation aid to Africa - 189 -

Table 25. Concessional education sector aid from OECDand OPECmembers by recipient country, 1981-83 average

Non-capital education sector aid t ] as a percentage of recurrent education budget [ 3 percent going to females [ ] Capital education sector aid I I as a percentageof capitaleducation budget ( 3 Total education sector aid t 3 as a percentageof total externalaid [ 3 Educationaid per capita

Table 26. Externaleducation sector aid from OECD and OPEC donors by level and type of education

Amount in dollars t ] Percentagefor primary [ ] for secondary[ 1 for tertiaryeducation t I Other

Table 27. Externaleducation aid from OECD and OPEC donorsby purpose

Amount in dollars [ ] Capital [ ] Technicalassistance t ] Fellowships ( ] Operatingcosts and supplies t 3 Non-specified

Table 28. Externalpublic debt of the educationsector

Externaleducation debt outstandingand disbursed[ ] Educationdebt service profile - 190 -

LIST OF TABLESIN PART Bs "ECOKOMICAND SOCIAL INDICATORS*

Table 1. Basic Indicators

Total population ] AArea C] GNP per capita [ ] InflationC 1 Life expectancy

Table 2. Lsnguages of Sub-Saharan Africa

Number of languages I] Principallanguag.is [ ] Languagesused for instructionin 1 ] lower primary t 3 upper primary t ] post-secondary education

Table 3. Populationgrowth and projections

Populationgrowth [ ] Populationsize [ I Hypotheticalsize of stationary populationI I Assumedyear of reachingnet reproductionrate of 1 1 I Population momentum

Table 4. School-age population growth and projections

School-age population growth C ] School-agepopulation size 1 ] primary 1 I secondary I 3 as a percentageof total population

Table 5. Demography and fertility Crude birth rate C I Crude death rate [ I Total fertitityrate

Table 6. Urbtdstation

Urban populationas a percentageof total population[ ] Growth of urban population[ ] Percentin largestcity I ] in citiesof over 500,000 persons [ ] Number of citiesof over 500,000

Table 7. Labor Force

Populationof working age [ 3 Labor force in agriculture ] iin industry t 1 in services t ] Growth of the labor force

Table 8. Growthof production

GDP I I Agriculture1 ] Industry ] MManufacturing C ] Services - 191 -

Table 9. Centralgovernent expenditure

Defense [ ] Education[ ] Health C ] Housing [ 1 Economicservices 1 ] Other I1 Total expenditureas a percentgeof GNP [ 1 Overall surplus/deficitas a percentageof GNP

Table 10. Disbursements of official development assistance

Net disbursements t I per capita t I as a percentage of GNP

Table 11. External public debt service ratios

External public debt outstandingand disbursed 1 ] as a percentage of GNP [ I Interest payments on external public debt t ] Debt service as a percentage of GNP t I as a percentageof goods and services - 192 -

LIST OF TABLES IN PART C: "SUPPLEMENTARYAND SUMMARYTABLES*

Table 1. SelectedComparative Statistics for Countrieswith Populations Under One-HalfMillion People

Population[ 3 GNP per capita I ] Percentof GNP devotedto educationt ] Percentadult literacyl 3 Percentgross primaryenrollment rate ( I Progressionrate from primaryto secondary[ ] Percent femalesin total enrollment

Table 2. Enrollmentsin sub-SaharanAfrica

Number of students [ ] Index numbers [ ] Average annualgrowth rate t ] Gross enrollmentratio

Table 3. EstimatedAverage Years of EducationAttended by WorkingAge Population

Total [ ] 1970 f ] 1975 ( ] 1980 ( ] 1983

Table 4. Literacy

Rates of literacy[ ] 1962 ( ] 1985

Table 5. Cross-National Comparisons of Achievement in Mathematics, Reading, and Science

Industrial market economies [ ] Upper middle-income economies [ I Lower middle-incomeeconomies ( I Low-incomeeconomies

Table 6. CountriesGrouped by Gross PrimaryEnrollment Ratios

Gross primary enrollmentratio [ ] Enrollmentgrowth rate minus growth rate of school-agepopulation t 3 Repeatersas percentageof primary enrollment( ] Femalesas percentof primary snrollment( ] Pupil- teacherratio [ I Public recurrentexpenditure per pupil [ 3 Public educa- tion expendituresas percentof total expenditures[ I Public recurrent expenditureson primaryeducation as percent of total public recurrent expenditureson education - 193 -

Table 7. Enrollment Characteristics and Education Expenditures by SecondaryEnrollment Groups

Gross secondaryenrollment ratio ( ] Rate of seconAaryto primary enrollments[ ] Progressionrate from primaryto secondary( I Femalesas percent of secondaryenrollments [ ] Public expenditureon educationas a percentof GNP t ] Secondaryas percentof public recurrentexpenditure on education

Table 8. Unit Cost Indicatorsby SecondaryEnrollment Groups

Public expenditureper secondarystudent ( 3 in constant1983 drlla,s [ as percentof GNP per capita ( ] as multipleof expenditurepet primary pupil [ ] Secondarystudent-teacher ratio [ I Secondaryrepetition rate

Table 9. Percentageof NationalsAmong TeachingStaff in Post-Primsry Education 0 z

I X' I r-4 V41 a i I - 195 -

PARr A

EDUCATIONINDICTORS ,-4

70 ri IT] 9t84 ae *iton 1891) Po )tioAh 59*13 9w90*8 W*tj -selo Ru16ul "I 01 *)3e313 1 seas)I 9518owl. *18 At fi "9n*8 of into Wrtj 9t

*his*# o} 04 us se, * *;epl 1ttOploUl" wM}lf1*J : VWAltillpft qIo AS top

a It 81 $- 4-11 K it te t 99 mu I AI

£t 1 tt t ' C- * 91t *UiU LIlt1 t ) 0q8 9 9t LU Ut 9' 5 09 9a U 1* I's * £1? 3? 10 11 1 ~ 4 I ) J It it et vt e - a to Lt zt t t tilt a gotg soltl uea ti a "SAN- _WAW

at 5 t ' Vs .I;t atz lot L; " a11 *&9 9I S91 9t I'S .t1t 14t1 11 a *V5t 89tN Lt 9* n 1 19 S *91ml 4tLit; 11S "*9 u 11 1 1'? II 9i19 8911? 5111 £? *LtAgl 1

89 19 to91 I'St tto ill 5 4tiL I1* StlOtt at54; lti £19t I SAVoiss t t1tD 159;91ft Ot *9 U is 611 1. S $0 ti t to U wit t

a11 45 4'9i 9' 9t41 1£; Ii to91ii

It 91 f? v; t' 0911 0801 tlU5 4 1? *t't l P315§ t Its 89 to I'S O'1 oil? got III Ott "NW314 it 9 ot '91 1'9 98?? SW? t4 OR #low? 89 11 4? 99 0*( OttU La Ott GS It£1i 01 5? 4? 6'S 0'?! 6L0? t4 It it 9IW)3Af 9

it t0t 591'@A"1 S 9m; Mt ItS 009 I AbI IS 11W1 t t;t51 19 *t it -' 1.1 3 t1 til L tt CZ

o) it it *t'I 11q ISQ Otst tt9 RI9 tt 1i t 4 1t l't *'9 £11 no9 t W Mt :U 99 99 5', 0'11 *tS9l 141 91 8 19 3ti,U 9 of e* St I,# t*l tot 11? 19 t18 Me a

Ut a it 61ta9? II5'? I.,0lt "ql W9ti la? 4I topp wantItm It 0 Ut i Oet oLU "119 l#lt *1A1 ;s 8t S t t,t *'0 1 Ot UAt Otl let *Al UI

41ntt t toIt @-U ft Stttlo1OCt IIIItS C£9 ttt981I St4? It 1 UI 9*1

U1 It 5 Io 101 9m 2 t ott4l t t off11It LiU8 KitKt ?'0ss9'? 9*t*et$It 1191tot t CtlLEW991:Ut. t?l911t l 9IUqid9 An1 @SUZStwAI II It*1t-l

59 11 91 5'? 9'Sl £99 "it 19 co? It ;;ltt" n 2M tit tit Os Itt tit""*1 91t 9?ttI *1 C9t-' t 1'SI5 11119 5£?£019 05*gm1 1?Mt tt lSI* aitt usi It ats itIS tIIt"t I'S t14.it * t9oolit tott15? Li1 t95.9tORI Iot1)tlo tt

W *t Ul 1 9' t

-K IS ti 9' S be? 41t tf1 ot ttPMS

st aIt U tS ItOltil-t 91'll OmOtt tow9 settl 960It 4 ADM)@ t

I'14It It4Lt 4t94 11-88I?I *8-114S-S it lOtt1941 004tot 414?tt 8)4?ils pttx I It In et t-0 ll t4t U4 U1 90 9 ti 0I to WI19 UI9 l't110" l-t9*'ts-k eZ asA t 6 f-Stu t out 131*ott 401 tX ^ 11 11 9?tI 4' *tt tl nt t I ttltW4

1991 1 9'? **s

l')9 it ~ Il*

. . .. . --- 161.--.. .- -- ...... --... ---...... - SW*03A$It ops #wA £tS ' @1891 uStm IN100111ue tA r6 S 'ttXl Al #A O"tl W #tA$Fl 'PC Ai t -NM*X@||t9t "I §>an-mtmo OA 01" 1 te a i ll A *a lnv late suti

aNtoa $I j " p No at ul mWp 393)36 an tlm b**n so tAllai VW9o9- tqisj t 95t 9 £* il 99 I3l Illt I NM ot It 9t t-s i-tl 9? it 9t *- 601 P6C' a K 5t 613 "a UK it; tt It .t $-II oil S tt U It 9.- *'tl 0 41111 "to an Sol I niAg ^WiP I 9? It ? 61'1 *-ltl

*tt tl * 91- Vol sli 1tb 1 5 l1 it * U1 *1t -t t-l WU R1 t at T*1 t 3It C r 1Ut s0t ml *f * *35*a 3lUom K 6t t I t-' 5-1 Kr t* St 9t tt 1t 1'* zt-t §tt tMt L$ $l *0 lta t et it 5'. z'.1 * @* *911 a5? t 9 I i t i1" 6? 05 al 91 WIWI 0 * 10 ii 3? 1 IIJ 3 U 3d * tt ut *-1 ,-*, a i t ItS ll-to * tq12 a ot ttJ l tt K IT n ot I cs 1it t I " a tt Ct $-t 1-t1 a *t I t wiloS a e 9* 0 5a13 ti u3? e IUU I it a a* 5* t'9 1.93 * Ki 3 it OM1iet s c5 91 951 601 9 1 I t 1n 6 a t: 1.0 $lt 6911 3tl IS 1 own a a u £1 e- 9 1to a it t *zn u *t 11 9 '-tl 3131 1t a I t 9t tJ9t

It II U C11 fli * 'K * t 3*s 6? ii 0. I'll . 93 19 1? 6 J93 S.31

t a 0t Olt ealt aowUl tl n St a tt t a o; plot tit 9, a tI _*1 it it I?t of L6 t'91 tot 90 ttli 5o u It it a PVt . 9 9* G9 1 161 w1 a Pt 1it o t11- '13 lb b 9 0 some It at t t11 5*9 013 5U a 91 9I OM A iUi 5 Z tt I'l 91 St it 0l 9 Wt1 1I OK n a t V11t 'i 11 et 5 fin UI 1 d3 tt 61 913 6 Om I1 tit 1t Jr. *s1of Ct 1 5? I 'S * 1 et31 15 1 I I" d1t I tt U tt *E0 ols tot tti U ft o1* tt It t99 S'o .*1 el a Wm t1 a t U I'l1 V it *1 a S tsm tl Clt tt tlra -t ts Al a .:rntu 11 1t 9t 3 fI-t * of3 * 3I Ugt4a91 3 t g U tt 5~It-1 - 53? or6 tI t t9115 t? tt53 PR11' att' Sort Pt Gs e *t U £1 3 t'tlO 99t Kt I1 t 9909 tt It it t-OI *-11 ou u o Iip0 1

*5 93 0 11 1 19 #9lo 11 t 9` KS K1 9? I3t Ii n0 I3 t #TM" C It ft LI 1-'1 I., WI 61 £ t 4 a3 141t K a Li t'* tlo 9 d II I w.jw t

t It t't I'll 0 931U 15t t Ml 0I p Il ot~~j rot )lot a i a tt It I'll 11#l "Z 11fit It I #IA-M NM$A oft It *1 t- t-o1 it Li 9t 6S #-it * 1o Li" 901 Ol a * i tni I t Of C Uloi I'll n 6 0u1e 0961 t1ol $11-0963I to t6 90631 0L61 096

)Ul33JO AA5625 't0339

- 861 - 199 -

Tole S.listrisutimo .4 wco0arywollaut e, tw. oEuafatsin Pvgst.t OIwoywwia,a0taof pwmtu dialf sl mean urel00 sat steAMs..s20so u. WtouallVmC YTAe, Vueauaal keeal Tralalso Tableda %eva Traseta, Tomeskc 297 293 Al 2970 241 at to7 291 a Iwo7 2933at 97 29 tmAl 970 19113Al

o a 93 I 3 0 4 90 9 I S 7 3 L2ssceac s. . 90 92 3 3 9 5 a 92 3 2 1 4 .5 92 3 2 3 7 39 3 3 3 I to L-snoeineaes-vsd le 92 4 2 7 4 90 6 3 2 4 I IUa 8 92 5 4 t0 5 U 8 4 3 9 23 2 ottaaFMe Is I 2 Is 12 13 a3 I W. 25 2 394Now 92 94 7 2 3 2 as 122 (.1 4Sakb Te, 94 93 3 2 4 , 94 92 3 1 2 6 SImlia 94 94 2 3 3 22 97 17 I 5 2 0 I cba 81 94 4 0 20 5 92 94 3 I.) S 4

es 93 3 I 4 U5 I4 3 1 3 I'aOeote 90 92 3 3 9 5 U 93 5 2 4 4

Ithispis4 2 I 92 I I I halt.e 75 24 to1 . 1 toI 9 YRIdhm 92 92 1 * 2 3 St 92 It I c.1 I.; tohbiu"a'4 73 92 9 7 23 3 32 97 9 3 9 2.) It tm.it. 92 a7 I 12 (.1 I a a7 12 12 2 1. 12sum2 49 57 a5 St 24 14 33 53 ot 25 2 t0 130aa0a 84 92 S 4 7 I * 1 S * tOftop 9 I0 0 t0 4 33 9 .2 I.) 14 5 ts C.tea)*Ertc.Asp. 01 9 2 2 22 5 i5 07 S 2 20 22 I* kqa.w 90 * 2 * 9 . 3 . 2 . IlImi. 94 I 3 * 0 .) I ISbs11011 72 37 24 37 12 24 81 21 to 49 24 32 it9r.eya9 94 S 3 2 2 93 9 I . 2 21d2iwa'Lae 94 2 * 2 * 97 * 2 2 213.4..d 9 92 2 2 3 7 91 69 2 2 3 9 22mm 0 94 27 2 2s 2 24 96 I9 2 25 2.) 230dm 97 94 2 2 I 1 OR 9 2 2 (.1 3 24 s.ega 90 91 I I 9 I U8 93 2 2.) It 7 2 Ub Aipe *2 37 3 3 51 Is 93 I * .93 94 2 5 5 2 93 90 3 1 4 I mide-tm ol loom" as92 91 3 1 4 3 4 91 4 1 5 2

a Nowits"s ~92 92 9 4 2,) 5 * 5 * 4 * 2 27Ubsra 92 91 2 2 5 4 9 95 2 2 S 4 2 IMMAi 9I9 3 3 2 9 94 4 4 2 2 it Laut* 02 97 o2 (.) I 3 30 97 2 I) I 3 3UCbedliur, 90 94 2 24 7 2I 0 3 I St liial 9 200 - . 5.) 2 LI) 97 200 I.) Id 3 I.; S2blSuaaIM95 3 2 2 2 94 * I 33ah ms 1 5 4 29 I 77 8 6 4 4

*U2t-taem002.~tW5.32 39 3 2 12 2, 94 73 3 2 12 23

p ft;M& 76u * 5 20 20 U9 U 3 2

Afrimc.nas .a 7 9 3 2 20 S U 0 3 9 20

PevMapmysottotl 20 ~ toruit toI.t ~. PpeS0t*6uttibatefin 99wots tisths Ipe$Ohetetesg)mt - 200 -

7030 4. rctlnu woltIu tamp ease Fad"aas low yutb taate pwotq Itbasad) 1wWet of tor

r- to097 to3 390ai 3It"--- 19oom 3940 197 90 a/

* 1.0 1.9 t Is 19 I-i_iu.ee z.it t 12 2 I9 I34 Ia O S.3 12 Is 21

, 314.9 . 12 .16 Las-ik= awwid t3 .d 2 I3 2 . U.3 13 17

I out .) 0.7 5.J 5.0 6.6 . s 12I 2 W4e fmo 3.) 0.2 1.6 3.4 . 2L6 * I 32 sg.w U. 1.4 2.L . 21.3 . . 22 4 kihi Te. WL .1. .d 3., . . . I itd 0.1 1.0 2.9 3.0 LS. 3.3 33 33 11 *~0 LI. I 2.043 5.04 , S5.7 * 14.6 2.2 l0 Is I Lnwl owa t 12 70 130 214 12.4 4.7 12 14 23 I MItta 0.9 4.5 14.4 16.0 14.9 3. I 1I *1 r 0.9 12.4 as 32.9 3.9 4.9 9 9 a.l 2.2LdI.1 2.40 . 4.4 2 30 lotaae -mm L. Li.) . . . . It?mamta L. 2.0 L. 6.2 , 1.4 . 37 17 !Ed 0t.5 3.9 2.3 3.4 . 6 29 ^ t4ftfi.b.3 4.2 5.9 7.S0 7.9 11.2 12 to 27 - p. e0.9 4,. 4.0 . .9 123 Is ! betva1M e 3.3 0.2 1.7 2.0 ILI2 Is3 10 I_aus 0.7 5.7 3.6 32.60 39.0 203 21 2 17Sa (.1 0.S 4.0 6.3. . 2. I 36l iso6 .d 0.6 1.2 1.4 . 5.3 . 9 14 It laq 3.0 7.0 3o 22. 13.7 19.5 6 IsU 9 2ueLwuSh 0.3 1.2 J Lo0t 9.4 3.6 It .1 25 r. mum (. 2.0 11.3 1332 -7.0 I3 22 Ua 1.5 .4 15.5 9 12.4 -5.9 It 14 32 22OW 4.0 14.3 ILI 37.50 10.4 14.3 5 Is 27V 24S.. 1.4 5.0 13.6 U1. 12.0 4.6 I7 I? 21 U2Ik iq L 2.0 1.0 3.3 3.2 44 36

a 19.2 .O 20 24 S1 SdI tuu at luptvg t I I5 47 16 12.7 0.0 22 22 30

a hitaa .I 3.3 0.4 0.6. . . . . t0o * twta 0.5 1.1 4.9 Ls IL2 -10.6 21 22 26 * lse 4.) 1.4 7.2 0.3 * 2. . 15 220 29LmS 0.2 0.4 2.2 2.7 12.7 7.1 22 34 9 30Cm dIvoti. 0.3 4.4 19.6 7." 2. -4.4 13 14 3" 23now 0.3 5. 0.3 19.0 10I 23 2 42 42 32 1 SoliN 1.3 0.2 1.9 3.1 -* 39 43 a btsms (.I 3.3 0.9 3.4 15. . . 44 34llztiti 0.1 2.0 1.0 0.7 ILI. -1U.2 . 30 o 13.4 16 Is Is 1104gms di 3eqtl, t 0 29 9n t1 . 12. 1L0 5 i

a RIr 7.0 2ao0 0.7 120.0 32.2 14.1 7 I5 3 t__s I. 2.7 11.1 13.3 . 5 0 14 I tap. w s o0.4 1.0 7.3 0. 35.6 7.9 7 to aMo t). 0.2 2.0 2.0 . 2L5 Is 36. W lbpl L0 2.3 2.2 2.7L 16.7 1 0.0 . 10 * 14.9 9.1 t0 36 23 M eti 321ta I33 27 437a 13.2 5.3 12 15 D

frmi o_vtae. t 4 t1 161 a 0. 4. 14 Is 1U * 3. 1L27 9 II 27 :04f ii t i i 1H 240 1. 7.1 12 36 2 I 35I.7 .6 6 23 13 t I 30 21 23 e 16.7 9 27 14 oft Fef dAa£eqNott O eoavp us tie tsfth*o ifta. Mab eq sas doS. t asIa_

a. tPIq pit a uelti as 4e 190 thm via a Wm w. utStuta.to thetSica Ws Rpm le d UuO we t 1w06. 201-

tell10)M S. S .4WIt wlmea .ua 34 6 1* uIsut1a5939a liem~~~U~~w a Is 9 trse9 OLIeb ml I unsO l

IthU 71 39 32 U0 29 *.) 4 ,. 1 3 .1 I Bsiu spa 72 .1 4 1 1 It 5 2 44 4 144.2) 3 Up . .. * .

.59 9 2~4 It 4 0 2 It U 0 1 3 3 IsesmelSw * U~ 3 39 9 37 9 I 24 7 4 75 5 4. itWaPs 41 14 59 7 1to I I) 9 4

SiZears . . . . . 9h1 MI'1 22 27 4 It429 U 3 3 3 a f.,a a1 41 24 to I2 14 7aI 1 UUtop 41 a9 359 1 a3 20 I I 4.2 29tac Afrtoeesp lb 24 59 7 4. 3 I . 9 . 16%imp" a~ 2 41 7 194 27 U 4.. . 27ko '72 14 47 Ua75 6 0 142 U I 59sd 59 9 3 1. 9S25 2a 5 2.) it 59 4.2 "9mm 2 12 9 5 8 5 7 59 23 .

am 34~~~5 s oe 41 14 to 3 7I . 7 a" ~~~~733 44 24 59 i 9 9 I 77 29 4 24 spl 59 4 29 2 29 t* 3 3 I 5 41 2 3M.4*hsu 52 14 29 I 4 3 a9 21 23 1.1 19 4 4.2

. 73 23 33 7 21 It a 3 tI S 1 *UIeswium silup'tp, .0 27 14 3 I 134.

27 LwiSta ......

59la14ws y. s .; a i 3124 .59 701 5 .2 It 2 4 2 4.24 4 . flU tis a12 32 a5 41 S9 414 .1 9 .J9 atoms 66 9 a 3 It it1 .441 1.2 LI 1.) 1 4 4.1 24UstUw 41 I9 3 7 to) 2 14 It I a a3 I

a0 a 5 4 7 9 a3 It to 444 . aUdo.4=mdlt a ton It 41 24 29 Us 9 13 2 I.)

DEpt. 19~~~S24 a I 02 14 22 to 4 14) 591mm 74~~~1I 42 1 a3 a) I S I) lot *4. 21bpiuWf tp,' 4w11134 a9 .1* 24 14 1 ** 2 4.2 is 32. 3Mm 72~~~~i9 Sa It 29 8 t I. 29bpIe 24~~3 2 Us PA49 7 5 2 44.2) .5 4 24 9 4 It 41 a I2 '70 25 2 M4mlsUrIsa .461 24 a9 ItI7 2 LI 59 a 1.1 frs100 1 a @7 It 47 It 29 S 9 tot a to94. .4 17 3 10 31 9 9 II 3 IN a 3 tqlqbmsmtnes s a. 19 23 Is59 7 IS1 l. a1 a04.2 a.44 12 a4 I9 1 a44lot 4 1.

Ferumap69.4 soe 0elopm to rc1*q.Od 'utmsm~.i q~ ow 1a" oP f I ItwSaow "m Iwt

U! II 11 I'! 91530tl >U 6I 336 a..

b~~~~~~~~~~~~~~~~~N au to I' a is i got @1t a 99 it a cc 01 m15 39 i a9 . i1

to b *t1t 3ta gtot-t C t- e99 111 N t 09 39 I91 I3S 31t99t LLP 1003I"S 113 913z m&£991 1 ,21490f3." d U 41 LI U1 31 19 t at ITT I nt 91 a 0 911 tIl l nt14e uam t 999 11t 94 Lu 1to t1 OK 09 99 9 Kff 31 1utt t9lit UKa an I u Ibiwgt"l * 9 1 3t *-11 90611 1391 41 St K 9 41-t 9'l at ol ti 1 atltt K #ts t1t U ' 91 a ft N LU " 6* 0 a5 93 6.9 "91 011 0 91 PW199U03 it £ * I'll 0 991 01t1 U41 tit I ,1!t it K tt 9? St 9'S 999 5ttt S 1U GJISI.pamU 11 tq a** Ol ba tl TV 6t " PM.- 99 99 61 9* Il litt li l 331 t Vt l a 61 31 ? 0'0 6.1 LU la ut 39 r3 ll it 91 9t of r1 atl 9t it 91 It t It a9 t o I's3 Ill99 3969 1 0091of 1 S1M ITUSm "llopil 39 2i It clot ri 9 It 9n U I't 1n 09l 1191 39 119 6 to 91 a 3tI C- S W? 11 I to634! .3 K K9 it 3 3i 1C9 49 13 at 9o) _ a It 6t It 9.' To 341 911 L A 901 an a ItKt 09itt 311 .119 Ii4* 1J91;' tot?ti919 t 30ItIl 1603t OWN991 U a! tKt 91 1 09e du Pt 96 v*ial to O lt tlt III Uq 1110 eAt e Om 6 1 to 41 19 $3 UL £4? Ut 0f "ON,t 91 ait Kft 99.got P4l619 193MI fi9mtll oltot NIeat nI ;t tS 41 t't 9-1 tt 9t oil ez M^IA"z P It =x Of tI t totlt cœto ts W *I of to 4t 3t Olt 0l 011 191 1611 339 ' t *, 3! a O1l U t 0e t At 141 96 W; 3n 09 41t K Ot r 1019 t11i3 0t6 Bit 5w91Uif it it it 31 t' CB 6 It 1t '"SoOm s el to it 91 9t t8 t t no t onPtel cc 9 a t t - 99 910 46 6* 9091 5J0 I Ui ot tt t'9 tit 160 Mg33. O6eu0n1m?

41 91 It 9'gt 1*9 5 61t 961 n t9 tSOf *915@Uat. it a Its Olt tot a 19t II 11 01 I' * S f 6 * " t 9K it Kt9 9l9 ft Ft 6 9t 39gm9 3t 11 t? 91'- 9'11 in Olt 65 1? PIP"O3 It 63 0t Ill 1' Si tK Ua 4 It It 9 11n ol t 691 94 at AM t 9t t 4O I'l O't 911 tlt 911 09 ou IIA.tt 3 in It 0.1 93 ut 9t 4u 01 1p1 I ca it t . 01 04 t6t 0e0t e3 PR? I 91A- ss 1ubl tt 0t tt t1. t 6 a

09 Lt i 0t 0-1 * It t 0161t6 09 £1 I 00941 to196 out 0161 094 ISIIA 1t115l t t t)r3M (88.510331 W wItloi ttW1 JI A1"a}41 s *was *s p* rewA t

zo0e55 9 1

_ _ _ _ . zo__- .____ *F61A l $A °q0StM "004 4M At"I am paqmaq 4;uwj4 a u.dww 6owmy; a

6t to tg it ta a Is u ai e a It 1 AN 61 aI et of La LIt It £9 a 6 1 a oa t asu ba: Ita a,aa alE a0* 11 ai asat * i asa _ ,;'a l e 000 Y6 U £9 II u 06 66 69 are93 UIt 66 K II@ U

"~~~~~~~~~~~ c u u n a @"_ IL IL 91 6£ LS a6 £a I 9 I 3' nit sit a 11 ml stb a Ot of k IItk *t19t 9999 91lt1 ccit "9w1 1 eL1 3a63l9l9 9* 191 99t1g91 a99 z6 1 CS a nI a 9t a1 ti 1a Op "I Si of "I 69 K9 to"

StI al 11 It tit oil U 1o SI of a a its *Ill SII 16 IIit 41!* It K11 @ I I U to

* 11 *I11C tt Ill So 91 si 66 a 1139 A M tiU;- 11 tI 31 U£ Ut1 L1i611it alot e S 3IIw Uit tit e ItS0 titnl 1 Lb tit tlI to "$"l ft of1telol lotLI 16I a6t tit totaW9 It 9t1 %fit96 n "I1 6 Sta* a6t" *tIM3elN 6to U99teI oh is a it l a U tC l1 5tIt

11 a 9 99 91 III It a Li a t1 "1061.1 t S £9 u0 tot 1o £6 6 1a u6t *UI , aeU ?'u a 9£ 6 e001 at to it 00I £0 0I It own a CC a it It a Ut 6t a it *6 66 it 1sona Oa it 1£ U 6 U IC ti 1 91 uof 91 " ll LII ?A I9 to 9* 6 31 a W 1 l I It ' .AZAdleaUfls4IPPlu U 1 9 it 9I 16 IL o9 66 0 6 1 a

6o t it 14 a 9 o 4 u 31 i t I6 c95t It 9t it oI 19 n9 13 it t3 It it a lob" a 61t 1t I IC a a a 69 ee a 6z ona 3 9n a 6 a tD U 6?u U 9 9 we tZ Ui it 1 0 it t es S SS a et M 1n to s it *1 of a et a a eT tt et MMvIalst 16 66 e 6 "I 661 It of 31 "I U9 o _ 611 0; 99 1~9 ~9p9 Cl61 Du0 0a9 u9a a69 *s019 09 p 9t ofIt096 -19 9* 3 WI 6ll 11 31 e1, to9 9 9C 5166 1 f6 t6 to it69 11 to to Os Ui 0 6 91 aa36 ea to a £0f1 t1 os of U1 9el9 Jp deh 66 a II £t 93 3 06 o9 *a I 0I It "6619I 9s a et el a9 a90 9 1 a a 19 wel a1 91 a0 1 U £991 S Is t6 in tS £ %" 1 t

SI 69 9£i it 69 I 99 a3 G 19 91 It 1PO616 U CL a to OOI oil 65 06 06 n is f 0116 6£ a 01 9 w 9t U O i It 1 91 116"

6u C 01 et 10 t1 iL 0 39 I t tt 01' =M;4M 19 to It I£ 61, 6 19 11 6 69 a1 t of 19U£it a LIttt 1i 1£ @ 1 Iee93 Is 91 II 6 Uo 69 91 0£ 99 a~ 9£ II Du S g 9 It 0£t 0 t it t1 hI 6 91 a t1 9 AIMt el UI e1 5 ot 9£ eIi at tl 6 " t 6t St 11 3 01 £1 es ti 1£ 13 a 6 It" 1

61 0£ II I it 91 9£ 11 a a£ 6l 6 a IO.iss tU@)1t.-O 41 t e£1 9t 9St a oI It tl61 C a et I elt 39 19 13 St a I 91 9£ !'W 91 13 91 41 11 II It 19 9 9 9 It '

It O606 616 01961 it16616out 0161 0961 tt t 1t C6t 016 096!

fA s6k9 ipeawM S ai p9tpSm JS9

- _O_ §~ ~ 5 5 I I

4*~*~~*4*4*4 44.44*4 4* . . .66. *. .4 . . .* . . . .- - - -- 44~*- -4* 4**f~-*.- - I-:i

------4 -

___- t -----_o 0-5_-~--- --4-- I Ii tYEI------I------"---- l--l------1-!I

W___ *~_.me __ t0_w_ __IW0~-~ j___ __,q}0

--- - o - -;------

; 0 * * -ddw -i o ; "4. ; 4*4 - - " * 4 * Z 4 * 4 *4 4 5 Al {~~4m "."" " o -; * - * - 4 4 4* s4a* 6 6 . 4* ~ ~4

_~4 _;__5e.§ii'~o$t6 ~§~t ;t -;~it ;Y -205-

Mow welled asA psrcstaqp of ap gras

1960 1970 5990 1993Al 1960 1970 1990 191 a/IN 1960190 1980 903Al

a .2 0.6 0.9 5.1 0.3 0,9 1.4 1.6 0.5 0.2 5.3 0.3 Los-sscaKcaOhM" a 0.2 0.5 0.6 0.9 0.3 0.7 5.0 1.5 0.5 0,5 0.2 0.3

a. 0.3 0.4 0.7 0.3 0.6 1.3 . . 0.1 0.2 L-Iisao Nsm-eWi a 0.3 0.3 0.6 * 0.3 0.5 5.0 . 0.5 0.1 0.2

Iul . 0.2 0.3 0.9* 0.3 0.5 1.3* I.) 0.5 0.2' 2 DWt3AaFMa 1.1 0.3 0.6 . 0.1 0.5 0.9 * .5 0.1 0.3 SIt . . 0.3 0.5 0.5 0.8 . . 01 0.2 * Stats.,h S5saaia . 0.4 0.; 0.; O.0.7; I. .; 0.1 0.1 0.1 6 6sd . .. 1.2+ 1.0. 2.4. . .$ 41.f

a 0.2 0.6 1.0 1.1 0.3 5.0 1.5 1.6 0.5 0.2 0.4 0.4 Las'uscaether * 0.2 0.5 0.9 1.1 0.3 0. 2.6 5.8 0.5 0.2 0.3 0.3

I fthiop3a . .2 0.4 0.5 * 0.4 0.6 1.0 * (.1 0.2 0.5 3 alars 0.1 0.7 1.2 1.2 0.1 1.3 2.0 2.1 5.) 0.5 0.2 0.2 9 HwaSts 0.3 0.4 0.4. 0.5 0.7 0.7* 0.5 0.2 0.7* 50euae-Bsm . . . . I1 Tawlat . O.; 0.; O.3 0.5 0.; O.5 0.5 0.1 12 uis . 0.2 0.6 0.6 * 0.3 0.9 0.3 (.5 0.3 0.3 i3 Ugaeia 0.2 0.5 0.5 0.4. 0.4 0.8 0.0 0.9" 0.1 0.2 0.2 9.31 14 la" 0.5 2.2 1.7 * 0.9 3.8 2.9 * 0.5 0.6 0.5 13 CastradOricte Isp. 0.1 0. i.2 0.02 1.7 2.2 . .) 0.5 0.2 IbNaiaoau 0.2 5.0 3.1 .3.5 0.2 1.3 3.0 4.2*, 0.5 0,6 5,7 2.6* 57Ilast 0.5 5.3 2.0* 0.3 2.5 .3.5 , 0.5 0.7*1 toboa"a CZ4 0.3 0.3 * 0.4 0.5 0.5 . 0.5 0.5 59 Kmya 0.; 0.3 0.9 0.9 0.2 5.5 5.5 1.5 5.) 0.2 0.4 0.3 20Sierra Laoe 0.1 0.5 0.6 0.6. 0.2 0.8 1.0 5.0. . 0.5 0.2 0.2. 25bado". 0.6 4.4 3.0* 1.0 7.2 4.8* 0.1 5.7 5.U 22 86.. 0.2 0.3 5.5 5.6 0.4 1.3 2,4 2.3 0.1 0.2 0.6 0.6 23 two. 0.4 1.2 1.8 2.1* 0.7 2.0 2.5 3.1* (.) 0.3 0.J 1.2* 24Imegl 0.5 1.5 2.8 2.2 0.3 2.4 4.5 3.6 0.2 0.5 5.0 0.9 2SHRn 4aiaa 0.3 0.1 0.5 0.3 0.1 0.5 0.2 0,1 0.5

a 0.2 0.9 1.9 2.1 0.3 5.4 2.8 2.9 0.1 0.4 5.1 1.2 fliosSs-secousoil aptete* a 0.1 0.9 5.6 2.5 0.3 1.4 1.6 2.5 0.1 0.3 5.1 0.9

26#WOU3aa. 0.4 0.4* 0.7 0.8* 0.1 0.5* 27 Liberia 0.; 0.9' 2.9 2.1 0.8 5.5 4.3 3.2 0.2 0.4 1.6 1.1 21Zas M3 0.4 1. 1.6* 0,7 2.3 2.5* 0.5 0.7 0.7# 29uLaueO 0.2 0.4 5.3 2.2 0.3 0.6 5.6 1.9 0.1 0.3 2.0 2.5 30Coto dIesars 0.5 0.9 2.9 2.4* 0.2 1.4 4.6 3.7* (.5 0.3 1.5 0.9*1 351154am 0.1 5.2 1.3 2.6 0.2 1.4 1.6 3.5 0.5 5.0 5.1 2.2 52s1itas . .6 3.9 3.0 4 0.8 4.8 3.6 0.4 3.5 2.5 33 kFM. . 1.5 1.6 . .4 1.8 0.9 0.4 34Usitsa 0.1 2.6 1.5 0.6- 0.3 5.1 5.4 0.8 O..2 0.7 0.4

a . .5 1.9 2.0 * 0.8 2.3 2.6 * 0.5 0.4 0.4 lsddie-sag ail l aserer e 0.5 2.2 2.1 * 0.3 3-5 4.0 * 0.1 0.7 1.0

23 lkesa 0.2 0.5 2.2 2.~ 0.43 0.3 3.7 . 0.5 0.7 3lcaw=sa 0.5 1.5 5.6 , .0 2.6 2,9 * 0, 0.4 0.; 37Clapm We's lap. 0.4 1.7 5.6 6.40 0.3 3.3 9.6 50.5* 0.5 0.2 1.6 1.6* UlOMD 0.2 2.2 3.3* 0.4 3.5 5.0* 0.5 1.0 5.7* 39boagl !.. 0.5 0.4* 0.6 0.5 0.6. 0.4 0. 0.1* a 0.2 0.4 1.2 1.4 0.3 0.9 1.6 1.3 0.1 0.2 0.4 0,4 k454wns Wfigas 0.2 0.5 5.5 1.2 0.3 3.8 1.6 2.0 0.1 0.2 0.5 0.4

* 0.2 0.6 1.4 5.6 0.2 1.5 2.1 2.4 0.5 0.2 0.4 0.4 frasuspt.sucasrse 0.2 0.5 1.3 1.4 ~ 0.2 0.9 2.5 2.7 0.5 0.1 0.4 0.4 a 0.2 0.6 5.3 1.5 0.4 5.0 5.6 5.8 0.5 0.2 0.5 0.6 klaglalus ceastra a 0.2 0.6 1.3 5.6 0.3 0.8 5.6 1.9 0.1 0.2 0,7 0.6 * 0.3 0.4 0.4 * 0.4 0.6 0,7 * 0.5 0.1 0.5 othe a . 0.4 0.4 0.5 0.5 0.7 0.8 0,2 0.5 8.1

SateFer iota aaaaalztyto cows"ag,me the tucisacalsitu. a.Figeroait asastera awefar 5992thes aith a crassaet estaatas. la thetauhaigal atos. Flgersfr Igatei NWEhidAu aro1904. - 206 -

Tall t0. Iel.cfs ad Klschm 1.acp NN4Wof 4410Wof (theeSBi frim" friar

Priar gas"" (thPurds) i Kw

9* 2 96i 90 93 ai 29*0 53 d 19*0 196 * 12 273 Lowism cmmesa t 1IJ U ? 29 14t 41 0e 141 246

* 247 201 Lwiatcamot-uid t 5 31 1 12 2 7 In3 220 2 Nai 2.4 7,9 6.3 4.5 0.4 2.26 162 225 2 lutisa Fo 1.5 4.1 0.2 1.5 0.4 1.2 142 231 3 lItp 0.4 4.9 0.1 1.1 0.3 1.7 39 147 4 OUi, Twh 0.2 2.4 0.1 0.e 0.1 0.2 70 303 1 bala 0.0 9.S 0.1 3.0 0.2 1.3 114 170 4 Xad 0.6 4.5 0.1 . 0.3 1.2 241 240

* 123 20 Lslugi thv t 14 22 20 122 3 n3 143 249

7 fthi*a 4.6 44.7 2. 13.2 1.1 7. 24 329 lafin 36.6 112.0 3.2 . . 9 Ue)ai 7.0 14.5 0.2 1.2 3.2 2.44 6W 235 toOkau0s.m-b 0.5 3.3 0,1 0.5' .66 * 97 It Timat 10.1 W.S 1.1 4.2 3.3 10.04 ISO 3 12l . * 6.1 .5 1.3 .9 it 3 13 _Vo t2.5 44.44 2.0 7.0. 4.0 4.99 330' 14 Top 1.4 10.2 0.2 4.4 0. 2.3 In 19 i5 Codat ffrca hp. 2.2 4.2 0.2 5.00 0.4 0.9 14 S24 16N bau.4 1.4.5 * 2.4 In3 I7 bkm 2.2 21.34 0.2 . 0.4 2.L t4 15" to wa . 14.1 I?2 3.2 1.4 a 474 1t gn" 2L. 227.7 1.4 20.4 5.2 22. to 34O 20 iUra me 2.4 10.2 0.5 Li 0.4 1.24 44 249 2 him l.t 7.9 0.4 5.2I 0. 2.44 145 109 221 24.2 52.3 22.0 3S 5.1 69 173 2 0d 0 47.0 2. 201.2 2.4 .36 143 21 24ns 3.0 22. . 4.9 . a2a 24S 25 bua00p 4.4 20.0 0.7 3.5 3.2 L. 130 I2t

I I" $14 1ldiI1iacue oil isptus t x2 43 *3 35 7 I t 131 2*0

24 lnta 0.4 2.LO 6.0 0." 0.2 0.4° 57 I79 27 Li*wta 1.9 7.0 0.1 1.1 0.4 IJ. 139 UZadaa 't.0 28.9 0.4 4.0 . 3.1 . U 29 LOM 2.5 5.30 0.2 1.30 1.0 1.10 134 253 20 be. riutre 5.0 32.1 0.4 0Q 1.7 5.L 141 240 St tlZiu 12.7 54.0 10.4 2.7 4.2 579 507 32 SmaI 0.9 3.9 0.1 1.10 3.3 0.5 15s 240 33 30m 1.2 4.4 0.5 1.4 0.2 0.5 182 397 343urUt 3.i 4.0 1.1 3.4 0.3 0.3 371 4

6 It7 234 Sddl.s.a d Nb. 11tw 1114 414 l4 19 28 151 23 21YAP"d, 97.2 399. 6.3 91.9 I5.5 to 3S o0.0 2e 3 . 0.7 9.0 5.5S2N 3 to"p l6We k. 2.2 7.30 0.3 4.44 0.4 12.44 2 S0e sSI" 5.3 3.66 0.2 2.70 0.5 O. 114 564 39Sia 2.9 32.00 0. 2.0 43 2 l6f

141 2n "-loorm,AMC$ t Su lm a 205 61 122 141 271 * 132 232 Fcaq hm Atrlm * 79 29 1 a 24 249 3 * 147 315 Iw =nitxttu t 20 52 351 211 47 5 I2 310 tU2r t U IU * 2S ' a e 1X20 218~~* O"D 2 25 122 4 20 inD2 167 aI"fV d dliatty ad cosqp, m tle toitcal ota.

a. lpu. de a astarid dat Tw otra the am qmfid. so the tetica MM - 207 -

Tabl II1 StuIoatstO&clirat&oh IbSt teeb ratios

Pr~au~ Sec4v w tertiary

two 17 1"90 1 190tt 197 l 9 13 a 1t70 193u/

* 42 42 43 0 19 23 23 23 to 11 L e l-ascecue.A a 43 41 42 39 2t 21 22 23 9 8

a 44 4 0 39 Is to 21 21 0 a Lw-ateecm s-vi4 a 40 44 42. St 2t 20 21 21 7 9

I WI 40 41 42 37 1 17 V 1 5 120 2 wbossoFM 44 47 54 42 23 20 23 24 1 10 3 liw 9 39 43 36 . . . 23 t 4 $W&1 , The 27 26 24 25 20 19 1 Is SUeeha 33 57 33 Zs 24 21 21 21 17 4 ,d 48 be 64 2: . . . 12 * 42 42 43 41 19 24 a1 25 10 It L-iacmethw a 44 43 44 41 20 21 22 21 9 7 I Etbipau 4 44 44 54 2a 3 44 44 9 It Slate, 4*3 . 42 23 . . . 9 I Ilei 43 41 45 9 16 ti 21 19 5 Be t0 fisa-fism 45 34 23 3 Is 10 10 19" 11Twao 41 54 42 42 13 20 t2 5' 4 12 Ihnl 37 31 35 4 It 1it toS 5 13 ,uia 34 34 34 340 23 21 * 21 9 * 14t3 . so 40 55 45 25 9 23 s 11# ICatorl Mrwcahp. p 44 47 40 49 22 . 42 25 . 4 16 _e"s 45 b1 55 . 29 . . S 46 1TeInt 41 53 4 S 23 31 . * 9 to h_si 60 50 59 54 13 1* 12 14 * 44 19""s 34 n 38 7 21 25 23 25 .7. 2 iiervLowu 32 32 2 34 29 1it 21 6' 7t 21GM$" 44 40 3 3 23 . 1" otO 22 ua6 30 31 29 23 17 22 22 21 4 23ow 47 37 34 34 17 21 20 25 12 40 24 toeq 45 41 46 41 . 22 23 17 1*4 231bziqw 0 a $I 54 17 . 3 35 9 34

O 42 4 41 39 21 21 22 3 1 4 tiIedaceuo oil tewertas a 40 0 41 34 21 21 21 24 4 9

26Beardtls 24 35 41 450 22 27 2i 31l 27Liraa 36 41 41 33 17 I7 3S 20b t 4n7 4 4 46 . . 21 24 t 10 29Lesotho 46 53 4 520 22 22 * 19 7 30Cate flvars 45 4 39 3S 21 . 30 20 3 lido 0 40 44 40 tI 20 45 325s lasse 40 3 34 33 11 21 It 111 I5 33Stm 36 33 32 31 15 17 1i 140 5 10 t NwutirU 32 24 20 23 25 so 26 22 4 3

so30 4 n7 2a N 35 14 9 14 tile.gmt oiloqeter a 40 4 4 44 21 S3 a 3 11 12 SIIgwt 34 34 . 21 . 37 S 150 36tera. 46 51 52 50 29 3 is 30 12 37Cmlp,Peels B.2. 5 54 5p 33 42 37 39 Is 250 x kba 46 4 45 440 19 22 19 19 s5 of 32 4 SI0 iS 1* 9 So

* 41 40 U3 S 20 24 26 36 9 12 be_lw "tAt F 43 41 42 t9 21 is 22 Is 9 11

e 41 51 49 44 22 29 31 27 12 IS FrW eqheMCotfrtl a 45 4A 47 4 22 23 27 24 11 to a 36 7 31 37 19 22 22 30 a 13 bgo e costna a 3 36 34 35 1I 21 21 21 7 7 a 51 47 54 46 21 34 31 31 Otter a 45 44 40 37 Is 21 36 21 9 7

OttowFw dtia Ctwality u0Z c , the ttdm repmtles.

a. Fioo sIths aetIfsk we fr VaS ethr thektOr *efi.s 1. US tOsotCitudMe. - 208

7dbo 12. studat floe andefiticomc Indicators Pr:wy kun~~~~~~~~~10od"r

Praowti.nof cohot Cutl pe coaglote Prruoo,os fromlast Tdsf rmboaiathe final $rads a sa altapl of. of priwayto first grads of Ragealo as a (cohot too00) cost Wmcouietwl bboatws as a wooo.drypwda odmtiti Ptortap of - -- if so drop'oot REcoUtIOof total mIlmaS Total Fadeo or ropitaios tota mowllmt Total Foans 19701993 a/ 19701913 hi 397019936 hI 970198 hi 1970391 Al 19701983 hI 19701993 hI

o 17 13 2.0 2.0 7 10 44 43 43 43 Losisoi moosols . 21 IS 4135499 536 530 2.9 1.4 to 14 29 36 3139 a 23 21 3.3 2.3 20 22 29 33 3o093 Lov"imcauomoos-id a 19 14 534 703 542 70 1.8 1.9 it 11 29 39 29 40

hilti 24 33a' 462 400 * .8 i.3 28 39 33 42 2 thlaaPFin 14 17 53, 703 542 102 3.5 1.4 31 33 13 22 13 33ipr 19 13 402 79 1.9 1.4 3 lB 24 33 20 23 4 Wis. The 13 Is 934 913 85 999 1.2 3.3 * 42 44 37 48 Shboallis . . . 2.1 to.0 so so49 4Ca 27 . 29 293 200 201 2.4 4.0 11 1219t I 34 32 2.0 2.0 6 10 45 43 45 41 I.iiao-inoho 22 Is 460 439 569 509 1.8 2.4 10 14 29 34 31239

l lthlnapl 12' .49 . 09 * .9 5ST 9 4991f I late 23 I9s. 2.2 3.4 It 0 9373 3742 9 oiml Is3 322 321207 2.0 2.3 1Is 7 l@boksumbarlo so3 * 1417 144 2.4 4.3 * 14 2946e .6 Ilt ula 2 19 342 737 * 44 1.6 1.4 . 3 3 34 7 12 itudi 22 14 34 943334 83 1.8 1.3 5 315 34 3 35 7 33UVW" 100 . 737 * to4 1.4 3.4 . . . i3 .3it 14top 34 36 67423734B193 4.2 3.3 19 34 2731t 2327 13 Cotrol uric"a0up. 29 U33. 93 44334 22 2.! 2.3 t0 235 23 36 2193 1I Udopuwt 29 . . 2.3 2.7 13s 41 * 1 17Iwoo 19 220 427 337 349 531 1.3 1.7 12 21s 239 I139 is oada 30 12 * 412 ,2374 3.9 2.0 3 5 7 4 6 2 l9 me" 3 130 744 607 479 1.3 3.5 3.) * 29 35 293 203iwva Lawo. .493 . 3 373 3672 21bo. * .40 a .29 1. . 8 32 49 7944 22am 3 2' 17347 444 2 64 3.2 143 3 25 303 too0 2356dm 2 (.0 . . 1.4 1.2 3 so5033 40 34 24 hp3 20 3s 460 835 .19 1.4 1.3 33 34 29 29 27 31 213Nmmoiluo 23 29 * 232 * 12 * 3.9 24 40 .39 12 8 1.3 1.4 14 14 29 44 29 50 33dd63ls-cuo01souibagtars 13 i3 773 032 39 639 3.1 3.4 it 12 30 45 33 47 24Nowstmal 15 17. .79 304 1.5 1.4 I1 31* 3339 .20 Z7Lbihra * .3 23 47 43 4033s 21 luka 1is2 1771829440720 1.0 1.2 . . 23 21 23 29LOA.h '20 230 35 375 490435 2.7 1.7 44 43 39 41 30 ColdloV oi 21 25 941339438 3M4 1.4 2.4 33 34 27 30 22 29 31Zlmb"i I I 1.4 . 0 74 31 n2 SIuuauslaod I 13 427 630 4A9 639 1.; .5 10o 474be 4347 a hotoma (.1 4 90 90" . . 1.3 .. . so303 a930 SIwtnitlu . 25 36 54 34 33

.2 a 32 . . 12 23 3093 30 39 lhddloimnol ugWt a 33 31 39963437763 . I1 l0 2739 23 43

ENligma . . . . . A 33 36 aCsroon2 309 S9i 70 577 645 . I 37 17 24 1323 37Caogs,We's sop. 33 3e30 72 674496 80 1.4 1.7 17 37* 37 73 36 13 3UOde 33 325 304 S57 439 53 1.7 3.9 10 205 39 2 20 30 fqda ~. 369 .. . so5 45 o 17 14 19 16 I 1 9 4 1 4 WWiOasftca a 20 36 4279373 77 599 3.9 3.3 It 14 30930 33 41 . 21 22 21 If 1 4 3 4 3 Frusspcoocoantrift 26 23 5936145453S 1.8 3.1 33 16 24 33 22 29 * 4 II 3.6 1.3 3 3 43 34 43 37 Anopophmcntas a 3 3 744 744432644 1.3 3.4 7 4 36 43 36 53 v 20 22 2.4 2.9 * 24 43 43 52 59 Mitb .3 2 22 144 2.4 3.0 * 20 37530 6949

MMtIVr data crowd.3stymd4 cwago o the toaicale oot.. - 209 -

tale I3. rcotae a#etudents wolled to privateuheools

Prtwy ae v

19i0 19n 1W9 293 ci 2970 1975 90 2903*V

* 9.5 7.8 3. 3.4 23.4 U8.0 14.5 8.3 Uo-jtome eeusu 7.4 6.4 3.3 1.9 24.2 129.4 10.2 9.3

t 8.5 5,7 4.1 4.5 17.2 7.5 20.2 24.3 Lwseone ies a 7.4 4.4 4.2 4.1 17.8 14.1 24.7 44.4

I gas 4.3 4.0 4.2 4.1 13.4 10.8 6.3 8.4 2 arkisieFase 3.5 7.2 8.4 8.2 34.1 44*.5 51.0 5.2 3 taw 4.1 4.4 3.1 2.13 22.0 14.1 15.5 4 6adea, teo 31.4 14.4 15.1 14.4 . 45.4 33. 44.4 5 ISnal 25.1 0.0 0.0 0.0 Chad 8.4 9.7 . . 6.8 4.1

a 9.a 7.9 3.9 3.3 23.8 18.0 24.3 7.8 Letoe s other a 7.8 *.4 2.0 0.4 34.3 20.3 6.7 0.4

7 Ethiopi 28.1 20.4 15.4 12.5 11.2 7.1 7.2 7.3 8 Zatre 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9 a1tt 11.2 10.2 4.8 13.4 1.4 20 huee-,s, , ...... 11 Tfanaa 3.3 M.? 0.2 0.3 24.2 30.1 41.0 44.2 12 hrudv 93.9 92.4 4.3 11.0 34.3 21.4 4.2 7.4 l30qeed ...... 14 Toe S.9 21.4 23.3 22.7 38. 20.8 12.0 22.4 IS CoeetraAtrcas R. 0.0 0.0 0.0 0.0 2.3 1. 0.9 9.3 12 Nadaqanz 24.4 23.4 . *70.3 9.21 32.3 17 klSue 34.2 4.4 3.4 .0 59 17.9 4.2 0.0 18 . . . 0.4 . 21.2 37.0 31.3 n9 KeNa 3.4 0.0 . . 42.4 49. 59.9 3L2 20 9i ra Low 78.5 77.5 . . . 21 Isma 0.0 0.0 0.0 0.0 0.0 0.0. o.0 O.0 22 64.u 2.1 I 2.8 . . 236d 4.3 2.0 2.5 1.9 24 S"al 12.4 12.3 11.1 9.5 . 30.0 S0.2 31.2 23 haubqn 0.0 0.0 0.0 0.0 37.5 * 0.0 0.0

* 4.2 44.4 S8.4 5.?2 32.9 S34. 34.7 55.4 de-.tesae oil swertrn a 29.3 27.8 25.9 2.0 51.2 34.3 35.4 35.7

26 Nwatarns 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 27Lbwra 34.3 34.4 35.5 29.4 51.2 38.2 44.2 35.7 28a eehe 21.4 24.4 0.4 . 29Lmnetb 100.0 20.0 46.4 ".5 . 8 o.9.9 30 Cateilnvur$ 21.8 128. 15.0 12.2 25.0 28.0 50.1 31.5 St liaiSeh6.8 $6.4 053.5 82 . . . 4.5 32Seilma 74.2 80.4 79.8 79.8 59. 45.40 40.4 3M. 33 btoa 5.2 5.4 4.7 5.- 5".0 30.3 27.4 32.1 34Nwittin 29.3 27.8 20.9 22.3

* 35.9 22. 17.4 1.21 30. 24.2 U9.1 22.4 Nhddle-ceK oil srtWrW a 37.7 21.5 18.2 17.4 29.4 12.5 22.0 22.8

35 litpria 3. S C4eus 3. 42.9 34.3 34.7 a5.8 54.8 47.0 46.3 37Cmq%, hole' . . 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 S M 49.4 45.2 39.0 3.9 39.1 32.9 43.9 45.5 39Eqde 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0

* 21.9 15.2 11.5 14.4 25.3 21.0 17.1 14.9 Ia-St.. Aric a 17.1 10.0 4.3 5L8 30.o 21.4 21.4 21.8

* 162. 13.7 9.4 t.2 24.0 127. 11.0 4.9 Frmaeeemconemn a 8.4 7.2 4.2 3.2 23.5 29.4 12.0 0.9 a 27.4 18.4 15.9 29.1 40.3 44.4 53.4 45. @3o1o94cstrlU a 29.3 24.4 11.0 22.3 12.2 41.8 40. 36. a 12.0 10.0 4.4 4.2 23.2 5.5 4.3 5.1 KW a 12.4 0.0 0.0 0.0 11.2 3.4 0.0 0.0 RetaiF dauti WOiIity vantem e M the tecOeacalnet4.

*. FiW", re Eurthe latest avaddln ea. Sm tk tKhnlCal estt. -210 - faw$34. Totalpubbe sopoiatvo on udgahtp Pubicampuitwoo om o hoawaa poresap oht Peti wuiitvo a $Msatio$ to cusstast3983a Us oagep"uml Tota""noew (wlinIo) "am rats m,utwus lI 1970 3975 to34/ 19063/ 19704 19600 1970 1911sIalm /IC Al 3970 19751960 / 39Ue&I a 2.1 -3.6 £7.0 17.2 16.1 14.6 4.1 4.1 4.0 3.2 LaWaKoi cfOMu 3 6.7 -7.0 I. 17. 7. II.6 35.3 3.8 3.5 3.7 3.3

* 9.4 -2.0 £1.1 15.3 s6.7 37.4 1.5 2.81 .3 2.9 Lowmsasuim-wtd t , . . .a 9.0 -4.5 10.6 35.4 19.3 EL8 2.0 3.2 2.2 3.6

Iu . 4. 6. -6.2 . 30.0 37.2 . . 4.6 3.7 2 &heaPo .FM 251.3 33.6 LI.9 . 19.8 23.M 2.4 3.2 Sniw 20.5 34.6, 57.2 44.9 30.8 -6.4 17.7 36.7 22.9 23.7 2.0 3.6 4.3 3.6 4 Saasa,t6o 2.1 5.7 6.4 11.86 9,0 I$1.6 * 12.53 2.3 3.2 3.2 5.90 S leiash 13.7 27,6 22.6 20.7 6.6 -2.9 7.6 12.50 3.7 6.3 3.0 2.1 1.7 t3.4

* 1.9 -3.8 17.2 17.3 15.1 14.3 4.2 4.2 4.1 3.2 UlawiMmostiNt 3 a 3.9 -2.6 37.7 17.0 1IL5 15.3 3.9 3.5 3.8 3.1 7Itbiqpia 36. 125.3 126.9 347.7 3.7 5.2 39.4 33.4 9.3 30.9 2.3 3.6 2.9 3.3 Slatro 29A. .261.5 * .. 9 * ?.4 . 22.0 * 7.0 . 6. 9 luaud 30.3 23,8 36.4 32.6 2.5 -5.1 13.2 9.6 LS6. 4.6 2.4 3.6 3.0 30Saiesk-Disou. 3.3 . . . 3.8 3.0 II 1mzaia 1336. 230.4 261.002.2 6.2 -2.4 it.;0 1.8 14.3'0 15.3 4. 3.4, 5.90s 5.6 t2 nMw. . 37.85 . . . . 3.40 13%of6 326. 214.4 4. 106.3 -57.7 33.4 3. 17. ?*0 33. * 3. 2.5 0.7y .3 34tops 32.1 23.5 46.3 41.9 14.4 -3.2 19.0 15.2 19.4 20.1 2.2 3.5 5.6 5.9 IS Cmtral"riceA"q. . 29.3 , . . 20.5 4.9 16 liqama . 99 316.2 43.5 -37.6, * 55M 3.2 3.7 2.3 to bams 19.2 25. 36.4 43. L.4 I.9 26.6 25i 21.6 24.0 2.3 3.5 2. 3.1 19Km"y 1IN. 244.9 346.3 27.6 11.2 -7.8 11.4 19.4 38 1.3 35. .0 4.3 4.9 4.6 20Serra Lam 26. 32. 31.9 3S. 3.9 -2.0 17.5 . 7.6 3.2 3.4 3.9 LS5 It Wson . 0.5 . 2.7 .. 4.0 22Oms 194.5 2B.0 340.5, 90.9 -2.7 16.8 39.6 21.5, . 1.2 4.3 5.9 3.3 2.0 23S" 359.1 221.7.7 7 * .8? 12.6 34.60 3.9 4.6 4.6 24 6sp1 67.6 . . . . 21.3 . . . 3.8 25 luzv2p......

* LS8 3.5 14.1 55.1 37.6 19.4 4.2 5.3 L.71.3 laiddS.tocasol s~tarov 1 547. a IL36. 0.4 14.6 34.4 16.0 37.6 3.4 3.7 6.3 5.S

2bUau'atasaa 39.5 2. 3.3 27Liberia 39.0 20.2 *1.5 51.i £3.i -7.7 9.5 11.6 24.3 13.2 2.5 2.4 6.3 5L3 23 mlas 326. 203.0 136. 176.26 0.6 3.3. £0.9 13., 9.7 35.2. 4.5 6.5 4.5 5.50 29UleutI 7.3 . 49. . 22.0 . 17.9 3.0 * .9 30Cotb d'Ivmr, 20.3 32.0 5.40 544.7 51.3 0.3 19.3 19.0 29.8. 23.2 5.4 6.3 0.40 9.3 3I lAisae. 53.2 170.2 329.1 426.3 9.9 9.2 . . 33.7 37.6 3.4 3,6 6.6 7.6 32l.nizili £4.7 19.9 27. 32.20 6.5 17.3 56.9 36.6$ 4.3 3.7 4.5 4.70 33Mommoa L.S 29.5 14.6 9.2 30.9 0.o 32.3 36.8 36.0 19.5 5.2 6.5 ?.3 7. 34Uawntlu 36.3 29.0 531.1 46.0 1.0. -2.4 113. 9.4 33.6 30.3 7.1 3.6 5.3 4.3

* 5.9 .14.2 39.4 16.7 15.9 9.8 3.6 5.3 6.0 4.3 UIEdlstms oil awter 2 SU5602 ao .6 -7.0 39.4 16.2 20.3 53.7 3.5 4.1 4.6 4.5

35Nboigei . 3071.5440.6. 2.64.5 * -36. . 36.5 35.7 9.30 5.5 6.40 4.35 36camam 309.7 147.1 102. 236.4 5. 6.7 £9.6 21.3 20.3 17.2 3.5 3.9 3.3 3.6 27Cops Poole'sIV. 44.6 694 92.6 302.0 7.5 , 23. 16.2 23.6 £9.20 5.9 6.3 6.9 4.0* 366"M 46. 69.3 83.9 LI5. * 362 . . 3.1 2.1 2.8 39 tqda ~~~~~430.403722 -7.0 . t0.1i . L5. 4.7 o 4.4 -9.2 16.7 36.6 16.2 35.9 4.1 4.9 5.5 4.3 M.Oierrs lfnea 2 a 6.8 -2.9 17.6 17.4 IUS 35.3 2.5 2.7 4.5 3.9 so?. 6.3 19.5 16.7 24.2 IL7 4.6 4.3 5. 35.0 frtxcme civnrw I 7.3 -13. 19.5 56.9 22.0 23.7 3.4 3.0 4.3 3.4 * 3,7 .31.j 34.4 14.7 34.3 15.3 4.0 5.5 5.7 4.3 A1aOM comte 3 a 7.5 -2.9 34. 36.7 14.3 35.3 3.9 3.7 4.4 4.6 * 4.1 -2.3- 13.5 £3.0 9.0 10.5 2.3 3.2 2.4 3.6 otba t ~. ~ ~ .So.2 -2.9 12.53.0 9.0310.5 1.9 2.6 2.3 3.0 a"st Pwdata comapa lssty ad cowap. us tkotosbocal sots.. A. PIpes lit6 a aitarik no IV vMs uter VW thMOrnup 1111.Los the tKstuCAl iNtOW -211- tableIs. brktrhtis 04total uditc tssu6tvoi ONo4uctao. by csmpoem (rocurrtst er cae,tal)

Rswrmt napsstoro Clptai s0sdatvor

lAs 20o totalpilic @*pSEtgtI Al"oa amuad isa I of total petistsapssteO bvarq umsaj enducatta fvout rate U5 OSIIIIti Vth rate

1970V 19319711923ai 197-U 192343 1979 tin 1923v1 tIw a 190-881923.

3.6 P.? 30.2 "0. 2.6 -1.2 14.4 12.3 U.1. 9.9 0.1 -31.3 Leu-K,NcaKn..) S 39.0 39.6 91.1 95. 5.0 -0.0 33,0 10.4 0.9 4.3 53 -30.3

a sy84.9 77.6 "4.I 4.1 4.5 4.1 13.1 2. 5.9I -37.I Lo-anagsmi-wad o 94.4 33.7 91.1 96.7 5.0 3.7 5.6 16.3 8.9 3.3 . -23.1

I11 . . 98.1 99.9 5.0 -6.2 . 1.2 1. -LI3. 2 basna Fag.. 93.0 99.2 4.. 3.2 . . 7.0 0.3 . 4L.0 3Niger 93.3 37.0 47.0 96.7 3.4 M9. 6.2 13.0 53.0 3.3 37.3 .42.9 4 $ovai, no 94.4 13.1 18.1 63.6' 8.2 * 5.6 21.9 11.9 31.4. 37.5 5 SMalI 10. 3.7 93.I I.9 5.3 -4.1 i.1 16.3 8.9 14.1 . 13.5

a 5.3 37.9 89. 89.6 2.4 -1.3 14.7 3.1 12 0.9lL 10.4 -1.3 -7.3 Lw-incusothw u 8.4 91.0 40.7 95.3 4.8 4.8 11.4 9.0 9.3 4.7 2.2 -23.3 7 [this"a 49.5 93.5 89.3 77.3s 30.0 32,3 50.5 16.5 30.7 22.7 31.3 76.4 8 lat, 88.1 . 96.0 * .3 * 1.9 * 4.0 * -31.1 9 slm 73.7 91.9 75.6 90.1, 2.7 0.7 26.3 3.1 24.4 9.9 1.7 -29.7 to .asn-1waw. 9. 21.3 -5.2 . LO3. I1 Trnaua 82.0 80.4 82.3, 67.2 L.1 -3.0 37.2 39.6 17.74 12.3 LS5 -t0.0 12b.Om . . ".I*1 6.0 I , . 1.94 13VW"i 82.4 93. 88.3 76.0 -17.1I 25.0 Ilih 7.0 11.7 '24.0 -21.0 67.1 34top 88.3 96.3 96.4 974 35.3 -2.9 11.2 3.7 3.6 2.6 2.2 -13.1 Is Cntra)Africa am. . 86.9 . 3. 40.4- 13.1 . 36Isaisaraw 95.2 93.5 go.; 0'.9 -36.5 . 4.0 6.5 2.0 -aa.44. 178wonto. 5.9 1o tuam9" 9. 84.7 99. 5.0 35.9 3.2 0.i 35. 0.7; 75 6. 19Kley, 939 9.4 92.1 96.0 10.9 -6.5 6.1 4.6 7.*9 4.0 14.1 -26. 20Sliwra Leoe 89.0 0.2 95. 9.3 4.6 -2.0 13.0 9,3 4.7 4.7 -4.6 -1.9 lumgoen 97.4 4.6 -0.6 . 2.6 22Ibe 0. 17. 9 73.;2 39.2 -4.5 -31.1 12.2 22. 26.8 10.3 .3 -3. 231.wo 9.4 89.6 92.2 * 6.6 * 6.6 10.4 7.8 . .5 24 8qa3 97.8I 3.5 . 2.2 . 23sguiqu. I

* 733.1 36.0 93.3 8.4 6.6 32.7 16.9 14.0 6.7 9.2 -17.5 13461e-inconoil etuotw o 8.5 33.4 65.9 95.6 30.0 4.3 I1.5 33L* 14.1 4.4 3.1.3 -27.4

26 lRawita 95.LO. 6.4 16.7 5.0 . . V7 iwtsa 8.9B 957 -4.6 3.Ci 4.3 -3.3 23Zamas 19.2 76.9 95. 99,94 2.6 16,4 20.8 23. '4.9 0.34 -12.7 -88.4 293.uoLc 91.3 . 93.7 . 8.7 . . 30Coto.f LIwir 35.6 84.7 "7.3l .39. 10.3 3. 14.4 15.3 22.24 10.5 36.3 -In.9 3t lajtus 9%. 93.3 97.4 95. 10.0 13.5 3.9 0.2 2.6 4.4 5.4 30.3 32Suallow 85.7 73.1 23.04 00.04 5.0 * 14.3 21.9 20.0. 20.0' 10.2 a3 otusma 84.6 54.5 75.3 33.9 19.5 4,3 15.4 45.5 24.2 1461 26.4 -1.9 34au,'tmoa 92.0 07.4 89.9 93.3 10.5 4.3 0.0 12.4 10.1 3.2 M3, -52.7

* 93.9 54.0 32.9 88.9 35.5 6.1 46.0 17.3i 11.1 37.3 MUdilo-ic oil onprtor a 95.6 83.1 82.3 91.2 5.6 4.4 16.9 17.9 8.3 14.81

U3Ikqora . 1.0 a2." 88.3 35.5 -1350*.0 0 7.10 11.7' . -2. 26ctwmr 92.2 33.5 31.3 19.3 3.9 7,7 1.0 16.5 33.7 20.9 14.8 12.31 37Cuap, Pewope, Up. 96.6 8.6 93. 94.0. 7.2 * 3.4 17.4 6.2 6.01 14.1 so0*"O 95.6 35.0 12.3 I 2.9 * 4.4 15.O 27.7 . 2. 39 bila 1 7.9. I 2. t#

SL36. 67.1 84.6 90. 9.2 -7.0 13.3 32.9 15.4 9.9 4.1 -20.7 344-Obo unto o 91.7 35.0 8.3 95.3 5.3 -0.3 8.3 35.0 31.2 4.7 9.3 -23.3 o 9.2 86.4 83.9 99.3 4.4 3.3 9.8 33.6 16.3 9.7 11.3 -37.3 frin"0s ctuatno a 94.4 16.9 93.0 97.4 4.6 2.0 5.6 13.1 7.0 2.6 15.5 -23.7 * 36.9 63.2 34.7 39.9 13.3 -30.0 33.1 36. 35.3 30.31 .3. -23.6 Uglopmecuatries a 2.4 93.9 88.1 93.9 7.4 40.6 33.6 36.1 33.9 6.1 .3. -26.7 * 55.4 W3. 19. 93.9 9.4 6.0 44.6 36.4 10.4 3.3 -33.3 42.6 04W o 7~~~~~~~~~4.03I.6 "J. 91.9 10.0 -4.0 25.3 36.4 9.3 L.I -11.3 450 itsle for tataceoprslity so cowl", an thetochuical nots. 4. P1gm. math44 utrok aStfwr yWgrothe than thea SPotafio. Sen thl tKnChInacllnt. *gwo Puw93; o.n as w0c am G U38s a a an " l mso l

upv mpauo an - IseAa Po4a ao ow api o

* * * 01 II 01 tl ' aA3 Ui tS IS 6 9 6t U * S I I * 1.2 LLI It31 ' 6U U3 Et t t 9 a3' Cl 6 it of it 1.1 ft It 63 a e ft a 66 * U I 91 91 II ti a ot tl 3 a it It 0 9t a it a * t 3 IIn Iz 03 41 31 Of a 55 63 it It b IS 6 'S t e 4 It 4 it I I t St S1t S ttit 0t v6 66v 0 t 6 a

11 it *1at el II t1 It at at e2 a it 0 t tl tl o1 * * . t1 a a 91 Vt Ct B2S a US . . S toa '~~~~~~~~~~.

* a 5 a * *" *at05

I itt 3 * 4V3 01 .n 63 a 95 soit 6t 06 1*,91d" it'SL @0 @6 2 9 1 3 t 0 93 @5I0t163 @12 06J *1 91S ° ' ' *3 * 35 ' ' 'tS * e W 9WI19

tl 31 ' 6 '. ' 'l 9 95 'i If 13 9i '6 *3 *9 UIIn565

11 9 9 * t3 f1 6 * 9S t It it w n33 t It 01 @ fttl 41 * a 5 t6 t 2 aatt i SS " " §1 11 ' tot tI bt on it it 'It fit 'I ' w tta tS O 9 4 3I II I I I 63 t1 6 6t 19 9 06 et " I? it5 O @ 9 a3 32 1 61 61 06 I a99 i t9 36 la5 in 6a *Jm'1.PPS2 @6 & c ft #C3 13 ;ii& a el"U a *0 11 21 *11it 51 11 tl Et 0K e 66 t g"39.a e3 I.t '*3 ' 4 'n Ol* a * * up"WIN I6 I t t 1 t.31 3 5t61 'i U t t 36ll' a6* 2

a 09 S Li 1 5i 32 ft 6 3 ttIt6t at 3 3 *atpPI £6**,

* 9 0S2 3 at It ' af a ' Itt it tt'

6I 3t 5 2 el it2t3t 3 S 2 a t a a6 ; tt5 3 ' ' t U 'it a3 * ' 3UI t1 ' 'L I t '' * ' * mm"Atl a * Si t 9 tI 61 It 62 tl 31 *1 2S 0t? e 0S " it 0 * * 9t St 1 61 ' I 6 0 il U 4 6. 1 4 9 5 St it tl 1 1 21 'I t1 93 Sl ' aU am L1 9 o tI I. I 61 a 9t it 1. 41 U US 0C atP it t1 6 of 91 0I ' Cl 91 31 1 lC tCupa 'wI a. 3 It 31 L01 33tOt I Itt to92 ' P 0 I1 it i t t et ' Ot 't ' ' tt9t t 1 il't1 et tl 1t K *t U tI oet * tt tZ *{ § 2 Wa 11

?3 e * .6 t* 03 @3 55tt5 '*t '6 ' oiCl* a *U ZIl tt t CIO e1 e1 U9 I t tt* 0" tloie

03 U 63 ' S 1 t of it 3t it Sa t W

**e of0 I0 * @3* I .3 .3 1 @I .391 . *t. 9 9.1 . a. . * M*JZ Q U32 n 1 a0 It Li i a Ot it P a t ^

40 I t 11 t1 61 62' 6n Li U6a 96 66 36 t496 9n - 01 a I1 1 I I Su tt Ot t Oa 36 n6 3 " tl ' ' 1 1 ' ' t ' C t I ' t ot o1 9 91 tt tt LIU t t9 i t a to *oi s ISs @ 11 it #I 31 C2 tt3 t U t 93 Li S9 ' Sf I

52 n t 1 Itt I t ' 1519 n ==WM f 61 U9l l 3 61 '3 ' It a 9t It t I t n MM14"

6 11 it 02 1t 53 It I a a3 9t a3 It a 0 3 1 to 01 i 3t 21t 3t l l U 6Ct tt U i t 6t w SI II "l0r 32 n41^ 62 1 ftl 43l Ul 9lLUt 33 f 66l 66 46'i9 WP@U1C80@1 1. 01 .2 4.3 4. .I 91 Li 3 5 L £ 6

1 S p@igIW" 5*a1l AAUP0 A801

0 P6 301 SOJI)1 PU J30 l12lt$980d0633

-41)060 -eaPAl ZTZ1P a 303008 -mj 211% 9 *IPS

- lz - - 213 -

Table 37. Pbic rtgwvt aeditwr w Pflsy wit

f pait wle tta ae ettal to gmtat 1 5 ofd p costa is cvrtr 0 7 191 1to si ia1 / 3910 3 1to0A 19041 16 4

o 5 45 3 32 . . . 2.3 Lo.-ap.cinCssamu a 61 52 41 40 2 23 35 13 0.1

w 10 5 51 49 . . . 2.4 l.sarte m aridlb a 10 62 53 48 4 31 2 2 0.9

lushi lb 50 5 41 72 * 41 25 t.3 2W*lzsF. . s 42 so 39 24 21 0.4 J Illpr ISI . . 65 3 , . 27 7.8 4 dis, Te 67 7i S 00o 26 24 21 279 S Imahia 147 66 47 48 45 20 11 19 ii 4had . . .

a 50 44 36 i . , . . 2.3 Lau riccndate a 46 43 36 30 so U 1 12 0.7 I Etbepis 37 7 22 26b 33 35 19 239 0.6 1tdre . 35 29 . . Ile . * Ilulai 26 15 14 13 1t I I I.) 0 kins o . . 29 . . 33 20 1. I1 tmaaa 4.0 369 S0 24 16 120 12 6.1 12 1 . 61 . *53 . 29 240 0.0 13 AA . 66 3 a 2D 4 4 1. 34 t10 S2. . U 23 12 . I D 1.) 15 Coatralhfria lho. 66 52 44 . 22 39 1l 1 dar 52 34 50 25 IS 9 14 S 3.) 17 koi. 35 it 54 * 36 2 20 to hhi 30 43 29 47 13 26 it 17 2.0 it tKsa 39 52 3 39 le 39 13 12 1.0 .- harnLa . . . 40 . . . 12 2Ihsua ga . . 75 . . 20 25 22Mm 41 21 16 13 30 6 5 1.4 23 m 72 . 92 . 24 . 26 24 a1 . 120 f l 30t . 30 7 23 3.6 Is kuas us ......

a 74 103 129 121 . . . . I.S lddIle.-a oil iqwurts a it1 is 132 106 11 1: I5 tI 0.6

2 Ouritemi * 196 134 143 . 45 305 3 2.9 P Wrt s . 43 63 . . 3 13 0.7 2 .6Z 64 10 57 lot 9 1 10 154 2.J 29 Lswth 22 . . Sl 30 . . 7 30Cotse lv roe 100 352 I19 3t9 13 20 21i 7 2.0 itSl60U _ 69 373 122 I1 12 24 11 (.1 32 3ualsd 56 I1I 1040 6 Be 11 0.6 333 t 50 62 130 306 is 13 14 12 3.) 34 hsiti. 77 i5 157 I33 It36 *3 0.1

_ _ . _

35 hmnawi 92 60 43. 55 14 3 5. 7s 36Cawa. . . . 4. . . . .' 37 Coq, 1ei's go3. I7 7I 7 14 10 0.1 39 hepIa . . . .9 . . .

67 36 so 2 . . . . 2.0 Ii-Iiw Afrlca a 67 61 51 44 36 19 I6 I5 0.6

* 66 56 5 6t , , 3.6 Fra 0omcattnbi a Is a 54 49 14 2 21 23 3.3 * 70 57 so 50 . . . . 2.4 dqluuu_ catnss a 65 41 52 55 16 I3 11 12 0.3 * 4 42 24 2 . . . . 0. llEsr * 2 5l 3 29 39 211 19 2 .) MMsfor data ct hilelitySW cow p, sd tie ttmteal Ates,

A. Flqvuwith Wastshsk es for yws ata r tos t1kmsuc tia.So. the tctUci ateL 6. Fig are fr the dt reat pitabIle year. SN the tmichal sate. ,Wm piW an am 6w3tt1*o tUr S$I * "a;j 6* '01110 113111311ul11upp M iGi sL 'AMkaAl £ t;aJ :0 uw1 i ' 119

6o isn a ill!woA *1 13 t *3 wlPonLk l

6.9 913 it It? ttR 1 36 AM t'l 01C t at tt ' 0' I* *3oft * t w atl U C' aet ti09 "I IC uC I mf tm t @ 1.fr *N wOt It*' "I 9t tot U? 13 J we 691 *61 9 u fit90 ti 11 ot9 it U WAP *-1 133 63t ItI ' ' '1e1 90t 9 M l t3 i

6'? 603 631? 1* ' ' LI Lit 01? 90 3

*I3* *3? tI 6 fitt I* I*6 I' 'll 3' 'I 193I3' 63 Mal tI9 in Ott I u tol Om la * ~ * tw as * * * C C btC nt5 *tbl *X U6933 9'l 00 ilt It * a 1?U s O'tt 1 tt * 613 'i5 tt ~*t 'L ' ' 'Li ' ' * * 133 itO? LS *IC l oi li *tU 39 9 933 *MSJIAam

* 1 Oa I I am I a9 s ulit ipm 6o c3m *31 3tw et3 11tl 1 t ILl USItot 33?tt 30 iO0i ) outP3 01390? #t1it 0?nI 16Ii *11aI Om0It9 Wion0n6? tisIl V uSn;*aut 1t t'l 911anu oalt53 iOlit 10i9 I 9 OIl103I nI 931o9f 16Oa3 3*I1 96I t I Ut.lItwiqul Ct 651 !9ni11 03? IL * ' Ni 361 ''1)UaI I'? 163 31 66 *101 09 01 38 993 LiZ 91* 3*5* l'l U3* 0351 III 3t 6 * * IV 0 on kg I't 363? *001 1* Ii 3l st3 t II I0 *tin1 me

30e 3*6 6331 3? 19 t9 U UtO 193 39* 96* LIt 6 *JP*1 Itt 4U,9!4Im

e'l 'II ' I . W . I limltqI i Ot ~~~* *9 * *3 * 03?

*I 3I3 161 It 31 LI 13 02 I 99 1t3 00lP it? lat "C11= Z& l tz Olt t l i t 61, 'n Awl tt *0 it19 t 3ttt lt ' ' "n g f'O~~~~ to i 9 S a ' t o W ett t

*., *'0' ' '* th1 tl3033* 6??t16I1 *1S' Oi I *9tit1t 191 i 60? 93I a III U?S19*3 093O5t f 'pq31_6931 I"1)S ttb0 Ub *91ll ti Ult i t w* Ui t'O '? '3 Ss Ot *3 t13* Olt I ~I bw 1t C on I gtt91 6U09J9 miAl tiItU1 31 131 "I I t I91*3 C1*01 *1 11- , Cl! tl3 31a 0 ot 9 9tt Pt 19lit * l e li to1 I 0Lt ' tC1I 19 I16 * It 1 1* itt I Itiln L0 ' 1** 03 '3 C 2t 31 9* '* till '4 31 ' n WLI

U0 P9 It? or3m tt 113 tit 31 9tt @ I111 90 o1*a '3 I t e01t *3 361 *fit?3 ISO ' *39 '9 69U1 *01 *5t U0 U 911 'iP1U I

9'0 03* 3 l3 1 UtO 11 tl latl 6 * UO3O14 , _~~~~ ~P 9 IIa iNI" 3*9 ILt6 09 I 96016Si1131n13te 93't hiC *9 It to to l9t 3 iI tOI t1t S3l o IIIA I rt 6l n96? 33I *I 0 n I b i Ott .t Ulo 99! IC113 I9* I'S*tl tl 9tt9 *! 9ll ' pi' tt atl * reOl 31Ut 301out 1:61*1 0a 611toCq1 atiti 919i 3911 Ut3I;tl tt' PAtM 41002I I

Cl9 on 199 IL at1 oil 0? sn 8 6 S*) St 1121393 hsspi1 'aui '16* '6 '9 01063 93m S Il t it ml 113ul 31 Il t06It3o61 316 0t6 IT 3 u1to ou LIII @I

"SMVICe 11M1112 kt933 APaWS vit*fhto .ato seq 6*qqissdof

*'S$ 4*9Mmeu JOE 'WJ**91PVWU * ' '4tl '111 94

_ - - 215

table89. p@bUtcr,cvvgst astimitwoOtt 1tiftAta stont

b,uiltv@ (aOiltere a a mitso c Mast 1993 e@4 p eWtott

1970 1975 1990A/ 199 a/ 1970 197 10 I 193

* 2791 7023 2313 1929 LorIKON KeCno . n 3207 30t 2167 2197 I I; 9 10

* . 22902 1548 Lu-acoe moor-mid. a . . 2149 13t . 15 ti

IPIl l99) 2517 2149 137 16 20 Is 9 2IWmt Fan 2w27 55 250 156 32 l4 PIbe . . . 399 . .s. 14 4 balea, the I Cloima . . 1 .9 . Chad~~~~~~~~~~~~~~~~32

a 2a82 2"2 2315 0 . ' Le.4 c6ONother * 4436 309 2210 29 11 13 8 80

itktiop . 2900 1496 13 19 13 84I four s 2316 2271 12 14 9 1LU1 5072 3120 3992 3365 35 I7 23 12 i0 haelm4a.et It Twula bet 6997 123 96 SO 29 47 3 12 1 . 3459 38150 I. 144 3VI" * 79"4 125 2197 . 2 7 t0 14i 2T9 . 27" 297 I 80to 85 Cast tUrc R.hp. . 3 224 1754 12 6 h114uar 443 1 . * . I 4 37 uno 185 0e 21t1 . 49 I3 * toboi . 245 0 434 15. 14 I t 19""a 8979 2910 32402 1521 t20 It 4 20a1srn Lse . . 26 . 9 21 nm a 8283 11971 , 4 4 22 * 36 1"1 619 . 9 3 2 23SOO90 . 2033 . 6 . 6 24Je_41 . 200 1949 . 2 neota p . . . ..

5 3679 3116 1416 . Idlloucea ml towtar I 242 9 3M 4097 t 4 4 6

2 baital a * . 7756 l0o9 17 25 27Llbet . 216U 4.445 4 9 29 lad$& 43)1 Ws0 I1I$ 3912' 14 5 e 7 2 lewtth 242 e it 4 30 Ctet 4 lvure 50 7l91 12914 '407 7 9 4 e SIlabo 8602 154 299 22m7 2 2 4 S 32 buleo . 2909 M49 154 I3 2 2' a btu . 423 6316 l25 , 9 7 9 3418aurtus 52i 2231 n53 440 1 2 3 4

* ...... R*ledIOcm@el ptrters a . . . . .

5 Nlwon 72t2 14621 139180 2198' it 20 150 50 6 c e . . . 3345 . 4 37 Coop$P0eIt's 39. 2307 65T 2992 . 4 1 3 IS li ...... 29Mo) . . . . .

0 433 641 52 236 . . h4allr ffrics a 242 309 279 2710 II 12 ? 9

o 44 3114 2375 2 . . fracAphosa ua 390 25 32112 9 12 . S 4004 7? 73 2314Me . . Atloh tcowtne a 242 3120 3123 262 11 9 6 6 "6 * . 2060 8364 143 . 3*6w a 2060 1204 8229 * 89 9 9110

Idtes For Me trWp nlsty e1 ctswp, o the teiael Wote

a. FiW' with a astorisk are hVr pwootht ,t tha S 1f011. Is til tatCalaOts - 216 -

tWe 20. Istrole of priey pdulicruru mpmiitr# by ewp , cirto 10

wct of prtwy towva spaCtre too

bot Teben' tUbAI If cut ufaeldrtdiat .i) s4 uteruil kelari)os "mce diiltnshtad *. , .7 n.o2 5.0 0.9 3*L 0.0 Lori-(mguiesias a 0.5 92. IJ33 . 0.4

.4 01.9 4.0 3,3 5.0 2.3 Lima earn-wad a 3.2 34.0 3.? , I.) 1.1 0.3 19) 0.5 97.2 2.4 1.1 1.1 2 Besul Famm3 8U.3 1.1 3 . 10.4 S 4wr *.0 03.1 10.2 . 1.1 3. 4 WiG- , The . . . . t SoWI 3.1 59.4 (. (.3 22.4 0.

* 1.2 07.7 4.0 1.0 3.4 0.O tao-ihm other a 0.3 93.9 I.J 3.) 1.1 0.

7 It4is 1.4 ".2 2.1 3) 0.7 0.2 3 lai. .. 9lataid 7.9 0.3 (.3 3,) 3.3 2.0 10 w s-Iu . 92. (.3 3.3 3.) 7.5 It t?m & 2.4 50.2 23.9 4.! 1.3 2.5 12b1S f.) 90.4 1.4 (.3 3.3 0.3 is VW"sE 0.0 95.4 .6. . 3.4 34Tap. 0.4 97.3 W.3 f.) 1.3 2.1 lustra)fri uulep . ', '. I4hagm 3.) 99.4 3.3 (.3 l .l 0.4 373 1. . . . .

1tKm" 2.1 US 2.4 34 10.0 ,., 21a l.m . . ... 21 bued. . 22ams 1.9.4 I.4 .;. . 23 . . .tA. 245ued .1. 91.4 3.4 0.4 4.5 25 lau . .

* 1.9 9. 1.0 1.0 1.3 4.3 134 Ka h Ol i90WtOf * 0.2 0.4 0.4 3.) 0.1 U4 2. Iwitua 0.2 97.9 1.9 W.3 (. 3. V Labsa L.3 3 1.1 .3 (.1 3.) 20hog 7.2 3.t 3.1 0.2 I.) 29 Loums ". 9 (.3 ( .3 0. Cat d'lbit 0.7 "0.0 1.5 5.0 0.7 2.1 31 1.1. 0.39.2 (. L.I2.3 3. 32basime ( 9..4 4.3(3(.1 32 MSOuS .3. (. .3E 14.2 34 UsrUit 3.0 U. (.1 ./. L.I 0.3

3 . . . . . 40 diatv 5 . . ..

U *. . . . . v Camp,Peto's Ip W.)99. 0.3 0.L00.

* 1.7 ilL $3. 1.3 2.4 2.4 3.4rnSPris . e0.3 91.2 1.1 (.) 3.3 .4 * 0.4 . 2.0 2.5 a.1 1.4 *. a 34 97.2 1.5 3.3 3.3 0.3 e 2.1 . 4.4 0.0 2.9 L. _qtu m a It.s 09.3 0.3 3.) 0.1 0.3 * 4.4 a7.7 1.9 3. 52 0.0 O .4 92S (. (. 0.7 0.?

bO ceweiluty n cwdeap we thetabla sats. .1. lapUluwe "msdIe vi theias lmwA. - 217 tOle 21.Oittnhutiao of seonmarypubcic rrcrrst epeedltureby pwow, citr 1 PerduatO9 WWAVYr ret Mtu tW* Toacher' Teghsep CUevea CU Aimaistratise bweaute stricials Scholarhips erWs dletnhu

u 3.3 7L. 6.2 0.9 3.1 3.5 Usacecouszsme a I 40. 4.1 Id 0.2 3.0

U *.1 52.4 31.4 I.) 6.l 3.4 tocoasOm-mrid a .5 56.9 .5 .0e)

I Ilai 1.4 54.1 6.3 I. 1,) 4.) 2 howiaFt7W 47.8 119 I1 13.9td 3 NIr 9.4 40.4 10.3 d.) d.) 9.l 4 laela, The . . . S Seli s0.759.3 12.1 I. ;7.3 0.1 6 Cha . .

* 9.? 7.5 6.1 0.9 5.0 e.5 tLousc other a (.1 74.7 4.0 1.W 0.2 6.7 7 IthiePsa 3.1 31.1 4.0 1.1 3.2 0.6 Zatire 9 las 4.0 n.3 9.1 Id 3.) 13.6 10Gusw-Mmu .. ) t . ..) ) I.) 20.5 It twau a 2.2 44t5 3.6 6.5 2.3 16.9 12two" (.1 67.3 2.3 I.) 255 1.5 IS Uga 1.1 54.0 . 1.1 W 46.o 14Tap 5.3 30.0 4.4 1.) 0.3 3.4 Is Catra Aric Rep. . . . .A". U6 nag 0.1 l,.5 1.2 I 3.6 2.7 I? Se, . Ut_cM.0 1.R 5.0 to.10.0;.0 19R"a. M.)6S O.) 1.3 I.) 20.5 20Sierra Low. 21taua . .S.. 22 i *.0.) 9.? I.) Id 0.3 25 .. . . . 24Smega) Id 16.1 11.0 0.a I.) I.) 23iau11 qo . . ... a 3.3 67.4 I.3 2.4 5.4 21.0 Itid1e-moeem leprter a 1.) 69.5 1.0 1.) 1.1 4.? 26 witaetaa 4.9 1.4 4.5 I.) I 1.1 27Llhris 13.3 6L2 3.5 6.3 .) I.d 23Z,Aa 16.5 39.2 .3 d.) 1U.0 231 29Ls 1.aI 99.9 I.) I.d 0.3 OCot#elwire I.) 69.5 I.d 5.0 .) 5.05 31tih Id 1.3. 1.0 d.) 11.5 9.0 32 waslan f.) VA.4 12.6 I.d I.d 4.) 33Staa I 4.2 I.) W.) 49.4 34nwrttc. 3.2 19.5 0.1 I.? 0.7 0.5

* . . . . . 16ie-iecu m1 ete . . ...

23Egna . ... 36 C_ae . . . . , 37CUSS, Pee')s Ile. I.) 33.9 1.8 .5 4.o 0.4 U Oat I . . . .

* 2.1 72.2 3.0 13. 5.2 33. lidsim ArIca a .) 9.5 3.4 .) 0.2 L3 a 0.9 1.4 2.3 2.9 l.t MI. frusephmecertries a 1.1 67.0 4.4 I.) *.$ 1.5 a 2.6 n.2 4,6 0. 7.2 IL3 bo1 _ cour a I.) 73.4 2.2 Gd to) 11.1 * 4.2 03.3 6.1 1.) 5.3 1.0 w' a 3.1 79.5 .0 I.) 3. 0.6 1tO FU dt (MPrnhlty Cd orwa, - the tKbtalulets .1. bPdture Mo ltiK e Vtt teOeS' eluotW. - 218 -

Table22. Diltruibtio it etiary publicrecurrat sepsditureby purpo, circa 193

Pwcestof tertiary recurrt n"eiture to$

teachrs' Teachiaq lufarm t Adbiaimtrati esumouts sturials kilarsips wsvice distibsted

u 2.0 40.0 1.8 44.9 1.1 8.2 Lew-inoueKOi i a t.l 52.0 0.2 44.9 1.) 2.4

1 .1 C.) 20.9 1.4 77.0 C. C.) 2 luwrbiaFaso 7.8 9.4 1.4 4.3 7.2 2.7 3 1i1r 1.C t.) 1. 59.4 I.) 40.4 9 8 0.4 91.1 4.1 1.) C.) 4.3 10kines-liu . 7.8 1.) C.) C.) 32.2 12 wm C. 59.8 1.) 40.2 1.) 0.1 14top 3.4 44.1 0.4 49.4 0.3 2.1 to IWGd .1. J2.1 4.) 27.7 C.) 0.2

q 4.5 54.0 1. 34.0 1.1 5.4 Nidde-iusC s us a C.) ?3.2 G.) 25.6 1.) 0.3

26Nitaia 13.0 17.5 7.5 42.1 .) (.) 281ai. C.) 81.5 C.) 18.5 .J. C.) 29Loatbo C.) 100.0 C.) .1* .) 30Cote dIveire 12.2 22.7 C.) 51.1 .) 13.9 31liUdi 1.) 92.4 0.2 .1. 4.0 1.5 32 uailad C.) 90.0 10.0 C.) 1.) C.d 33 htia Cd 74.4 2.)5.4 . 1. Cd) 34 Nwitis C.) 100.0 CI) 'I' 'I' Cd 37Ceo, hole's Rs. 1.1 17.3 10.7 45.4 1.) 4.4

u 4.3 49.7 1.9 37.2 1.2 4.5 8h-SabarnAfrica a ,7.8 0.1 33.9 I) 0.9

u 7.1 24.7 2.7 55.1 0.4 10.0 Frucapkb ceutri a C1 20.9 0.4 59.4 1.) 2.1 a 4.) 87.3 0.6 9.2 2.5 0.9 Ibngl_m cmtuiue a C.) 91.1 0.1 C.) C.) C.) * C.1 47.8 C.) I.) C.) 32.2 Other a C.) 47.8 4.) C.) C.) 32.2 bts Fwrdata cipwarility ad coearep, thetchaical sets. .t. b lto a lea l umith tachwssesolumts. - 219 -

1tl 23. irai e. WurWYtI vs awvep tusul , ret 19o

Prtmy mhr's uW* testw's ewap $m1ary airy

a a wtioshe a awItIple of urg of asaw vrt s* icapata vruto Wp. ta

M ISO) . 3 1 ceeme$ IS" 6. St 2. a 2m .3 7 ofnt-aid e 217 11.; 1v02. I hIl t0 125 13 0.3 2 W1hz Pm 2255 12,5 251 13.9 3 I31W 2382 9.0 29192.3 , Wm, I"* dlsta S.; Li Oc . . .

D 8332 * 3591 -iuu *a 23 4.9 2.59 12.4 ilthipla 11ip 10.0 . 0 lre . Idam 744 3.5 29 1L4 to50.-i48 822l 0.5 2111 10.3 It tm 0 2.8 U 2L. III Ins83 4.9 4o10 8.2 au,ed Inm.s 24 I1.3 14iqe 8329 4.7 277 9.2 ia CtI lri ce p. 14qm . . 278.1 ... 0111 2423 9.0 4 15. 1 ISO" 1232 3. 13= L. 3SibarsUi 21 adca ;2 Sam 9.9 $142 1.; 238 . . 24_qu 3211 7.S 0 10.0 U " l .st . .

ellg% l tupete 40 041*S1.o Ukeh Li mL

Laub t01 4.0 34 110. 2c Iaw 1taia 40721 85. 82328 21.2 38 li_i. 48110 5.4 73 10.0 3huzilud 3077 3.3 a2n 5.0 a 024mm Mt8 3.2 4"0 5.4 S3vet_n 34 2.1 430 a.

o . ..

25 UIt ... S0 C_m . .. v COOp, we's k 4.0 WI7 4.9 3S9_ . . . au .~~ . ~ ~. _ mm a 207 . 44., 8.-in dana , 2155 5.0 460 89

* 3748ls* * 100 . 4340 dagiepe mit a 230 5.4 30 80.2 1sIt * 2139 Othr 13 I130 0.

we par datetcwOltty ad £ qon tin t0U2d lati_ - 220 -

table 24. Estwaalaid to MAricastatsoc mo trua 10by d#w, 1981*113oratw

mill., of doliles

Edvataomsector al Total a a Projoct cod of wcuta obf Omrn S9- related bkOAlAfrica ll ad to *ellaobluo Xlkw total traintas bi ttdait atro c/ Total Anfica edtatle

1. ED ld WIC denwrs t I68.2 657.3 2,5 394.4 1".5 1369.7 a 58.5

A. ConceoiLul ad t 56.2 559.2 757.4 366.9 Me9.8 1314.1 u 14.5

55) sulaturoS t 149.4 137.8 507.2 246.6* 159. 94.6 , 40.7

Fro 23.4 182.8 204.2 . 3.2 279.4 58.8 lelpus 54.6 43.9 8.5 . 7.4 44.9 4.2 Unatodlifadoa 15.7 24.2 39.9 . 23.8 63.7 4.1 United Slote 7.0 29.3 ;6.3 . 64.3 18.4 6.5 wowav,fraI Rhp. 4.5 14.7 19.2 * 7.5 26.7 1.7 Italy 2.4 15.4 17.5 . 3.7 21.5 1.4 soad. 0.3 23.0 23.3 . 0.0 2.3 1.5 t1r ad ot diotritbtod 8l.5 24.5 106.0 244.6 ' 9.9 363.5 23.3

(2) bItilatral t t8.# U1.4 20.a2 120.3 0.0 3e.5 u 2.

Inbtartioal l edt AsoIats . 123.6 152.6 54.3 0.0 12.9 11.8 Africs ewlapmt Fed 0.0 35.7 33.7 54.6 + 0.0 40.3 3.1 Eure Obolepmt Frd 13.2 26.0 39.8 . 0.0 9.8 2.4 UnstedAstumo hvleat PrWa, 2.2 15.8 18.0 . 0.0 10.0 1.2 5IC5F 0.6 7.5 5.4 . 0.0 8.4 0.5 Oros ad wt diutnihtcd 2.5 18.9 21.7 51.4 * 0.0 73.1 4.7

L. ru-cogeuol ad t 0.0 25.1 25.1 27.5 0.0 2.4 u 3.6

In 0.0 11.9 I1.9 7.9 0.0 19.5 1.3 Afnrc vlgit Sak 0.0 14.2 14.2 59.6 0.0 3.8 2.3 it. £i5tfEromo0-wkstc scoaoeaadCuh 1t 32.8' 7.2+ 40.0 . 50.8' 9o.a L5.8*

311.Otbsr (Ept, Inioa) t 0.0 0.0 0.0 . 4.7 4.7 . 0.3

IV. ha-qowvrot wouat:sh t 9..8 + 8.2 * 90.0 * . 0.0 * 90.0 . a 5.8 . tow t 210.8 704.7 915.5 394.4 245.3 I5M.2 a 100.0

Mfr dat cratahsty ad corp, t. tic tKui l ott. Flp. into a cros es sotsat, W tii tocticol aces.t

S. Hsa uote to ccaro) depatrg ofadvotum. t. Aid fwr traiuiq in actlrs oew tha bd tJlm. c. Hat cdtbry sobidutloiatof Arica stodats tays. droad, a r d damu y fellwoispo. - 221 -

T461o3. Cromustsa1inatisn sactwaid fro MM1ad 07P2agatur by racaPagst 011tty, 193143aVWwaq

Matdialbwus of Is-aptal Cipatal ~~~~~~~~~~~~total As Weastsp Aspwcmtagsfdnaa of totaW toftal is a5 romwvut Puautqs £401t51 fUCt4ot lEKctios hUvys siastas. goasqto buysi siaatam kblurs of total didpa (mills,) bed" l4osR Isaillao.s bdwe (sallass) estorelaid caplt. a 30.8 35.2 * 1.2 10.4 2.43 Lainocmscuman t 446.9 a 19.7 32..4 t 107.2 *47.7 t 574.1 s 11.0 2.94 o 29.4 51.0 s44.7 7.9 3.30 Lou-lagos arn-iE t 73.5 . 29.5 32,5 t 29.5 * 54.9 t M9. a 9.9 3.42 1 Hall 12.3 21.9 24.1 3.0 84k.4 15.3 7.3 2415 2 Iktsa Fags 15.3 32.4 32.5 4.3 70.9 19.4 9.2 3.02 SlI 15.1 22.1 52.7 12.5 34.7 27.4 30.4 4.61 4 Oa*lia,The 5.0 24.46 43.4 2.4 50.6 5.4 13.3 7.11 SImaia 34.4 55.2 2.0 2. 54.2 14.7 3.4 3.71 4 Cad 4/ 33.4 55. * .0 300.0 14,4 22.3 3.1 17.4 34.0 a 26.3 11.1 2.20 t.l.ossot4MM t 3M.4 a 37.0 32. t 01.7 a 47.7 t 477.1 a 13.1 2.54 I Itblpia 23.4 14.4 33.1 4.5 10.4 26.3 14.1 0.53 I late, 4.2 30.4 24.3 5.0 24.2 45.2 33.0 1.I 9 hIm. 10.0 9. U3.? 7.0 53.4 37.3 14.7 2.74 t0hoaI 03ssrn Si ..... It11oamas 37119i . 33.4 33.0 S.7 1.94 32kiwi 33.4 564 31.3 5.1 56.3 23. 36.5 5.47 is lobs 2.1 2. 37.61 2.2 11.1 4.3 2.2 0.32 34tos 10.7 36.7 27.9 . 74.3 14.3 M3. 5.11 I5 Cot AricMawgop. 13.7 32.9 22.1 2.2 34.4 35. 17.7 4.43 lb hiaqama 20.0 19.7 40.7 3.9 14.4 21.9 3.7 2.33 Ii Isi LI7 12.9 27.2 4.3 47.7 32.8 35. 3.44 to be"a 9.9 17.0 29.4 5.9 52.3I 15.3 NJ. 2.37 19to"s Si 324.5 26.9 47.2 . 24.5 25.7 . 20hews turn 7.4 37.1 53.9 3,2 43.3 30.3 33.3 3.53 21Odass I.9 3.9 32.3 4.9 79.0 3.3 9.3 3.54 22amn 9.4 31.1 31.1 1.6 9.3 33.2 7.9 0.92 2336w. 7.0 0.0 41. 4.0 37.0 13.0 1.8 0.84 24ftusg 50.4 53.3 22.4 9.3 "5.7 23.0 9.95 Uhsuk"w 3.4 . 4.1 2.3 10.7 5.1 0.63 * 4.5 43.4 a 32.0 s 33.4 4.14 11001#4oomoil aopwtvs t 95.9 s 31.1 44.0 t 21.9 s 26.0 t 117.3 s 10.7 4.14

24 Nettslua 4/.. 27 lUkn L.5 31.3 v.3 2.0 24.3 L5 .3 4.25 a hogl 21.3 10.9 . 2. 51.3 24.4 7.9 4.07 29Lauts 9.5 27.9 54.7 2.4 21.5 32.3 33.5 L.4 30Cmt d'1wlrs if 24.3 4.3 33.3 0.7 0.4 25.0 13.3 2.33 3125.4* 30.4 2.2 49.5 45 31.0 34.9 7.3 2.29 52butlas 3. 3.3 44.3 1.6 19.9 14.3 34. 34.71 55btins 12. 20. 47t7 5. 55. 17.7 37.4 19.63 54lewitAIs 2. 4.5 33.7 0.4 29.1 2.9 4.1 3.2

a 1.2 34.4 * 1.3 a 13. 0.53 NOMMO discmsllrptws t 53. a 3. 53.7 t 10.2 a 2.9 t 43. a 14.0 .2.

U Napt. 36.3 0. * 0. 0.1 11.1 30.2 0.32 24tuar. 19.0 10.1 33.; 7.3 11.3 24. 12.4 2.03 V7caps, FayWslop 12.3 30.3 0.2 2.9 13.0 14.0 7.45 31*.M3.Li 3.9 33. 0.4 3.4 3.5 5.4 5.00 5960 Lia6. 1.9 37J 1.5 14.2 9.4 34.0 3.20 a 7,3 3.3 a 6. a 10.7m 2.01 Mkbgslfrt"es1t3LI1 14.7 535 t 139.3 . 30.5 1717.4 I 33.1 3.33 * I3. 30.7 a 39.0 a 13.2 3.15 frrongopaemutra t 3.4 a 16.7 51.3 t 13.2 a4.OA 342.4 a 13.0 233 a 4.9 . 44.5 a L.I a 9. 3.41 AnO3OMosmotts t 2742 a 33.9 40. t 53. a 24.2 t 329.7 a 31.3 3.3 * 7.2 ILI3 u 39.7 a .0 3.32 ow I 52.5 341.4 34.3 t 32.4 016.4 146.1 9.4 1.03 aMMUW Eam sayu*lsty SWtoerop, us tbstothcal b.Lo a. mtodatoTIdleIL 4. CosInosfor r*ag sallortla iswts ot: saedusatas sad to avaslilo.Fipas" ""Mato autal suatta no6. - 222 -

table24. Eataraultati ctor aid froa,OmC A OPClons bylout ad tm of oducatia Pwcotapof tota

Dollars skoudy (Gtilian) --- - - Othxr 1981-03 tla Vocational am aot averae Priwy emaal training tothical Tertiary distributod

A. Co.cos aa ad t 75.4 a 1.4 14.0 4.2 16.9 34.2 19.3

MI tlatorl t 507.2 a 3.4 20.9 3.0 14.9 42.4 15.4

fruco 204.2 0.9 34.1 0.9 20.2 39.4 4.5 Polgiw 58.5 0.3 30.4 0.3 17.1 40.1 11.2 Unito KIndo 39.9 0.3 19.8 5.0 5.0 54.4 13.3 UitudStat" 34.3 6.9 10.5 17.1 4.9 29.7 21.9 Ibroy, Fowdhp. 19.2 . 4.2 1.4 14.4 59.9 19.7 italy 17.8 15.7 5.4 . 7.9 51.7 19.1 _d. 23.3 29.4 2.1 3.0 18.5 15.5 31.3 ot0r 104.0 2.4 3.7 3.7 10.2 49.4 30.4

12) 4ltlatord t 250.2 a 15.1 4.1 12.7 20.9 17.5 27.1

atoaitm DeeopubtAoiatiue 121.t 27.1 *.1 10.4 241 11.4 14.1 frica Oelow t Fud 33.7 4.) 6.9 4.5 31.C 2.4 12.2 Efor u aseOpmtFad 39.6 I.) 4.0 5.) 14.3 60.1 21.4 Unitodatim eIbnl_utPro 18.0 2.C 1.7 10.4 12.6 25.0 47.1 WCIIF 6.4 39.3 I.) 13.1 2.4 f.) 45.2 Othrs 21.7 1. 4.) 4.) 4.) (A. 100.0

. lhmcosmsgal aid t 26.1 a 22.4 3.4 ILS 211.1 20.4 6.4

low6 11.1 27. 8.4 I0.9 26.1 11.6 15.1 frica DevelopmontIank 142 11.5 I.) 24.7 29.4 27.2 W.)

Total t 7I.5 a 8.0 15.6 t.7 17.3 33.7 16.7

SacilantgrWWI ai

Lawiueo ad-arit t 99.0 a 4.1 1L3 1.4 13.4 39.3 25.4 _'ium ohr t 471.1 a 4.2 12.4 4.3 2.3 31.3 20.5 IAddle-too oil iwto t 117.6 v 10.4 13.8 6.4 11.2 30.1 25.9 iddle-imaoil urtere t 3.S a 3.4 15.2 13.2 28.0 22.1 16.1

Fr ola iecnatries t 342. a 6.3 5.4 6.0 28.4 26.5 23.4 gookmecatria t 329.7 a 2.1 21.5 4.5 17.0 34. 1113 Owr t 48.1 a 7.2 4.9 2.4 27.7 23.4 32.2

Wt Fw datac ability ad covwap, o the taiucl stav.

a. Cemia Aid only. - 223 -

Table27. Eaterul eati mooterad fro G b PC darn by pwV=

6411,, Purustaqe@4 total (otilissl Ihilt~~ ~ ~ ~ ~~~~~4wto Catoat kn-at IS)4IZteduscal Opuatsacott Mu- aweq Capita mnietm falhoglps a uplij apcifted

. Cuicral aud t M.4 a U.4 44.3 14.9 10.9 1.5

11) ral tilatt 50.2 a 7.4 57.5 20.5 12.7 1.9

Friot 204.2 2.4 62. 11.4 3.6 1.) 3Igim 5i.5 0.3 64.3 25.0 4.4 01 UnitedIangdu 39.9 2.0 95.4 39.3 3.3 I.i UIitedStats 34.3 7.2 547 lv.3 1.86.) kra, , FedoralAep. 19.2 9.3 41.7 23.4 24.4 1.) Italy 17.9 1.7 4.3 13.5 16.5 1.) Saed" 23.3 24.1 21.4 0.7 49.9 I. otr 104.0 20.3 12.4 34.4 24.2 .11

(2) bltilatwral t 250.2 64.9 17.4 9.4 7.2 0.9

lteri.ala IMel t ASktOCiad 128.4 03.2 14.6 1d ) W Id fri*ca lepunotFud 33.7 64. 9.3 1.9 3.0 1.) Euraa Ooeput Fund 39.8 12.5 7.3 33.2 .5 1. nlited atios hvelepmt Prtoe 161.0 1d M4.2 12.2 23.4 I.) IMiCEF 9.4 9.5 23.8 7.1 9.5 0.1 0th. 21.7 20.3 12.4 34.4 24.2 0.7

L. Ncoacuusemalaid t 20.1 * 95.3 11.9 1.1 1.7 1.)

ImD 11.9 3.2 14.3 1.) 1. 1.) Afrncasevolupewt lan 16.2 941 93 1.9 3.0 W.)

Total t 795. a 26.5 43.1 14.4 10.4 1.4 lctipiut gerup a)

Lu-iae is-artd t 9.0 e 2D.8 53.1 12.2 5.4 3.3 Low'inomether t 477.2 a 17.1 34.1 14.0 9.2 25.6 liddloacoe oil iWpt.rs t 117.3 a 34.3 42.3 10.7 11.5 1.2 Niddlo-iecmoil ptorts t 43.5 a 44.4 34.5 11.7 0. 0.7 fraoephowcontries t 362.4 e 19.4 20.7 15.1 9.9 34.9 b lopbmcoutrie t 329.7 a 24.1 542 12.5 5.3 1.9 Other t 45.1 31.6 37 11.4 20.4 3.5

mateF datacmnarillty ad covrwp,uo the techuicalMtn.

a. Cocatissl,l aid sly. -224 - table2. bstarudpehl&c Mat a# theelutahe setw

lot soarvc Om$Ie kkootsat ditaad 1365111milio) Mo diatarsd (on6ullaPel Actua Presectel/a ave a9m am4 19. a9m a am am .9w. IIm a,u s90

Lotm-accKcoooetu t 42.6 364.1 577.4 1.0 31.4 30.1 6.9 20.0 16.3 23.6 i1.5 U1.S

LU-iscomsue-utd 1 0.2 42.4 101.3 .0 0.3 0.1 2.2 4.2 1.6 3.1 LB 5.9

I Asia 0.1 20.4 24.0 3.) 0.1 0.3 0.2 0.3 0.3 0.3 0.4 0.1 2 herbs.Fmr 0.0 3.1 9.6 0.0 I.) (. 3.) 0.1 0.1 0.1 0.1 0.1 3 1iw 0.0 0.2 22.6 0.0 4.) L) 1.6 U. 0.9 3.4 3.a 2.9 OSUatsa,the 0.0 1.3 5.5 0.0 3.) 1. a. .) 4..3 a. 4.) W. I Smalls 0.0 1l.6 33.6 0.0 0.1 0.1 0. 0.2 0.2 1.7 2.1 2.2 6 Chud 0.1 4.9 4.8 0.0 0.0 0.0 Ed .) 0.2 0.1 0.1 0.1

Lw-lo. oce t 42.4 321.7 476.3 11.0 II.) 10.1 4.1 35.8 14. 21.9 15.7I 5.7

7 £t610366 6.5 49.9 77.0 . 0.5 0.7 0.6 0.9 .1 3.2 1.5 1.7 * lIir. 0.9 6.7 6.4 0.3 14) 0.3 0.2 0.3 0.2 0.1 0.3 0.3 94.1aw 'I.9 21.0 57.3 1.3 0.2 0.2 0.3 0.4 0.6 0.5 0.7 0.6 50kuerBtsau 0.0 1.2 I.0 0.0 3d) 0.2 0.7 (.1 3.) (.1 (.) 0.1 It Tamale 9.5 31.6 40.0 0.1 0.5 0.4 0.1 0.6 0.7 0.9 J.0 I.0 I2b"iU 0.0 9.9 29.6 0.0 0.1 0.1 0.1 0.1 0.2 0.3 0.3 0. is up" 9.6 21.0 22.9 9.3 0.3 0.1 0.3 0.5 0.6 0.6 0.6 0.6 14Iqe I.) 0.I 12. (.1 0.0 3.) 3.) 0.1 0.1 0.1 0.1 0.1 Is Cot's"frice hsp. 0.0 3.9 11.3 0.0 3.) 3.) 0.1 0.1 0.1 0.2 0.2 0.2 16 uiagmcw 1.6 79.6 10.5 0.1 4.0 4.9 0.5 II.! L.S 7. 3.7 0.5 17koss 3.1 0.0 3.0 3.) 0.0 0.0 0.0 0.0 I.) 0.3 0.3 0.1 to AmuG 0.0 2.7 6.7 0.0 3. 3d1 I (. 0.1 0.2 0.3 0.3 it Km" 6.4 22.3 43.6 1.) 0.5 0.1 0.6 0.0 1.0 1.4 1.9 2.1 20Starve Lea I.) 3.9 34.4 0.0 0.3 0.1 0.1 0.3 0.2 0.3 0.3 0.3 21luu 0.0 4.7 22.3 0.0 3.) (.3 0.1 0.1 0.1 2.4 2.4 3.1 USiDe" 2.0 2.3 2.3 0.3 0.0 0.0 0.1 4.) 0.1 0.3 0.1 0.1 23hor 0.0 22.2 21.7 0.0 0.2 0.2 0.2 0.2 0.2 1.4 0.6 0.6 24S.q.l 0.0 32.7 49.1 0.0 4.7 1.9 0.1 0.3 0.6 4.1 3.6 3.2 25 fim lpe. . . . . fflUIe-$sceoalliqarerb" t 5.3 231.6 214.7 0.3 43.7 32.0 39.3 39. 22.9 36.7 32.0 32.6

26htmtams 0.0 3.2 6.9 0.0 3.2 0.9 0.3 4.) 3.? 0.2 0.2 0.2 27Ubert 2.3 11.0 37.1 .1. 0.3 0.2 0.9 1. 1.2 1. 1.4 1.4 * lawsa 0.4 39.6 52.5 0.2 4.9 4.3 4.9 6.3 5.9 5.6 5.3 5.3 29Laue4 0.0 0.0 4.9 0.0 0.0 0.0 0.0 3.1 L.1 (.1 0.1 0.1 20lete Iivars, 1.6 110.7 61.2 9.1 35.9 24.0 30.6 27.3 31.1 24.6 15.5 15.9 31Ziabatu 0.9 0.0 0,0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 32Su55lad 0.0 6.9 14,3 0.0 0.4 0.2 0.6 0.7 1.1 1.9 l.i' 1.6 33 Mutse 0.0 L.1 22.1 0.0 0,5 0.6 0.6 1.2 1.7 3.3 5.4 5.3 34lawitlus 0.1 31.9 13.4 LI 0.6 0.7 1.3 2.1 1.9 3.6 3.6 1.9

idile-smd,eal seterts t 4.7 135.3 164,0 0.3 19.5 17.4 13.4 30.7 15.4 21.9 40.0 49.3

35agas. 2.6 16. 96. (.3 2.6 3.1 3.6 5.7 10.6 20.. 33., 43.0 36cauea 1.7 47.1 10.6 (. 1.4 1.6 3.3 2.4 2.1 2.4 3.6 3.9 37coup. upils P. 0.0 5.1 10.7 0.0 0.1 0.1 3.0 3.0 3.1 1.3 3.3 0.9 3NIa 0.2 23.1 6.1 0.1 14.3 10.1 7.5 13. 3.5 1. 3.6 .3. 39 #eeAd.& . . ..

s-tiaraAfrica t 526 731.0 91L. 1.4 74.2 AM. 5.7 69.7 54.6 90. 93.1 103.4

Frascoqhaeestine 4.7 333.9 3VL. 0.6 19.0 44.3 44.1 39.3 240. 43.0 34.9 31.6 b,lqhc "scm og 43.4 336.2 406.7 0.6 14.6 IU. 13.1 .29.3 31.6 46.1 51.0 63 BIbs 6.1 62.9 113.4 4. 0.6 3.0 3.6 3.2 3.2 3.0 3.6 4.0

Mm Fw ddatcoimpr latyofd cowerap,em thetaMchca este.

a.PrOoiW Mt0 WOK i mase Omcuwetractuil k59t3 a do btstfble at the medof 39. It02eld..11 theectof mubeqetdoht relie qreeete. - 225 -

PART B

ECONOMICAND SOCIAL INDICATORS -~~~~c I

t-4~~~~~~~~~~~~~~~~l wv~~~~---- IrZ ~ V~~im I9 i-L - 227 -

Table~~~4pSubas 5. uui W w Copts, Ufe heape1 v~*apqmuda upon"ac W aut"Ithaeis Pao ego rat of taflati at birth (laWSi) 44"we hIlt vst_ -t- 8i4-IS Wilnm I'114 19 13 1970 , 210 -04 4.3 19.4 46 Lu-IsCs asamias t 21. 559 a 24 0. 3. 51.2 to a 52 0.2 *.1 12.2 4 Leuiacow wai-wad t 30.9 0114 * 590 1.1 3.9 oea0 I Hai 7.3 140, 14 1.1 7.4 10.4 4 2 wkie Fmo L. 274 154 1.2 2.4 10.4 45 3 Uilq 4.2 1247 19 U.S 4.0 11.5 43 4 Subd, ThW 0,7 11 20 1.1 3.0 10.4 42 5 Sli 5.2 130 2411 3.4 20.2 44 & ch 0.9 520 . . 4.4 .4I

* 22 .3 4.3 20.1 9 L aim ather t 220. 19 a 2Us 0. 3.4 11.5 a

7 Ibiasta 42.2 im I1 0.4 1.0 4.4 44 * lare 29.7 230 100 1. 18.7 0.2 a 9 Wuasi *.0 U9 10 1.7 4.C .4 a5 10esuma-fa 0.9 3 1 . . *.1 3s 11 Tama 21.5 4 210 0.4 342 11.5 52 12DU IN 4.4 21 30 1.9 2.9 52.2 0 1 s 55.0 21 210 -.29 5. 44.5 1 14 top 2.9 5 20 0. .i 0 5 IS tial Origs ago. 2.5 UJ 4.1 3.0 13.0 4 14 flaae 9.9 0 2Us .4 0.1 14.4 52 1ism S.9 113 20 1.0 .i 10. 49 to %N, 5.0 2 3 2L3 7.7 10.5 41 It Koo" 19.5 33 310 L2 2.3 10.0 54 200law.^aa 3.7 n Jl 0.4 1.9 15.4 30 21SOima Li 204 3.5 3.0s 4.5, 22am 12.3 3 35 -1.9 0l $2.2 53 231Wm 21.3 2#t 3iO 1.1 7.2 19.3 0 l4 aspal a.4 1% 3S 0.5 3.0 9.0 04 35 saSqae 13.4 US -L. 4.5 11.0 4

30 3.2s a 1.1.l.t 5 V4#0e-teg aS m4WtW1 t 24 31 41I 1.5 4.3 11.1 54

24 lwstma 1.7 1031 0 0.3 3.9 1.7 04 27 Lilvia 2.1 111 40 0.5 1.5 4.7 so 20 Imeau 4.4 15 070 -1.4 5. 10.5 t2 29Letha IJ 30 S LI 4.4 11.9t 54 3s tet,lvar 9.9 u 410 0.o so 11.1 52 31la " Li 39 70 1.5 1.1 11.4 7 32 taslmd 0.7 17 79 4.0 4.3 14.0a 54 IIUsta_s 1.0 40 40 04 4.4 9.0 a 4 Ihitt 1.0 2 lOe 2.7 L ILI X

Uli ae,i. oila"to.c1 t .1 3254 a 0 2.9 I 13.4 51

21 liwu 94.5 2 730 2.7 1L.3 I.4 50 $4 Cw 9.9 475 m L9 54L 12.7 54 v tomes No 1.9 Su 146 LI J .6 IW v a Sts 0.8 24 4100 5.9 0 15.5 51 S3RApla LS l7 . .A 1.4 U

a 3S 1.0 L.S 152 4 Wir-Mv Mraig t 40.2 2207 a 350 1.1 4.2 11.5 4

a 3a I. * .2 141049 frMqeis m i, t 120.7 to725a 1. 4.5 12.1 X1 *520 ~~1.7 5.9 1442 S0 At1q) am"nee t 210.3 0.2 30 10.5 04 * . . 2.e . 45 Shea t 70.2 S3 3 2.1 4.0 I4t 04 wsesFe doteeeoswih1tg no aumrqaon *1.toge ese a. Ftp's atIb a ltwtab eam Eu 9e -U, ad I 4.". - 228 -

TiFl 2. Laq_ps df 1riwu ifttt Laqp suts of loompwutofpqla*

- Ptga btbw sgi official Lis" L_ VW Pot- total Priuiiepal i uno a/ tugw laquag laqug tramc prnwy prnay ,r,wv

I.FauMm ucmuma *0dPltiK*"l l #_ b I 20 1 t wltt l 13.11 4Ii laura 31 2_ Fufudsi 20 Arabic Free lat*

2 krtm Fan 0. NMI 50 FroK lay w X

1. 1 7 t tM& 44 24 lroomt Itna * e - ., ne @ $ jw-as 19 4 Iuia,lb I Nftuula 4' It 11l4 13 Falfuids 14

boiac EagUa La

Italt

* Oai 15. S fit IS U a t a

i .vo to 4 st s~~~Lu 13 a tI ftbs~~~a 1~0 4 *Ainwc33 4a a a flpa.ya 14 halle S Eugli too

03...'. 2 0 5 Slubib 34 IS * a Liaal 23 41 . L*a 17a am IS 10 . a Kw 12 Is*

911.111 4 Soutli 50 1 LOMLs It1* Yu 14 faqlr d 5 *

pwt_pos Iow . . a ,lielw} ~--! -I -I -I -! -j -4 -!

,,IsIXi :~ i-. : j j,:

lle'l-l~~~~~~~~ - -Ill-!1 I

I ~~~~~~~ . EiTrl s I - I Is > | } [ [ , ~~~~~~~~~~~~~~~~- - !|g!--I - !1Xsi- -I- st n ,-

1~~~~~ -i '1 f1'

U 1 -II ^j js 1i + 1

} 1 ~1 I i 1 [i. , 1 i~ ~3 IJ~f !i zX |P1 -s | - 230 -

Idle 2. W_ df t4*w fEnc (ct^.iI

Purtatdf geltta ~~~w~|l of tu a igeli _iKipd low .. PfI_iul Lm Ln vpe Put Wt Priecpod lupmu tag lapa 'now neue otW, pflOy rimy

20Uln WmO 1t 4 XIlUd 1 5 To 25 4

21daa Is* 4 SO tIloeU a BinlGes a2l

2ru 54 * 1 to 12 a ramh

km 0. Se. t

Iotic Is

I."

K** Is t4slep S S . 42 * Seen It

$W Is

Fruit lae£ usimevt I0 Pstupmo to aa - 231 -

Idle3-2. Ln*n ol autica (cuot.'i

Loo ad m Pct.4 guatlmmdat-ada epatiaqIn" as te owsUa *itlntsu i,e ,k of latp"e _ ~~~Pnmcpli Pett Nli Wicta Us Lw Wa' Purt Total Pnmlpat 1.aguausal tug 11al Ot_11° Ue p pAhWV hla.

RiUleiainc oil tpwtere

a ItFto s * *4*V.U.le I 7 blefJ IJ Irmb Ie a a a

VLsitwt 21 4 km 16 23 KWeI 20 0 In 40 50 lgliish Isa a uais n e 31 25 * 11ula It 43 X a Tsp 11 too It

loagls toa iLam 2 2 st" * * Wi 1025

60~wuan 36 5 Frec~~~~~~2 1s 35 $Ore 15 t * X

[Iela6"si 1416 a *

la X - 232 -

tdl 2. 11ps of hb 4*v #fttc (c4st.'il Lagea NW- Psrasstof pqs)atam lwuktq tenqs. asa thesodim of Inatnctiomtoo Rubeof lampa" -- * Prasel s11e' Sac4 Official ULw Lw vow pgst* Total Prinuld lapa"* i teap Iaqu IaVg frames prisy prisvy prawy *Uls-amemall spwtars iSlagma S1W It 1 NM 0 20* Veuk 20 a Psifaldt t

IJKi a I

lie U

Edits 20

S" 13* 4 Ft"~~1115 Is. a llcssqs,sspisO0 4 Kp 52 P;iiqsnhbK_sk_*_ 2_ a Freedi logi

Shamus 15. t4 Faq 30 20 d~~~~~~~~~~~~ut, w1s 1. 35 a a a Mel~ ~ Stva~ Fu~Otetceclsts aacsalt d a.bytemtmaasfuh aams v**sur ai.R h eluclahsfrohrpicpllsas

tii 30 0 . . - 233 -

Tablo3. lstasrpOWst a prsioStal

Of,pu1tles Pqulatau Ual of V of lwcosta laitt * u tat y truAka at pquatas

_ _..-..--.---.-----. ------14S11)tti A'Oit2a motts 4-73 M913-0419*2900% 14 2990 2600 (allas raot of I t

LLb 2.8 3.l LOor-tac ocomoa * 2.4 Loa 3.0 t 29 sit 419

* 2.5 2.5 2.7 Lrs-aoine" St-Wad a 2.2 2.7 2.t t Jl 34 47

hali 2.4 2.4 2.4 t 9 It 34 20 1.0 2klabaa fb. 2.0 1.e 2.0 1 7 9 32 20t0 1., 30W 2.3 3.0 3.2 t I 10 34 200 .9 *SSli TM 2.1 3.5 2.4 1 I 1 3 2040 IsEfw L.I 2., 3.0 5 & I s0 204 1.9 Chad 1.9 2.1 2.5 5 4 1 22 2040 1.8

a 2.6 2.9 3.1 LUs-ioes stoi a 2.4 2.8 3.1 t 223 274 37

7 ftbalopa 2.* 2.3 2.7 42 4 a 2" 04 1.9 9leais 2.4 3L0 3.2 30 3 47 130 00 2. IJ Wam 2.0 3.1 3.2 1 I it a 204 1.9 10SNOwNSo 1.2 3.9 2.1 I I 2 4 2045 It tWazua 3.2 3.4 3.5 21 2 37 123 20 2.e 22 laad 1.4 2.2 3.0 5 5 7 24 2t 1.9 l3 uqpil 3.4 3.2 3.3 15 to 24 0 2035 2.0 14top 3.0 2.0 3.3 3 4 5 1 203 2.0 lSCatraI *ta OR. 1.4 2.3 2.6 3 3 4 12 20 2. 1bUma 2.4 2.6 3.1 10 22 14 20 1.9 11hut 2., 2.3 3.2 4 5 4 20 205 2.0 to NM" 3.1 3. 3.4 4 7 1t 0 20 2.0 I9Koo" 3.3 4.0 3.9 20 23 3 Ill 2030 2.1 2 hwIv a Low 1.1 2.2 2.4 4 4 5 17 20 1.6 21loss" 1. 2.0 2.1 & I 0 24 205 . 22ama 2.2 3.5 22 is 20 54 2030 1.9 2s Im 3.0 2.9 2.9 21 35 34 201 2 2.9 24 wol 2.4 L.6 2.9 4 6 10 30 m5 1.9 25 1m1 uu1 2.3 2.4 3.0 13 12 21 47 2035 2.9

o 3.3 3.5 3.4 lidsle-l"b oil Iowta a 3.0 3.2 3.3 t 32 0 54

2b Revuita 2.3 2.1 2.7 2 2 3 * 2035 I.3 27 lwaa I* L33 3.2 2 3 4 11 205 2.9 23 lUa. 3.0 3.2 3.4 4 0 11 x 20 1.2 29 Loo 2.1 2.4 2.4 1 2 2 * 2030 1.6 SOCdo d*lres¢* 4.6 4.5 3.7 t0 13 11 44 2035 3.1 3tlimb" 3.4 3.2 3.4 6 20 13 33 2025 2.0 32b .lusa 3.2 3.4 3.3 I 1 2 3 2030 S bt_M 3.3 4.4 3.4 1 1 2 5 20 2.0 343ms'itsu 2L0 1.4 1.5 I I 1 2 2020 2.1

L2.S 2.6 3.3 041l. sl ssrtg s 2.4 3.1 3.3 t 113 144 I"

35bona 2.5 2.L 3.4 94 tie 143 Su 20 2.0 3 Comm 2.4 3.1 3.3 20 12 l1 51 2030 2.9 37Cmp ftk s lip. 2.4 3.1 3.7 2 2 3 9 2 1.9 s sN 0.2 2.5 2.4 1 1 4 20 39lADO 2.1 3L. 2.7 9 1o s 43 204

a 2.L 2.9 3.2 lo-ftwes anca a 2.4 2.3 3l t 40 4 04

o 2.4 2.4 3.0 Frugqb ogatnm * 2.4 3.6 3.2 t hi 2144 29 o 2.1 3.0 3.3 bglh cusstrie a 2.1 3Li 3.2 t 212 247 W7 o 2.4 2.L LI NWr * 2.0 2. Lb t 71 1a IJ

UNSWfet llt aiprlity VA coa, u the tesacal st. -234-

1*1,g4. komI-qa ppuatuu yutb oadpro eetm kOdI op pqaslal

lesap usmadvanO la psimt of pnswvofd smudy praimy tummy ~ of total oekool-apoWlrs lulirns) saiUieI po1*Abm 3966-019706 ausas two4a9m 2090 w tow 2. 196 20* * 2.8 2.9 3.2 a 31.2 31.5 LrN-iissaascsusl a 2.7 2.8 3.1 1 45. 54.4 75.7 54.0 42.4 3.1 a u.S 31.9 * 2.9 2.4 2.0 39.9 30.9 3.auaacMgaal-ud a 2.4 2.4 2.0 1 5.0 5.9 0.3 4.3 5.0 4.7 30.1 33.2 I Hall 2.7 3.0 2.i 1.3 I.5 2.0 1.1 63. 1.4 31.4 33.5 2 otu.a FMs 2.2 2.4 2.1 1.1 1.2 3.5 3.0 ,1.1 3.4 U33. 31.9 3 Now. 3.0 2.1 3.5 3.0 1.3 1.0 0.9 3.1 1.7 30.5 32.7 4 04*a, 1,0 2.5 L.9 3.0 0,3 0.1 0.2 0,1 0.1 0.1 27.1 2n.4 SkdIt 4.7 3.4 3.4 0.9 3,3 1.5 0.5 0.4 0. 25.4 27.2 6 cm$ ~~~~2.31.9 2.4 0.7 0.8 3.2 0.7 0.0 W. 29.0 30.0 o 2.0 3.0 3.3 a 31.3 31.4 Lau~~taurnstOp a ~2.7 2.9 3.1 4 40.4 40.5 47.4 30.5 37. 53.4 A3. 32.2 I EttOqia 2.0 2.9 2.I 4. 7.4 1.1 5.S 6.4 .1 29.3' 29,4 * aIro 2.0 2.1 3.1 5i.1 40 LO0 4.0 4.9 LI7 30.4 29.5 9 isII 2.0 2.9 3.3 1.S 1.1 2.3 Lb4 0.7 3.0 33.0 31.1 IOhuSo Isas 2.0 2.0 3.9 L.i 0.2 0.2 0. 0.J 0.1 2.7 24.0 It Immnia 3.4 3.5 3.1 4.2 L.0 7.5 2.7 3.5 0 341. 33.9 lZbmil - 1.7 2.5 3L0 0.? 0.9 1.3 0.7 L.0 1.2 31.1 33.3 13VP*t 4.3 3.5 3.6 3L0 LI' 5.3 2.0 LS 3.4 33.4 34.2 14tops 3.2 2.7 LS 0.5 0.4 0.9 0.5 . 4.0k 33.1 34.4 IS Ctral mricm Sq. 1.7 3.2 2.4 0.4 0.5 0.7 0.4 0.4 0.4 32.0 32.1 34 iapw 2.I 2.7 3.1 3. 2.0 2.4 1.5 1.0 Lb4 32. 32.2 ? on"e 2.9 2.9 3.4 0.7 0.9 1.2 0.4 LI0 1.1 34.4 35.4 lIItwo 3.5 L.9 3.5 1.4 1.5 2.3 0.7 0.0 3.3 37.2 34.9 19IiFM" 4.3 3.0 4.3 4.5 4.0 0.1 2.7 3.4 LI7 34.4 37.7 20 tlsws Uame 3. 1.3 2.4 0., U. 1.0 0.5 Li7 0.0 33.0 34.3 21hins" 1.4 1.7 2.3 0.9 3.0 1.3 0.7 0.8 1.1 27.2 27.0 22ohms 2.4 2.5 3. Li1 2.4 3.3 1.9 2. 3. 33.1 20.7 n bii 2.2 3.2 3.0 3.5 4.1 5.5 2.7 3.3 4.4 20.0 29.7 20limp) Li 2.5 3.1 1.1 1.2 1.7 1.0 1.2 3.4 31.4 32.4 35 oxhw2.5 4.4 3.0 1.9 2.2 3.1 1.4 1.9 2.a 3. 24.2 a 3.3 3.0 3.5 a 31.9 33.0 *41f-lscom oil aqers a 3.0 3.5 3.4 t 5.9 7.3 10.3 4.4 5.5 0L0 a 31.2 32.0 24 larltlata 2.7 2. 2.0 0.3 0.4 0.5 0.2 0. 0. 31.2 33.1 2? Labkna 2.9 2.3 3.0 0.3 0.4 8.4 0.3 0.3 0.5 27.6 32.4 23 Iowa 3.0 3.4 3.S 1.3 .46 2.2 0.7 0.9 1.3 30.4 31.4 29 Laat2 2.4 2.4 3.0 0.3 0.3 0.4 0.2 0.2 0.3 27.3 29.9 30 Cse E'lvmir 3.3 5. 3.0 I.? 2.1 2.0 3.5 1.9 2.7 31.4 31.9 31 lid". 4.0 3.0 3.7 1.4 2.0 3.0 1.3 3.4 2.4 35. 37.1 UknaMlnd 3.4 5.0 3.6 0.3 0.2 0.2 0.3 0.i 0.2 33.4 32.0 a5 Istsm 3.2 4.2 3.4 0.2 0.3 0. 0.1 0.2 0.2 33.0 33.5 30 sttiu 2.4 .0 0.3 0.1 0.1 0.2 0.2 0.2 0.2 29.0 24.3 a 2.S 2.0 3.S a MI. 34.2 *dik-t s441,sWWtWsu a 2.S 2.I 3LS t .2 24.7 35.0 110.3 25.0 US o 30.3 34.5 a It 4 2.4 2.7 3.5 37.0 20. 29.2 35.2 10.9 27.0 53.3 34.5 54 Cmum 2.0 3. 4.3 3.4 2.3 3.0 1.4 U. 2.0 30.3 34.0 v how tqA el' 2.4L 4.0 4.0 0.3 0.4 0.4 0. LA4 04 3.9 35.3 U Si. 3.0 U. Li 0.3 0.1 0.2 M. 1 0.2 2M. 3.3. 39boo 2.3 2.7 2.9 1.2 1.4 1.9 1.3 1.4 2.1 30.1 30.0 2.0 2.9 3.3 v 31.7 32. ""uweMta * 2. 2.8I Li3 1 71.7 I4 131.9 57. 70.7 90.4 * 31.2 32,0 * 2. 2.7 3.2 a 31.3 31.0 Irmscomi CaLMIr a 2.4 2.7 3.1 t 0.4 24.3 33,4 17.4 21.9 29.3 a 31. 32 a 2.9 3.0 3.1 . v $L.8 53.9 blboomcGeOtrl .2.93.1 LS51t .40.749.431.J U.SU.9 31.4 52.2 a 2.0 LO 2.9 20.9 30. Ottar 62.02.8~aI L 2.9 t 0L912.43.1 9.0 ILI913.7 %04 27.2

VeA Psi wet cowerillty M eweva. am too astoned "MLu - 235 -

Idle S. h_egb m4 ut23akr Opustapcamp Si"

pIwatSm ttr at I9N5 sq. 1t5 1e 1W-41 94 1 I* 2m

* Z4 47 22 t3 tI.5 -30.3 4.4 LS Lrtsa ""wee e 4 47 24 I9 0.4 -24.0 4.5 L.

* 47 40 24 20 I.J -21.2 4.5 6.0 Lwascu 4 4 27 21 2.5 -23. .5 4.6 13 It 27 20 -$.I-2.7 4.5 5.9 2 braa fP 4U 47 24 22 2.2 -14.4 4.5 4.0 3 40w 31 29 2 4.1 -24. 1.0 6.4 4 eggs, TN 47 4 20 22 La -19.9 i.5 U 5 usl 50 09 2 20 *3.4 -23.7 4.0 4.2 6 tw 0 43 24 23 L.7 19.4 5.4 LS

a 4 47 22 to -2.1 -(7.4 146 LS 0 . U 47 23 I7 4.4 -2L0 .5 Lk4

I 44 43 I9 24 -1 2L.3 4.1 5.5 I lase U 4 23 I5 -S.0 *3.3 01 4.9 1idm 54 54 2 22 -. S -17.9 1.4 4.4 to kSu-e u.a U6 U6 0 24 0.9 -33.5 40 5.4 It timamla 4 U 22 13 L6 -54.0 7.0 LI t2 *md4 47 4 24 t9 4.4 -24.0 1.5 .1. l3 'pmi 4 U 19 1t 2 1 L4 1.9 LI7 14Top U 4 23 l *2.0 -$.S h9 tSCatrS MritamP. 34 42 24 37 -3. -32. 5.4 .4 34 eaWX 44 47 2t 3s 4.4 -29.2 1.5 0 17bms 49 4 25 I? 0.6 -29.3 4.5 5. to bob 52 52 1 I9 0.1 L4 0.0 647 U9K M 53 21 Is 9.0 -3.4 1.9 L5 M eaV? Wm49 33 24 3.0 -20.3 4.5 4.0 21$d.. U6 47 30 24 3.3 -12.0 tb 5.4 bam U U 20 34 4.6 -29.5 6.4 4.? 23 dm 47 24 17 -3.4 -2L0 4.4 LS 24 _siI 47 23 9t -2.0 -37.9 44 LS 23ms sla4, 45 2 3o -7.n -S32.2 1.3 5. 4 It 34 -5.4 -2.3 4.4 .0 Itdel.-su elt x"urtes e U4 U 20 34 -*2 -2 6.5 4.0 24 hwatma 44 25 39 3.5 -25.3 1.2 1.9 2 Ubrna U 4 22 37 6.1 -2*2 4.9 LI 2 tolsU 49 20 I5 -2.1 -2. 4.8 Lb Z9 uee 42 41 to II -4.4 -19.7 La8 4.1 Oktediwsu 44 45 22 4 5.4 -3.3 4.5 4.0 JliedU 55 47 17 ii -14.2 -31.0 1.3 4.0 Ubmilf U 4 U21 4 -1.4 -34.4 4.9 L0 St e 53 U It I2 -3L3 -33 17 4.7 so 1m4uj 37 21 0 I -43.5 -21.9 2.1 5

* U 49 23 37 -3.4 -27*20 .9 5.7 mose-sumol " tWu" 41 41 32 4 9.2 -30.3 4.4 L.4 DIem 53 50 23 34 -3. -33 619 5LI 3 Cmum _0 *4 20 14 55 -35 a.? Li V CtOMp Oe 41 45 to 32 9 -31.4 4.2 566 313m 32n 22 tA 39.4 -24.4 4.9 5. 3b . 49 47*t 29 22 1.3 -24. 4.4 LS

* 40 47 23 l7 -2.5 3 25. 4.4 5.5 fca * U 7 23 17t 4.4 -24. 4

q _1~cosutrts a 47 47 22 14 -5.0 -2.e 4.4 5 * U 43 23 22 -LS -. 3 4 .4 5 5i 27 22 *1. -17.9 49 t.9

Ift Pe NU Wu5RtY mECOWp, o ttugludas Oe. - 236 -

lkbaa901t10isi Pc'cuateqOf ~uMo pepulatiom

As a porceeteq AMWeqWWIu In cattu e# cities of of total greut rate In lrqoest ame 50 O 0,000oee 0,0 powaatiom Iperot) ctty sers Wu"m

19651 19946 1965-731913-84 1960 1960 1960 1900 1960 1960

o 12 21 6.2 6.1 it 41 3 36 Lwoin eeocvos' a I 21 6.0 6.1 37 39 0 31, 1 1 14

* I 20 o.2 5.5 . 33 0 * Loiriecoc ats-uld a II 20 5.6 8.7 . 34 0 0 1 0 0

INil 13 1t 5.4 4.5 32 24 0 0 0 0 2Bork Fasoe 6 it 6.5 4.8 . 41 0 0 0 0 3Niqer 7 14 7.0 7.1 . I 0 0 0 0 46aebia,The 14 31 4.2 50.2 . 5Seal I a 20 33 *.4 J.4 . 34 0 0* 0 616A4 9 21 6.9 6.5 39 0 0 0 0

* I2 21 6.3 &.2 31 42 3 42 Loo-ieceaothor a II 2! 5.9 6.1 37 47 0 38 1I 14

IEthiopia I Is 7.4 6.1 so 37 0 37 0 1 hair.o 19 39 5.9 7.1 14 23 14 38 I 2 914JAbs ~ ~~~~S12 0.2 7.3 19 0 0 0 0 10baae-lisses 16 26 4.1 6.8 . . It TAoeosia 6 14 6.1 8.6 34 so 0 500 12 114radi 2 3 1.4 3.3 . 0 0 0 0 is *"a 6 7 8.3 -0.1I 3o 52 0 52 0 I 14top II 23 6.4 6.5 * 0 0 0 0 0 IS Cotratl Afrima 3Ap. 27 45 4.4 4.6 40 36 0 0 0 0 16 U!ASaaCa 12 21 5.3 5.5 44 36 0 36 0 1 I?leoa II is 4.5 5.0 . 3 0 *3 0 I to Roma 3 5 6.0 6.6 . . 0 0 0 0 19Kenya 9 Is 7.3 7.9 40 57 0 57 0 I 20laterraLer 1o5I 24 5.0 3.5 37 47 0 0 0 0 2lbseoa ~~~~~~~~~~1227 5.0 6.2 37 go 0 90 0 I 22Shcam 216 39 4.5 5.3 25 35 0 48 0 2 235Swan Is 21 6.3 5.5 30 31 0 3t 0 1 24beeqal 27 35 4.2 3.8 53 65 0 *5 0 I 2sfezespe5 16 8.2 10.2 75 63 0 83 0 I

* 19 ~~~387.5 6.9 . 38 0 31 NadOlo-uoseoil ia7eters a 14 27 7.6 6.4 37 0 17 t 0 3

Zb9owitueus 7 26 16.0 5.1 . 9 0 0 0 0 27 boru 23 39 5.3 6.0 . I 0 0 0 0 lIllolls 24 48 7.6 6.4 . 35 0 35 0 I 29Leeot 2 13 7.I 20.1 . . 0 0 0 0 SOCoWed vozre 23 46 8.2 6.3 21 34 0 34 0 I I1 Uu6ehv 14 27 6.8 6.1 40 50 0 so 0 1 32 SM1al.a 6 19 5.0 12.9 . .

33htouaa 4 24 19.0 11.3 . , 34 fwataoe 37 56 4.6 3.4 . ,

Is1 31 4. 5.2 Il 19 it 52 lhdie-4locm oil uorter a 16 40 4.7 5.4 26 21 0 21 t 2 10

350iqerx Is 30 4.7 5.2 13 17 22 60 2 9 36 Cuwm ~~~~~1641 7.3 8.2 26 21 0 21 0 1 3l CoeqePavias RVe. 35 56 4.4 5.4 77 56 0 0 0 0 a w=o 21 40 4,0 4.6 . 39 Aoqo1 13 24 5.8 6.6 . . .

o 13 25 6.0 5.9 28 36 0 47 Soli*maras ntca a 13 24 6.0 6.1I 37 37 0 26 3 I 27

ii Is 23 6.0 7.2 30 36 4 31 fraerep4scacaotrio, a 12 25 5.7 5.5 37 5 0 0 I I I a 13 25 5.9 7.5 22 29 1It 4 biaqehooecentrioo e 14 24 6.6 6.3 36 41 0 4 t 2 I? a 9 to 7,0 4.4 37 46 0 33 Other o~~~~~~~~1S 24 6.4 6.6 . 37 0 37 tO0 2

NtotePot date coearashsty SWcoverag, to thetehcfl ateIct00. - 237 -

Tdle 7. Law twce

PwConmtof Pwutaqs of lw f#WC mi

mdttin 0 ------w Pth Vw$" I AtiWtwo attry%mem f Ftt"iUd

1 1904 1945 190 uI 1945 9 aI 1945 3930 19 l333 193-2X0

6 53 StU 4 79 7 3 9 Is 2.2 2.2 2.3 Low-a eon 53 52 9 02 5 7 10 14 2.5 2.3 2.9

* 53 SI t1 is 4 4 S If 2.2 2.0 2.4 witAaw ma-tid a 3 52 90 5 IS 7 ll 2.9 2.1 2.4

I Ads 53 so 90 3 1 2 I 13 2.2 1.9 2.4 2 brwlhsaa 53 52 9 37 3 4 7 9 1.4 1.4 1.7 334e 51 S1 95 91 1 2 4 7 2.1 2.3 3.0 Ohaia, tk 53 55 a 34 5 I 17 9 1. 3.4 2.1 5 1Ial 49 52 of 74 4 3 13 1I 3.a 2.b 2.4 4cmI 55 54 92 33 3 5 5 12 1.8 2.3 2.3

* 52 St 33 78 7 3 10 14 2.2 2.2 L.o Lwia moootkr a 53 31 37 II t I a 12 2.0 2.1 2.3

7 ftbtogp 52 51 4 30 5 3 * 12 2.2 2.2 2.5 Oales 52 51 82 72 9 I3 9 14 3.9 2.3 2.3 93.. 51 4 92 i3 3 7 5 9 2.3 2.5 2.7 IObluw ius 52 89 32 S 4 4 34 Ilt saaa so 9n 34 3 5 e 10 2.4 2.4 3.2 12$VON 53 52 94 93 2 2 4 5 1.2 1.7 2.5 SIj LE 53 49 91 3 3 4 4 30 3.1 2.2 3.2 14toT 52 S0 i3 73 3 to 33 17 3.2 2.0 2.9 lS Cutral4frincmal . 57 55 9 72 t3 3 21 1.1 1.4 2.4 t4I aown 54 50 sao . 3 . * I.9 2.0 2.9 17bfg 52 50 S3 tO 5 7 12 23 2.1 2.0 2.b 1t ib51 51 N9 93 2 3 3 4 2.7 2.3 3.t 19KM" 4 4 4 o3 5 7 9 12 3.3 2.3 3.5 20 Uans Lou 54 14 t9 70 11 14 It 14 3.0 1.0 1.e 21 Ibu 53 3t v7 St * 9 * 10 1.2 1.2 1.3 Z2iWm 52 4 b 1 5S 15 Is 24 24 1.4 1.5 3.5 231No_3 52 82 71 5 7 33 22 2.3 2.4 2.3 24 Stoqu 53 52 is 31 5 4 II Is 1.7 2.2 2.4 2lhUMMb0 55 1 7 35 S 7 7 8 1.3 1-* 2.4

a 53 44 77 52 0 7 Is 21 3.0 2.5 2.9 IiIe- ou "al ortwtr a 52 52 31 70 5 9 I 37 2.3 2.3 2.7

24Uhwmitam 52 53 tO 3 9 7 22 1.9 2.3 2.1 27 Liboi SI 52 79 74 30 9 I It 2.3 3.. 2.5 2Ilsaa St 4 14 73 I 3o0 IS 11 2.3 2.1 t.1 29LObe 5. 53 92 34 3 4 * 1o 1.7 1.3 2.3 30iCa te ie 54 53 31 45 5 3 34 27 4.2 3.9 3.3 31lt_aa 51 45 79 53 a IS* 13 34' 2.7 1.5 3.4 t2 bltng 53 49 e5 14 5 9 1o 17 2.4 2.2 2.7 t3 20tu 50 3 7e0 4 13 7 17 2.2 4.2 2.9 t4aitsua 52 42 37 23 25 24 3 2.8 2.3 2.1

U 53 5s 40 49 32 11 20 20 1.? 2.0 3.2 *deI p.aoolstlotwor a 5 It 79 70 I II I3 20 1.7 1.9 30. Fiwis 51 4 72 40 10 12 1t 20 1.7 2.0 1.1 6 caw=n55 50 N 70 4 3 9 22 3.9 13. 3.0 n coop, _twe't60. 5 51 4 a2 11 t2 23 24 1.9 I.t 2.7 x I" a 5 a 75 3 It 9 14 * 0.2 0.3 2.0 P 3 552 79 74 3 10 Is 17 1.5 2.4 2.7 a Hs au in 7 t 12 16 2.1 L2 2.a 3.I*-" mM "$ a 53 51l 34 7s 5 3 9 ia 2.0 2.2 2.7

* 54 a a 7s a 3 9 15 2.0 2.1 2.1 Jraa tne ntes a 5 52 8 It 4 17 14 1.9 2.0 2.4 * 52 4 74 70 9 10 14 Is 2.1 2.4 S.l q _ cmsttru a 52 4 84 74 5 9 11 17 2.3 2.3 2. o 5I 31 12 80 I I 11 12 2.3 2.2 2.5 Otler a 54 52 34 s0 5 3 3 14 2L0 2.4 2.4

Stat Fo tac4 bWity SW IM * o te twomol Mtet. a. FiWo utt s atonsk e fw yes otkr tam ths eWif cu. - 238 -

tl3e 6. kaml of pramg.a

*vpap ammaWowg rit. 1w t)

3965-1It7"4 A/ '9*73 197-64e/ 196-7 3944 a/ 1965-7397549*, 1965-73am-0 a

a 4.0 3.6 2.6 6.4 . . 4.1 La.loaaeaamcmesa a 3.6 2.6 2.6 1.4 8.3 2.6 7.4 40.3 4.5 3.4 a 0.3 3.5 L-tuon u-wxi 2.0 2.9 3.9 .6 5i 5.i 45 4.5

I ril 3.3 4.40 0.9 5O60 Li 6.6 0 . . 4.1 4.50 2 kt1ooFM 2.4 2.9 * .3 LI5. . . 32 3 Upv .0.5 5.2 . -2.9 3.4 0 35.2 10.90 . . -1.5 5.9 a 4 IubtaTno 4.5 2.0 4.5 * 4.4 . . 4.5 Skelah 3.5 2.5 6Lit 0.5 . , . . . . ,

a 4.3 L.i. . 63 . . Lam-urnsgt~~~~~ 0 3.6- 2.3 2.9 1.2 6.2 2.6 7,l 40.3 5.6 3.0 7Itbtepos 4.3 2.3 ' 2.1 3.2 4 .Li 2.60a L3 3.5 6.Li 3.60 Sfire 3. .3.0 0 . 1.40 *o.00- -5.60 * 3.1 a I lhad 5,7 3.3 . 2.* 3.3 . . 4.0 to ioA4sgs . 2.10 . *3.30 . 1.30 . . 12.10a ItIlmaa L.0 2.60 3.1 , .9 Li67 6.2 13 IV" 4~~~~.63.4 4 4.7 2.30 146.4 6.3 . .0 LS3 t3 lsda 3.6 -1.306 3.6 -0.7 3.6 .0 * . 3.6 L4,4 I4futop3 2.304 2.4 3.30 6.2 Lb6 . .3 3.0o to Caanui 3tIamU 2.1 0.7 2.3 1.1 7.1 1.2 . .4 lMa, w" e 3.5 0.0 . 03 -3.41 O.9 SleW,u 2.2 4.4 * 2.7 * 7.9 . 5.Li lbto 6.3 5.4 . , . . 19if p 7.9 4.4 6.2 3.i 13.4' 4.3 IL;4 6.9 7.6 4.9 20 SANarLow. 3.7 3.0 3.5 2.0 1.9 -2.5 3.3 1.6 7.3 3.7 23kMdu3.8 3.d * .2.4 * 5.7I -2.0 * 2.3 22 ohme 3.4 40.9 4.5 0.2 4.3 4.9 6.5 .6.9 3.3 0.4 23 S" 0.2 5.5 6.3 2.7 3.0 6.4 * 3.1 0.5 7.5 23uland 1.5 2.6 0.2 0. LS5 6.0 I.5 2.3 Uhaia 7.4 *2.20 4.5 -2.4. 33.3 -6.5 3.6 -LOo 9.2 4q40

LI5. LI . . . 3.3 *Idle-ugu oil togws" a 5.5 3.6 6.4 I.; 9.3 0.; 3.3 7.4 3.3

24eWitWla 2.6 3,3 -2. 2. 4.3 0.9 . 7.4 Li1 27ubwta 54 0.20 65S 2.00 6.2 -. 35 S 33.2 0.5*f 3.6 0.304 U11*a 2.9 0.4 2L0 6.9 Li7 4.3 9.3 0.6 2.3 0.6 29Lamas 3.9 5.LO. . . 30 CA*6iv,et Lb6 4.2 . . . . R sk$ai 9.4 3.7 . O.; . . 33 5ma3.d l.a 3.7 L.6 4.9 . 3.3 3.4 * LI7 32.2 3.20a 23Boom& 34.0 30.7 6.4 -4.0 30.2 35.4 L. 10.4 10.3 so LSuo2. 3.6 * -3. 4.4 * 4.3 LS6.

6.U 6. M. 90 * 3. . N*iiI-III Nouaton Lo3 3.5 LS5 4.3 34.5 6.5 33.5 9.9 5.2 L.3

31isMIpt 9.7 1.36 2.o 4.9 0 39.1 0.30 15,6 0. 0L6 4. 60 S6cmwm 4.3 7.20 4.7 1.30 4.7 35.2 7.5 9.9 3.4 7.20 37Cmep.PUs-Ow* 6.3 LI 4.3 0.4 9.5 I2.L? 6.7 6.9 U I"t 7.4 I.5 .1. 394aod Li4 4.7 0. -Li50 3.3 4.0 31.5 4.5 2.;

o 6.4 3.3 Lb6 16. . . 5. MI*wo Melee . . 2.6 3.0 3.2 6.2 .4 6.6 2.3 4.6 3.2

* 6.4 1.3 Lb6 5. 5.Lt frusmcomom, 3mls; a 3.3 3.i 2.3 3.5 6.2 Li5 -2. 3. 3.2 o 7.6 3.3 3.6 * 4.o . . , 6.7 booment$ a 4.6 2. 4.1 3.3 4.4 0.4 9.6 4.3 5. 3.7 * 4.6 -L.4 1.5 4. 6.6 LI5. obw a~~~~3.9 La 2.3 -2.6 31.3 -3.9 6.6 6.6 6.7 3.4

an Foram ao kiatty ad "smop, it. tabtal O"es

a. Pips dtn iuahitWe #O I97545 Mt 39734 L6. " 1,tmamlt. lb e.laatin dIw Of . - 239 -

Tole.9. Cuttal fewmat 4*su*tvre Pwimtap .4 tota asiatvo ?oal

IaetisIm Iboti Ibotq an.ta 5w 04 WIS 'WcINW* of UP)

1m672 a9. i97n 'mm3 1972 19021a 19 191.'s72" 97 193 197m a13. i9m 565s a 1972 1963.1 13.2 10.3 15.1 I.I 5.2 4.5 5.7 5.0 20.9 21.5 39. 42.0 21.0 20.1 *3.9 *4.4 LmnumKom" a0 11.7 7.9 M5. 16.3 5.7 LI 3.3 3.6 31.9 59.2 36. 41,0 19.5 26.6 -2.9 -4.6 a IL?7 58.3 12.6 25.2 6.3 5.9 5.3 11.2, 5.5 11.? 366 2V. 14.0 40.5 -0.5 -0.5 up-mmm'arld * 23.3 14.6 * 12 * 5. * 2.6 * 37.6 IM3. 0.3

5I1 . 79 10.1 . 2.5 * .. * .1 . 47.6 469 . -16.4 2 6WUSoFMo 11. 0 20.7 20.6 I 9. 6.2LI 6.6 6.6 . 6.0 55.50 6.3 37.6a 25. I0. * 3.6 0.3 v. 3 05gw . . . . .

ISamula 23.3 - 5. 51 * 2.;It. 40.5 13.5 . .e A, oi 24.6 . 14.0 4.4 * 1.1 . 21.8 33.7 16. . -. 2

V 53.0 9.6 15.6 11.6 5.1 4.5 5.0 4.i 21.0 25.9 39.5 43.0 21.5 56.6 -4.' 4.2 Los-aim§16w S~~~11.5 7.9 154 16.3 LI7 5. 3.9 3. 22.9 19.2 36.6 45.6 59.6 26.6 *3.9 -4.6

7151so 14.3 * 14.4 I 5.7 4.4 . 22.9 . 16.3 Ml3. * -1.4 I luear 51.1 7.9 a 55.L 1.3 2.3 3.2 0 2.6 0.4 a, 53. W.I* 51 55.4 $6.6 27.5 -7,5 -3.0 9 imam 3.1 6.3 15. 53.4 5.5 6.6 5.I 5.3 37. 1 3.2 36.7 37.1 23.5 32.0 -*.2 -7.7

Sittmauia . 1. 72 * 21 * 3.;P59 223.6 It.; -. 526w.61 50,30 23.40 6.0* * .70 $Lo.9 IL3.I 1139* I) Ugsk 23.1 17.0 1141 12.9 £3 4.6 7. 2Lk 12.4 9.5 36.6 53,4 U1. 4.5 -L.5 1.2 SOIqTop. , 9.6 LI. LI6. I I54 . 41.6. * 34.5 -2.5

59 son" o.6 53.6 21.9 20.6 7.9 7.0 3,9 Li 30.1 24.6 30.2 333 21.0 266 -3.9 -L.1 30 wra5asLo 4.2 . 5.6 6.2 . .5 33.1 45.2 * 21.2 . -sLO4 3213Mm 7.9 6.i2* 20.5 16.7* 6.3 Li80 4.i 6.6. 15.i 59.2. 4. 43 59. 7.5 -5.6 -2.. 23 kwS 24.1 9.5 a 9.3 6.1 5.4 1.3* 5.4 LS30 15.6 23.506 44.1 57.3 * 19.2 WV69 -4.0 -4.6*0 24 1qa5. 97 . 5. 4.7 * 6. * 19.2 * 40.3 17.4 2. .0.8 -4..

o * 54.5 ~~16.5tL5. 7.6 7.5 5.9 7.9 23.4 3L8 43.0 26.5 2t.. 36.2 -51.1 -9.2 618.5,-ams W1 avWmn a 7.5 56.3 56.6 7.7 7.3 I3. 5.3 23.e 25.7 4.35 41.2 23.2 3L., -7.5 -6.1

V7 kn -w 7.9 IL5.; * 7. 2.7 23.6 37.7 . 34.9 -to1.; U0laws. 19.0 ILI2 7.4 6. * 1.3 1.50I ILI7 23. 0 W5 1. * 4.0 45.54 -53.6 .19.0 39 LoosUs. 19.5 17.4 6.0 1.2 6.5 1.3 34.5. 29.4 41.5 44.7 1g.6 27.6 4.9 -2.5

55 Sotoim , .0 5.0 19.4 6.4 S.; 25.7 9.i 3.0 27.4 4.5 31. 33. 44.; -23.5 15.; l4Swsta. 0.64I 0.9 ILI5 1.6 16.5 7.6 15.0 31.1 1339 941. 43.4*0 45.3 WS.3O 9. -51.2 -9.3

ESIgrI. 402 * 45 3.6 LI5. * 9.6 * 3.4 * 14.2 . l camus 9.6 IL13. Li3. LS6 23.0 * 90 1. 1.3

39 boo . , . . , .

* 292 51.4 55.3 56.6 5.5 5. LI3 6 -21.2 2.8 ILI5 37.2 57.6 35.0 -3.4 -4.0 lut6dw Mnce a 1t.7 7.9 15.5 56. 6.6 6.0 3.9 Li7 23. I3.2 3 45.4 19.4 VA. .1 .4.6

* it.? 16.5 54.7 5IL4 LI3 4.0 4.5 7.3 19.4 2.1. 47.5 4. 26J 29. -4.9 -2.3 fteapm Cmk*t a 51.3 6.6 17.9 51.0 Li5 4.2 37 6.5 31. 17.5 33I 41.0 1ILI 27.5 -L3 -3.0 .26.5 I130545. t LO66.5 6.46.2 25.623.46 .5335.4 1A.0 23.3-33 -6.0 kleailw smlrta a 9.9 7.5 a5. I5.7 6.3 6.5 4. 2.4 24.1 24. 36.7 433 19.7 26.5 -50L 6.

Sees FW &Mt emw*5,ity ad oeasp. m tie toibs adiUt,

0. Ropm dota utui we 6#r197 W 592 - 240 -

Tab),tO. fhufiaata 04 @sE4Ia44wlapUt IseztlItea ___~~~~~ ...... -- _...... ------... .__ __----.- _.____ Mt Euawiecats of 0m fromall w a , Pu capita pucEt

1979 19" 1910 19"1 13 I 1914 1934 1914

U 21.7 9.0 Lattacm wucau t 3517 4715 597 5U14 5543 5619 2b.0 10.2

79.9 22.4 Lowiacm us-ard t eT 97 t1i3 1142 1257 IOS 12 . a 34.2 17.3

I WIla 143 193 267 230 210 21 n20 43.6 32.0 2 Sbai Fa iF 1" 212 217 213 184 1t 29.7 19.7 3 11w 157 174 170 II9 259 175 142 26.1 14.0 49a01a, Tb. 3 37 54 45 40 42 5 77.9 $51 S S"oa 212 179 433 374 442 327 3 9.4 ClAd 125t 4 35 0 4 5 115 23.4

* 19.3 5.2 Laaaacoc otir t 264 34 4225 4425 43604 406 4414 . 22.1 6.3

7 Etiopla 140 191 216 21 200 34 34 9.4 7.7 1 Zaire 3II 414 429 394 4 317 314 10.4 10.1 9 a 99 142 143 I 121 117 15 23.2 13.8 30ohaa-9s2 50 53 hO 0 U5be 4 55 U3.S It Taula 424 5 479 702 413 421 55 21.0 14.7 12 krudt 75 95 III 122 127 142 141 30.7 15.0 3IS _ 23 44 114 13 133 137 144 10.9 3.3 14 tos 103 110 91 43 77 112 110 37.3 14.7 15 Central Afnca a". 5 1 84 111 102 90 93 114 45.1 11.8 ihtaiagaw 91 139 230 234 251 1t 156 15.9 7.0 17ih 42 15 91 92 0 07 77 19.7 8.0 1 AMa 125 149 155 154 151 151 It 5 29.2 10.2 19Ka" 20 35 397 449 45 402 431 22.1 7.5 20 e rrts n 40 54 93 41 32 44 41 14.5 4.2 2lau ,a 40 5 90 107 9 9 123 20.9 4.3 22 bal 114 149 193 14 142 11o 214 17.5 4.4 23 a4rn 319 47l s5 49 740 97 414 29.9 6.5 24*9aq 223 307 262 397 23 322 333 52.2 14.S 25 readlqw 105 144 lI9 14 206 211 259 19.3 4.9

a 77.4 4.9 Nddg-iscoaseoil iewrtrs t 919 944 1246 1200 1231 1153 1219 a 37.1 9.4

26 l ritaoia 238 147 174 231 193 172 14 101.5 24.& 27 Uiberia 4 I 9I 109 109 119 133 2.2 1S. 28 7iud 195 277 319 231 30 214 239 7.1 9. 297Lswb so *4 91 101 0 104 97 *5.9 17.4 39 Cotbd'Ivlr, 131 142 210 124 137 157 129 13.0 2.3 31 ZaeaO 9 13 144 212 214 20a 299 3.7 5.9 32 tSo iId 45 50 SO S7 29 34 19 24.5 3.7 33 leta 40 10o 104 97 102 104 103 9.2 11.4 S4Nritaas 44 32 53 5 a 41 3* 31.1 $.5

4.1 0.5 dileuS all meetsn t M3 4*2 Sol 425 45 42 W7 19.0 2.5

35I gW 43 27 4 41 37 a n 0.3 0.0 S6Catm 170 270 2u I" 212 130 1to 19.0 2.5 37 cap, toe' Sa. Ot 91 92 A 93 1o l 53.9 5.3 399mb. 44 37 54 44 42 U4 714 3.1 2.5 391dm 47 47 53 41 t0 74 93 10.9 1.3

17.9 4.1 W _ierasAfrita t 479 6133 7145 7192 7i1 7122 7325 a 2a.2 9.3

a 24.7 . Frmcop. conttao t 23 2917 305 3032 2 2774 2974t 29.5 10.2 a 14.7 2.9 qlebm cantro t 173 27 315 3247 3370 2 3214 .5 9.0 16.1 4.2 044w t 93 415 931 I93 99 1023 1134 19.3 4.9

WbmFor ataau iaOlity.4 carsrp, - tk tu acc tteato. - 241

TableIt. EtatwSpbacdb riert

eisttnaseand dlewso Interco$ayeus Deatm lgsas a pecutqeof:

polart Asa persetap pukecdMt kiwI,of 0*4 lsellis.) 4 of ali duSia.1t.a@4 81 ad uWasn 1970 594 19 t 70 to8 1970 598 5970 o 191490 I98A) a 17.4 54.4 e 1.4 3.5 5.2 LoviKm cc 'go t 3567 29054 a 14.8 59.7 t 40 M9 a .0 2.2 4.2 11.31

* 25.8 * 0.4 3.3 lhtacu s461-ward t 399 338 15.9 41.9 5 2 O5 a . 2.3 3.8 8.0 S Nab 23 940 9.5I 95.9 5.1 0.3 5.7 5,4 8.0 2 F.kiuaFaa 21 407 4.4 42.6 3.1 0.4 2.3 4.2 3 94gm 32 488.7 41.9 1 27 0.4 4.1 3.8

Wegall.1 1a 24.4 . .3 * 25 3. 4 Chad ~~~~~~50 ~~~~325.9 . . .0 * 3.9 5.1

* 5.7 53.4 1.3 3.5 5.4 Lamincas other t 27s8 25627 * 5.2 59.5 t 77 748 a .2 2.0 CS8 13.2 I ithiep& 149 5384 9.5 29.5 4 31 1.2 5.8 11.4 53.5 8 Zaire 351 4084 17.4 532.0 9 250 2.1 15.4 4.4 7.7a 98gm. 122 713 4.3.2 43.5 3 32. 2.1 7,2 7.2

It raaaia 250 2594 I9.5 4004 3012 1.;. 12 hrad 1 334 3.5 35.8 (.1 8 0.3 5.9 ISISgami 138 475 7.3 13.5 4 32 0.4 5.7 2.7 54too 40 459 14.0 100.1 I 37 0.9 10.1 2.9 24.3 lSCutraIAOmana". 24 224 13.5 37.5 I 4 1.4 2.0 4.8 to0 1Sladahaa 93 544 50.8 73. 2 35 0.8 5.2 3.5 57mma 41 58 1640 59.8 Ii 17 0.1 3.9 2.3 Is bairn 2 244 0.9 ILI C 3 0.5 0.4 1.2 3. 59Ka"o 339 2433 20.4 45.8 12 544 1.8 4.1 5.4 21.5 2oiawrvLow. 59 342 54.3 34.7 2 4 2.9 5.4 9.9 9.3 21haus. 312 11id 47.1 59.5 4 21 2.2 5.3 22Sam 495 1122 21.9 22.9 12 24 5.5 5.7 5.0 53.2 23S44a 307 545 15.2 78. 53 45 s.1 s.5 10.4 13.4 246.9505 50 555 51.9 44.9 2 53 0.8 4.1 2Ls

25 ueau . .

a 24.5 67.5 m 2.4 7.1 * 5.4 N4ddl.-ncAm.1l uirtWS t 5352 15755 14.4 54.4 t 51 474 a 1.5 5.0 4.5 52.5

24Nhmitials 27 1573 13.9 575.2 I.) 23 3.1 4.2 3. 12.8 27uibera 159 15 49.9 77.4 4 20 5.5 4.3 * 8.4 21ladia 623 2779 35.7 109.4 24 43 3.4 4.J 5.9 11.3 29Lstbom 8 534 7.7 24.3 5.5 4 0.5 3.8 . 5. SoCobd'lwsro 254 460 58.7 88.1 it 404 2.8 51.3 O.11 20.9 it Zieleb. 233 5444 15.7 28.4 5 139 0.4 5.4 . 2.0 32 kmnlamd I 33Itstma 274 17.9 35.3 1Is 0.7 3.8 La3. 34Ptistil 32 354 54.3 35.3 2 as 1.3 7.5 3.0 14.6

NIddlolnguoulumpttas t 755 In"4 12.s 23. t 27 I35 a .6 4.4 iLi.

35Niger 480 018151 4.8 54.3 20 1572 0.4 4.4 4.2 25.4 i4cuwsi 531 5736 52.1 23.2 4 to? 0.8 3.0 $.5 8.4 3?cing, PWLssfto. 144 1394 539 4.2 3 70 3.3 13.7 to1.5 39dupa

IL55.36.3 a 52 4.4 5.1 5I.7 Im-alweb Mrts 1 5294 55754 15.2 59.5 1 In8 2824a .0 4.1 4.5 s2.8 * 9.2 4.5 s .5 4. 4.3 54.0 Fr.qsm mtn, t 5810 21790 13.5 454 1 40 509 a . 4.7 3,3 8. * 5.7 27.9 s .2 4.0 5.4 25.5 bgiqhm ambles t 323 35357 a 5.8 35.0 t III 1755 a .3 4.1 5.2 13.2 WV t 246 all~~~~~~~~~ t a oft* Fw dat coeWabs5atyad coapurg, us theteemacca mots.

a. figMs atot a astenso wefor 83,rAo 190. XNV7IH SI

3OVd SIHI

Z ti 7 - 243 -

PART C

SUPPLEMENTARYAND SUMMARYTABLES I P;' .e >~N

r*V - 245 -

Table1. SelectedcoWrative statistics for cwntrieswith population uxdr one-halfml I i IonPeMie

Percentgross Progression Percent Percentof Percent primAry ratefrom females Population GOP/capita W devoted adult enrolleant prlmaryto In total (milions)a/ (USdolars) to educatlon lIteray rate secondary enrollment

CameVerde 0.30 $348 10.8, 36.91 - 14.7X 48.0%

Comoros 0.40 S349 5.41b/ 49.0%c/ 103.0% 32.1X 40.0%

Dilboutl 0.40 $480d/ 11.5% 10.0Xe/ 42.0Xf/ 24.9% 39.4%

Equatorialbunea 0.40 S190 - - 78.11 14.41 42.01

SaoTom andPrlncipe 0.10 S708 5.91 b/ - - 23.21 47.0%

Seyctells 0.10 $1,938 5.91 57.71 95.01 Wr 21.91 50.7%

Note: Gneral souwceAfrican Oevelopent Bk*, 1988.See the Tetmical Notesfor oter, spec Ifc soIrces (a/l-). - 246 -

Table2. Snary tabl: e:rol l6ent levels, ratios, andgroath rates, selectedyear8, 19E0-83

1980 1970 1975 1980 1983

Primaryeducatlon

Nwberof students 11,853 20,971 30,117 47,068 51,345 Index(19860 . 100) 100 177 254 397 433 Averagpamnual growth rate from1960 to: 5.5 8.41 7.11 8.6X from1970 to: 7.5X 8.45 7.1X froe 1975to: 9.31 6.B9 froe 1980to: 2.95 Grossenrollment ratlo 365 48X 589 786 75X Secodarye.t ion

Nuberof students 793 2.597 4,284 8,148 11,119 indre(1980 a 100) 100 327 540 1,027 1,402 Averageamnal growth rate from1960 to: 12.65 11.95 12.45 12.23 from1970 to: 10.5 12.15 11.85 from1975 to: 13.75 12.75 from1980 to: X10. GrossenrollQent ratlo 35 75 10C 161 20X

Tertiaryeducation

NMur of studants 21 118 218 337 437 Index(1980 a 100) 100 552 1,029 1,60 2,081 Averageamal growthrate from1980 to: 18.68 18.85 14.95 14.1% from1970 to: 13.35 11.35 10.85 from 1975to: 9.3X 9.2S from 1980to: .915 Grossenrollment ratio 0.2O 0.85 0.85 1.2S 1.45

All levels

Nuber ofstudents 12,687 23,684 34,617 55.551 82,901 Inde(1M0 a 100) 100 187 273 439 497 Averae annualgrowth rate from 1960to: B.5X 8.9 7.7X 7.2X from1970 to: 7.9X 8.5% 7.85 from1975 to: 9.9 7.8X from 1980to: 4.29

Note:Nudefs of students9h thusands.Based on Amm TablesA.1-A.9. - 247 - Table3. Estisatedavera" vws of ediutien attendedby waring Ag p ltion ...... _...... , , ..... _.. Avera wars f idetaon atteded

IU5 I970 I975 3980 393 ...... _...... _~.__...... _...... a9 0.6 1.25 1.64 2.31 2.9I Lauinc ecemien a 0.44 0.79 1.20 1.7t 2.20 ...... ___._...... _._...... _. __...... a 0.11 0.33 0.51 0.80 0.96 Lu-incoa. aaei-vd a 0.11 0.30 0.42 0.75 O."0 ...... _... _._..... I ali 0.11 0.30 0.59 0.94 1.12 2 KarkinaFa 0.09 0.22 0.38 0.56 0.67 3 Niger . . 0.29 0.4 0.64 4 iia . . 5 Somalia 0.17 0.30 0.42 6 Chad . 0.56 0.91 1.26 1.47 __ ._ __ - .. ------_ ._ ...... __...... _.... ------*l 0.94 1.36 1.31 2.56 3.31 Lwinctm other a 0.49 0.I8 1.38 I."0 2.94

7 Ethiopia 0.06 0.17 . * a lair 1.15 1.96 2.E9 3.99 4.71 9 Wa)ai 1.77 2.26 2.69 3.10 3.41 10 kine-isa . . . 11Tanzania 0.31 0.70 1.13 1.56 12hwudi 0.47 0.7S 1.11 1.43 3.53 13Uganda 0.42 0.93 1.43 1.0 2.16 14Tow 0.50 1.08 1.88 3.06 3.85 15Catral Africa Ibp. . 0.75 3.36 2.14 2.62 16Nuagascar 0.41 1.16 2.13 3.2e 3.97 17heni 0.40 0.70 1.22 1.77 2.22 IS baNa 1.49 2.t1 2.67 3.07 3.27 19Kenya 1.19 1.92 2.67 3.57 4.26 20Sierra Len 0.31 0.69 1.17 1.71 21Gaina 0.49 0.10 1.20 1.56 22 6han 5.33 5.96 23 dan 0.55 0.84 1.19 1.70 2.03 24Senl 0.27 0.71 1.20 1.64 1.93 25Robzmique 2.02 2.32

a 1.24 1.64 2.61 3.18 4.50 Niddle-iaconail isptnrs a 1.31 1.81 2.73 3.65 4.27

26 uitania . . 27Liberia 0.51 1.00 1.60 2.30 26t ia . 1.54 2.51 3.65 4.27 29Lasote 2.62 3.68 4.72 5.63 6.17 30Cot. d'Ivire 0.47 0.9 1.47 2.10 31li^de 1.31 2.20 3.06 3.76 32Suailuan . 1.61 2.95 4.24 33Bot . . 2.03 2.64 3.37 34flauritins 5.00 6.11 7.50

a l.99 2.36 2.64 3.03 3.531 zdd'is-inctnoil eonartena 0.67 1.43 2.36 3.06 3.56 ___.------_.------_ -- - .. . __...... _..__ ...... ___ 35 Nigeria 2.40 2.72 2.85 3.08 3.46 36Cawern 0.65 1.43 2.36 3.36 3.97 37conw 0.69 1.67 3.50 5.63 33S1.21 . 1.99 2.93 3.56 39Angola 0.10 0.40 0.88 1.57

* 1.21 1.60 2.06 2.63 3.19 SubaburanAfrica a 0.50 1.04 1.60 2.30 3.32

* 0.65 1.19 1.78 2.52 2.86 Francolonacwsntrie a 0.47 0.87 1.38 2.10 2.42 * 1.83 2.23 2.34 2.75 3.29 Anglophtcamntrin a 1.25 1.85 2.59 3.06 3.43 a 0.42 0.58 0.72 1.57 othef s 0.14 0.35 0.65 1.57

Notesgo the TechnitclNItns. - 248 -

Table 4. Adult literacy ------__------__------Literatesas percent of adult population 1960 1985 or Latest Year ------__------__------Low-incomeeconomies 7 33.8 Low-income semi-arid 2 15.4 ------__-----__------1 Mali 2.0 16.8 2 Burkina Faso 2.0 13.2 3 Niger 1.0 13.9 4 Gambia 6.0 25.1 5 Somalia 2.0 11.6 6 Chad 6.0 25.3 ------__------__------Low-incomeother 10 40.7 ------__------__------7 Ethiopia 1.0 55.2 8 Zaire 31.0 61.2 9 Malawi 41.2 10 Guinea-Bissau 5.0 31.4 11 Tanzania 10.0 12 Burundi 14.0 33.8 13 Usanda 25.0 57.3 14 Togo 10.0 40.7 15 Central African Rep. 7.0 40.2 16 Madagascar 67.5 17 Benin 5.0 25.9 18 Rwanda 16.0 46.6 19 Kenya 20.0 59.2 20 Sierra Leone 7.0 29.3 21 Guinea 7.0 28.3 22 Ghana 27.0 53.2 23 Sudan 13.0 24 Senegal 8.0 28.1 25 Mozambique 8.0 38.0 ------__------__------Middle-incomeoil importers 19 72.2 ------__------__------26 Mauritania 5.0 27 Liberia 9.0 35.0 28 Zambia 29.0 75.7 29 Lesotho 73.6 30 Cote d'Ivoire 5.0 42.7 31 Zimbabwe 39.0 74.0 32 Swaziland 67.9 33 Botswana 70.8 34 Mauritius 61.0 82.8 ------.------__------__------Middle-incomeoil exporters 16 56.2 ------__------__------35 Nigeria 15.0 42.4 36 Cameroon 19.0 56.2 37 Congo 16.0 62.9 j8 Gabon 61.6 39 Angola 41.0 ------,_------___------_------__-----__------Sub-SaharanAfrica 8.5 41.8 Francophonecountries 6.5 40.2 Anglophonecountries 17.5 58.3 Other 3.5 38.0 Note: Values for country groups are medians. See the TechnicalNotes. - 249 -

Table5. Cross-natlonalcomparlsons of achievement In mathematics, reading, and sclence

Indexof achievementa/

Reading General Mathematics comprehension sclence

(a) (b) (c)

Industrialmarket economies (IMEs)

FrancophoneBelglium 0.81 1.00 0.79 Englandand Wales .. 0.96 0.92 Finland 0.76 1.00 0.93 France 0.84 Japan 1.00 .. 1.00 UnitedStates 0.72 0.91 1.00

IMEaverage (country sample size) 0.79(14) 0.94(9) 0.90(11)

Uxkmrmiddl-ln=ome ecnomiles (UMEs)

Chile .. 0.82 0.59 HongKong 0.78 Hhngary 0.89 0.95 0.87 Iran .. 0.53 0.52

UMEaverage (sample size) 0.80(3) 0.77(3) 0.66(3)

Lowermiddle-Inome economies (LMEs)

Nlgerla 0.53 Swaziland 0.50 Thalland 0.67 .. 0.77

MIEaverage (sample size) 0.57(3) *- (0) 0.77(1)

Low-yoanseconomles (LIEs)

India .. 0.72 0.59 Malawi .. 0.46 0.69

LIEaverage (sample size) *- (0) 0.59(2) 0.64 (2)

p/The average o.iber of Items ansmered correctly bythe sample ofstudents Ina particularcountry Is expressed asthe proportion of Itmes answered correctly bythe students Inthe hlghest scoring country. For further explanatlonandscurces, seethe Techtical Notes. - 250 -

table6. Cwtrie gWd b rliy violit retie V

PWIICregrui firoiluwt Pubiic NItUls on guIl ras bhutr f_le amtlcn pr ley sucatlpu Oro" ai gram a as ml ito wditwur as pasap of Driury rates wnetp Prap Pl II- regrsit as W ea total pie rtofr- mrllaw sooal- oft gtr of grimy taI v w lture toUtl ICe rantidesiitwes ratio Suliatlon vioit volint ratlo pw 0ol I tires an eiAian (a) (b) (a) (d) (a) (t) (t) (i)

Loworl WOllstnt a V 11 39 42 40 17.3 3I c4ttrife (LFIB) a P 1.1 1i 37 41 49 11.2 34

ouaila 21 -10.2 . 38 28 46 6.3 so 10*i 23 -2.4 33 37 37 41 37.2 33 NIger 2 40.7 Is 38 23 35 21.7 w3 &rkinafo 27 6.6 17 37 62 33 23.9 31 Ou'tu 38 0.1 29 52 36 75 12.7 31 Outanila 97 6.0 17 3g 4 143 . 23 Ethlopia 38 2.4 12 38 54 26 10.3 48 mm8 38 2.0 . 27 64 lkwundl 45 16.2 14 40 43 a3 . 39 in 49 -0.1 (.) 41 34

Mledumrihy snroli- u 65 16 42 39 54 15.2 43 mtg t rai (1E1 ) a 87 0.4 1S 41 23 55 15.2 45

Su9 a 53 5.4 15 40 41 101 . 43 SierraLo 54 2.4 . 41 34 40 17.6 40 Uh* 57 -0.4 10 43 3 * . *iaul 51 -1.6 15 42 53 13 6.5 31 alon-Iism a2 -1.7 30 as 33 29 11.6 6? 612 -1.3 12 48 54 4 24.0 74 enin 67 2.9 22 31 38 lIa.11g so 6.6 13 a as s8 . 49 Libela 70 -3.4 . 4e 3J S3 13.2 30 IvoryP t 77 1.? 25 41 26 1to 28.2 45 CentralirleoAM. 77 2.7 385 3 8 44 55 Mzalim 79 -4.9 29 43 58 1 "au 79 -2.7 2 44 26 . 15.2 32

Hil prlmry oroll. a gS 14 47 3t 53 10.3 41 Mgcsm ItrlSB(if) a 10 0.0 1s 48 42 29 16.3 48

TariSla -1.6 1 48 42 30 16.3 47 Ylprl 6 -2.4 . . 3t 55 9.3 33 Zaire 90 -0.3 1 43 42 sotas 96 1.5 6 53 31 108 19.5 43 za2ia 10 1.2 1 47 48 39 15.2 48 K"a 1C0 -0. 13 49 37 7t 15.3 es ToP 102 -6.7 31 33 45 23 20. 25 madpo 104 1.3 . 48 . 25 . 56 Cauros' 10t .0 30 48 50 4 172 41 LoW" 110 3.5 23 56 52 31 17.9 37 Soualuw 111 1.6 13 50 33 100 . 47 guitius It2 1.t . 49 23 151 10.3 48 GCb 116 -0.6 32 49 44 . Z3l8 131 11.2 1 48 40 122 17.6 e1 Anpla 134 -7.6 3t 4t 37 . 10.1 Cntr. Fhle*ll w. 103 .0.1 31 48 a . 19.2 48

9*-S s rlnA a 75 14 44 39 52 11.9 41 a 77 0.0 18 42 39 48 15.3 43

NOto: n km TablesA.1. A.?, A.11,A.12. A.U. A.1l, uw A.17.

V UEPC:WMos Prismy rWoilmItratio alos 5 inI193. WIM: ratlo t0t uim 551u1 MLt. : ratio ooseUS. - 251 -

Table?. Ir@lluet ZrtarlglSI SW tit sWlturo w 0 y rollWltts '

FPi Ic Seadry as Grow Ratioof ProPualos Fuels as expnlture wrcent of pOllc ssowry seommy rats fro W Xtof on ediation feare,nt as- Wolitnt tO prisw primyto gOsuYay asa Wct ItUgean ratlo rrollnts onrY urollawts of SIP tajatim (a) (b) (C) (d) (e) (f)

Lows0ry g oulyInlt 5 .0? 18 30 3.0 29 cntrlel; (LSDa) a 5 .10 1S 32 3.3 30

hwIa 2 .m 4 34 3.1 14 Taril 3 .02 1 35 5.8 30 bnkinAFMs 4 .13 22 34 3.2 17 &rindl 4 .GS II 37 3.4 35 I1lal 4 .03 7 29 3.0 is Olad a .1S 1S 15 IIWtlq. 6 .10 40 30 Niger 6 .1S 35 27 3.6 31 Wtil I .2 42 28 3.7 37 8 .OS 13 33 1.3 40

1ba UWVY Wtolwlta t 15 .15 48 3 4.9 31 On&trls (MEC) a 15 .21 4f 33 4.0 30

Ethipala 11 .23 83 36 3.1 29 laulss 11 .13 18 19 3.0 1S Npla 12 .11 50 33 4.7 uritanml 12 .2 3S 24 . 32 serpi 12 .21 29 33 . 2 s:MalIa 14 .2 50 34 1.4 25 ilasn 15 .34 as 28 4.0 3? kpscar 1S .15 . 44 2.3 36 SlerraLon 15 .22 73 28 3.5 31 Canl MrloanP. 16 .20 38 29 . 15 Zla 17 .10 21 3S 5.5 34 Gaala, lhe 19 .24 44 31 5.9 28 Kg" 19 .12 35 40 4.8 1S Luat is .10 45 S0 3.9 35 -2n s19 .32 53 41 oat a1volfe 20 .23 30 29 9.1 40

Hlig aamr yS"ofl t a 30 .2 d 57 33 4.4 49 omaltres O1Ce) a 24 .23 54 39 5.3 35

Liberia 21 .23 a3 29 5.3 is Cerowi 21 .18 28 38 3.6 35 Eotsue 21 .13 31 54 7.2 29 Usnin 22 .29 40 28 6Fn 23 .20 27 40 Ulgaris 28 .24 . . 4.3 56 TOP . 24 .22 31 25 5.9 26 6tna 38 .52 . 37 2.0 41 Zt 39 .20 74 40 7.6 24 SozIlIm 43 .22 f6 49 4.7 35. lWritlwS 51 .51 54 47 4.3 36 Zaire 52 .48 71 26 CaDo,FPlo's b. 7 .50 73 41 6.0 25

hr-SeumnAfrla lm 20 .22 43 34 4.1 42 16 .20 40 33 3.9 31

Mt: Basedon MeM TablIs A.1. A.2. A.$. A.12. A.14,at A.l1f.

I/ lSECS:gross sgodary envolita ratio low 10 In 183.,vSECM: ratio bstu IS and20. t6ECs:ratIo aos'208. - 252 -

Table8. Uniltcost Indicatorsby seondaryerrol ent groups

Variable Secondaryenrollment groups a/ 1970 1975 1980 1983

PbliIcSxpenditure Persecondary student (In comtant 1983 $)

LSECs 452 348 207 256 MSECs 250 255 186 163 HSECs 397 230 204 225

Sub-SaharanAfrica 382 308 195 223

PUblicexpenditure persecondary student (as S ofGNP per capita)

LSECs 249 207 122 115 ISECs 88 78 64 80 HSECs 78 38 35 37

Sub-SaharanAfrica 111 93 62 62

PubiIcexpenditvre Per secondary student (as multipIe of exprenditureper primaryptpii) LSECs 10.3 8.1 11.1 9.6 MSECs 3.5 5.2 3.2 3.1 HSECs 4.8 4.8 4.0 3.8

Sub-SaharanAfrica 5.6 5.6 3.8 4.1

Secondarystudent-teacher ratio

LSECs 17 19 23 19 MSECs 21 21 22 24 HSECs 22 22 21 23

Sub-SaharanAfrica 21 21 22 23

Secondaryrepetition rate (X)

LSECs 9 12 10 12 USECs 11 13 10 1B HSECs 11 11 19 17

9ub-SaharanAfrica 10 12 10 18

Note:Figures shown are medians. Based on Amex Tables A.11, A.12, A.17. andA.18

Al LSECS:gross secondary enrol lunt ratio belo, 10QIn 1983.MSECs: ratio between101 and 20S. HSECs:ratio above2Q. - 253 -

Table9. Percentageof nationals aog teachingstaff In post-rlmary eduatlon

Cowntry YW Secodary Hiwr

Benin 1979 97.3 66.7 Botea 1976 25.5 21.4 8wuIni 1978 37.2 CentralAfrican PRepblIc 1978 . 34.2 EquitorlalGulnea 1978 69.0 Gabon 1977 16.4 16.5 Gabla 1979 75.9 85.7 GuineaBissau 1978 83.4 ivoryCat 1978 47.0 51.2 KeWa 1978 80.0 Liberia 1978 77.0 Malawi 1977 71.4 Mali 1978 97.1 40.5 Mauritania 1977 39.7 Niger 1977 40.0 PWanda 1979 73.9 52.3 Sepal 1978 73.7 49.0 Szi land 1980 74.7 Towa 1979 92.0 77.0

Mea - 65.1 49.5 Median 74.3 50.1

Source:EUNESCD, 1982J. en H

Q ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~t.

179 I z+r 0 - 255 -

TECHNICALNOTES

The technicalnotes providegeneral informationfor groupsof tables, furtherinformation on specifictables within the group, and countryspecific notes on individualdata items. When possible,definitions for the technicalnotes have been taken from officialpublications, such as The World DevelopmentReport, published annually by the World Bank and the UNESCO StatisticalYearbook.

All monetarygrowth rates are shown in constantprices. In Annex A growth rates have been computedusing the compoundgrowth rate equation,

Xt = Xo(l+r)t' where X is the variable,t is time, and r is the growth rate.

Growth rates in annex B have been computedusing the least squaresmethod. The least squaresgrowth rate is estimatedby regressing the annualvalues of the variablein the relevantperiod using the logarithmicform,

log Xt = a + bt + et,

where Xt is the value of the variableX in year t, a is the intercept,b is the slope coefficient,and et is the error term. The growth rate is equal to (antilogb)-1. This is equivalentto the logarithmic transformationof the compoundgrowth rate equation.

PART A. EDUCATIONINDICATORS

Tables 1, 2, 4, 6. Enrollment

Enrollmentincludes students of all ages enrolledin both public and privateschools, unless otherwisenoted. Enrollmentgrowth rates are calculatedusing the compoundgrowth equationdescribed in the introductorynarrative to the technicalnotes. The percentageof females enrolledat each level is computedas the number of femalesenrolled dividedby total enrollment;the means are weightedby total enrollment.

Mean enrollmentgrowth rates are computedfrom group enrollment totals of those countrieswith data for the first and last years of the growth period. Means for the percentageof femalesare computedfrom group totals of femaleand total enrollment.

In order to obtain a close approximationof enrollmentin 1983, only data for the years 1982-1984are reported. In cases where data for these years were unavailable,the least squaresregression Line was fitted to enrollmentdata from 1970 to obtain enrollmentestimates. These estimatesshould not be interpretedas precisequalitative indications of enrollment. They are includedonly to *btainmore valid aggregate measures for 1983 than could be obtaineain their absence. - 256 -

Other notes: Table 1. Primaryenrollment Liberia: Data includepre-primary enrollment.

Table 2. Secondaryenrollment Botswana (1970): Data refer to governmentmaintained and aided schools.

Table 4. Tertiaryenrollment Cote d'Ivoire (1970): Data refer to institutionsunder the Ministryof Educationonly. Nigeria (1970): Data at the tertiarylevel includeteacher education. Swaziland: Data on studentsenrolled include students studying abroad.

Table 3. Distributionof secondaryenrollment by type of education

The term general refersto educationin secondaryschools which requireat least four years of primarypreparation and which do not aim to preparestudents directly for a specifictrade or occupation. Such schoolsmay be calledhigh schools,middle schools,lyceums, gymnasiums, etc., and offer coursesof study whose completionis a minimum condition for universityadmission. Schoolsproviding both academicand vocational trainingare also classifiedas generalsecondary. Teachertraining refersto educationin secondaryschools that train studentsfor the teachingprofession. Technicaland vocationaleducation covers education providedin secondaryschools that preparestudents for a trade or occupationother than teaching. Such schoolsvary greatlyas to the type and durationof training.

Mean percentagesare weightedby total secondaryenrollment. The data shown are for the years 1981 to 1983 with the followingexceptions: Nigeria and Chad (1984),and Mauritaniaand Swaziland(1980).

Other notes: Ghana (1970): Middle schools are classifiedat the primary level. (1980): Commercialschools are includedin generaleducation rather than vocational/technical.For teachertraining, data are not comparableto previousyears due to reorganization. Uganda: Data refer to governmentmaintained and aided schoolsonly. Tanzania (1970): Vocatic al is includedwith generaleducation; (1983): Generaleducation data includeonly part of vocational education.

Table 5. Distributionof tertiaryenrollment by field of study

Definitionsof levelsand the compositionof fieldsof study follow the InternationalStandard Classification of Education(ISCED) system. ISCED defines20 fields of study in higher educationwhich here have been reducedto two broad areas and five fieldsof study. The broad field Arts includeseducation, the social sciences,and commerceand business. Educationincludes education science and teachertraining; the social sciencesinclude fine and appliedarts, humanities,religion and theology,social and behavioralsciences, law, home economics,mass communicationand documentation,and servicetrades. - 257 -

Natural sciences,medical sciences,agriculture, mathematics and engineeringmake up the broad Science field. Medical sciencesincludes health and hygiene. Agricultureincludes forestry and fisheries. Includedwith mathematicsand engineeringare: computerscienzes, architectureand town planning,transport and communications,trade, craft and industrialprograms.

The other categorycovers other or non-specifiedfields of study.

The means for the distributionof enrollmentby field of study are weightedby total tertiaryenrollment; those for the percentof all femalesenrolled are weightedby tertiaryfemale enrollment.

Due to the unavailabilityof data prior to 1980, only the most recentyear for which data were availableis shown. Figures for Nigeria, Cameroonand Malawi are for 1980; for all other countries,data are for either 1982 or 1983.

Other notes: BurkinaFasos Educationis includedwith social sciences. Burundi: Social and behavioralsciences are includedwith commercialand businessadministration. Cameroon: Data refer to universitiesand degree grantinginstitutions only. CentralAfrican Republic: Educationincludes humanities, religion and theology. Congo: Commercialand businessadministration, mass communicationand documentationare includedin social sciences,mathematics and computersciences are in naturalsciences. Guinea: Law is includedin commercialand businessadministration; humanities,religion, theology, social and behavioralsciences, math and computersciences art includedin education. Kenya: Humanities,religion, theology, other and non-specifiedsubjects are includedin education. Malawi: Data refer to universitiesand degree-grantinginstitutions only. Rwanda: Commerceand businessadministration are includedin social sciences. Swaziland: Data refer to universitiesand equivalentdegree granting institutions;math and computersciences are includedin natural sciences. Togo: Mass communicationand documentationare includedin social sciences;humanities, religion, theology, fine and appliedarts, in education. Zambia:Humanities, religion and theologyare includedin socialsciences; architecture,town planningand trades,crafts and industrialprograms are with engineering. Zimbabwe: Data refer to universitiesand equivalentdegree granting institutions.Commercial and businessadministration are included with social sciences,math and computersciences with naturalsciences. - 258 -

Tables 7, 8 and 9. Enrollmentratios

The data on number enrolledrefer to estimatesof total,male, and femaleenrollment of studentsof all ages. They are expressedas percentagesof the total,male, and femalepopulations of schoolage to give gross enrollmentratios. Thus, the gross primaryenrollment ratio describesthe capacityof a school system relativeto the size of the officialschool age population. For example,a ratio of 100% indicates that the numberof childrenactually enrolled, including those outsidethe officialage range, is equivalentto the size of the officalprimary school-agepopulation. It does not mean that all childrenof official primaryschool-age are actuallyenrolled. When misinterpretedin this way, the ratio overstatesthe actualenrollment picture in those countries where a sizeableproportion of studentsa underageor overageowing to early or delayedentry and/or to repetition.

While many countriesconsider primary school age to be 6-11 years and secondaryschool age, 12-18 years, othersdo not. The differencesin nationalsystems of educationand durationsof schoolingare reflectedin the primary and secondaryratios while for tertiaryeducation, the total enrollmentof all ages is dividedby the populationaged 20-24.

The means are weightedby the school-agepopulations.

The enrollmentdata used to computethe gross enrollmentratios appear in Tables 1, 2 and 4. School-agepopulation estimates are from the U.N. PopulationDivision and World Bank data files.

Table 10. Teachersand schools

In general,data in this table cover both public and private schools. Data on teachersrefer to both full- and part-timeteachers and excludeother instructionalpersonnel without teaching functions. Because the proportionof part-timeteachers varies greatlyfrom one countryto another,data comparabilitymay be affected,particularly the pupil-teacherratios in Table 11. The means for the averagenumber of studentsper school are weightedby the number of schools.

Figuresmarked with an asteriskare for 1981 or 1982, excluding Chad, Zimbabweand Nigeriawhose figuresare for 1984. For the following countriessecondary data refer to secondarygeneral education only: Mali, Ethiopia,Burkina Faso, Togo, Mauritania,Swaziland, Mauritius and Zambia.

Other notes: BurkinaFaso (1983): Data on primaryteachers refer to public education. Tanzania: Primaryschool data refer to governmentmaintained and aided schools. Kenya (1983): Secondaryteachers includes tertiary teacher training. Malawi: Secondaryteachers includes tertiary teacher training. Senegal (1982): Data on secondaryteachers refer to publiceducation. Botswana(1970): Data on secondaryeducation refer to government maintainedand aided schools. - 259 -

Table 11. Student-teacher ratios

The student-teacher ratios are computed as the number of students enrolled divided by the number of teachers, both full-and part-time. The means are computed from group totals of enrollment and teachers. The years of data coverage are the same as those for tables 1, 2, 4 and 10.

Other notes: Chads Data on primary education exclude Moslem private education. Uganda: Data for primary and secondary education refer to government maintained and aided schools.

Table 12. Student flow and efficiency indicators

Repeaters as a percenta2e of enrollment is computed by dividing the total number of repeating student3 by the number of students enrolled for both public and private institutions. In cases where data on repeatersin private institutions was not available, UNESCOestimates of the number of repeaters in these institutions were used. In some cases, the estimates reflect the proportion of repeaters in private schools for previous years. When no information was available, data on repeaters were adjusted 4.;2 reflect the incidence of private enrollment. In no case, however, did the missing data constitute more than 5% of total enrollment.

The survival rate is an indicator of the holding power of schools and represents the proportion of entering students who can be expected to reach the final year of primary school given prevailing drop-out rates. Reconstructed Cohort analysis was used to derive the rates. On the basis of promotion, repetition and drop-out rates, the method simulates the progression of an entering cohort (n - 1,000) through the primary school cycle, deriving for each grade and year the nAmberof pupils who are promoted, repeat the same grade the following year, or dropout.

Cost per primary school completer is an efficiency index defined as the ratio of the totdl number of student years invested in each primary school completer to the number of years in the school cycle given prevailing promotion, repetition and dropout rates. The index value shows the average cost per completer relative to the prescribed cost and was derived by means of cohort analysis (see paragraph above). For example, an index of 1.5 indicates that the cost to the system per graduate was 50% greater than it would have been if there were no repeaters or school leavers.

Progression from the last grade of pzimary school to the first grade of secondary school indicates the proportion of pupils enrolled in the final grade of primary school who enroll in the first grade of secondary school in general education the following year. The rate is derived by subtracting the number of repeaters in grade one of secondary general education in year t from the total enrollment in year t+l and dividing this figure by the enrollment in the final trade of primary school in year t. - 260 -

The survival,cost and progressionfigures shown are for the most recent year for which promotion,repetition and drop-outrates were available. In general,the data cover the years 1980-1983.

The means for percentrepeaters, cost per completerand progressionto secondaryschool are weightedby total enrollment,the number of years in the school cycle and total enrollmentin the last grade of primaryschool, respectively.

Table 13. Percentageof studentsenrolled in privateschools

The number of studentsenrolled in private schoolsis expressed as a percentageof all studentsenrolled in both public and private institutionsfor primaryand secondaryeducation. Care shouldbe taken when interpretingthese figuressince some countriesclassify aided schoolsas privatewhile others do not.

The means are weightedby primaryor secondaryenrollment. All data shown are for the most recentyear, generallybetween 1980 and 1983.

Table 14. Total public expenditureon education

Total public expenditureon educationincludes both capitaland recurrentexpenditure at every levelof administrationaccording to the constitutionof the country,i.e., centralor federalgovernment, state governments,provincial or regionalauthorities and local authorities, unless otherwiseindicated. Recurrentexpenditure refers to expenditure on administration,teachers' emoluments and supportingteaching staff, schoolbooks and other teachingmaterials, scholarships, welfare services and maintenanceof school buildings. Capitalexpenditure covers expenditureon land,buildings, construction, equipment, etc. This item al-a includesloan transactions.

Total public expenditureon educationis expressedin constant 1983 U.S. dollars. Local currencieswere convertedby applyingthe GDP deflatorand 1983 officialexchange rate. Educationexpenditure expressed as a percentageof Gross NationalProduct (GNP)and total government expenditurewere computedwith constantdollar GNP and total government expenditure. GNP data are from the World Bank, total government expendituredata are from UNESCO,supplemented with informationfrom World Bank data files.

The weighted total expendituregrowth rates reflectthe real growthof total educationexpenditure for those aountrieswith data for both the first and last years of the growthperiod. The growth rate was computedby summingacross all countrieswith data for both years and applyingthe compoundgrowth formula. The means for educationexpenditure as a 2ercent of GNP and of total goversmentexpenditure are weightedby cotnstantdollar GNP and total governmentexpenditure respectively.

All 1983 figureswith an asteriskare for 1982 with the exception of Nigeria (1984)&ad the notableexceptions of those for Gambia,Burundi, Swazilandand those for the Congo which are 1981 figures. Total education - 261 - expenditurefor these four countriesare shown but are not includedin the calculationof growth rates for 1980-1983. All 1980 figureswith asterisksare for 1979, excludingthose for Nigeriaand Angola which are for 1981.

Other notes: Angola: Data refer to expenditureof the Ministryof Educationonly. Ethiopia (1982): Data includeforeign aid. Kenya (1983): Data refer to the Ministryof Basic and Higher Education.

Madagascar(1970): Data includeforeign aid for tertiaryeducation; from 1980-1983,expenditure on tertiaryeducation is not included. Somalia (1980,1983):Data refer to the Ministryof Educationonly. Tanzania (1983): Data refer to the Ministryof Educationonly.

Table 15. Distributionof total public expenditureon educationby component(recurrent or capital)

For the categoriesof expenditureincluded under recurrentand capital,the methodologyfor computinggrowth rates and the years and sourcesof data, consultthe notes to Table 14.

Table 16. Public recurrentexpenditure on educationby level of education

Public recurrentexpenditures on primary,secondary and tertiary educationare expressedas percentagesof total public recurrenteducation expenditure. The non-specifiedcategory is treatedas A residualand includesexpenditure on other types of education(e.g., adult education) and all expenditurewhich could not be attributedto any of the three levels of formaleducation. The lattermay includeexpenditure on administrationfor which there is no breakdownby level of education.

The means for primary,secondary and tertiaryeducation are weightedby recurrentexpenditure in constant1983 dollars;means for the non-specifiedexpenditure is a residual,i.e., 100 minus the sum of the percentagesfor primary,secondary, and tertiaryeducation.

Other notes: Gambia: Except for 1980, only teachers'emoluments are distributedby level of education. Mali: Scholarshipsand allocationsfor study abroadfor all levelsof educationare includedwith higher education. Swaziland: Data refer to expenditureof the Ministryof Educationonly. Uganda: Data refer to expenditureof the Ministryof Educationonly.

Tables 17, 18 and 19. Public recurrentexpenditure per primary,secondary and tertiarystudent

Recurrentpublic expenditureper primaryand secondarystudent is expressedin 1983 constantdollars and as a percentageof constantdollar incomeper capita;tertiary per studentexpendIture is e::pressedas a multipleof constantdollar GNP per capita. Per studentexpenditures on - 262 -

teachingmaterials are expressedin currentdollars. All means are weightedby total enrollmentat the respectivelevel of education.

The 1983 per studentexpenditure figures with an asteriskare for 1982,with the exceptionof Nigeriawhich is for 1984, and the Gambia,Burundi, Swaziland and the Congowhich are for 1981. The 1980 figureswith an asteriskare for 1979, excludingNigeria and Angola which are for 1981. Per studentexpenditure on materialswas computedusing data for the most recentyear for which they were available. See notes to Tables 20 and 21.

Tables 20, 21, and 22. Distributionof public recurrentexpenditure by purpose

In these tables,public recurrentexpenditure on administration, teachers'emoluments, teaching materials, scholarships, and welfare servicesare expressedas percentagesof total recurrentexpenditure on primary,secondary and tertiaryeducation.

Administrationincludes emoluments of administrativestaff and other expenditureof centraland local administrations.Teachers emolumentsrefers to salariesand additionalbenefits paid to teachersand to other auxiliaryteaching staff. Teachingmaterials covers expenditures directlyrelated to instructionalactivities such as the purchaseof textbooksand other scholasticsupp'.ies. Scholarships includes all forms of financialaid grantedto studentsfor studiesin the countryor abroad. Welfare servicesrefers to expenditureon bvardingschool meals, transport,medical services,etc. Expenditureswhich cannotbe classified in one of these categoriesand other expenditureattached to the operation and maintenanceof buildingsand equipment.arecaptured in the amountnot distributed.

The means are weightedby total public recurrentexpenditure for the respectivelevel of educationin constant1983 dollars.

Figuresfor Mali, Somalia,Guinea-Bissau, and Zambia are for 1982. Figures for Ethiopia,Burundi, Togo, Ghana,Senegal, Swaziland and the Congo are for 1981. Tanzania's figure is for 1979.

Other notes: Burundi: Data refer to expenditureof the Ministryof Educationonly. Senegal: Data refer to expenditureof the Ministriesof Primary and SecondaryEducation only. Swaziland: Data refer to expenditureof the Ministryof Educationonly.

Table 23. Primaryand secondaryteachers' average salaries

The teachersalary figuresshown are expressedin currentUS dollars convertedat officialexchange rates. Cautionshould be exercised when interpretingthe averageteachers' salaries. The figureswere computedby applyingthe percentof public recurrenteducation expenditure allocatedto teachers' emolumentsshown in Tables 20 and 21 to the recurrentpvblic educationexpenditure on primary and secondaryeducation - 263 - for the same year, and dividingby the numberof teachersshown in Table 10. Data for the lattermay not, in every case, pertainto the same year for which expendituredata are reported.

The means for averageteachers' salaries are weightedby the total number of teachers.

Table 24, 25, 26, and 27. Externalaid to African educationand training

Externalaid to educationand trainingconsists of loans and grants,made on concessionaland non-concessionalfinancial terms by all bilataral,multilateral and privatesources. Bilateraldonors not listed but includedin the other categoryare Canada,Denmark, the Netherlands, Switzerland,and all other OECD and OPEC memberswhose aid representsless than 3% of the total aid to educationand training. Includedin the other categoryof multilateraldonors are UNESCO,WHO and other donorswhose aid representsless than 1% of total educationaid. Educationaid from the USSR is capturedin the tocal for the East-Europeannon-market economies,that from privatesources is capturedin the total for iron-governmentorganizations.

The value of educationaid is the averageannual aid disbursedor committedbetween 1981 and 1983, convertedfrom local currenciesby applyingsingle year officialexchange rates. Aid from the International DevelopmentAssociation, IBRD, the AfricanDevelopment Bank and African DevelopmentFund are averagecommitments. From all other donors,average net disbursementsare shown.

The breakdownof direct educationaid between capitaland recurrentexpenditures is inconsistentbetween Tables 25 and 27 owing to differencesin the sourcesfor the two tables and in the way these sources treat technicalassistance. Technicalassistance that is used for capacity-buildingpurposes is classifiedby some sourcesas an investment item ratherthan as recurrentexpenditures.

Educationsector aid refersto the value of capitaland non-capitalaid disbursedto centreldepartments of education. Capital aid representsthe value of externalassistance for constructionand equipment. Non-capitalaid includesthe value of overseasfellowships, technicalassistance, and financialsupport for education'soperating costs, the purchaseof suppliesand for miscellaneousexpenses. Fellowshipsinclude external assistance for tuition,fees and living expensesto supportAfrican students studying abroad. Aid for technical assistancerefers to the cost of foreignnationals working for the recipientcountry,

Aid for prolectrelated training (PRT) reflectsthe estimated value of externalfinancial assistance for trainingin sectorsother than education. Becausedata on assistancefor project relatedtraining is only availablefor the World Bank (InternationalDevelopment Association and IBRD), all other figuresare estimates. The estimatesare calculated by applying the ratio of PRT to total aid observedfor the World Bank to the total aid of the other donors. - 264 -

The cost of hostingAfrican studentsabroad is an estimateof the educationalcosts of all African studentsstudying in foreigncolleges and universitiesthat are not coveredby tuitionand fees. In countrieswhich chargeno fees, the host countrysubsidization of Africanstudents representsfull teachingcosts. In the U.S., tuitionand fees cover approximatelytwo-thirds of an estimated ,900O annualper student cost. The level of subsidizationestimated for the U.S. is also applied to estimatethe cost of hostingAfrican studentsin the U.K.

The data on externalaid to the educationsector from OECD and OPEC countriesare drawn from annual reportsprepared in each countryby the local UNDP office on externalaid, and from financialreports of USAID,the World Bank, the AfricanDevelopment Bank and the African DevelopmentFund. A more completedescription of the data and sourcesis available[Millot, Orivel and Rasera, 1986].

Table 28. Externalpublic debt of the educationsector

The data on debt in this and other debt tables in Annex B are from the World Bank Debtor ReportingSystem. The dollar figuresshown are in currentUS dollarsconverted at officialexchange rates.

Outstandingand disbursedexternal public debt represents.he amountof public and publiclyguaranteed loans that has been disbursedto the educationsector, net of repaymentsof principaland write-offsat year-end. Public loans are externalobligations of public debtors, includingthe nationalgovernment, its agencies,and autonomouspublic bodies. Publiclyguaranteed loans are externalobligations of private debtorsthat are guaranteedfor a repaymentby a public entity. The data do not cover nonguaranteedprivate debt due to the unavailabilityof countrylevel data.

Debt serviceis the sum of interestpayments and repaymentsof principalon externalpublic and publiclyguaranteed debt.

The data in this table shouldbe used with caution. Sufficient informationis not alwaysavailable to enable accuratesectoral allocationsto b% made. Frequently,debt can only be assignedto an unallocatedsector. All values shouldthus be consideredas lower bounds.

PART B: ECONOMICAND SOCIAL INDICATORS

Table 1. Basic indicators

The estimatesof populationfor mid-1984are based on data from the U.N. PopulationDivision and World Bank sources. Refugeesnot permanentlysettled in the countryof asylum are generallyconsidered to be part of the populationof their countryof origin. Data on area are from the FAO ProductionYearbook, 1984.

Gross NationalProduct (GNP)measures the total domesticand foreignoutput claimedby residentsexcluding deductions for depreciation. The 1984 GNP and GNP per capita figuresare calculatedaccording to the - 265 -

World Bank Atlas method. To smooth the impact of fluctuations in prices and exchangerates, a conversionfactor computed as the averageof the actual and deflatedexchange rates for the base period 1982-1984are appliedto 1984 GNP convertedat currentpurchaser values. The resulting GNP in US dollars is then dividedby the mid-yearpopulation to derivethe 1984 per capitaGNP.

The averageannual rate of inflationis the growthrate of the gross domesticproduct (GDP) implicitdeflator and shows annualprice movementsfor all goods and servicesproduced in an economy. The GDP deflatoris first calculatedby dividingthe value of GDP at current purchaservalues by the value of GDP at constantpurchaser values, in nationalcurrencies, for each year of the period. The least squares method is then used to calculatethe growth rate of the GDP deflatorfor the period.

Life expectancyat birth indicatesthe number of years a newborn child would be expectedto live assumingthat mortalitypatterns prevailingat the time of his/herbirth were to remainconstant throughout his/her life. Data are from the I.N. PopulationDivision, supplemented by World Bank estimates.

The means for GN? per capitaand life expectancyare weightedby population. Those for averageannual rates of inflationare weightedby the share of countryGDP valued in currentU.S. dollars for the entire period.

Table 2. Languages of Sub-Saharan Africa

The number of languages refers to the estimated number of indigenouslanguages spoken in each country. The figuresshown are not exact. Informationprovided is sufficientonly to distinguishbetween linguisticallyheterogeneous and homogeneouscountries.

Principallanguages are indigenouslanguages spoken by at least 10 percentof a country'spopulation and any other language(s)serving as an officiallanguage, lingua franca,or medium of instructionin the country'ssystem of education. Severalnames are sometimesused for the same language. Equivalentlanguage names are listedat the end of the notes for this table.

The percentageof populationspeaking a languageis broken down into the percentageswho speak it as a mother tongueand as a second language. The total percentageof the populationestimated to speak a languagecan be ascertainedby summingthese two percentages.

The officiallanguate(s) is the languagerecognized and promoted for use in officialdomains such as law courts,national speeches, public documents,and is often the languageof instruction. The percentageof speakersof the officiallanguage is only roughlyestimated, when estimatedat all. This percentageis usuallydirectly related to the percentage who complete several years of formal education. A lingua franca is a common language used as a means of communication between groups of people whose mother tongues differ. - 266 -

A languageis definedas a medium of instructionwhen it is used to teach a varietyof subjectsin the curriculum. A languageis not considereda medium of instructionwhen it is taught only as a subject. The medium of instructionis dividedinto four levels: lower primary, upper primary,and post-primary. The languageof lower primary refersto the medium of instructionused in the first three/fouryears of primary education;upper primary,the last half of primary instruction. Post-primaryeducation includes secondary schools, universities, colleges, and other post-secondaryschools.

The data in this table are based on [Bruhn,1984; Center for AppliedLinguistics, 1984; Morrison et al., 1972; Tadadjeu,1980; and Skinner,Table A, 1985]. The table was compiledby a consultantto the World Bank, Dianne C. Bowcock.

Other principallanguages

BurkinaFaso : Mande Mozambique: Sena Senufo Shona Fulfulde Ndau Mallua Cameroon : Bassa Chope Duala Tsonga Hausa Makonde

Gambia : Arabic Niger s Fulfulde Tuareg Guinea-Bissau: Balante Kanuri Fulfulde Tamahiq Malinke Sudan : Nuba Kenya : Nandi Fuv Dinka Mauritania s Berber Nuer Soninke

Mauritius : Creole (French) Bhodjpuri Chinese Hindi Telegu Urdu

Languape Names

Name used Equivalent Akan Twi-Fanti Bambara See Manding Bemba CiBemba Dagbani Mole-Dagbani Dyula See Manding - 267 -

Fulfitlde Fulani,Peul, Fulbe, Fula, Ful, Pulaar, Adamawa Galla Oromo Ganda Luganda Gaya Gbaya Ibo Igbo Kabiay Kabre Kanuri Manga, Lare, Kagama,Bornu, Sirata Kongo KiKongo,Kituba, Kingala Lingala Bangala,Ngala, Mangala Luba CiLuba,Tshiluba Luhya Lukuya Malinke See Manding Manding Bambara,Dyula, Mandinka, Malinke, Mande, Jula, Dioula,Mandekan, Kangbe, Mandingo Mossi More Nganguela Ganguela Ngbandi Banda Nyanja Chinyanja,Chewa, Chichewa Rundi Kirundi Sango A pidgin "basedon Ngbandi," (Morrison,p. 2001 Song'ai Songhay,Songhai, Sonay Sotho Sesotho Swahili Kiswahili,Mbalazi, Tanga Swati Swazi,Siswati Tamahiq Tamacheq,Tamasheq Teke Lali, Lari Rawana Setswana Wolof Oulof,Olof, Jolof

Tables 3 and 4. Population groowth and projections

The population growth rates are period averages calculated from mid-year population estimates. All means are weightedby each country's share in the aggregate population.

The estimates of gopulation for mid-1984 are based on data from the U.N. Population Division and World Bank data files. In many cases, they take I.ntoaccount the resultsof recentpopulation censuses. Note again that refugeesnot permanentlysettled in the countryof asylum are generallyconsidered to be part of the populationof the countryof origin.

Povulationgroiections are made startingwith informationon total populationby age and sex, fertilityand mortalityrates and internationalmigration in the base year 1980. These parametersare projectedat five year intervalson the basis of generalizedassumptions until the populationbecomes stationary.

To project fertilityrates, the yaar in which fertilitywill reach replacementlevel is estimated(see definitionbelow). These estimates are speculativeand are based on trends in crude birth rates, total fertilityrates (see notes to Table 5 for definitions),female life - 268 -

expectancyat birth, and the performanceof familyplanning programs. For most countriesin Sub-SaharanAfrica, total fertilitvrates are assumedto remain constantfor some time and then to declineuntil replacementlevel is reached. For a few countries,they are assumedto incteaseuntil 1990-1995and then to decline.

Mortalityrates are projectedas a functionof the femaleprimary enrollmentratio and life expectancyin 1980-1985. For a gi,en life expectancyat birth, the annual incrementsduring the projectionperiod are larger in countrieswith a primaryschool enrollmentratio greater than 70% and a life expectancyof up to 62.5 years. At higher life expectancies,the incrementsare the same.

Migrationrates are based on past and presenttrends in migration flows. The estimatesof futurenet migrationrates are assumedto be zero by 2000, but for a few countries,they are assumedto be zero by 2025.

The 1980 base-yearestimates are from updated files of the U.N. World PopulationProspects as Assessedin 1982, the most recent issuesof the U.N. Populationand Vital StatisticsReport, World Bank countrydata, and nationalpopulation censuses.

The net reproductionrate (NRR) indicatesthe numberof daughters a newborngirl would bear duringher lifetime,assuming fixed age-specific fertilityand mortalityrates. An NRR of 1 indicatesthat fertilityis at replacementlevel. At this rate, child-bearingwomen, on average,bear only enough daughtersto replacethemselves in the population.

A stationarypopulation is one in which age- and sex-specific mortslityrates have not changedover a long period,while age-specific fertilityrates have simultaneouslyremained at replacementlevel (NRR -1). In such a population,the birth rate is constantand equal to the death rate, the age structureis constant,and the growth rate is zero.

Populationmomentum is the tendencyfor populationgrowth to continuebeyond the time that replacement-levelfertilizy has been reached (NRR=1). The momentumof a populationin year t is measuredas a ratio of the ultimatestationary population to the populationin the year t, assumingthat fertilityremains at replacementlevel from the year t onward.

The estimatesof the hypotheticalsize of the stationary populationand the assumedyear of reachingreplacemeint-level fertility are speculativeand shouldnot be regardedas predicticns. They are includedto show the long-runimplications of recent fertilityand mortalitytrends on the basis of highly stylizedassumptions. A more completedescription of the methods and assumptionsused in calculating the estimatesis availablein the World Bank publication,Wlorld Population Prolections- 1985: Short- and Long-TermEstimates By Aga and Sex With RelatedDemographic Statistics. - 269 -

Table 5. Demographyand fertility

Crude birth and death rates indicatethe numberof live births and deaths per thousandpopulation in a year. The total fertilityrate representsthe number of childrenthat would be born per woman if she were to live until the end of her childbearingyears and bear childrenin accordancewith prevailingage-specific fertility rates.

The means for all indicatorsare weightedby each country'sshare in the aggregatepopulation. The data sources are the same as those cited in the notes to Table i.

Table 6. Urbanization

The data on urban populationas a percentageof total population are from the U.N. Estimatesand Proiectionsof Urban,Rural and City Populations1950-2025: The 1982 Assessment,1985, supplementedby data from various issuesof the U.N. DemographicYearbook, and from the World Bank.

The urban populationgrowth rates are calculatedfrom the World Bank's populationestimates; the estimatesof urban populationshares are calculatedfrom the sourcescited in Tables 1, 3 and 4. Data on urban agglomerationare from the U.N. Patternsof Urban and Rural Population Growth, 1980.

Becauseestimates in this table are based on differentnational definitionsof what is "urban,"cross-country comparisons should be made with caution.

The means for urban populationas a percentageof total populationare calculatedfrom countrypercentages weighted by each country'sshare in the aggregatepopulation; the other means are weighted in the same fashion,using urban population.

Table 7. Labor force

Estimatesof the populationof working age refer to the populationaged 15-64 and are based on World Bank populationestimates for 1984 and previousyears. The means are weightedby population.

The labo. force refersto economicallyactive persons aged 10 years and over, includingthe armed forcesand the unemployed,but excludinghousewives, students and other economicallyinactive groups. The agriculturalsector comprises agriculture, forestry, hunting, and fishing. Industrycomprises mining, manufacturing, construction, and electricity,water and gas. All other branchesof economicactivity are categorizedas services. Estimatesof the sectoraldistribution of the labor force are from the InternationalLabor Organization(ILO), Labor Force Estimatesand Projections,1950-2000, 3rd edition,and from the World Bank. Means are weightedby the size of the labor force. - 270 -

The labor force growth rates are derivedfrom the Bank's populationprojections and from ITL data on age-specificactivity rates in the source cited above. The applicationof the ILO activityrates to the Bank's latestpopulation estimates may be inappropriatefor some countries in which there have been importantchanges in unemploymentand underemployment,in internationaland internalmigration, or in both. The labor force projectedgrowth rates shouldtherefore be used with caution. The summarymeasures for 1965-73and 1973-84are countrygrowth rates weightedby each country'sshare in the aggregatelabor force in 1973; those for 1980-2000,by each country'sshare in the aggregatelabor force in 1980.

Table 8. Growthof production

Gross domesticproduct measures the final outputof goods and servicesproduced by an economy,i.e., by residentsand nonresidents, regardlessof the allocationto domesticand foreignclaims and without deductionsfor depreciation.Sectoral definitions are describedin the notes to Table 7.

The figuresfor GDP are dollarvalues convertedfrom local currencyby using the single-yearofficial exchange rates. For a few countrieswhere the officialexchange rate does not reflectthe rate effectivelyapplied to actual foreignexchange transactions, an alternative convers'on factor is used.

Growth rates were calculatedfrom constantprice seriesusing the least squaresmethod.

Table 9. Centralgovernment expenditure

Centralgovernment expenditure comprises current and capital expenditureby all governmentoffices, departments, establishments, and other bodiesthat are agenciesor instrumentsof the centralauthority of a country. The shares of governmentexpenditure by categoryare calculatedfrom nationalcurrencies.

Defense comprisesall expenditureon the maintenanceof military forces;including the purchaseof militarysupplies and equipment, construction, recruiting and training. Expenditure on strengthening publicservices to meet wartimeemergencies, the trainingof civil defense personnel,research and development,and on the administrationof military aid programsare also included.

Educationcomprises expenditure on the provision,managetaent, inspection,and supportof pre-primary,primary, and secondaryschools; of universitiesand colleges;and of vocational,technical and other training institutionsby centralgovernments. Also includedis expenditureon the generaladministration and regulationof the educationsystem; on research;and on such subsidiaryservices as transport,school meals, and medical and dental servicesin schools. The differencesbetween the percentagesshown in this table and those in table 14 are attributableto differencesin the years of the data. - 271 -

Health covers public expenditureon hospitals,medical and dental centers,and clinicswith a major medical component;on nationalhealth and medical insuranceschemes; and on familyplanning and preventivecare. Also includedis expenditurefor the administrationand regulationof relevantgovernment departments, hospitals and clinics,health and sanitation,national health and medical insuranceschemes and research and development.

Housing and communityamenities covers publicexpenditure on housing includingthe supportof housingand slum clearanceactivities, communitydevelopment activities and sanitaryservices. Social security and welfare coverscompensation to the sick and temporarilydisabled for loss of income;payments to the elderly,the permanentlydisabled, and the unemployed;and family,maternity, and child allowances. The welfare categoryalso includesthe cost of welfare servicessuch as care of the aged, the disabledand children,as well as the cost of general administration,regulation, and researchassociated with socialsecurity and welfare services.

Economicservices includes public expenditureassociated with the regulation,support, and more efficientoperation of business,and economicdevelopment; redressing regional imbalances, and the creationof employmentopportunities. Among the activitiesincluded are: research, trade promotion,geological surveys, and inspectionand regulationof particularindustries. The five major categoriesof economicservices are fuel and energy,agriculture, industry, transportation and communication, and other economicaffairs and services.

The other categorycovers expenditure on the general administrationof governmentnot includedelsewhere; for a few countries, it also includesamounts that could not be allocatedto other components.

Overallsurplus/deficit is definedas currentand capitalrevenue and grants received,less total expenditureless lendingminus repayments.

The means for the percentagedistribution of centralgovernment expenditureare computedfrom group totals for the expenditurecomponents and centralgovernment expenditure in currentdollars. Those for total governmentexpenditure as a percentageof GNP and for overall surplus/deficitas a percentageof GNP are computedfrom group totals for the above total expendituresand overallsurpluts/deficit in current dollars,and GNP in currentdollars, respectively.

The data on centralgovernment expenditure are from the IMF GovernmentFinance Statistics Yearbook, 1986, IMF data files, and World Bank countrydocumentation. The accountsof each countryare reported using the system of definitionsand classificationsfound in the IMF Manual on GovernmentFinance Statistics. However,due to differencesin data coverage,the componentsof centralgovernment expenditure may not be comparableacross countries. - 272 -

Table 10. Disbursementsof officialdevelopment assistance

Net disbursementsof ODA from all sources consistof loans and grantsmade on concessionalfinancial terms by all bilateralagencies and multilateralsources with the object of promotingeconomic development and welfare. Net disbursementsequal gross disbursementsless paymentsto donors for amortization.Net disbursementsare shown per capita and as a percentageof GNP.

The means of per capitaODA are computedfrom group aggregates for populationand for ODA; those for ODA as a percentageof GNP are computedfrom group totals for ODA and for GNP in currentU.S. dollars.

Table 11. Externalpublic debt serviceratios

Externalpublic debt outstandingand disbursedrepresents public and publiclyguaranteed loans drawn at year-end, net of repaymentsof principaland write-offs. To estimateexternal public debt as a percentageof GNP, public debt is convertedinto U.S. dollarsfrom currenciesof repaymentat end-of-year officialexchange rates. GNP is convertedfrom nationalcurrencies into currentdollars by the method describedin the notes to Table 8. The mean ratiosof externaldebt to GNP are weightedby current dollarGNP.

Interestpayments are actualpayments made on the disbursedand outstandingpublic and publiclyguaranteed debt in foreigncurrencies, goods, or services. They includecommitment charges on undisburseddebt when such informationis available.

Debt serviceis the sum of actual repaymentsof principal (amortization)and actualpayments of interestmade in foreigncurrencies, goods, or serviceson externalpublic and publiclyguaranteed debt. The mean ratiosof public debt and debt serviceto GNP are weightedby GNP in currentdollars. The mean ratios of debt serviceto exportsof goods and servicesare weightedby exportsof goods and servicesin currentdollars.

PART C. SUPPLEMENTARYAND SUMMARYTABLES

Table 1. Countrieswith PopulationsUnder One-HalfMillion People

Unless otherwisenoted, these data come from the EducationSector Policy Paper,African Development Bank, January1986. All such entries are based on the most recent informationavailable, generally from the years 1979, 1980, or 1981.

Other data sourcesare: a/ The World Bank, 1983. b/ UNESCO StatisticalYearbook 1986; figuresare for 1982. c/ The World Bank, 1987; figure is for 1985. d/ The World Bank, 1984. e/ Literacyrate; The World Bank, 1984. - 273 -

f/ Includes20,065 and 1.782 studentsin public and private schools,respectively, Excludes 2,500 studentsenrolled in independent privateschools. g/ The World Bank, 1986; figure is for 1977.

Table 3. EstimatedAverage Years of EducationAttended by Working Age Population

The averageyears of educationis the ratio of "educ._.aal stock" to workingage population. The educationalstock serieswas constructedas a weighted sum of the enrollwrntof studentsby level (primary,secondary, post-secondary) during the previous58 years. The term "workingage population"includes ages 15 through64. For details see Lau-Jamison(1987).

Table 4. LVteracy

The "adult literacyrate" is the percentageof personsaged 15 and over who can read and write. The 1960 data come from the World DevelopmentReport 1983 and other World Bank data sources. These data are based primarilyon informationfrom UNESCO,supplemented by World Bank data. Becausethese data are normallygathered in large-scaledemographic surveys and censuses,they often are not availablefor the more recent year. For some countriesthe estimatesare for years other than, but generallynot more than two years diptantfrom, those specificied. Thus the seriesare not comparablefor all countries. Data for 1985 or latest year data are from The CurrentLiteracy Structure in the World, UNESCO, 1985.

Table 5. Comparisonsof Achievement

Column (a) is based on data from the InternationalAssociation for the Evaluationof EducationalAchievement (IEA) Second Study of Mathematicsas reportedin [Livingstone,1985]. Tests were administeredin participatingcountries to 13-year-oldstudents in five mathematics subfields.

Columns (b) and (c) are based on (Heyneman,1980]. Scores for all countriesother than Malawiwere obtainedunder the auspicesof the ILEA.Respondents in these countries were ten years old. The scores for Malawi are not strictlycomparable. The tests were administeredby a World Bank mission in just two districtsof the country (Mulanjeand Nsanje). The questionswere in Englishrather than in the vernacular languagesof the districts(although instruction past the first few years of primaryschool in Malawi is also in English). The Malawianrespondents were older (16 years old on average)than the IEA surveyedstudents.

Table 6. Gross PrimaryEnrollment Ratios

All figuresare for 1983. All are percentsexcept those in column (f), which are 1983 dollars. The averagesshown are weightedmeans (w) and medians (m). - 274 -

Table 7. Enrollment Characteristics and Education Expendicures by Secondary Enrollment Groups

A:.1 figures are for 1982. The averages shown are weighted means (w) and mediano (m). - 275 -

Bibliographyof data sources

Education U. N. Educational9Scientific and CulturalOrganization indicators tape. _ Variousyears. StatisticalYearbook. Paris. _ 1985. Female Participationin HigherEducation. Peris. Naitionalsources. World Bank countrydocumentation. World Bank data files.

External Miilot, B., F. Orivel,and J. Rasera. 1987. L'aide education exterieurea l'educationen Afrique sub-saharienne. aid DiscussionPaper SeriesNo. EDT65. Educationand Training Department,The World Bank. U.N. DevelopmentProgram data files. U.S. Agency for InternationalDevelopment country documentation. World Bank documentation.

National InternationalMonetary Fund. 1985. Government Finance accountsand StatisticsYearbooks Vol. IX. Washington,D.C. econc2ic Sawyer,Malcolm, 1976. IncomeDistribution in OECD indicators Countries. OECD OccasionalSt-udies. Paris. U.N. Departmentof InternationalEconomic and Social Affairs. Variousyears. StatisticalYearbook. New York. * 1981. A Surveyof NationalSources of Income DistributionStatistics. StatisticalPapers, series M., no. 72. New York. e 1985. NationalAccounts Statistic Compendiumof Income Distribution Statistics. Statistical Papers, seriesM, no. 79. New York. FAO, IMF and 'JNIDOdata files. Nationalsources. World Bank countrydocumentation. World Bank data files.

Balanceof The Organisationfor EconomicCo-operation and Development. payments and VariousYears. DevelopmentCo-operation. Paris. debt . 1986. GeographicalDistribution of FinancialFlows to DevelopingCountries. Paris. IMF balanceof paymentsdata files. World Bank Debtor ReportingSystem. - 276 -

Languages Bruhn. 1984. '"AfricanLingua Francas."Center for Applied Linguistics. Washington,D.C. Center for AppliedLinguistics. 1984. CountryStatus Reports. Washington,D.C. Morrisonet al. 1972. Black Africa. The Free Press. New York. Skinner. 1985. "Degreeof Multilingualism."University of Wisconsin-Madison,unpublished table. Tadadjeu. 1980. "A Model for FunctionalTrilingual EducationelPlanning in Africa." UNESCO. Paris.

Population U.N. Departmentof InternationalEconomic and Social Affairs. Variousyears. DemographicYearbook. New York. . Variousyears. Populationand Vital Statistics Report. New York. 1980. Patternsof Urban and Rural Population Growth. New York. 1982. 'Infant Mortality: World Estimates and Projections,1950-2025." PopulationBulletin of the United Nations,no.14. New York. Bulletinof the UnitedNations, no. 14. New York. 0 Updatedprintouts. World PopulationProspects as Assessedin 1982. New York. * 1983. World PopulationTrends and Policies:1983 MonitoringReport. New York. __. 1985. Estimatesand :'rolectionsof Urban,Rural and City Populations,195G-2025; The 1982 Assessment. New York. World Bank data files.

Social Food and AgricultureOrganization. October 1985. Food indicators Aid Bulletin. Rome.

_ Food Aid in Figures. Rome. ProductionYearbook. . Trade Yearbook. 'Standard"Computer Tape. U.N. Departmentof Internationaland EconomicAffairs. Variousyears. DemographicYearbook. New York. Variousyears. StatisticalYearbook. Paris. FAO and World Bank data files.

Labor force InternationalLabour Office. 1986. Labour Force Estimates and Prolections,1950-2000. 3rd ed. Geneva. InternationalLabour Organization tapes. World Bank data files. - 277 -

Student Heynewan,S. 1980. "Evaluationof Human Cajital iD achievement Malawi." World Bank WorkingPaper No. 420. The V-7rld Bank, Washington,D.C. Livingstone,I. 1985. 'Perception of the Intendeda'nd ImplemAentedMathematics Curriculum.' Universityot Illirois(M4meo). NOTES NOTES * MAPSECTION SUB-SAHARANAFRICA TOTALPOPULATION 1984

$WRA~~

CAM~kt N I MAU_. ¢ ; o. ,v 'A''

wAL ; j_EMATOK4OUNU-c SIERRAt3COTAi,

A NNR

Millionsof People: [11 1 0-19 20-39 f 40-59 60 andabove [l |No data

na mh hsbn pprdo by The tISWBan*', sfA 'eck4*i W Vi onvfld Of "ts vas adcA,OA tr VWk ua of TheW,8 ftd SWe " Senbww fRw4 Coto The o bo MeWAa ~ on On Mmp do n #A*. on ft pa*of Th* MulB andV It O4ueftna Fiencotwbo an sd0nv onRDiat ataoAOf an tWNWOtfoN WrMent or atn of Aib n IBRD19825

SUB-SAHARANAFRICA POPULATIONDENSITY 1984

#ItA SUDAN

I AFRICANREP. uaa~ - S O M NRA

SAO704I ANDPMFOPJ aAWONP./

ZAMtE ~~~~~~~~~SIYC*UUES

- ~~~ |jJOE F~~MADAG;ASCARMAAURWS w squarekilometer: : 0-19 g 20-39 40-59 60-79 M80 and more J Noclato

WWAV**~fto medta000 we of TheV*dd 8W* end mekA*nobwu a Thed00cultow wd *erd sAow~ W al 4 nom VITh Wbt8'*ad theivoeatlt~Fae CoeaEn £''~ist antwym a" 'NWety enda'amient oaepant SWoAa

NOVEMBe.r19 SUB-SAHARANAFRICA LANGUAGEGROUPS 1984

CARE

MAUNO LanguageGroups by Country:

Francophone a AnglophoneE: Other No data

0 10?0 KAOMIE

f mn hanO 6ev rmaEby The &bS0 *Wsxduway AV t comnvi*msof ti? N W NxWokJsAI.IOt ford* O*eme us of The ftv Bakt* a hbovw FhwI Coino he dtw tu M ud en " boIi t on ensmp do not *VW.an ft of the fo* VWan tiWnt et Few" corpabon. dgwmjnt af DI ItQ$W atoy laD or to &WwwOew or *cc4PMm* of itn botose IBRD19826

SUB-SAHARANAFRICA INCOMEGROUPINGS 1984

ThEOAM ~ ~ BQATO O

ADMOC 0AI

Typeof Economy:

Low-incomesemiarid F Low-inconmeother Middle-,ncomeoil import [I I Middle-incomeoil export No data

KIlOMETERS

to.s has S been a nr%eed forythe e VwtWa 8as** e o stafThe .xtwf Vto Ba*fo adh (t canowc knh oft Ctpcrhov Thedeabo n usedand BdmoI- Wn on t ma d not tM Itfle d4 T H wn and te k*awbw Frne COW1n, NW _u.mant 't tWt of aw taintwvyor an er4n or #cEtn of dLvhb&DCxW

NOVEMBER1986 SUB-SAHARANAFRICA ENROLLMENTIN PRIMARYSCHOOLS 1960

;~~~~~~~~~~~'

CAPt CA QWNUPtX /AU gW

w~~~~~~~" 'i.;g._

MANo d

SAC_ TOME_ < I E WA^

Percentage4e ofby Age emB.O!e. Group:.

e e| f8and e above e 50-79 eo [=0-49 No data

nsn4m A" G# ewDqfTdt Wn SW*dR$Sbirft * lt t wn Of t_We Core fW dworAvkb ue ed fhv1bouad&m &ie an cr mendo not bpn bean, rt o Te merM8e ithdtvn if n Fc Ctp torshn.wwpmt OR e gft ethe ofden *wyto of VWendmovegt or ecogtec of &oihbosidene IBRD19824

SUB-SAHARANAFRICA ENROLLMENTIN PRIMARYSCHOOLS 1983

'~~~~~~~~~~~I

gaiX'~~~~~~~~~~sr MAURIT*J4IA~~~~~~~~~~~~~~~AUI

LESOTO~~~~~~~~~ECHLE etctgeof Age Group:\

80 and above. UIA - FM 50-79 eO/ KIOM" COOlO 0-49

Noz dato

.en prWeX by TheI WA7 aw* '$ Odt &xC&v kf gme COnmre ot 8flagf oXCA4fAg Growp: Th a" Me 80Wb aodts andatbvdo Wo fabt e C shono no

Aof dNWtwrrt or WVrwow"w eV edp o sewhb4ws

NOVEMBER1986