ReportNo. 11191.PE Peru PovertyAssessment and Social Policies andPrograms for the Poor Public Disclosure Authorized May 5, 1993 HumanResources Division CountryDepartment I LatinAmerica and the CaribbeanRegion

FOROFFICIAL USE ONLY

MICROFICHE COPY

i Report No.: 11191-PE Type: (SEC) Title: POVERTY ISSUES AND SOCIAL SECT}

Public Disclosure Authorized Author: JONES, T Ext. :35309 Room:I7181 Dept. :LAIHR Public Disclosure Authorized

4~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~( W ;~' , , . ..__; .x..- , P~~~~~~~~~~~~~~~~~~~~-- .e~ Public Disclosure Authorized U q. This. ,.has.40 r.stricteddis.tton afdmay--- be used.byrecip, s soff.tafyvt- d ?c -f Iiek duis co, nte,of ewise -oth W'~~~ W * WodBn oi~ati, . CURRENCYEQUIVALENTS

Currency Unit prior to 1985 Sol SIS.) 1985-1990 Inti (/.) = S/1,000 1991-present New So (S/. = 1/1,000,000

ExchangeRates US*1.00 equivalent (period average)

Official ParallelMarket 1986 1113.9 1/17.8 1987 1/16.8 1/40.2 1988 1/128.8 1/314.8 1989 1/2,666.2 114,394.9 1990 1/187,885.6 1/206,441.1

1991

October S/0.92 November S/1.02 December S/1.00

1992

January S/0.99 February S10.97 March S/0.96

From August 1990, the exchangerate was unified. FOR OFMCIALUE ONLY

ABSTR4CT

1. The objecdveof this reportis to identifythe componentsof a povrty alleviationstrategy for Peru, witha focuson socia services. The reportanalyzes Peru's previouseconomic performac and the main eementsof the reformprogram whichcould affect the poor. It also analyzesthe characteristicsof povertyin Peru. It examinesthe curret stats of the socia sewicesmost critcal for the poor, identifyingkey issuesand examiningthe utilizationof services. Ibe repoit makes recommendationson priorityprograms. Fmally,it looksat issuesin the implementationof a povertyalleviation strategy, and it examinesthe Government's progressthus far in definingits strategy. 2. Peru's potenti wasfrUste duringthe decadesof heavyGovernment interventionIn economicacdvity. Oneof the legaciesof poor economic pefmae is the persisnc of highpoverty levels. In August1990, a new adminis on announcedthe first measu in a strongstabilization aa adjusatent progrm. The initationof majoreconomic reforms provides a uniqueopportuity for Peru to considera programto alleviatepoverty. Theseeconomic reforms are layingthe groundworkfor broad4baedeconomic growth, a reducedrole for the state and greateropportunities for the poor. A targetedpovety alleviationprogram includingsocial services can proect the mostvulnerable groups and assistothers to take advatage of new opporites. 3. Low spendingand poor outcomescharacterz Peru's socialsectors. Infnt an mater mortalityrates armamong the highestin LatinAmerica. Chfld malnutiton peists at 1971levels. The 1991cholera epidemic demonstaed in a dmatic waythe prcaiousess of sanitatonconditions. In education,Peru performswell in tms of primaryand secondaryschool enrollment, yet reptiion rat are highand educationalquality is low. With sharpdeclines in Government expendiureover the past fiveyears, publicresources for socal serviceshave becomescarcer. Whileprotected from the extremereductions experienced in other sectors,social spdiw is lowerthan in neigboing ries. 4. Healthand educationsector issues infuence directy the qualityof life for the poor. In health,the benefs of Peru's extensivenetwork of primarycare facilitiesare not realizedbecause of a scarcityof equipment,poor distution of pesnnel and a lowproportion of total resoces devotedto primarycare. Quality and utilizationare low. In nutrition,widespread child malntrition has not been reducedsignificantly trough foodassiace. Numerousagencies wortc on nurition but in the absenceof an oveal fiamework,adequate norms or surveillance.In education,prepimary schoolingquality is poor. Multigradeschoois in rural areas iack suppliesand well-trainedteachers. Illiteracypersists at a relativelyhigh rate, withlittle progressin eradicatingit amongnrual women.

restricteddistribution and may be used by recipients only in the performance This document his a authotietion. of their oflicialdutis. Its contents may o otherwis be disclsed without World Bank 5. Whoare the poor in Peru and howdo theyuse socialservices? Ts report constructsa povertypro9fie drawing upon a 1991Living Standards Survey. Halfof all householdssurveyed are poor and one fiflhare extremelypoor. The poor send their childrento schoollater anda largerproportion either never attend school or drop out veryearly. Fewerof the poor attendpre-pimary schools. The poor are less likelyto receivevaccinations, seek media care or haveaccess to public seweragesystems. Theyare more likelyto use primaryhealth cenrs and posts. 6. Thisreport recommends that education programs for the poor improve conditionsin rural multigradeschools, provide textbooks and teachingmaterials and train teachers. Moreemphasis needs to be placedon preschooleducation. Non- Spanishspeaking women should be trainedIn functionalskills including literacy. Healthprograms for the poor shouldconcentrate on providingmothers and children with adeuate primaryhealth care services. Suppliesof basicequipment, mateials and medicinesneed to be ensuredfor primarycare services. Preventionand treatmentof choleraand improvementof hospitalemergency services are other high priorities. Well-designedsocial compensation programs can complementimproved socil services. Employmentprograms can reduce poverty by providingincome and, creatingand rehabilitatinginfrastructure. Food assistance can protect the nuritional statusof vulnerablegroups. 7. Implementationof a povertyalleviation strategy needs to buildin approprlite' targetingand institutionalarrangements. In Peru, geographyis the singlebest indicatorfor reachingthe extremelypoor. Bydirecdng resources to thb rum sierra, one in everytwo extremelypoor householdscan be reached. Themagnitude and seriousnessof povertyin Peru call for exploringa vaiety of insdtitonal arrangements.Among those worth considering is the socialinvestment fund model. Peru establishedFONCODES, its NationalFund for SocialCompesation and Developmentin August1991, and by mid-1992,this organizat bega to show encouragingsigns of its potential. NMOsshould also be calledupon to participate actively. Theycan mobilizeand utilizesubstant levelsof resourcesto reach underservedpopulations effectively.

8. Ihe Governmentrecendy prepared a povertyalleviation strategy which is generallyconsistent with the findingsand recommendationsof this report. The Governmentstrategy recognizes that povertyalleviation demands essentily an economicstrategy. It aimsto improvethe livingconditions of the extremelypoor throughprograms in food assistance,health, education and employmentgeneration. Financingfor this strategymust come from governmentand donorresources. To attractsufficient donor resources, the Governmentmust demonstrate its own commitment.In 1990,total government spending on healthand educationreached only2.2 percentof GDP. As its revenuesincrease, it is recommendedthat the Governmentincrease the resourcesit allocatesto healthand education.The challengeand testof the Government'sresolve in redrelog povertywill be the extent to whichcurrent and futurepolicies and programsfollow its poverty lleviation.. Contents

vii

latroduetle ix

Summand Cya (English and Spanish) xi aopWt 1: From Emosic ad Soa Crus to Rm 1 cotrodm1on Economi Stanion EconomicCdisis 3 Soca Crids S The Econoic lteformProgram 9 Initia Responseof the Economyand ExpectedBenes 12 ConclusIon 12

Chapter 2: Pow"ertyNeI13 b Trnds in Povety 1970-1990 13 Poveftyrfle 14 Smmary 24 Imicaions for PovertyAleviaon Stratey 25 ahapter 3: Soca Srvks 27 27 Statusof Services 27 ham 28 Wealthmad Nourh 30 Health Condo 30 Statusof Sevces 31 Issues ~~~~~~~~~~~~31 NutritIon 32 Child NutritionalStatus 32 Micronutrients 33 NutritionalServices 34 Issues 34 Conlusfions 36

Chapter 4: Towards a Poverty AlleviatIon StrateW 37 IJtroductlon 37 Socia Services 37 Education 37 Health 39 Nutrition 41 Socal Compensaion Programs 42 Food Assistance 44 CurrentStatus Magnitude,Agencies and Progrms 44 Major Issuesin Food Assistance 47 Major Recoramendations 47 Program SpeclftcIssues and ReendaIons 49 Concluion 49

ChapterS: ImplementInga Poverty AlleviatIonStrate 51 Itoduction 51 Targeting 51 InstitutionalArrangements 52 SecurityConstraints 55 PovertyAlleviation Program of Govemment 55 Concluion 56

ibliogaphy 59

Tables 3 Table 1.1 PerformanceIndicators of the Economy 4 Table 1.2 Indicatorsof EconomicCrisis 5 Table 1.3 Compl ative Social Indicatorsin SelectedCountries 6 Table 1.4 Struchre of CentralGovernment Spending, 1970-1990 8 Table 1.5 Social Setor Spending,1970-1990 9 Table 1.6 Coefficientsof Vulnerability,1985-1990 9 Table 1.7 GovernmentSpending on Educationand Health 14 Table 2.1 LatinAmerica: Poverty EstimatesAround 1970, 1980and 1986 15 Table 2.2 Distributionof Total Expenditureby Decile: All Sample Areas 16 Table 2.3 1991 PovertyHeadcount 17 Table 2.4 RegionalDistribution of Populationby PovertyGroups 17 Table 2.5 HouseholdSize and Per CapitaConsumption by PovertyGroup 18 Table 2.6 Sex of HouseholdHead Amongthe Poor 18 Table 2.7 Distnrbutionof the Poor by Langage of Interview Table 2.8 Distnrbutionof the Poor by Employerof Head of Household 19 Table 2.9 Food Expendituresby PovertyGroup 20 Table 2.10 Food Consumptionof Poor Householdsby Type of Acquisition 20 in Rural Sierra iv Table 2.11 Non-FoodExpenditure Pattem by PovertyGroup 21 Table2.12 Public Servicesfor the Poor 22 Table2.13 EducationLevel of HouseholdHead 23 Table 2.14 AgriculturaLand Owwnshipin the Rura Sierra 23 Table2.15 Crops Grownin the Rua Sia 24 Table 3.1 Wage and BudgetIndies 28 Table 3.2 Indicatorsof IntemalEficiency of PrimaryEducation 29 Table3.3 InfantMortality Rate by Region1989 30 Table3.4 Coverageof the Maternal-ChildProm 1989 31 Table 3.5 Intensityof Contactsin the Maternal-ChildProgam 1989 31 Table 3.6 Nursesper Doctor 32 Table 3.7 Disea Trends 33 Table 3.8 Changesin HouseholdFoodshare by Region ard ExpenditureLevels 34 Table 3.9 Food Consumptionss Percentageof Total Consumptionfor Lima Households 34 Table 4.1 SchoolAttendance by Age Groupand PovertyLevel 38 Table4.2 Kinderaten Attendanceby PovertyGroup 38 Table43 Percentageof ChildrenReceiving All RequiredVaccinaons 39 Table4.4 Percentageof Ill Who seek MedicalAttention 39 Table4.5 DrugPurchases by PovertyLevel 41 Table4.6 Unemployedby PovertyGroups 43 Table4.7 Major Food Assistce Programs,1991 45

Anexes

Annex A Social Securityand the Provisionof Health Services The Private Provisionof Social Semvices Annex B RecentResearch on Povety in Peru Annex C Calculationof PovertyLines Table 1: PovertyLines for reme Poor and Poor by Region Table 2. PovertyLnes Relativeto ExpenditureDistibution and Mleanby Region Table 3: Basic Food Basketfor Limaand the Coastof the NationalInstitute for Food and Nutrtion (INAN) Table 4: Basic Food Basket for the Sierra Annex D living StandardsMeasurement Survey, 1991 (SelectedTables)

SlatsicalAppendix

v PREFACE

Thisreport is based on the ndings of missionswhich visited Peru in June and September1991. The missionsconsisted of The-esaP. Jones(Chief of mission), EduardoAtalab, Karen Cavanaugh, Ernesto Cuadra, Magaret Grosh,Bruce Herrick and RichardMoore. 'Me sectionon non-govermenalorganizations draws upon a backmund paperby Tbn de Wittbased on findingsof an April 1991mission led by ChristopherHennin. 'he sectionon the povertyprofile draws on a background paper writtenby GustavoYamada and a recentLiving Standards Measurement Surveyconducted and processedby the firm Cuanto.The vnissionswould have been unableto preparethe reportwithout the assmanceprovided by staffof the Ministry of Economyand Finance, the Ministy of Health,the Ministryof Education,the Cetr ReserveBank, the NationalPlanning Institute, FONCODES and other agenciesin Peru. Tbis reportbenefitted from the Government'sreview of an earlier version,which was discussed with representatives of the Ministriesof EReazh, Education,Economy and Finance,and the CentralReserve Bank in October1992. Tlese discussionsconfirmed the Government'sagreement with the report's ov. -all conclusionsand recommendations.Tbis version has beenrevised to take into accountthe Government'srecent formulation of its povertyalleviaton strategy. The rport also benefittedfrom an exceptionallyhelpful group of peer reviewerswithin the WorldBank, comprised of PaulGlewwe, Margaret Crosh, Phil Musgrove,Larry Wolffand HelenSaxenian Acronyms and Abbreviatons

M^CA 6 _RwalCanAId EO_na FOdASbd fLC

APRCL Fomodn4oafUpta4Apcaofin PA

DAIZOD DeaurofodslaAgdcukaydeIahfaeshuc- Ag ancRliunand Reo tumftslmdeaZomDepbniu Developwensin DmudAm=as

DESCO Centrodoe ad6ol y_Picido1Duszolo Cunx forth Stuidyand Ptomtio of Devem wsPM Famb de Bme Social 3o" Emwv7 social Fund (Bo"ivia)

PONCODES FamdoNaionadeC_sp yDeamyDnOo NaiuSoci and De"dePmmkd

INAN T NacNatond AlonaldioyNut Ntlnsftu onFeEJigandNutMotb

MCIN maternaladMddtldeakh

MOE bsodeBdwaci h fEducaam

NGO __

OPASA Obia F yAuthciaSocial AvstDvyelqpmwandRaiAgecy

vil PRONAA nNAIwddAyo_ NaPoodAs_P

PAD Prmg dOAiseDn ef DheIaD4xePa

PAHO PanAm4hb H _ zo

PANFAR p oy Nukle IeaIcFamiNa Fading ProgmfbrH4ghRiskFamls enAltoRao

PAT-PAMg cotoWoS G sPUmy School0ildm andTB}

PLSS FemLivivg Stadars Survy

PRW;MA PmacosenhftadcSalud,aM y PwjeclnsImunionaSgs

PRODA PtoVm ddeDarrolo Ingal lJtbo cm FoodAad I ontadUrban

MON"d Natonl FoodSupport Progrn

PRONABCE Progam Naconalde Contrl deldocby Naionl C3okwand danicism CkdisnoRadmibos aaaSPtmg

PRONOEE qdm NoEsh de niBdlca NlEProgrmm

SISVAN S_ste deVIlmolA y N Iuiol Foodand Nutitim Surveilace Syslem

alDE Uided DepSnaldeSad I k 'I nit

WFP WoddFood Program

V]il Introduction

1. In August 1990, in the midst of plummet- advantageof oppaortnes, not only the few able iug output and hypnla on, a new adiniz't- to gamer ecomic rents. Second, the new admin- tion in Peru announced te first measures in a istaion has decidedto reduce the role of the Gov- strog stabilizationand adjustmentprogram. The ement in economic activity. As the state pulls main tenets of the pgam are tight fiscal and back from many eoomic activities, it will free moneary policy, libalzation of makets and a resourcesto improvethe prwision of public ser- reductin of the previous pervsive role of the vices such as educationand health Govemments overnmentin the economy.Mhe objectives are to traditionallyhave an importantrole in providing restore croeconomic stability onorder to pro- these services to the disadvantaged.Third, the mote sustainabe, broad-basedgrowth, therebyre- pmotion of growthwould beneit the poor, but it versingPeru's legacyof economicstagnation and will take sone timefor the ecomy to respondto deterioation.Thatlegacy is reflectedin many eco- the new incertive structe. Also, although sore nomic and social indicato, but pehaps most im- reformsare likelyto help the poor, othersmay lad portaty, in the prsistee of a higb incidenceof to tempory losses for some. In addition,certain poverty In cerai regionsand ethnic ps, and a groups in the country - those lacking sufficient decras in welfare to below the poverty line for educato and those withpoor health,for example may who had not previouslybeen poor. -may not be able to take advantageof emloyment opportu-ni withoutspecial assisUce. Ihe needs 2. The initiation and, thus far, impressve of thesegroups shouldbe addressed. implementaion and broadeing of the reforms pwvide a unique opportuniy for the Govement 3. The purposeof this reportis to identifythe and the agencies providing financial support to components of a poverty alleviadon strategy for Peru to consider the needs of the poor. It is a Pe, focusingon social sevices. Of course,a pro- unque opportunityfor three reaso. Fira, inttt- gram to reduce poverty would encompass other nationalexpedence shows that the firstprong of an aspects as well such as rural developmen and in- effectivestategy to reduce nationalpet Is the fsuctiemprovemens. However,those areas promotion of broad-based growth, partculaly are not the focus of this document The report be- economicexpansion dhat encourages the use of la- gins with a brief analysis of Per's previouseco- bor. For many years the requisie polcies and nomic performanceand the main elementsof the incentive strucue for growth have been absent reform program, especially those aspects which in Peru. That appearsto be changingand qukidy. could have the biggest impact on the poor. The If this program is succesfl in achieving its second chapter analyzes the chaact cs of goals, alare groupf people will be able to take povery m Peru.mhe third chapter exmines the ix curt statu of thsocial sCevi mostcritidl cpter looksatpissues sh asthe forthe poc. J te fourthchapter tohe i ldc-t oleof socia ivestmentfunds and non-govern- floindie dXoion ~ofsocia vis pluslfor- me Sal ios.a X o nes p mlon on thOeuse of ths servcesby thepoor, are madeby the Oovermet In defi a povet al- used as thebasis fowrecammeng the priority leviatonstategy andin iXdfying proets to programswhich need to be supportd by both the spport tiht satg. Govenment an exti asistnce. Th fia

' Summaryand Conclusions

1. One of the legaciesof poor economicper- groups and assisting others to take advantageof formancein Peru is the peistence of high poverty newwoppouties is calfedfor andis the focs ofthis levels in some regions and ethnc groups and the repot declinein living andards of others.Tho initiation of major economicreforms provides a uniqueop- 2. The promisingpotential of Peru was fes- porunity for both the Governmentand thosepro- trateddurng three decades of heavy Government v tag financial supportto Peru to considera pro intevention in economicactivity which rested in gmmto alleviatepoverty for several reasons.Fist, a legal structure and policy environmentinimical internationalexperience shows that a meaningful to growth.The overexpansionof the state led to a and sustained reductionof poverty is impossible waste of resourcesand economicstagnation. Be- without broad-basedeconomic growth. Up until cause of the lack of growth and rigid, closed mar- now the macroeconomicstability, administrative kets, few groups in the country had access to the frameworkand incentive structure conduciveto meansto improvetheir livelihoods.Public policies growthwere absent in Peru. That is changingand diorted most of the markets in the economy. quicldy. Second, one reasonfor the poor growth Trde policy reliedon an import-substitutionstat- was the over-exteion of the slate in economicac- ey with very hightariffs and many non-tariffbar- tivity. A key element of the reformsis the reduc- rers. Decliningtax recipts contibuted to the de- tion of the role of the state. As the public sector teriorationof fiscal performanceand reduced the movesout of mm whereit adverselyaffected out- resourcesavailable for operating,maitaining and comes - labor and financial markets and produc- investing ine al public seices. Evenpolicies tion - it will free resourcesfor areas where it can and progrms establishedto benefit workers and and should make a positivecontribution, namely famners,such as labor stability laws and agrarian the provisionof public servicessuch as education refonn, proved to be detrimentalto those groups and health. Third,in the mediumand long-runthe and to the economy.Real wagesfell afer 1970 and implementationof the reformsshould reduce pov- employmentin the fomal sector hrank. Agricul- erty by providingmore opportunityfor the poor to toregrew at only half the rate of populationprovid- use their assets, such as labor and land. However, ins little opportunityfor an improvementin rurl in the short-runsome measuresmay lead to tempo- incomes. rary difficulties. It will also take time for the economy to respond to liberalizationand some 3. PresidentGarcia's explanationof cbronic groupsare not able to participatefully. A targeted inflation and low growth was that Pem was in a poverty alleviationprogm, includingsocial ser- "debt trap":servicingthe external debt would ac- vices, and aimedat protectingthe mostvulnerable celetaft'devaluation and inflation,erode domestic xi savings,and curtail investment. Consequently, his Government'sability to providebasic services in administration(1985-90) put a ceilingon foreign the long-run.Growth in urbanemployment and debtservice of 10 percentof the valueof exports, wagesis a majordeterminant of the pace of pov- andlater stopped payments altogether. It instituted ertyreduction both through its directinfluence on widespreadforeign exchangecontrols and in- the exng wbanpoor and through the opportuni- creasedprotection. The Governmentalso raised ties it creates for migrationfrom rural areas. the minimumwage, expanded credit and subsidies, Greaterneutrality in thetrade regime can support a decreasedthe valueadded tax rate and lowered morelabor-intensive pattern of industrialexpan- publicsector prices. This strategy led to a highfis- sionin import-competingas well as exportingsec- caldeficit, financed by moneycreation. After 1987 tors. The Governmenthas movedquickly and the marketlost confidence,inflation rebounded boldlyto liberalizethe trade regime.However, and economicoutput and real wages declined tradereform is mosteffective if factorsof produc- sharply.Economic mism ment reacheda cli- tion are free to respondto marketforces. In the maxin 1989-90and was reflectedin hyperinfation. caseof Peru,state policies since the 1960'shave severelyrestricted factor movements. In orderto 4. The poor economicrecord is one factor promotemore opportunitiesfor urban employ- behindthe poorhealth conditions in Peru.Among ment,a keyarea for reform is labormarket deregu- countriesin the region,Peru ranks near the bottom lation.Policies to promoterual developmentalso on healthindicators. The 1991cholera epidemic arean importantcomponent of thepoverty allevia- demonstratedin a dramaticway the precariousness tionstrategy. of sanitaryconditions nationwide. In education, however,it ranksnear the top on the ratioof stu- 6. Growthis important,but is not the sole dentsenrolled in primaryand secondaryschools componentof the povertyalleviation strategy. Pro- althoughhigh repetition rates and lowquality are gramsare neededto ensurethat the poorcan take issues.The publicresources available for social advantageof job opportunitiesand to protectsome serviceshave become scarcer with the sharpde- vulnerablegroups that are not ableto participate dine in overallGovernment expenditures during fullyin the economy.In orderto designthese pro- thelast five years Thedecline is explainedby sev- gramsinformation on thepoor is needed.The pov- cralfactors including the severerecession and the erty profileof Peru constructedin this reportis nearvanishing of the country'stax system.The basedon datafromaUving Standards Survey car- socialsectors were protectedfrom the moreex- ded out during October/November1991. Two tremereductions experienced in otherparts of the povertylines are calculated. Households are classi- Government.Nevertheless, because of the decline fledas extremelypoor if their totalexpenditures intot resourceavailabilit, Peru generally spends are less than the cost of a basic food basket. lessthan its neighborsin esear Householdsare classifiedas poorif theirtotal ex- pendituresare less than the cost of a basicfood 5. In August1990, the Fujimoriadministra- basketplus an estimateof non-foodexpenditures. tionlaunched an orthodoxand sweeping stabiliza- Abouthalf of the populationsurveyed fal below tion and structul reformprogram. During the the higherof the twopoverty lines and a fifthfail fir stage,the Government took dramatic steps to- belowthe lowercutoff. wardsstabilizing the economyand rebuildingties withthe internationalfinancial community. A sec- 7. Wheredo the poor live?Among the four ond stageof policyreform began in March1991 regionssampled (lima, urbancoast, and urban and andwas based on a seriesof broadxsed stctural mralsierra), the incidenceof povertyis highestin measures.These reforms aimed at promotingcom- therual sierra.Two-thirds of the householdsthere petitivenessand deregulatingeconomic activity. are poor and47 percentare extremelypoor. Only Severalreforms are crucialfor povertyreduction. 10 percent of the householdsin Lima are ex- Peru's abilityto reversethe sharpdecline in tax tremelypoor, comparedto about 20 percentin revenueslies at the heartof the restorationof fiscal boththe urbancoast and sierraregions. In eachof balance.It is also vital in orderto guaratee the thesethree regions the share of poorhouseholds is xli about half. The rural sierra is estimatedto contain 12. Among land-owning houeholds, the the largest numberof extremelypoor (1.6 mfllion poor have roughly 50 percent less land than the people). Because of its size Lima contains the non-poor.About a third of poor householdsown highest numberof poor (2.5 millionpeople). irrigatedland, comparedto about half for the non- poor. Poorer householdsin the rural sierra tend to 8. The typical poor person lives in a larger have more workers engagedin agnculture than is householdthan the non-poor,on averge contain- the case for better-offhouseholds. The most im- ing 1.5 more individuals.Each worker in a poor portant crops produced in the rural sierra by the householdsupports four familymembers. For the poor are potatoes, barley and wheat The crops non-poorthe numberfalls to three. Per capitafood which are producedmainly by the poor are white expendituresin an extremely poor householdare maize,barley, potato,oca and olluco. only 40 percentwhat it is in other householdsand the similarao for totalexpenditures is 30 pecet 13. The povertyprofile providesbroad guid- ance on a poverty alleviation strategy for Peru. 9. Only in Lima are female-headedhouse- Jiaty, targetedprograms should focus on the ex- holds found disproportionatelyamong the poor. trem poor whichrepresented a fifth of the popula- Indigenouspopulations are dad in Peru. tion surveyed.Second, prograns to alleviatepov- Taken together, householdsheaded by Quechua erty among the extremely poor shouldtarget the and Aymaraspeakers account for 40 percentof the rural sierra and the indigenous population. Im- extremelypoor and abouta quarter of the poor. In provements in access to public services, particu- Lima, migrantsare slightlymore likely than non- lady for ducation,are neededto reduce extreme migrantsto be founJ amongthe poor.The poor are poverty. These householdsare at greater risk for found largelyamong two occupationalgroupings, poor healthbecause they lack dean waterand sam- the self-employedand privatesector workes tation facilities.Third, working in agricultureis positivelyassociated with poverty. There may be 10. The food commoditiesaccounting for the potentialfor agriculturalpolicies to reduce pov- largest share of expendituresamong the poor are erty, such as reformsrelated to land titling, the re- rice, bread, sugar, tubers and meat. The most im- habilitationor constructionof small-scale irriga- portant products which are consumedfrom own tion works, and measuresto increasethe produc- productionare tubers, followedby meat and veg- tivity of the crops most important to the poor. etables.The mostimpont non-foodexpenditures Otherpoverty reduction meaes are unrelatedto are detergents, public transporation, soap and agriculture.For example,these householdswould clothing.Although the poor stand to benefitfrom benefit from opportunities for non-farm work. any price decreasesin these commodites and ser- Fourth, the poor (as opposed to the extremely vices rewting from deregulationand trade liberal- poor) have relatively good access to public ser- ization,there is no productor service for which a vices. Thesehouseholds are poor becausethey are subsidywould be well targetedtoward the poor. either in low-paying,low-productivity jobs in the private or informal sector or unemployed.They 11. Slightly over half of extremely poor need more productivejobs to raise their expendi- householdshave access to a publicwater systoem, tures above the poverty level and they are rela- comparedto 80 percent for the total sample. The tivelywell-equipped to take advantageof employ- extremelypoor who have access to a publicwater men opportunities.The executionof the economic system have been more affected by cutbacks in reform program and liberalizationof the labor supply than others. Fewer than 30 percent of the market are what is most important to reduce extr.'nely poor have public sewerage semrice, povertyfor this group. PfMh,the poor wouldben- comparedto slightlyover 60 percentfor the total efit from declines in the price of wheat, rice and sample. Education levels are strongly related to sugar. They wouldalso gain from the elimination poverty.Heads of householdswith less than com- of the ban on second-handclothing imports and plete secondary schooling are more likely to be lower prices for detergentand soap. Finally, the found amongthe poor. Government should reconsider its surcharge

zili scheme on agricultmalproducts. With the excep- the intensityand qualityof servicesis inadequate. tion of dairy products,the scheme eitherhurts the The compositionof professionalskills is not opti- por (sugar, nce, wheatproducts) or benefitsbet- mal, mainlybecause the ratio of nurses/doctorand ter-off farmers(yellow maze and sorghum) nurses/populationis too low. There is also worri- some evidenceof low utilizationof primaryhealth 14. Socal servicescan play an impotant dual care services. This low utilization is likely to be role in alleviatingpoverty. First,improvements in due largelyto the lack of effectivesupply. Part of health,nutrition, and educationdiecty addressthe the explanationmay be user perceptionsof limited worstconsequences of beingpoor. Second,there is value because of poorly traied staff and lack of ample evidence that investingin human capital, equipment,supplies, etc. in healthfacilities. especiallyin education,attacks some of the most importantcauses of poverty.The progmmswhich 17. In nutrition,Peru suffers from relatively have the greatest impact on the poor are primary higb rates of malnourishedchildren and low-birth educationand basic health care. weight babies. Food security generally has de- clinedfor the poorest.The significantincreases in 15. In education,there has been an imptes- the incidenceof acute diarrhealdisease and acute sive expansion in priy enrollments in Peru. respiratoryillnesses also suggestsgreawe precari- Howeverthis was coupledwith a declinein expen- ousnessin livingconditions, decreased effective- ditureslstudentand an exodus of trained teachers ness of preventivehealth measues andthe incmeas- fromthe system.Interal efficiencyis low. Repeti- ingly vulnerable nutritionalstatus of the popula- tion rates are high, particularlyin the early grades. tion. Micronutrientdeficiencies are common in Students spend an average of 13 years in each some areas of the country. A national survey in grade so that by grade 6 almostthree-quarters are 1986/87esfimated that goiteris endemicin 88 per- overage.On average,for each graduatethe system cent of communitiesin the sierraand selva. Small- spendsresources equivalent to 8.4 yearsof school- scale studiesindicate that anemiais a publicheath ig, 40 percentmore than whatwould be necessary care problem among pregnant and lactating in the absenceof repetitionand prematre leaving. womenand children.Ihe responseto malnutrition One problem is the poor quality of pre-primary has excessivelyrelied on food assistanceand given education.In addition,there are wide disparitiesin insufficient attention to complementaryhealth educationalattainment and efficiencywithin the measres, micronutrientsupplemenion, and the country. The lack of supplies and poor teaching promotionof better healthand nurition practices, practices are a patcular problem in multigrade includingbreast-feeding, and child growth andde- schoolsin rural areas.Aother concern is the per- velopmentmonitoring. The Child Growthand De- sistence of a relatively high rate of illiteracy as velopmentProgram of the Ministy of Health has little progress has been made in eradicatingit in comprehensiveand well-foundedobjectives, but rur areas and amongwomen. As a consequence, not the budget or activitiesto achievethem. Other the number of illiteratesincreased by 1.5 percent problems are poor coordination among the between 1980 and 1986. plethoraof agenciesactive in nutrition,a lack of an overallframework for nutritioninterventions, the 16. In health, in spite of an expansionin pri- absence of norms and guidelinesand weak nutri- mary care infrstrue, Peru has some of the tion surveillance. worst indicatorson the continent, particularlyin rural areas.The infant and maternalmortality rates 18. How do the poor make use of social ser- are among the highest in Latin America,particu- vices in Peru? Poverty influences school atten- laMyfor the poor. Many healthproblems could be dance in two ways. First childrenof poor families improved with more concered efforts in primary begin attendingschool at a later age than children healthcare. The full benefitsof the networkof pri- of non-poorfamilies; second, a larger proportion mary care facilitiesare not being realizedbecause of childrenof poorestfamilies never attendschool of a scarcity of equipment,poor distribution of or drop out very early. One of the factors behind personneland a low portionof expendituresspent the relatively high repetition rates for children on the progm. There are gaps in coverageand from poorer households is lower attendance at xlv kindergarten.Altendance rates for femalestudents to purchase drugs, particularlyin the sierra. The are lower than for male students among the ex- lack of access to a public sewerage system is tremelypoor and amongall householdsin the rual dearly relatedto problemsof diarrheaamong chil- sierra dren. While only 17 percent of householdslack seweragefacilities, they accountfor 35 percentof 19. In a programto alleviatepoverty the first childdiarrea casesm priorityfor educationis to increasethe interal ef- ficiencyof schoolsin poor areas.Impwving condi- 21. Programsfor the poor shouldconcentrate tions in rural multigrade schools would particu- on providingmothers and youW chidren with ad- larly benefit the poor. Second, a progam to pro- equate primay health care. It is importantto im- vide textbooksand teachingmaterials and to train provethe supplyof basic equipment,materials and teachers can improve the quality of teaching in medicinesnecessary for carryingout these activi- other types of schoolsin poor areas, increasestu- ties, particllarly in poor areas. Cboleraprevention dent time dedicated to school work and provide and treatmentis a secondpriority and the thd pri- them with more relevant learning experiences. ority would be to strengthen hospitalemergency Third, more emphasisneeds to be placed on pre- service. school education. Steps need to be taken to im- prove the preparationof childrenfor schoolin the 22. Even with improvedsocial services,in the PRONOEIand Government-runkindergartens and short-runsome groupsmay remainvulnerable be- to promote greaterawareness of the value of kin- cause they are alreadyliving on the edge - for ex- dergarten among the poor. One way to atract ample, householdslacking employedworkers or poorer students would be to offer school feeding whose heads can only find low-productivityjobs programsat this level in some localities.Fourth a in household services or agriculture. For these program to teach non-Spaish speaking women people well-designedsocial compensationpro- functionalskills includingliteracy shouldbe part gramscould complement the two mainelements in of poverty alleviationeffors Progams to reduce the poverty alleviation strategy - promotion of illiteracyin this group must considerthat they are broad-basedgrowth and improvedand more equi- widely dispersedand thus hard to reach and that tablydistributed social services. their free time is minimal.Experience from other countriessuggests using NGOs. Tey have the ca- 23. Employmentprogams can reducepoverty pacity to reach deprivedareas and to incorporate by providing employment(and thus inwme) to literacy programsinto ongoingactivities, thereby those in need and can generate additionalbenefits reducingcosts. through the rebabilitationand creation of infra- stutre. Sincepoor peopleare wfllingto work for 20. In health, the coverage of vaccinations a relativelylow wage theseprograms can be cost- among children is low, particulary among the effective.The impact of an employmentprogram poor and in the rural sierra. In general, about half on poverty will be greater if unemployed-headed of all ill people seek medicalcare. Amongthe ex- householdsare found to agreater extentamong the tremelypoor, this dropsto about one in three. The poor than among other groups and if the project gap between the poor and the nonpoor is paticu- improves infrarctu or facilities used by the laMylarge in the sierm Where are the poor likely poor. In Peru, about two-tbirdsof the households to go for care when they arei and whomare they headedby uwmployedare poor and 30 percentare likelyto consult?The poor are morelikely than the extremelypoor. However,households headed by non-poorto use healthcenters and poss, as well as an unemployedperson represent a smallnumber of pharmacies.Nevertheless, the highest proporkion the poor, generally3-4 percent In the urban coast of the poor (30 percent)use hospitals.The poorare and sierra the proportionof the extremelypoor is more likely to consult paramedics, pharmacists somewhathigher - 6-8 percent. It is importantthat and traditionalhealers than the rest of the popula- these programsbe cost-effective.The personsem- tion. But across all expenditure levels the main ployed by a temporary employment program sourceof medical consultationsare healthprofes- should be poor. This can most easily be accom- sionals.Extremely poor householdsare less likely plished if the wage rate is set sufficientlylow that xv that persons with higher paying (and higher pro- derstanding of the nutrient gaps of participants. ductivity) alternatives are not tempted to leave Food assistance coverage among patients with them. The wage rate ensures that an automatic chronic diseases,especially tuberculosis, is inad- self-selectionwill occur. The average wage rate equateand needs to be increased.The coverageof recommended for any emergency employment the schoolfeeding program should be concentrated progam in Peru is the legal minimumrate. An- on a smaller number of children and assistance other design featurethat would promote cost-ef- should be increased to levels which will have a fectivenesswould be the locally demand-driven positiveimpact on absenteeismand desertion. selectionand administationof projects. 26. Two issueswhich arise in the implements- 24. Poor householdsspend an averageof two- tion of the poverty alleviationstrategy are target- thirds of their expenditureson food, and those in ing and institutionalarangements. Targeting seeks the rural sierra spend three-fourthson food. Thus to distribute transfers or other benefits only to protectingthe food security of the poor is a useful those identifiedas the beneficiarygroup in the in- way of protectingliving standards. Food assistance terest of efficiency.The administrativemechanism can also protectthe nutritionalstatus of vulnerable mustnot cost so muchto operatethat it effectively groups such as pregnant and lactating women, absorbsthe savingsfrom excludingthe non-needy. young childrenand the sick and the elderly.Food In Peru, geographyis the indicatorwith the greatst assistance is a significantresource for the social value as a targetingmechanism. By directingre- sectors. In 1991 donor food assistanceresources sourcesto the rral siea, onein everytwo extremely totaled at least $150 million. Food assistance poor householdsin the countrycan be reached. programs provided over 226,000 metric tons of food to at least 9 millionbeneficiaries. Food assis- 27. Some characteristics closely associated tanceto the urban poor is widespreadand substan- with povertysuggest the typesof assistancewhich tial in Lima.Transfers in cash and goods (includ- would mainly benefit the poor. Nearlyall house- ing food assistance)from non-profitorganizations holdswhich draw their water from wellsare poor. represent on average one fourth of total expendi- Other housing characteristicsclosely associated tures in extremelypoor householdsand almostten with povertyinclude the use of latrinesand the ab- percent of expendituresfor poor households in sence of householdlighting or sanitationfacilities. Lima. However, in other areas these transfers aount for a negligiblepercentage of household 28. Peru seems to be an appropriateenviron- expenditures of the poor. The fact that food ment for a socialinvestment fund. Experiencewith assistance does not reach the rural poor suggests social funds in othercountries indicates that effec- that other mechanismsshould be explored,such as tiveness dependsprimarily on the capacity to se- food coupons. Food assistance appears to be lect the right projects,which in turn dependson the relativelywell targetedto the poor in Iima; among qualityof staffand their independencefrom politi- the non-pooronly 6.5 percentbenefit, compared to cal pressures.In August 1991 the PeruvianGov- three-quarters of the extremely poor households ernmentestablished a social investmentfund, the and nearly 40 percent of the poor. Outside Lima, National Fund for Social Compensation and however,more than 40 percent of food assistance Development(FONCODES). Progress in making beneficiariesare from the non-poor. FONCODESoperational was slowand initiallythe 25. The most nutritionallyvulnerable groups confusionoverwhether the agencywas going tobea are pregnant and lactatingwomen and pre-school "demand-driven"or a "top-downeoperation ad- aged children,particularly those under the age of veiselyaffected the qualityof someprojects. How- 2. The coverageof directfood assistanceto moth- ever, in mid-1992performance began to improve. ers and childrenappears adequate, but the low lev- els of assistanceprovided to mostbeneficiaries and 29. The RegionalCompensation Fund which the absenceof complementarynutrition promotion is operatedwithin the Peruvianbudget has the po- and health care activities makes it unlikely that tential to contnbuteto a poverty alleviationstrat- they havea significantimpact on nutritionalstatus. egy. However,currently no attempt is made to di- Also, beneficiary rations are not based on an un- rect more resourcesto poorer areas. A change in xvi the allocationfomala would only be feasibleand 33. There are two sourcesof financingfor the meaning with higher tax revenues and fic povertyalleviation stategy, govemnmentand do- stabilization. nos In orderto attract sufficientdonor resources the Governmentmust demonstate clearly its owt 30. The magnitude and seriousness of the commitmentto poverty alleviation. In 1990 it is povety in Peru call for the active participationof estimatedthat totalgovemment spending onhealth all esources,including Non-Governmetl Orga- and educationreached only 2.2 percentof GDP in nizations (NGOs) They have several strengths. Peru. As tax reformand the measuresto strengthen They can mobilizeand utilize substantiallevels of tax administrationimprove revenue generation,it resoures They are also able to reach undererved is rcommended that the Governmentendeavor to populationseffectively. NMOs are a strong con- increasethe resourcesallocated to healthand edu- stituencyfor the promotionof povertyalleviation cation.Spending on educationand healthof about goalsand progm here are manyNGQs in Peru 4 percentof GDP is the averagefor LatinAmerica coveing diffient fieldsand easof the counry. and about the level that was spent in Peru during the first half of the 1970s. 31. NGOs also bave some weaknesses.They are often vulnerable to donor influence and have 34. The Governmentof Peru has taken an im- difficultyformulating policy proposas. Collabora- portantstep to reducepoverty. Over the mediumto tion with the Govemmentis constrainedby real long-runwhat is neededto alleviatethe high levels and perceivedbarries The dispersionof Govern- of poverty is broad-basedgrowth That is the ob- ment responsibilitiesover a broad range of state jective of the stabilization and adjustment pro- gencies makes NGO eny difficult.Also NGOs grm. However, this is not enough to alleviate differ as to their mangement and adnistrative poverty,particulary in the short-rtn Measuresare capacities. Some Government agencies view needed to protectvulnerable groups and to ensure NOOs as competitos instead of potenial execut- that the poor are able to take advantage of the ing agencies.Mistrst hinders collaboration.The greateropportunities in the reformedeconomy. In Giovernmentshould establish clear and efficient order to address these latter needs,this report has mechamn for NO cMollaoatio mng NGO examinedthe profile of the poor in Peru and rec- consorta provideone mechanismwhich should be ommendedpriority policies and progms targeted explored for fostering greater coordination,dia- to them. The Oovernmenthas made a start in this logue and joint plannin with the Government. area. Its povertyalleviation strategy sets clear and aoprit prioritiesand goalsfor povertyreduc- 32. The Government ecentlyprepared a pov- tion efforts within a fiameworkof a continuation erty alleviationstrategy. The purpose of the docu- of economicpolicies that would maintainmacro- ment is to set the pricritiesfora povenyalleviation ecnomic stabilityand promotegowth. The chal- programin Per. It is generallyconsistent with the lenge and test of the Govemment'sresolve will be fidngs and recoendatons of the analysiscon- to what extentcutrent and futurepolicies and pro- tamed in this repot The stratea reogizes that grams are governed by that strategy. The recent the allevation of povertyin Peru demandsessen- improvementin PONCODESis an encouraging tially an economic strategy. The objective is to sign. But in order to ensure that the poor reap the improve the living conditions of the extremely benefits of adjustment, the Government should poor tbrougb spm in the areas of food assis- strengthen the social services most critical for lance, health, educationand employmentgenera- them and providemore effective safety nets for the tion precs mostvulnerable.

xvii PERM.Socia Seito Cmuponeas fer PovertyAflwlatloa Sb*at

Low attendane rates for poor Offer schoolleeding programsi in Sdw eoeFuftn Prf*a children,partculary in the rura theseareas Sieram MaaingadeSdwoo Project Highlepetiton rate, pdatiulary for Inceas the Internalefficiency of MiVJ*lrd $dwdoProjea* fthpoo and in rural aesschools in poor areas, in particularby Improvingconditions In rural Textook PM*jct mutgaeschools.Provision of textbooks;training mafteras, and teachertraining geared to the reqirmetsofmuligrad. schools and for other schoolsin poor -waaurban and rural areas& Improvequality and attendance,in &StrengateigIWPre nWy pre.scoo program. Improve Educati Series Pr*jct preparationof chidren for schoolin the PRONOITand Govefnment-r

persistentr higllitry rate among Supporta programto teachthem LMteCaProjec non-Spanishspeaking woume basic functionalskills, ielyingon N100s.

Qualit of car for vulneable grops - Providepre-natal anti-Otetnus mothrs and hiuldren- Is low, and hronimunizton uplements, prticully in terms of the mber of monitorweiSgt gain and providefood contacts supplement,if needed,and trie personne for high-riskdeliveries. Train birth attendantsfor low-risk deliveries. Inrease the covaeageand ftequeny Chid Growthand of community-based growt Developmet Program"R) moitrig,partiuary amon chiodrenunder two.

l dest.in Italcs refr to projerpvUafleproposed by the Gownwwu.m hogein iardrwdpim referto dng Gatea pGeiwprogracL.

xiI PEU: Socl Sector Copoeut for Poverty Aleviion Staiq Smnmay Table

Jasuesddh Gormalogrm

Decliningand lowcoverage of Effors shouldbe made to increase ChildGrowth and vaccinationsfor chdken underone the vaccinaioncoverag for children DevelopmentProram (MOE) underone hruwghheaflt faciitie Low coverageof familyplann Stroengthnfamily planning educai p as and incrs accessto contraceptive. Lowutlaton of prma healthcme Rehabilitatedeterioraed faoilites ma &R8*0*89 PbMWYCnm servicesby the poor, partularly In improvethe supplyof basie Bsc He and MNuAm the siera equipmentand drugs in hoalth facilides. Poor ditrbution of professionl Examinemanagement and incentive Hma Resore resources.There are not enoughhealfth isaus for staff.Upgade staff housg Devkpn fAre profssionalsin the least-developed for pnmary healthcare facilides. EJlru, Po" regionsnor ough wkg in primaryhealth care. There are not Promotegreater us of local BasicHMMt and NMwo enoughnurses rsoues troupg the provisionof tainig for nonprofesionals for example. Sincepoor households have lower Promotelow-co systemsfor water Lad, and Edaadon in accossto publicwater and sewerage supplyand sanitation.Where no Sankado systems,their childreanare more likely sanitationfacilites exist, latrnes to suffer from diarwhea shouldbe built. Cholerais a serioushealth problem Strengtheneducation campaips, Lats and Educaon in that affts parily the poor ensureavailability of suppliesto treat Sadn acueases,wj and rehabilitatewater systemsand intl latines. in someareas, a substantialproporton Strengthenhospital emergency of the poor use hospitalservices servicesin poorareas

Absenceof normsand sndards for Tne Governmentshould develop and nutritioninterventons monitorcompliance with sandards in three key areas: O unit costs for intervention;(i) foodassistance programmingpractices; and (i) informationcollection and impact evauation. No agencyof the Governmentis The Governmentshoult assignthis responible for mobiliing resources fucion to the Ministryof Health.

*Il in itaics referto proct pres proposedby dw Govenmeant hosein sndwd ptn referto exisfnS Govwnmewprrmr.

xx~. PERU: Soda Sector Componet for Povet Afevim Stratey -y Tab

I ssues Reconumendatim Govenment Prorm'

Lack of a sysem for nutrition SmallsentiDe populaton suveilance on a naionallevel represontaveof importnt taget groupsshould be monitored.

Absenceof a frameworkfor nutriton Governmentshould implement inteention leads to duplicafionof nutritionpla functions,a lack of attentionto acheving major improvementsin nutrit statusalid less than optimal resourceuse Patternof growth tring amoong EC campaignshould be developedto ChildGrowth and childen duringweaing period promote ppropriatewaning DevelopmentProgram (MO)D prctios. Goiteris endemicin a8 pereut Of Goverment shouldconider NationalPogram to Control communidesin the sierra and selva. subsidizingthe incerasein production Goiterand EndemicCretinim Anemiais also a publichealth problem costs of iodizedsalt aad involvesmall (PRONABCE) as is vitaminA deficiency produetionfacilities which market salt fromnatura sources.

Iron supplementationfor pregnant women;semi-anal vitaminA distributionfor children.

Food As e Foodassistance does not reach The Governumentshould examine Vasode Leche important targetgroups amongthe alterna mchasms to reacbthese poor, particularlyhouseholds in the groupssuch as foodcoupons and rural sierra wherefood security has introducepilot program. The Vasode declined.Also, the largestproportion Lecheprogram should strengthen the of beneficiariesof the Vasode Leche qualityof managementand control. programare school-agedchildren. Nearly20 percentof its beneficiaries Foodasdssnce to mothersand PANFAR are not poor. childrenshould be expanded. Somecommodities in foodassistance The leastexpensive sources if programsare not the leastexpensive nutrientsacceptable to benefi.;aries sourcesof nutrients,for example,milk shouldbe selectedas commodities. in the Vaso de Lecheprogram Alsostaple foods used in assistanc programsshould be fortifiedwith iron.

r*ae.r in ltalics refer to projectprofiles proposed by the Gowrment. hose in standardprint referto wxisng Gowernmntprograms.

xzci PERU: Soca Sector Componeb for Povr AblviaUonStrate -UMMTable

Jmes GovermenProam Impactof foodassstance on Beoficiary raions shouldbe based nuridonalstatus is likelyto be on an undersandingof the nutrient negligibl, gapsof participants.

Food asistance needsto be PANFAR acompaniedby complementay healh and nutrition ducation measures,Including breast-feeding and childgrowth and dovelopment monitouing.

La Escola DefiendeLa Vidadoes not Programshould rt chidren in La BEcuelaDefiende La Vida reach poores and mostvulonerable and grades 1.4 and thosein pow areas, has mi nutriional impact insteadof concetratg in I3ma. Prgram shouldonsmr otnuity i feeding.

Employmnt Program Howto ensurethat programreaches Programshould be designedto FONCODES poorwhile meetingpoverty alleviation encourageself-selection by the poor. objctdveswith minimum Whileless than the tinimum wage administrativecoat wouldimprov targeting,it has some disadvanaWs.Povrty alleviadon objectivesmay not be met and the participationof pdvatecontractors wouldbe discouraged.For thos reasons,the minimumwage should be offered.Codsderation should also be givento geographicatargeting. Adminstrationof employment Demand-drivenand promotionof the FONCODES programs use of privateconactors.

* Malesin Italcs rfr to projectprofies proposedby die Govermwn Those n sard prpntrer to edsLng Goewnentprogranm.

xxii Resumeny conclusones

1. Unode los legadosdel pobredosempftio econdmico del Perd es la persistenciade eledos nivelesde pobrezaen algurasregiones y grupos tnicos, y la dismlnucidnde los nivelesde vida de otros. For variasrazones, la inlroduccidn de importantesreformas ocondmicas ofre una oportunidad niucatanto al Goblerno comoa quienesproporcionan apoyo financieto al Perd pars consider un programa de allviode la pobreza. Primero,la experienciainternacondl muestra quo es imposiblelograr una reduccl6nsignificativa y duraderade la pobrezasin un crecimientoecon6mico de base amplia. Hastaahora en el Perd no exstfanla estabilidadmacroecon6mica, el marcoadministrativo l la estcra de incenos que conducenal crecimiento.Esta situacionesta cambiando,y conrapidez. Segundo,una de las razonesdel escasocrecimiento registrado era la participacion excesivadel Estadoen la actividadecondmica. Un elementodave de las rformas es la reduccidndel papeldel Estado. A medidaque el sectorpdblico se redre de los soctres en que su intervencidnafectaba negativamento los resultados-trabajo, mercadosflncieros y produccion-,liberara recursos para otraseras en que puedey debehacer un aportepositivo, a saber,la prestacidnde serviclospdiblcos tales comoeducacion y salud. Tercero,en el medianoy largoplazo, la aplicacid de las reformasdeberfa reducir la pobrezaal ofrecermas oporxnidadesa los pobres para utilizarso activos,como la manode obra y las tierras. Sin embargo,en el cortoplazo es posiblequo algunasde las medidasque se apliquenocasionen dificultadestransitorlas. Tambien le tomaratiempo a la economfaresponder a la liber;izaci6nde los mercados,y algunosgrupos no estanen condicionesdo participarplenamente. Se requiereaplicar un programade aliviode la pobreza- temacentral del presentoinforme- dirigido a gruposespecfficos, con incluslonde serviciossociales, destinado a protegera los gruposmis vulnerablesy brindar asistenciaa otros,a fin de que puedanaprovechar las nuevasoportunidades que se presenteny estees el temacentral de este informe. 2. Las posibilidadesprometedoras quo tenfa el Perl se vieronfnstadas durantetreinta afilos de fuerteintervencion estatal en la actividadeconomica, que se tradujoen la institucionde una estructurajurfdics y un entornoen materiade polfticasque eranincompatibles con el crecimiento.El crecimientoexcesivo del Estadocondujo a un despilfarrode recursosy al estancamientoecon6mico. Debido a la fWtade crecimientoy a la existenciade mercadosrfgidos y corrados,muy pocos gruposdel paftstendan acceso a los elementosnecesarios para mejorarsus mediosde vida. Las polfticaspdblicas distorsionaron casi todos los mercadosde la economfa. La polfticacomercial se fundamentabaen una estrategiade sustitucionde importaciones,con elevadosaranceles y muchasbarrens no arancelarias.La disminucidnde la recaudaci6nde impuestoscontribuyd al deteriorodel desempeho fiscaly redujola disponibilidadde recursospara el funcionamientoy mantenimiento de los serviciospdblicos esenciales, y la inversionen los mismos. Inclusolas polfticasy programasque se establecieronpara beneficiara los trabajadoresy agricultores,como las leyesde estabilidadlaboral y la reformaagraria, resultaron pejudiclalespara esosgrupos y para la economia.Los salarlosreales cayeron despudsde 1970 y disminuy6el empleoen el sectorfonnal. El sectoragrfcola creci6tan sdloa la mitadde la tasade auirentode la poblacidn,brindando muy pocas oportunidadesara que mejoraranlos ingresosde la poblacidnrural.

3. La explicacin que dio el PresidenteGarcfa de la situacionde Inflaci6ncrdnica y lentocrecimiento era quo el Perd se encontrabaatapado en la wtrapa de la deudae:el serviciode la deudaextrna acelerarfaIt devaluacidny Ia inflaci6n,acabarta con el ahorrointerno y reducirfala Inversi6n.En consecencla, su Goblerno(1985-90) establecid un topepara el serviciode la deudaexterna, ijindoloon el 10%del valorde las exportaciones,y mis adelantesuspendid totalmentelos pagos, Impusoamplios -controles cambiarios y aumentoel proteccionismo.Asimismo, el Gobiernoaumento el salariomflimo, el creditoy las subvenciones,redujo la tam del impuestoal valor agregadoy dispusouna rebajade los preciosdel sectorpiblico. Esta estrategiaprovoco un elevadodeficit fiscal, quo se financi con la creacionde dinero. A partirde 1987el mercadoperdio conflanza,se produjoun rebroteinflacionario y el productode la economfay los salariosreales disminuyeron drusticamente. La desacertadagestion econdmica alcanz6el puntoculminante en el perbodo1989-90 y se reflejden la biperinflacion. 4. wospobres resultados de la economfason un factorque explicalas malascondiciones de saludexistentes en el PenS. Entretodos los palsesde Ia region,el Perd ocupacasi el lltimolugar en los indicadoresde salud. En 1991la epidemiado c6lerademostro en formadramatica las precariascondiciones sanitarias imperantesen todoel pafs. Sin embargo,en materiade educacidn,esti cercadel primerlugar en lo que respectaa la proporcionde alumnosmatricutados en escuelas prtmariay secundaria,pero el elevadoporcentaje de repitentesy la malacalidad de la ensefla representanun problema.La disponibitidadde recursospdblicos para prestarsrvicios socWalesva en disminuciondebido a la bruscareduccidn del gasto pdbico globalen los iUtimoscinco aflos. La declinaci6nse explicapor varios factores,entre los que figurala graverecesi6n y la desaparicioncasi por completo del sistemaimpositivo del pats. Se protegi6a los sectoressociales de las reduccionesmais extremas que experimentaronotras esferas pdblicas. Sin embargo, debidoa la disminucionde la disponibilidadtotal de recursos,el Perd suelegastar menosquo sus paksesvecinos en estossectores. 5. En agostodo 1990,el Gobiernode Fujimoripuso en marchaun programade estabilizaci6ny reformaestructural de tipo ortodoxodo gran envergadura.En la etapainicial, el Gobiernoadopt6 dristicas medidas desdnadas a estabilizarla economfay restablecerlos lazoscon la comunidadfinanciera internacional.La segundaetapa de la reformade polfticascomdz6 en marzode 1991y se baso en la adopcidnde una seriede medidasestructurales de base amplia. El propositode talesreformas era promoverla competitividady desreglamentarla actividadecon6mica. Para reducir la pobrezase requierenvarias reformas fudamentales. La capacidaddel Perd para invertiria bruscacabda de las recaudacionesimpositivas es el puntocentral para recuperar el equilibriofiscal. Tambidnresulta vital a fin de garantizarla capacidaddel Gobiernopara proporcionarservicios bisicos en el largo plazo. El aumentodel empleoy los salariosen las zonasurbanas es un factorimportante que determinarIl01 ritmo que tendrala reduccidnde la pobreza,tanbo por su influenciadirecta en la actal poblacidnurbana pobre como por las oportunidadesque crmepan la mlgracidn desd. las zonasrurales. La mayorneutralidad del sistemade comercopuede respaldarun nwdelode expansionindustrial con un uso m4sintensivo de manode obra en los sectoresque compitencon las Importaciones,comno tambien en e1de las exportaciones.El Goblernoha avanzadocon rapidezy arrojopara liberaliw el sistemade comercio.Sin embargo,la reformade este sectorresuta mis eflcazsi los factoresde produccidntienen libertad para responder a las fuerzasdel mercado. En el casodel Perd, las polifticasestatales existentes desde el deceniode 1960han restringidofuertmemnte el movimientode los factores. Parapoder ofrecor mis oportunidadesde empleourbano, una reformadave es la desreglamentacdndel mercadode trabajo. Las poifticasde fomentodel desarrollorural tamblen constltuyenun elementoimportante de la estrategiapara aliviar la pobreza.

6. El crecimientoes un com onenteimportante de esta estrategia,pero no el 1nico.Es precisoaplicar progranu que permitanasegurar que los pobres puedanaprovechar las oportunidadesde arabajoque se presenteny protegera algunosgrupos vulnerables que no estAnen condicionesde participarplenamente en la economfa.Para poder disefar estostipos de programasse requiereinformaci6n sobrelos pobres. El perfilde pobrezadel Perd quose elaboraen el present. informese basaen datosrecogidos en una encuestasobre.niveles de vida realizada en los mesesde octubrey noviembrede 1991. Se calculandos Ifneas de pobreza. Las unidadesfamiliares se clasificancomo extremadamente pobres si sus gastos totalesson inferioresal costode una canastab$sica de alimentos.Se clasifican comopobres si sus gastostotales son inferioresal costode una canastabasica de alimentos,mis una estimaci6nde los gastospor conceptode artfculosno alimentarios.Aproximadanente la mitadde la poblacidnestudiada cae bajo la lIfnea superiorde ambasIfneas de pobreza,y la quintaparte cae bajoe0 Ifnite inferior. 7. 4Pdndeviven los pobres? Al compararlas cuatroregiones del muestreo(LJima, litoral urbano, y sierraurbana y rural),la incidenciade pobrezaes mayoren la sierra rural. En esa zona, dostorcios de las unidadesfamiliares son pobresy el 47% del total sonextremadamente pobres. Tan sdloel 10%de las unidadesfamiliares de Limason extremadamentepobres, en comparaci6ncon alrededordel 20% en el litoralurbano y en las regionesde la sierra. En cadauna de estastres regiones,el porcentajede unidadesfamiliares pobres alcanza a aproximadamenteel 50%. Se estimaque la sierra rural albergala mayorcantidad de personasextremadamente pobres (1,6 millones).Por su tamalho,en Limavive el mayorntlmero de pobres(2,5 millones).

8. Por lo general,una personapobre vive en una unidadfamiliar mas extensaque una personano pobre, que en promedioincluye a uno y medio individuosm4s. Cadatrabajador perteneciente a un hogarpobre sustenta a cuatro miembrosde la familia. En el casode los no pobres,esta cifra baja a tres. El consumode alimentosper ctpita en una unidadfamiliar extremadamente pobre representas6lo el 40%, en comparaci6ncon el de otros hogares,y la misma relacionen lo que respectaal consumototal alcanza al 30%. 9. Entrela poblaci6npobre, s6lo en Limase registrauna desproporcion deuriidadesfamiliares encabezadas por una mujer. Las poblacionesindfgenas en el Perd estAndesfavorecidas. Consideradas en conjunto,las unidadesfamiliares zeabezadaspor hablantesde quochuay almar representanel 40% de la poblaci6n extremadamentepobre y alrededordel 25% do la poblacidnpobre. En Lima, es algo mis probableencontrar inmigrantes quo no inmigrantesentre los pobres. Estos so encuentranmayoritariamente en dos agrupaclonesocupacionales: los trabajadores independientesy los trabajadoresdel sectorprivado. 10. Los aimentosbsioos quorepresentan la mayorproporclon de los gastosdo la poblacionpobre son 0l aoz, el pan, el azIcar, los tubEculosy la came. Los alimentosmas importantesque so obtienenmediante produccidn propia son los tuberculos,seguidos de la came y las hortalizas.Los gastosmis importantespor conceptode productosno alimentarioscorresponden a detergentes, transportepdbilco, jab6n y vestuario. A pesardo que los pobresserin los boneficiariosds cualquierrebaja do lospreclos de estosproductos bisicos como resultadode las medidasde desreglamentaciony liberalizacion del comercio,no hay nhngdnproducto ni servicioque puedasubvencionarse con objetodo beneficiaren verdada los pobres. 11. Pocomis de la mitadde las unidadesfamiliares extremadamente pobrestienen acceso a una red p6blicado abastecimientode agua,en comparacidn con el 80% de la muestracompleta. La poblacion madamente pobrequo tiens accesoa una red piblica se ha vistomis afectadapor las reduccionesdel sumnistro quootros grupos. Menosdel 30%de los extremadamentepobres cuentan con serviciopdbllco de alcantarfllado,en comparacioncon pocomis del 60% de la muestracompleta. Los nivelesdo educacionestin estrechamontevinculados a la pobreza. Es mis probableencontrar jefes de hogarque no han completadola enseflanzasecundaria entre los pobres. 12. En lo que respectaa las unidadesfamiliares quo poseen tierras, los pobrestienen aproxhnadatente 50% menossuperficie quo los no pobres. Cercado un terciode los hogarespobres poseen tisrra do regadfo,en comparaci6ncon alrededorde la mitadde los no pobres. En el casodo las unidadesfamiliares mis pobresdo la s3errarral, todoso al menosla mitaddo los trabajadoresestan ocupadosen actividadesagrfcolas, en tanto quoen el do los hogaresquo gozan do mejorsituaci6n, una mayorproporci6n de los trabajadoresrealizan otras laboresen formaindependiente. Los principalescultivos quo producen los pobres de la sierra rural son la papa, la cebaday el trigo. Los pobresproducen principalmente ma£z blanco,cebada, papas, oca y olluco. 13. El perfllde pobrezaofrece una orientaciongeneral para la formulacidnde una estrategiade aliviode la pobrezaen el Perd. Primero, los programasfocalizados en gruposespecfficos deben centrarse en los extremadamente pobres,que representanla quintaparte de la poblaci6nestudiada. Segundo, los programaspara aliviar la pobrezaentre la poblacionextremadamente pobre deben dirigirsea la sierrarural y a la poblaci6nindfgena. Pars reducirla extremapobreza es precisomejorar el accesoa los serviciospdblicos, en particularla educaci6n. Estasunidades familiares estin mis expuestasa presentarun estadoprecarno de saludporque carecen de aguapotable e instlacionesde saneamiento.Tercero, el trabajoagrfcola se asociapositivamente a la pobreza. Es posibleque la adopci6nde ciertaspolfticas agrfoolas permita reducir la pobreza,como lIs reformasrelativas a los tftulosde propiedadde tierras, la rehabilitacidno construccidnde obrasdo riego en pequefiaescala y medidaspara aumentarIs productividadde los cultivosms importantespara los pobres. Otts medidasdesdnadas a reducirls pobrezano so relacionancon la agricultura.Por ejemplo,esas unidades famuliarms so beneficiaran. de oportunidadesde trabajono agrfcola. Cuart, los pobres(a difercia do los extremadamentepobres) tienen un accesoreladvamente bueno a los serviclos piblicos. Sonhogaes pobresya sea porquesus integrantestienen trabzjos mal remuneradoso pocoproductivos en el sectorprivado o el sectorInformal, o porque esthndesempleados. Requieren trabaos mEsproductivos para elevarsus gastospor encimadel nivelde pobrezay estin relativamenteblen preparados parn sprovechar las oportunidadesde empleoquo se presenten.La ejecucidndel programade reformaecondmica y liberalizacidndel mercadode trabajoes el elementomAs importantepara reducirla pobrezaen estegrupo. Quinto, los pobresse beneficiarfansi disminuyesoel preciodel trigo, el arroz y el az4car. Asimismo,se verfanfavorecidos si se eliminarala prohibicionde importarropa de segtnda mano y se rebajaranlos preclosde los detergentesy el jabda. Per Iltmo, el Gobler=o debe reconsiderarel sistemado sobretasasque se aplicaa los productosagrfcolas. Conla excepcidnde los productoslicteos, el sistemaya sea perjudicaa los pobres (azdcar,arroz y productosderivados del trigo)o beneficiaa los agricultoresquo gozande mejorsituacidn (malz amarllo y sorgo). 14. Los serviciossociales pueden desempeflar una importantedoble funcidnen el aliviode la pobreza. Primero,las mejorasen las esferasde la salud, nutci6n y educacionabordan directamente las peoresconsecuencias de ser pobre. Segundo,existen abundantes pruebas de que la inversi6nen capitalhumano, especiaWmenteen educacion, permite atacar una de las causa is Importn de la pobreza. Los programasque tienenmayores repercusiones sobre los pobresson los de educaci6nprimaria y atencidobsica de la salud. 15. En materiaeducacional, la matrtculaon la ensefaiazaprimarla ha umeontadoen formaimpresionante en el Perd. Sin embargo,esto incrementoestuvo acompafiadocon una disminuci6nde los gastospor alumnoy un exodode profesorestitulados del sistema. La eficienciaintna es baja. Los porcentajesde repitentesson altos,particularmente en los primerosgrados. Los alumnospasan, en promedio,1,3 afiosen cada curso,por lo que en sextogrado casi tres cuartaspartes do los estudiantesestEn por sobrela edaddebida. En promedio,por cadaalumno que egresa,el sistemagasta recursos equivalentes a 8,4 afiosde escolaridad,es decir, 40% mis que lo que serfanecesario si no hubieserepitetes ni desercidn escolarprematra. Un problemaes la malacalidad de la educacionpreescolar. Ademnis,existen marcadas discrepancias en cuantoa los logroseducacionales y la eficienciaa travesdel pafs. La faltade materialesy las malastdncas pedagdgicas constltuyenun problemaparticular en las escuelasde enseftanzasmultnea de varios gradosen las zonasrurales. Otro aspectoque preocupaes la persistenciado una tasa rlativamentealta de anaifabtismo,ya quoso ha progresadopoco en SU srradicacitnon las zonasrurales y entrelas mujerm. Comoconsecuencia, el ndmerode analfabetosaumento en 1,5%entre 1980y 1986. 16. En la esferade la salud,a pesarde la expansionde la infraestructura para la atencidnprimaria, el Perd registraalgunos de los peoresindicadores del continente,particularmente en las zonasrurales. Las tasasde mortalidadmaterna y de nifiosmenores de un aflose cuentanentre las mas altas de AmericaLatina, en especialen la poblacidnpobre. Podrtanresolverse muchos problemas de saludsi se concentraranmayores esfuerzos en la atencidnprimaria. No se materializantodos los beneflciosdel sistemade establecimientosde atencidnprimaria debido a la Insuficienciade equipo,la maladistribucion del personaly la escasaproporci6n de recumsosque se destinana los programas.En general,la coberturano es adecuada, tampocola intensidady calidadde los servicios. Lc composicidnde especialidades no es la optima,principalmente porque la relacionenfermneras/m6dico y enfermeras/poblaciones demasiadobaja. Asimismo,existen datos inquietantes relativosa la escasautilizacion de los serviciosde atenci6nprimaria de salud. Es probableque tal situacionse debaen granmedida a la faltade una ofertaefectiva. En parte, la explicacidnpuede estar en que los usuariosle atribuyenun valor limitadoa tales serviciosdebido a personalno bienpreparado a la faltade equipo, suministrosy otros en los establecimientosde salud. 17. En lo que respectaa la nutricion,el Perd registratasas relativamente altasde malnutrici6ninfantil e insuficienciaponderal del reciennacido. En general, la seguridadalimeonaria para los mis pobresha disminuido.El importanteaumento de la incidenciade enfermedadesdiarreicas agudas e infeccionesrespiratorias agudas tambienrevela un mayordeterioro de las condicionesde vida, menoreficacia do las medidasde saludpreventiva y un estadonutricional de la poblaci6ncada vez mas vulnerable.En algunaszonas del pafsse observauna deficienciade micronutrientes. Segdnuna encuestanacional realizada en el perfodo1986187, se estimaque el bocio es una endemiaen el 88%de las comunidadesde la sierray la selva. Estudiosen pequefiaescala indican que la anemiaconstituye un problemade saludpdblica en las mujeresembarazadas y en perfodode lactancia,y en los niflos. La respuestaa la nalnutrici6nse ha basadoen medidaexcesiva en la asistenciaalimentria, prestdndosepoca atenci6na la adopci6nde medidassanitarias complementarias, la suplementaci6nde micronutrientesy la promoci6nde mejoresprkticas sanitariasy nutricionales,incluida la lactanciamaterna y la vigilanciadel crecimientoy desarrolloinfantil. Los objetivosdel Programade crecimientoy desarrolloin&ndl del Ministeriode Saludson ampliosy bienfindamentados, no asf el presupuestoni las actividadespara lograrlos. Tambidnexisten problemas de mala coordinaci6n entre los numerososorganismos que actianen el campode la nutricidn,falta de un marcoglobal para las intervencionesen esta materia,ausencia de normasy directricesy deficienciade la vigilancianutricional. 18. LC6mousan los pobreslos serviciossociales en el Perd? La pobreza incideen la asistenciaescolar en dos formas. Primoro,los nifiosde las familias pobrescomienzan a asistira la escuelaa una edadmis tardfaquo los de las familias no pobres;segundo, una mayorproporci6n de niflosde las famuliasmfis pobres muca asistea la escuelao abandonalos estudiospo. Uno de los factoresque influyeen los porcentajesrelativamente altos de repitentesprovonientes de las unidadesfamiliares mas pobreses la baja asistenciaa las guarderfasinfantiles. Las tasasde asistenciade las niflasson mAsbajas que las de los niflosen la poblaci6nextremadamente pobre y en todaslas unidadesfamiliares de la sierra rural. 19. En un programapara alivlarla pobreza,la primeraprioridad en materiade educaci6nes aumentarla eficienciaintema de las escuelasde las zonas pobres. El mejoramientode las condicionesen las escuelasrurales do enseffanz simultdneade variosgrados beneficiarfa de maneraespecW a lospobres. En segundolugar, medianteun programade entregade librosde textoy materiales didicticosy de formacifnde profesoreses posiblemejorar la calidadde la ensefianzaen otros tiposde escuelasen las zonasrurales, aumentar el tiempoquo los alumnosdedican a tareasescolares y ofrecerlesexperienclas de aprendizajemis pertinentes.En tercertdrmino, es precisohacer mis biucpid en la educaci6n preescolar.Deben tomarse medidas para mejorar la preparacidnde los niflospara la escuelaen las guardertasinfatiles estatalesy del ProgramaNo Escolarzadodo EducacionInicial (PRONOEI) y fomentaruna mayorconciencia entre los pobresde la importanciade la educaci6npreescolar. Un ostrategiapara atraera alumnosmns pobresserfa ofrecer programas de alimentacidnen eatenivel en alganaslocalidades. En cuartolugar, comoparte do los esfuerzospor aliviarla pobrezase deberfacontar con un programapara impartirlnstruccidn funclonal a las mujeresque no hablan espaflol,con inclusi6nde la slfabetlzacidn.Los programasdirigidos a reducirel anafabetismoen estegrupo deben considerar que estaspersonas estin muydispersas y quo,en consecuencia,resulta diffcil llegar a ellas,y tambienque su tiempolibre es mfnimo.La experienciarecogida en otrospasses sugiere recurrir a las ONG. Estas organizacionestienen la capacidadpara Ilegar a zonas marginaas e incorporar programasde alfabetizaci6nen sus actividadespemanentes, lo quopermite reducir los costos. 20. En materiade salud,la coberturade vacunacin para los nifioses reducida,particularmente entre los pobresy en la sierra rural. En general, aIrededorde la mitadde las personasque se encuentranenformas solicitan atencid mEdica. Entrelos extremadamentepobres, esta cifradisminuye a alrededorde una de cada tres. La brechaentre los pobresy los no pobreses especimentegrande en la sierra. ,A ddndesuelen recurrir los pobrescuando se enfermany a quiensuelen consultar?En comparrcidncon los no pobres,existen mis probabiidadesde quo los pobresrecurran a los centrosy postasde salud,al igualque a las fmacias. Sin embargo,la proporcidnmis alta de los pobres(30%) recurre a los hospitales.Es mis probableque los pobresconsulten a personalparamedico, rmactcos y curanderostradicionales que el restode la poblacidn.Pero a lo largo de todoslos nivelesde gasto,la fuenteprincipal de consultasmedicas son los profesionalesde la salud. Hay menosprobabilidades de que las personaspertenecientes a unidades familiaresextremadamente pobres adquieran medicamentos, particularmente en la sierra. La faltade accesoa un sistemapdblico de alcantariliadoestL aamente relacionadacon los problemasde diarreaen los niflos. Si bien s6loel 17%de las unidadesfamfliares carecen de alcantarillado,reprentan el 35% de los casosde diarreainfantil. 21. Los programasdirigidob a los pobresdeberfan concentrarse en proporcionara las madresy a los niflospequeflos una atenci6nprimaria de salud adecuada. Es importantemejorar el suministrodel equipobisico, materialesy medicamentosnecesarios para llevara caboestas actividades, particularmente en las zonaspobres. La segundaprioridad es la prevencidndel colersy su tratamiento,y la tercera,el fortalecimientode los serviciosde urgenciade los hospitales. 22. Inclusosi so mejoraranlos serviciossociales, en el corto plazoes posibleque algunosgrupos sigan siendo vulnerables debido a queya vivenen condicionesextremas -por ejemplo,las unidadesfamillars en las que no hay ttabajadoresempleados, o las encabozadaspor personasque solopueden encontrar trabajospoco productivos en el serviciodomestico o la agricultura.Para estas personas,la existenciade programasde compensacionsociales blen concebidos podrfacomplementar los dos elementosprincipales de la estrategiapara el aliviode la pobreza-promocidn de un crecimientode baseamplia y mejoramientoy distribucidnmgs equitativa de los serviciossociales. 23. Losprogramas de empleopueden reducir la pobrezaal proporcionar trabajo(y, por lo tanto,ingresos) a los necesitados,y generarbeneficios adicionales a travesdel mantenimientoy creacidnde infraestructura.Estos programas pueden ser eficientesy de costomfnimo porque las personasestan dispuestas a trabajarpor un salariorelativamente bajo. Su efectoen la pobrezasera mayorsi las unidades fbmiliarescuyos jefes de hogarse encuentrandesempleados pertenecen en mayor medidaa gruposde poblacidnpobre quo a otros grupos,y si con el proyectose mejorala in aestucturao las instalacionesque utilizanlos pobres. En el Perd, alrededordo dos torciosde losjefes de hogarque se encuentransin empleoson pobres,y el 30% sonextremadamento pobres. Sin embargo,las unidadesfamiliares encabezadaspor una personadesempleada representan una pequeflaproporcidn de ios pobres,por lo generalentre el 3% y el 4%. En el litoralurbano y en la sierra, la proporcionde poblacidnextremadamente pobre es algosuperior -entre el 6% y el 8%. Es importanteque estosprogramas sean eficientes en funci6nde los costos. Las personasempleadas en el marcode un programade empleotemporal deben ser pobres. Estose puedelograr mas facilmentesi la escalade salariosse fija en un valorsuflcientemente bajo para quolas personasque tienenalternativas de mejor salario(y mayorproductividad) no se sientantentadas a dejarlas. La escaa de salariosasegura una autoseleccidnautomatica. La escalapromedio que se recomiendapara un programade empleode emergenciaen el Perd es el sueldo m&imolegal. Otracactftica de diseffode un programade estetipo que oimentarfala eficienciaen funcionde los costosserfs la selecciony administrad6n de los proyectosimpulsadas por la demandalocal. 24. Las unidadesfamiliares pobres gastan, en promedio,dos tercios de sus ingresosen alimentaci6n,y las de la sierra rural, tres cuartaspartes por ese concepto.En consecuencia,la proteccionde la seguridadalimentaria de los pobres es una formaii de protegerlos nivelesde vida. La asistenciaalimentaria tambien puedeproteger el estadonutricional de ciertosgrupos vulnerables, como las mujeres embarazadasy en perfodode lactancia,los nifiospequeflos y los enfermosy ancianos. La asistenciaalimentaria constituye un recrso importantepara los sectoressociales. En 1991la asistenciaalimentaria de los donantesalcanz6 por lo menosa US$150millones. A travesde programasde este tipose proporcionaron mas de 226.000toneladas metricas de alimentosa por lo menos9 millonesde beneflciarios.La asistenciaalimentaria para los pobresde las zonasurbanas es ampliae importanteen Lima. Las transferenciasen efectivoy bienes(incluida la asistenciaalimentaria) provenientes de organizacionessin flnesde lucrorepresentan la cuartaparte de los gastostotales en todaslas unidadesfamiiares e pobres, y casiel 10%de los gastosen todoslos hogarespobres de Lima. Sin embargo,en otraszonas estas transferencias representan un porcentajeinsignificante de los gastosde las unidadesfanilares de los pobres. El hechoque la asistencha alimentariano lleguea los pobresrurales indica quo deberfanexplorarse otros mecanismos,como la distribuci6ade cuponespara alimentos.La asistencia alimentariaparece ostar relativamonte bien focalizada en los pobresde Lima;entre los no pobres,s6lo so benoiciael 6,5%. en comparacifncon et 75% de las unidadesfamilare a pobresy casi el 40% de los pobres. En otas zonas, sin embargo,mis del 40% de los beneficaros no son pobres. 25. Los gruposmis vwlnerablesdesdo el puntode vistanutriclonatl son las mujeresembaazadas y en pertodode lactancia,y los nifiosen edadpreescolar, paticularmentelos menoresde dos aflos. La coberturade los programasde asistenciaalimentaia directa para las madresy los niflosparece ser adecuadaen las zonasurbanas, pero debidoa los bajosniveles de asistenciaquo so brindaa la mayorfade los beneficiariosy a la fadtado actividadescomplementarias de promocion iutriclonaly cuidadode la salud, es improbableque tales programas tenganun efectosignificativo en el estadonutricional de la poblacidn.Ademis, las racionespara los beneficiariosno so basanen un conocimientode los deficitde nutientes obsarvadoson los participantes.La coberturade los programasde asistenciaalimentaria entre los pacientescon enfermedadescrdnicas, especialmento tuberculosis,es insuficientey so precisaaumentada. La coberturadel programade ainerntcin escolardeborfa concentrarse en un grupomis poquefiode nifiosy deberfaincrementarse la asistenciaa nivelesque tenganun efectopositivo en el ausentismoy la desorck$n. 26. Dos cesdones que surgenen la aplicaci6nde la estrategiapara el aliviode la pobrezason la focallzaclny los arreglosinstitucionales. La focalizaci6nprocura dsitbuir las transferonciasu otros beneficiossolamento a quieneshan sido ideAtcados comoel grupobeneficiario en interisde la eficiencia. Los costosdo opevaift del sistomaadministrativo no debenser tan elevados,de maneraque no absobnoel ahorroquo representa la exclusidnde los no necesitdos. En el Peri, el factorgeorico es el indicadorde mayorvalor como mecanismo pan el enfoqueselectivo, Al dirigirrocursos a la sierra rural, se puedellegar a una de cadados unidadesfamuliares extrmadamente pobres del pafs.

27. Algunascaracstleas erchamentos relacionadascon la pobreza indicanlos tipos de sseadcia quebeneflciarfan principalmente a sj-pobres. Casi todaslas unidadesfmliares que consumenagua de pozoson pobres. Entreotras cancterticas en materiade viviendaestrechamente relacionadas con la pobreza figuranel uso de letrinasy la ausenciade electricidado instalacionesd saneamieon en los hogares. 28. El Perd pareceset un entornoapropiado parm el establecimientode un fondode inversidnsocial. La experienciaen materiade fondossociales en otros pafss indicaquo su eficaciadepende principalmente do le capacidadpan seleccionar los proyectosadecudos, lo cualdepende a su vezde la calidaddel personaly so independenciade las presionespolfticas. En agostode 1991el Gobiernoperno establecidun fondode lnversidnsocial, el FondoNacional de Co cin y Desarrollo(FONCODES). Los avancespara poner en marchael FONCODEShan sido lentosy, al principio,la confzsifnreinante sobre la formaen quooperarfa (*impulsadopor la demanda'o 'de arria abajo")af%ct adversamente la calidadde algunosproyectos. Sin embargo,su desempeflocomenzo a mejorara mediadosde 1992. 29. El Fondode CompensacidnRegional -quo opera en el marcodel presupuestodel Perd- tienela posibilidadde contribuira la formulacilsde una estrategiapara aliviarla pobreza. Sinembargo, actualmente no se estubaciendo ningdnesfuerzo por dirigirmas recursoshacia las zonasmas pobres. Solo oa factibley significativoun cambioen la formulautilizada para la asignaciondeolos recursossi aumentaranlas recaudacionesimpositivas y existierauna mayor estabilizaci6nfiscal. 30. La magnitudy gravedadde la pobrezaen el Perd exigela participacidoactfva de todoslos recursos,incluidas las ONG. Estasorganizaciones tienenvarias ventajas. Estifnen condicionesde movilizary utlizar elevadosnivales de recursos,y ademaspueden Ilegar eflcazmente a las poblacionesmarginadas. Las ONGconstituyen una instanciaimportante para la promocidnde las metasy programasdo aliviode la pobreza. En el Perd existennumerosas ONG, que cubren diferontescampos y zonasdel pafs. 31. Las ONGtambien tienen algunos puntos debiles. Suelenset vulnerablesa la influenciade los donantesy tienendificultades para formular propuestasen materiade polfticas.La colaboracioncon el Gobiernose ve limitada por algunosobstculos realesy otros que se perciben.La dispersio&de las responsabilidadesdel Gobiernoen una ampliagama de organismosestatales dificulta la entradade las ONG. Asimismo,existen diferencias relativas a la gestiony las capacidadesadmihistratlvas de las distintasONG. Algunosorganismos estatales las ven comocompetidores en vez de posiblesorganismos de ejecuci6n.La desconfianzaobstaculiza la colaboracidn.El Gobiernodeber(a establecer mecanismosclaros y eficientespara la colaboracidncon las ONG. La reciente formaci6nde consorciosde ONGconstituye un mecanismoque deberiaexplorarse a fin de alentaruna mejorcoordinaci6n, dialogo y planificaci6nconjunta con el Gobierno. 32. Recientementeel Gobiernoelaboro una estrategiapara el aliviode la pobrezacuya finalidad era fijarlas prioridadesde un programade aliviode la pobrezaen el Perd. En general,concuerda con las conclusionesy recomendaciones del anS1isisque se efectia en esteinforme y reconoceque para logrardicho alivio, se requiereesencialmente una estrategiaecon6mica con el objetivode mejorarlas condicionesen que vivenlos gruposexremadamente pobres mediante progrms de ayudaalimentaria, salud, educaci6ny generacidnde empleo. 33. Existendos fuentesde financiamientopara estasmedidas tendientes a aliviarla pobreza: el Estadoy los donantes. A finde atraersuficientes recursos provenientesde los donantes,el Gobiernodebe demostrar claramente su propio compromisocon el aliviode la pobreza. Se estimaquo en 1990el gastopidblico totalen saludy educacionen el Perd alcanzdtan solo al 2,2% del productointerno bruo (PIB). A medidaque la reformatributaria y las medidasque se adoptenpan fortalecerla administraciontributaria mejoren la generacionde ingresos,se recomiendaque el Gobiernohaga esfuerzos por incrementarlos recursosque se asignena los sectoresde saludy educacidn.El gastoen educaciony saludde altedodrdd 4% de PIBocrresponde a&promedio do AmEicaLAdt y a spoxmel nivelque so gast6 enel Perden la primeamdtd ded deeao de 197,0. , - 34. El Gobiernodel PFrd ha dadoun pasohmpo_ pra reducirb poboa ae d pus. A medianoy largoplaz so roequl us crecido do ba empoipia alivWlos devadosuveles de pobrea. Es as d objotv detprogram. do stabizada y eusto.Sin embago, esto no es suf,lonto_ a cOrOf plazo.,Es prso adwopwmedidas pam proteger a losgi vulnl y ar quolos pobm pueda aprovecarlas mayom p s quoso presoan eana ocomfarefouada A finde abordaress ditinm nec s, and pelst. infomsso ha eumlmado.d perlil de los pobres del Pord y o ha recomendab alunas polfca prHoria y programasdirigidos a elios. El Gobeo yaha dado upaso bilW m e ta ate. La estrategiapar d aliviodo pobrm a ownma y priados clam y apropiadasen el marcodo a ap lia do polia ecoicas quomantengan la estabilidadmacroonimica y promuv d ar mkoi . El desafo-y a prueba-respecto del grado de deds del Gobleno sad b m_da an quolas polftlcas y programasMos sorijan pot em st.ategL Les roset. avancesen FONCODESson alentadores. Peo paraasuWar quo los obres oa torecibonlos beneficios del aj, d Goble. ha do Italecer los servidossodaes mIS fdamenales paraellos y proporcdor eWddo proteodd m;isecaces pa losgupos mis vulnerablesdel pals. PUU: Couponentesdel sectorsodal pas una stratega de allio doIl pobroa CuAdNomrnes

Ed~wd6

Bajasta de asistoncia Ofiecerprogramas de Proyetode mdormleto escolarde losnifiOs allimentaci6neolar an di la alinueacd escolar pobres,on particular en la es zonas sierra,rura Proyectosobre esceas de enheflaa sbnuld a de wio gaduos

Alto porcentajede Aumentarla eflciencia Proyectosobre escuas de repitne, en particularde internade las escuelasde enhela sbnulrdeade losalumnos pobm y en las zonasruraes, en wveosgrados las zonasrures paricularmediante di mejorarmientode las Proyectode entregade condicionesan las escuelas librosde exo ruralesde e_nsenza simultAneade varios grados.Entrega de libros de textoy materies didacicos,y formacodnde profosoroorientada a las necesldadesde est. tipo de escuelasy otrasubicadas en zonasubas y ursies maginalespobres Mejorarla calidady la Proyectode mejorwmuemo asistncia escolaren el de los senicos de programapreescolar. educacidnpreescolr Mejorarla preparacidn para la escuelade los nifios 'ue asistena las guarderfas infantilesestales y del PRONOEI

Persitncia de una alta Respaldarun programsen Proyectode aWabtadci6n tasade analfabetsmoentre que se les imparta y capatacIdnlaboral las mujeresque no hablan instuccidnfuncional espaflol bisica, apoygndoseen las ONG Satbd Maa caldaddo b atmici PWporcionarinmnizacda a lts gruposvulnoables - antltiet ca prenataly madimy nis-, suplementode bierto, especlm on tdmlnos vlgHlarel aumentode peso, dd uemro de contatos y proporcionarsuplementos atimentlclos,si es neces_rio,ast como personalcalificado para atenderlos partos de alto riesgo. Caaitar a pauerm para lospartos de bajoricsgo

Aumentarla coberturay Program de caudemo y frecuoecnade la vigilancia desawolo aUM dc crecimleatoan la bae (Mnr de Suod) paicularmentode los nibosmenores do dos alos Escasacobertwa de So debenhacmr esfueros Poam de acroimo y vacenac6ndoe los nifuos pamaaumntr la cobertuma da i monm do dos aft, y de vacunacidnde los nifbs (Mini_b d Salud) diminucidado ats menoresde un abo en los establecimientosde satud Eca cobenurado los Foralecer la instruccidnen pgams do pbaflcacldn materiade planificacion de la fmilia familiary aumentarel accesoa mtodos anticonceptivos Escasautflizacidn de los Rehabilitarlas instalaciones Proyct de mo serviciosdo atenkn deterloradasy mejorarel de la a pilmauapww e primariade la saludpor los =ministrode equipo mm de ecm poabra pobres,partcuarntu an bEsicoy medicamentosean la siena los establecimientosde salud Maladistrlbucidn do los Examinarlas cuestiones Proyectode recursos profesionales. No relativasa la gestidn e pe*fccIoniveato de los so cuentacon suficientes incentivospara el personal. recursoshwnaos en zonas profesioralesde la saluden Mejorarlas condicionesde de exma pobreza las rgiones menos alojamientodel personalde desarlladas y son los establecimientosde insuflclenteslos que atencidnprinaria de la trabaJanon serviciosde salud atencift ptimariade la saud. No hay suficiente Promoverun mayoruso de personalde enfermerfa los recursoslocales mediantela capacitacidnde personasno profesionales, por ejemplo

Dadoque las unidades Promoverla construccion Proyectode revncin del familiarespobres tienen de redesde abastecimiento C6leray otas menosacceso a redes de aguay saneamientode eqfermedades taW pdblicasde aguay bajo costo. En los lugares por el agua alcantarillado,sus nifios dondeno existan estin mis expuestosa instalacionesde sufrirdiarrea saneamierto,deberfan construirseletrinas

El cd6eraes un grave, Reforzar las campahas Proyectode prevencidnde proble de saludquo educativas,asegurar la cdera y otras afectaprincipalmente a los disponibilidadde efemedes srn pobres suministrospara el por el agua ratamientode casos agudos,y rehabilitarlas redesde aguae instalar letrinas

En algunaszonas, una Reforzarlos serviciosde Proyectosobre servcios de proporcidnimportante de urgenciade los hospitales urgenclaen zonasde lospobres utiliza los de las zonaspobres exrmnapobreza servicioshospitalarios Nutrli6n

Ausenciade normasy El Gobiernodeberfa crear pautasrelativas a las normasrelativas a tres inbervencionesen materia esferasfundamentales y de nutrici6n vigilarsu cumplimiento: i) costounitario de las intervenciones;ii) practicas de programacidnde la asistenciaalimentaria, y iii) recopilacidnde informacidny evaluaci6n de las repercusiones Ningln organismodel El Gobiernodebeda Gobior tienela oncomendaresa fimcidnal responsabiidaddo Ministeriodo Salud movlizarrecursos Carenciade un sistew de Deberfavigilarse a una vigilancianutricional a pequeflapoblacidn nivel nacional centinelado reprentates de importadtesgNpOS de beneficiarios La inexistenciade un El Gobleo deberfa marcopara las implomentarel plan do intervencionesen materia nutricidn nuricionalconduce a la duplicaci6nde funciones, al descuidoen lograr inportantesmejoras en el estadonutricional de la poblacidny a una utiflizacidnmenos que dptma de los recursos Curvade crecimiento Se deberfainiciar una Programade crecimientoy inadecuadade los niilos campafiade informaci6n, desarrolloinfantil durantoel perfodode educaciony comunicacin (Ministeriode Salud) destete destinadaa fomentarel uso de tEcuicasde destete apropiadas El bocioes una endemiaen El Gobiernodeberfa ProgramaNacional de el 88% de las comunidades subvencionarel aumento Controldel Bocioy de la sierra y la selva. La de los costosde produccion CretinismoEndemicos anemiatambiEn representa de la sal yodaday (PRONABCE) un problemade salud comprometerpequeflas pdblica,al igualque la plantasde produccionquo avitaminosisA comercialicensal provenientode fuentes naturales

Suplementode hierropara las embarazadas; distribucidnsemestral de vitaminaA tiaralos nifios Asistuida aimentarla

Los programs de El Goblernodeberfa Vasodo Labo asistenclaallimentaia no esudiar mecnismos Rlegana Importantesgrupos alternatlvospas legar a debeneficiaros pobres, en esosgrupos, Como la particularunidades distrlbucldnde cupones familiaresen la siermarur para alimemos,o Introduclr dondela seguidad un programapiloto. alimentadiaba disminuido. Debrfa mejorarsola Asimisrno,la mayor calidadde la proporci6ndo beneficialos adm cion y del del programaVaso de controldel programsVaso Lecheson nifiosen edad de Leche escolar. Casiel 20%de sus benefclcarlosno son Se deberfaampliar la pobres asistenciaalimentaria a las madresy los ufios Programado Alimenackn Algunosproductos bEsicos y Nutricifna la Familiado de los programasde AltoRiesgo (PANFAR) asistenciaalimentaria no Comoproductos basicos so son las fuentesde deberfanseleccionar las nutrientesmL econdmicas; fuentesde nutriene mEs por ejemplo,la lecbeen el economicasque sean programsVaso de Lsche aceptablespara los beneficiarlos.Asimismo se deberfanfortificar con hierrolos alimentos basicosutilizados en los programasde asistencia

Es probableque los efectos Las racionespara los de la asistenciaalimentaria beneficiaiiosdebedan en el estadonutricional de basarseen un conomieoto los beneficiariossean de los deflcitde nutrientos insignificantes que se observanen los participantes La asistenciaalimentaria debe ir acompafldade medidaseducathvas PANFAR complementariason materiade saludy nutricidn,incluida la lactanciamaten y la vigilanciadel crecimiewoy desarrollode los niflos La EscuelaDofendo La El programadeberda ostam La BasculaDefindo La Via no llegaa la ditigidoa nifiosde 14 Vida poblacid mdspobre y gradobAo y a los do las vUlnerbleS,y au efecto n oZaspobres, on Yezdo materlanutricional as concent0a on Lima. minlmo Deberfaasegurar la continuldadde la alimhntacidn Programa de emploo

Cdmoasegurar quo el El programadeberla FONCODES programallegue a los ftrmularsede maneraquo pobres, al tiempoquo logro foment.la autoseleccion los objetivosde aliviode la por los pobres. Si bien pobrezacon un mfnimode con una cantidadinferior al costosadministrativos salariomfnimo se mejorara01 ofroque sectivoe e11otendrfa algunasdesveotajas. Es posibleque no so logren los objetivosen materiade aliviode la pobroza,y se desalnarf la partlcipacion de coutratists privados. Por estasrazones, se deberfaofrocer ol salario mfnimo.TambiEn se deborfaconsiderar una selecciongeogrica de los beneflciarios

Adniinistrac6nde los Utilizacionde conratistas FONCODES programasde empleo privadosimpulsada por la demanda,y promocionde dichouso

* Los tlos que aparecenen bastala se refierena peSifesde proyectos gubernamenales.Los tAdos en mpresidnnormal corresponden a progrnmas en marchadel Goblerno. 1

From Economic and Social Crisis to Reform

Introductdoi its divetsified and hatd-working labor force, a broad endowment of natural resources and dy- 1.1 Poverty is a long-standing problem in namic export performance.'This potential, how- Peru. During the last twenty years a higher inci- ever, was frustratedduring three decadesof heavy dence of poverty has persisted there relative to Government intervention in economic activity mostof its neighborsin the region.The reasonsfor whichresulted in a legal stmctureand policyenvi- Peru's lack of success in reducing poverty are ronmentinimical to growth.The overexpansionof many. But without doubt the dominanton4 is the the state through a growing number of public en- dismal perfotmanceof the economy over an ex- terprisesanu administrativeand regulatorybodies tended period. However, in August 1990, the ad- led to a wasteof resources.Because of the lack of ministrationof PresidentFujinori initiateda major sufficientgrowth and rigid, closed markets, few stabilizationand adjustmentprogram. To date its groups had access to the means to improve their implementationand the broadeningof economic livelihood.For the vast majority,economic oppor- reformshave been impressive.The starting point tunitieswere few and livingconditions either stag- of a strategy to alleviatepoverty in Peru is the ex- nated or deteriorated. ecution of these economicreforms since interna- tional experienceshows that the first prong of ef- 1.3 Publicpolicies seriously distorted most of fective and sustainablepoverty reductionfor any the marketsin the economy.Trade policy reliedon country is the promotionof broad-basedgrowth, import-substitutionwith very high tariffs and particularlydevelopment which provides opportu- many non-tariffbarriers, depriving the countryof nities for the poor to use their labor. This chapter the benefits of concentratingon its comparative briefly discussessome of the main factors behind advantage. Although progress was made in the the poor economicperformance of Pem and its re- early 1980's in liberalizingtrade, the reform left flectionin social conditions.It then highlightsthe many tariff exemptionsin place. Moreover,it was main elementsof the reformprogram, focusing on not sustained.Expansionary fiscal and monetary thoseaspects which wouldbe expectedto contrib- policies, combinedwith a deterioratingterms of ute most to povertyreduction. 1. This and the followingsection draw on Peru - Economicand Stagation SectorReform to SustainStabilization and Lay the Foundation Economic for Development.Green Cover Report (No. 10361-PE, February 1992),and the President'sReports for theTrade Policy Reform 1.2 In the 1960s,Peru was one of the most (No.P-5666-PE, January 1992) and the StructuralAdjustment prnising economiesin LatinAmerica because of (No. P-5714-PE.March 1992) Loans. 1 tade, and an appreciatingreal exchangerate led to basic wages. They included social security and latge curent accountdeficits Financingthe defi- other payments (21 percent), vacations (10 per- cits through external borrowing becme increas- cent), bonuses (20 percent),profit sharing (4 per- inglydifficult. Instead of adjustingto these shocks cent), and paymentto trust funds for yearsof ser- by letting the exchangerate depreciate,the Gov- vice and "labor stability"(17 percent).These ben- erment chose to impose trade controls Reversal efits significantly increased labor costs and de- of the earlier libealization began in 1982 and, by terred employment without improving workers' 1984, tariffs were inceased and many imports incomes. Real wages fell after 1970 and an in- once more requiredlicenses. creasing share of the labor force became unem- ployed or joined the informal sector. Over the 1.4 The tax code becamecomplex and full of three years since 1987 the minimum wage has loopholes and tax administrationwas ineffective fallento 44 percentof its 1987value and the aver- and evasionhigh. Extensivetax incentivesalso re- age salary and wage to roughy 30 percentof their ducedrevenues. These factors causedthe tax sys- value at the same time. A longer comparison tem to lose its buoyancy.Over the period 1976-89 shows an even greater decline. The minimum no revenuefrom any taxincreased pai-passu with wage in 1990 was little more than one-fifthof its income.As a result, tax revenuesdropped from 16 value,measured in real terms, in 1970;the average percentof GDP in 1980 to 14 percentin 1985 and salary was less than one-fifth of its 1975 value; to 9 percentin 1988. In other countriesat a similar whilethe averagewage had fallento lesstan one- stae of development (i.e., per capita income quarter of its 1975 amount. These declines are around $1,200),the averageratio of tax revenues measuredin the formal sector. At the same time, to GDP is 17 percent.Declning tax receipts con- the formal labor marketwhich complieswith the tnibutedto the deteriorationof fiscal performance laborregulations shrank drastically. It is estimated and reduced the resourcesavailable for operating, that only 778,000workers, or about 10 percentof maintaining,and investg in essentialpublic ser- the labor supply,are whollyprotected by the laws vices such as healthand education. (1990).The limitedformal sector employmentre- flects the highercost of labor, which not only has 1.5 Even policiesand programsestablished to encouragedthe substitutionof capital for labor,but improve conditions for workers and farmers has reduced the competitivenessof formal sector provedto be detimental to thosegroups and to the firms and prompted their shift into the informal economy.Job swurity in the formal sector was sector.Between 1961and 1988the shareof the la- constitutionallyprotected, and strictly enforced. bor force workingin the informalsector rose from Workersbecame entitled to labor stability subse- 15 percentto 22 percent,while the shareworkung in quent to a probationary period of only three the prnvatesector fell from36 percentto 25 percet months.A worker who was laid off couldseek le- gal redress and, if successful,could choose either 1.6 In the case of farmers,the land reform in to be reinstatedor to receiveseverance pay. Judge- the early 1970sseverely restricted the marketabil- ments were rarely made in favor of employersbe- ity of land, includingits use as collateral,and made cause the definitionof "justcause' for laying off a land tenure more uncertainwhen collectiveplots worker was ambiguousand difficultto prove and wereparceled out. In addition,the import-substitu- the clim subjectto verificationby the Ministryof tion trade policy contributedto the persistentdete- Labor. The Labor TribunalCourts created further riorationof the agriculturalterms of trade as a re- problemsbecause they were autonomousentities, sult of high trade barriers for industrialproducts, separatefrom the Peruviancourt system,and un- chronic overvaluation of the currency and dis- der the strong influence of the Labor Ministry. criminatory tariff exemptions and multiple ex- Firms also needed "Neauthorization of the Labor changerates. Also, agriculturalservices were inef- Minisry to suspend,dismiss, or lay-offworkers or fectiveand several large scale public agricultural even to redace workingdays or productionshifts. investmentswere unproductive.The Government Fringe benefits,either legallyrequired or custom- altempted to offset these disadvantagesthrough ary in the formal sector,added about 70 percentto maive subsidiesprovided through the grain trad-

2 TABLE1.1 PERU: PERFORMANCEINDICATORS OF TIE ECONOMY

REAL R8AL GDP WVEStMENT AGRICULTURE INFIATION REALWAGE GROWTH PRODUCTIVrYf GDP GROWTH RATE OROWTIH

1968-1977 39 0.175 1.6 16.6 0.6

1978-1982 2.6 0.097 1.6 64.9 .03

1983-1990 -2.0 .0.001 0.7 1397.8 .6.1

1. Ratio of GDP growthand investmentto GDP. Unitsof real GDP incroaseper additionalunit of real investment.A decreaing trendmeasm a lowerproductivity, that is a less efficientuse, of capital.

Source: Peru - Economicand SectorReforms to SustainStabilia*tion and Lay the Foundationfor Development,Green Cover Report,(No. 10301-PE,Februy 1992).

ing companiesand the Banco Agrario.However, prises play an important role in the economyand administrativeinefficiency and operating costs their poor performance is a major cause of the were high and mediumand large-scalefatmers in country's disappointingeconomic and financial the better off coastalareas receivedthe bulk of the recod. subsidies.These policiesfailed to promoteagdcul- tural growth. From 1970 to 1990 agriculturegrew 1.8 Mainly as a result of these bad policies at only half the rate of populationproviding little and the overexpansion of the state, the perfor- opporunity for an improvementin rural incomes. mance of the Peruvianeconomy after the 1960's In particular,production has stagnatedfor the pfn- was poor. The growth of outputslowed so that by cipal crops grown in the sierra - potatoes, wheat the late 1970's and into the 1980's it was well be- and white corn. The policiesalso contributedto a low the rate of increaseof the population.Invest- worseningof poverty.Better off frners were the ment productivitydropped steadily and inflation- ones who receivedthe subsidy.The poorestfarm- ary pressuresmounted (Fable 1.1). ers, however,paid a net tax because they suffered the consequencesof the worseningof the agricul- EoonomnkCrisis tural terms of trade, but did not have access to credit, water and pricing subsidies.2 1.9 PresidentGarcia's explanationof chronic inflationand low growth was that Peru was in a 1.7 Domestic regulatory policies provided "debt trap": servicingthe externaldebt would ac- specialprivileges to incumbents.The proliferation celeratedevaluation and inflation,erode domestic of public institutionsover the last two decadeswas savings,and curtail investment.Consequently, his accompaniedby a sustainedincrease in publicem- admiisraton (1985-90)put a ceiling on foreign ployment.In Pem today the public sector employs debt service of 10 percentof the value of exports, around 1.3 millionpeople, dmost half of the non- and laterstopped payments altogether. It instituted agricultural formal labor force. The number of widespreadforeign exchangecontrols on the cur- publicenterprises rose from30 in 1968to over 180 rent and capital accounts,with multipleexchange in the late 1970s,and today is 133. These enter- rates which were frequently changed. Protec- tion was also strengthened.By late 1987 foreign

2. Foradiscssoanofthepgiumactrofagiutand exchangeallocations were being used to control an analysiof reent msfae er -uAgiultm Polciesfor aU imports. A host of non-tariff barriers - import Ecnmiccincyq, (No.10605-PBS),Setmber1992. monopolies, minimum-domestic-contentrules,

3 and bans on second-handimports, for instance - managementand finances led to a declinein pro- supplementedthe officialnegative lists. Tariffsre- duction.Relations with the internationalfinancial mainedhigh and a varietyof prefetrentialschemes institutionsand bilateral and commerciallenders - allowing half of imports to enter duty-fiee - were suspended because of the accumulationof benefittedselected sectss and firms.To offsetthe large arrear Ihe governmenthad to reducepublic heavyprotection offered to import-substitutingac- investmentto historicallylow levels.Inflation con- tivities,non-traditional exports continued to enjoy tnued out of corol, furthereroding tax revenues. heavy subsidies.The impact of these ptotective policies was to close the Peruvian economy:by 1.11 Economicmismanagement reached a cli- 1990 exports and imports were eight and seven max in 1989-90.In attemptingto controlinflation, percent of GDP. These policies also generated a the Garcia Administrationfell into the trap of re- distorted set of incentives to the tradable-goods- current sporadic adjustments to administered producing sector. Effective protection rates to prices and wages followed by short-lived pnce manufacturing averaged around 80 percent by freezesand other ineffectivemeasures, while fiscal early 1990 and werequite variable.The complexi- and monetary imbalances kept escalating. The ties of importprotection and exportpromotion led country's fiscal base almost vanished and most to enormousdiscrimiaton and use of discretion, public sector prices and tariffs approachedzero in not only between sectors but also between firms real terms. The fiscal imbalance,including Central within sectors. Bank foreign exchangeand financial losses, sur- pased 10 percent of GDP in 1987 and 1988 in 1.10 The Governmentalso increasedthe mini- spit'eof a steady decline in Goverwmentexpendi- mum wage, expanded credit and subsidies, de- tures after 1985. By late July 1990 priceswere in- creasedthe value added tax rate and loweredpub- ceasing at an annual rate close to 36,000percent. lic sectorprices. This strategyled to a highand in- Hyperinflationbrought about a collapseof finan- creasingfiscal deficit, financed by moneyceation. cial intermediationas shownby a fivefoldincrease While the initial results were excellent, the in the velocity of circulation of money between "boom"proved to be uinable. After 1987the 1987and 1990.Crable 1.2) market lost confidence, inflation rebounded and economicoutput decined shaiply. In key sectors, 1.12 The distortedand continually changing in- such as petroleumand mining,the breakdownof centive system - with as many as 11 exchange

TABLEL2. PERU:INDICATORS OF CONOMICCRIS

1987 1988 1989 1990

Gmwthof GP (%) 10 -7 -12 -4 inflation(%) (end of peiod) 114 1722 2775 7650 Publicsectordeficat'(%OFGDP) 11 13 8 7

oevelocf*ryofmoney2 8 16 24 42 1. Includesnon-financial public setor and Central Bank sses 2. xcludesdollardeominations depoit

Soure: Peru- Economicand Sco Reformsto Susta Stabilizationand lay the Foundationfor Deveopment,Green Cover Report (No. 10361-PE, Februy 1992).

4 TABILE13. COMlARATVE SOCIALINDICATORS IN SELECWEDCOUNTRIES

Infant Under5 Fertiloy Primy Seconday Pet Ca%ts Mortaity Mofality Life Rate School School GNP RaFt Rate Expancy (BirthsPer Enrollment Enrollment Coutry (US$ 1989) (Per 1,000) (Per 1,000) (Years) Women) % %

Argetina 2,160 30 36 71 2.8 110 74 Bolivia 620 106 165 54 5.9 91 37 Chile 1,770 19 27 72 2.6 102 74 Colombia 1,190 39 50 69 2.9 114 56 DominicanRepublic 790 62 80 67 3.6 113 47 Ecuador 1,020 61 85 66 4.1 117 56 E Salvador 1,070 56 90 63 4.7 79 29 Guatemala 910 56 97 63 5.5 77 21 Honduras 900 66 103 64 5.3 106 32 Mexico 1,990 45 S1 69 3.4 118 53

Source: 'Socialindicaors in LatinAmerca and the Caribbean: A Compilationof Satstis from 1970 to the Present" by Gemp Psanopoolos mandBill Wood, a viewfrom LATHR, No. 23 andWod BankDevelopment Report, 1991. Dataae for 1989or, if notavailable most recent year (pre-1989) fates.widespread pricecontrls, 15 import tariff gional and ethnic discrimination.Progress in im- levelsranging up to 160 perce - togetherwith the proving health indicatorshas been slow. In the recessionay efects of hyperinflationitself - led to case of infant mortality- one of the prime indica- an unprecedenteddecline in production. By July tors of a nation's health - the improvements i990 real GDP had plunged25 percentsince 1987. achievedby Peru wereone of the lowestin the re- Investment, both public and private, was ex- gion even thoughit started from a less favorable tremely depressed. The country's inrastructure ps T 1991cholera epidemc demonstrated had detenoratedseverely because of iappToptiate in a dramaticway the precariousnessof healthand imvestmentand iadequate matenance. Dispos- sanitary conditionsnationwide. Since its appear- able intrnational reserveshad been exhausted.In ance in February1991 throughJune 1991,223,00 addition,violence and tenWrismwere escalating. cases were reported, one of the most serious epi- demics of this century. in education, Peru ranks Socal Cris near the top measured in terms of primary and sec- ondary enrollment ratios. However, a gross pti- 1.13 SOCIAL INDICATORS.A compatisn may enrollmentratio well in excessof 100 (a situ- of social indicato among countriesin the region ation that has peristed for sometime) is indicative shows that Peru ranks near the lowestin most ar- of relativelyhigh repetition rates (see Chapter3 for eas related to health (Table 1.3). However, poor furtherdiscussion). economicperformAne does not explainall the dif- ference because in many cases Peru falls below even countrieswhose per capila GDP is lower.For exape,the emimple,infant mortality mrWityd rateMe iis higbahigher dwthan 3.'oetDuringVqand the 1970's Inqualityand 1960s: in Lti An ADeicOverview andoftheCrbba the wouldbe expectedgiven its levelof percapita in- Evidence',Dominique van de Wae, pa 7, a Viewfrom come? Other facts include long-standing re- tATh No. 22. 5 TABLE 14 PERU:STRUCTURE OF CENTRALGOVERNMt SPENDING,1970.1990 (% OFTOTAL SPENDING)

1970 1975 1980 1985 1990

Crent 62.6 61.9 49.8 43.0 493 (ExcilIt. ) of which:

Wages 28.7 24.9 17.5 16W6 13.9 49 Goodsandservices 5.3 4.6 2.6 38 12.5 Defense 1S.8 19.3 17.5 14.9 83 Capital 20.5 21.3 192 115 Debtpaymen (Amort.& Int.) 17.0 16.8 31.0 45.5 424

Total 100.0 100.0 100.0 100.0 100.0 Memorandumltems Govt spending(%of GDP) ExcI. socal secity 17.1 19.7 22.9 235 144 I socil seuiy 18.7 21.6 25.0 25.1 n.a.

Index of real Govt eding (1987_100) ExcL oascuri 57.0 87.0 1170 112.0 41.0 IncL socialsecuity 56.0 86.0 114.0 107.0 as

Souce: Govermmeof Pem staffestimates

30 1.14 SOCIAL EXPENDITURES.The public were 65 percentlower than the 1980 level and resourcesavaiable for socia srvices in Penuhave percentbelow the 1970 level. bem scacer with the sharp decline in overall Government expenditues duirig the last five 1.15 Changesin the structure of expenditures on years. The declne is explaid by several fictors givesorx indicationof the competng demands nearly includingthe severerwession and the nea vanish- shrnking resources Wages accountedfor of the ingof the country'staxsystem-in 1989tax rev- 30 percent of spending at the begining by the enuestotaled only 4 percentof GDP. Centralgov- 1970's, but had fallen to about 20 percent for eomient spendingwas equivalentto 20 pecent of end of the decade. In 1990 wages accounted al- GDP during the 1970's, rose slightly during the only 14 percentof totalexpenditures. The share at first half of the next decade, but by 1989-90 its located for goods and services remainedstable late sharehad droppedto lesstan 15 peent of GDP, 4-5 percent of total spendingexcept for the about low in mparion withother LatinAmerican and 1970's and early 1980's when it droppedto middle-income developing countries. By 1990, 3 percent.However, the share goingfor capitalex- most central govemment expenditures in eal terms penditu fell from about 20 percet during of the 1970's to about 15 percentin the 1980's and to only 8 percent in 1990. At different points of timedefense and/or debt service represented large 4. A-bdesciponofthpovioofscasvicesbytb the pivateact is induMin Anua . shares of total spending the former durng

6 mid-197l 's and early 1980's and the latter in the capita spendingon educationhas fluctuated con- mid-1980'sand again in 1990 (Table 1.4). siderably,but in generalincreased from an average of $26 in the 1970's (withparticularly high expen- 1.16 How did centralgovernment expenditures dituresduring the middleof the decade)to $34 be- on the social sectors fare? The data indicate that tween 1980.86before falling to $25 in 1989-1990. the social sectors were somewhatprotected from C(able 1.5) the most severeexpenditure reductions. Neverthe- less, because of the decline in overall resource 1.17 Real centralgovernment spending in Peru availability,Peru spendsless than its neighborsin fell consistentlybetween 1984 and 1990. In order these areas. In the case of healkh,its share of cen- to see how vulnerable the social sectors were to tral governmentexpenditures has remainedfairly budgetausterity a coefficientof vulnerabilitywas stable at 4-5 percent since 1970, equivalent to calculated.(Table 1.6) This is a simpleratio of the about 1 percent of GDP.- On a per capita basis, percentage changes in sectoral expenditure to spendingby the central governmenton health in- those in total expenditures or the elasticity of creased from about $5 at the beginning of the sectoralexpenditures with respect to the govern- 1970's to $13 in the early part of the 1980's. Al- Mentexpenditures calculated only for observations thoughthere have been some fluctuations,in gen- in whichgovernment expenditures are falling.6 In eral there has been a decliningtrend since 1986 the case of health,the coefficientwas 0.5, indicat- with per capita expenditures standing at $8 by ing that on averagehealth expenditures decreased 1990. The WHO estimatedin 1981 that spending by about half as much as total expenditures.The of US$10per capita could achieveprmary health annual data showthat while in 1985-86health ex- care for all, showinghow constrainedgovernment penditures continuedto expand in real terms, in healthexpenditures are in Peru. The Government's 1987-88they were subject to slighty less or pro- emphasison educationappears to have lessened portionalspending reductions,and by 1989 they somewhat.In the first half of the 1970's expendi- were experiencingcuts that were about20 percent tures on education accountedfor between 17-19 greater than the average for total central govern- percent of total central government spending. ment expenditures.In 1990 health expenditures However,this droppedto about 13 percentduring were reduced three times as much as were total the last half of the decade and the 1980's. Mea- expenditures.The treatmentof educationhas fluc- sured as a share of GDP the figures tell a similar tuated, although the coefficientof -0.4 indicates story. In the first half of the 1970's educationex- that on averagespending expanded in real terms in penditureswere equivalentto 3.5 percentof GDP. spite of the decline in overall central government This sharedropped to 2.5-3percent thereafter. Per spending.However, this is due to a sharp increase in expendituresin 1986, while in both 1985 and 1987 educationsuffered larger than proportional s. Healhexpendituresdonotincludesoecurty.A.nanalysisreductions in spendinglevels. In 1988-89educa- of the socialsecudty system in Peruis outsidethe scopeof tis tion sufferedspending cuts whichwere lowerthan paper. Howevm,analysis of social secuity SY s i Lati the average.In 1990,education expenditures were Americaindices thatgenerally the poorare excluded fom the subje to larger than propotional (th times Stn, althou dependg on theincidence of thetx wM subject toelarg tha portional (thrs time financesocalsmcity, the poormay bar someof tem Dta higher) spendingcuts. For the social sectors as a for Petushow that socialscuity covage in poorprin is whole,the coefficient(.0.2) shows that spending negigible.(see -SocialSecouity m LatinA :ercIssues and increasedin real terms, in contrastto the rest of the OptionsfortheWorldlBank"byMcoevy,WoildflnDiscusson Papr,nfe No. 110.)W one pyMv by tbe skcmty govermment.While this is primarily a result of systm in Per are also concenuatedat the terdaaylve o spendingdecisions made in 1986, for two of the hospitals.Of course,even if the poor are excludedfom sos remainingfour years (including 1985 and 1988) security,they may be affectd by cha_ in it._I For example, the reductions suffered by social services were worker lose socialsecurity benefits becaw theywoak in the infon economyor are laidoff fromtheir jobs, tbey andtheir familieswiU have to tur to the hat sevies provied by tbe Minisuyof Health.This could reduce the resoure availablefor 6. For methodologysee Hicksand Kubick (1983), "Cutting programsfor the poor. Therole of thesocial security system in GovernmentBipendire in UDCS.' Financeand healtbcnPasanditsfislsSladbsai ariadinA eA. Develpment 21:37-39,1984.

7 TAIBLELS. PERUt SOCIALSECTOR SPENDING, 1970.1990

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979

Slweotoalo peaal (S) HMealh 5.5 49 S.1 43 4.5 4.7 S.4 4.0 4.5 48 Edlio. t18.8 118.0 185 18 182 175 16.8 13.7 11.6 123 Total 243 22.9 23. 22.9 22.7 222 22.2 17.7 6.1 17.1

Hualth 52 53 6.2 6.4 7.7 10.1 10. 7.9 6. 89 Educados 17.7 19.8 225 27 31.4 37. 34.1 27.1 17.7 22.7 Totd 22.9 25.1 28.6 343 392 47.7 44.9 34.9 24.5 31.6

Shweof GDP(%) HMal 0.9 09 1.0 0.9 09 1.1 09 1.0 1.0 1d_cMlo 3.2 33 35 3., 3.4 34 4 3.0 2.6 25 lbTow 42 42 4.5 4.6 4.4 45 3.9 3 35 I960 1981 1982 1983 1984 1965 1966 1987 96 19 1990

Shamoft tota uealg (%) Heddt 4.7- 5.1 4.4 43 43 44 4.7 48 4.9 4.4 32 Wmmtioa 133 14.9 13.7 12.1 11.9 11i 16.4 126 13. 130 9.7 To.al 18.0 20.0 18.1 164 16.2 16.2 21.1 17.4 17.9 17.4 12.9

HUa 12.7 15. 12.4 107 14 7.9 10.2 95 10.0 8.6 83 1d_ctIo 35.7 44.9 389 30 287 213 35. 25.9 26. 25.2 2.0 Total 484 60. 513 40.7 39.1 293 4S.7 35.7 369 33.7 33.3

SHARBOPGDP(%) 22.9 21.6 21.0 24.6 246 235 213 182 M5 132 14.4 Health 1.1 1.1 0.9 1.1 1.1 1.0 1.0 0.9 0.8 0 05 Bdaca 3.0 3.2 2.9 3.0 29 2. 35 23 2.0 1.7 1.4 Toald 4.1 43 33 4.0 4. 43 4S 32 2. 23 1.9

Smwoe:GM at of Pan d otaff.admtea TABLE1. PERU:COEFlCIE OFS VULNERABILITY, 1985-190

1985 1986 1987 1988 1989 1990 AVERAGE

Health -13 -1.6 0.9 1.0 12 3.1 O.S

Education 1.7 -12.8 3.6 0.9 1.0 2.9 0.4

Totalsocial sector 0.9 -9.8 3.0 0.9 1.1 3.0 42

Source: Staff estimates.

TABLE 17. GOVERNMENTSPENDING ON EDUCATIONAND HEALTH (% OF GDP)

EDUCATION HEALTH TOTAL

1975 -79 1980-85 1989* 1975-79 1980-85 1989* 1975-79 1980-85 1989*

Arentina 2.8 3.1 1.9 0.5 0.4 0.6 3.3 3.5 2.5 Bolivia 3.8 3.2 3.1 0.4 0.7 Una 4.2 39 Dna Chile 3.7 5.0 3.6 22 2.9 a 5S.9 7.9 na Colombia 2.2 2.7 2.7 0.8 0.8 n.a 3.0 3.5 1.8 Dominin Republic 2.1 2.1 1.5 0.6 IA 1.0 2.7 3.5 2.5 Ecuador 3.7 4.3 2.8 0.9 1.9 0.9 4.6 6.2 3.7 El Salvador 3.4 3.7 n.a 1.3 1A na 4.7 5.1 na Guatemala 1.7 1i nua 0.8 1.1 1.0 2.5 2.9 nLa Honduras 3.6 4.0 4.9 1.0 1.6 2.2 4.6 5.6 7.1 Mexico ~ 1.0 3.5 2.1 0.6 0.3 0.3 5.0 3.8 2.4

Average(Unweighted) 3.1 3.3 2.7 0.9 1.2 1.0 4.0 4.5 3.7

* 1989ramost recentdataavailable post-1985. Sour: Govemmentof Peruand "GovernmentExpendi on SocialSectors in latin Americaand the Caribbean Stisial Trends', by HongyuYang, A view from IATMR.No. 13

slightlyless than the cutbacksin total spending, lhe Economic Reform Prorm indicatingthat thes secors were protected. 1.19 In August 1990,the FujimoriGovemment 1.18 Comparing Peru with other countries in launched a sweeping stabilizationand structutal the regionshows that in education,central govern- reform program.7 Duringthe first stage, the gov- ment spendingas a share of GDP was about aver- emnmenttook dramatc steps towards stabilizing agebetween 1975 and 1980, but dropped below the averagethereafter, slightly during the first half 7. Thaissection diaws on thePresdens Reportsfor te Trade of the 1980's, but significantlyby the end of the Pohcy ReformLoan (No. P5666-PE,January 19) and the decade.The same is true for health(rable 1.7). sa Ao. 9 the economyand rebuildingties with the intema- fiscal balance.It is also vital in order to guarantee tional fincial commumnity.The main components the Government's ability to provide basic social of the program aimed at eliminating the chief services in the long-run. The Government has source of hyperinflation- the monetaryfinancing implementeda new tax systembased on five types of the fiscal deficit. A tight domesticcredit policy of taxes on: income of individualsand corpora- was maintained.The govemment placed strict con- tions, wealth of firms and individuals, value- trols over public sector wages and salaries, in- added, selective consumption,and imports. Ex- creasedfuel pricesby about 30-foldand tariffs for emptionshave been reducedand rateslowered. Al- electricity,water and telephoneservices by about thoughthe averageimport tariff rate has been low- ten-fold.Price controlswere significantly reduced ered, other actions havebeen taken which are ex- Severaltax exemptionswere eliminatedand some pected to result in an increase in net trade taxes. emergency taxes introduced.Strict cash controls For example,many preferentialregimes, as well as were placedon centralgovernment finances. One export subsidies, have been eliminated. The of the first measuresof the new Governmentwas agency which administerstaxes is being reorga- to liberalizethe foreignexchange market and abol- nized and it has been given funds to purchase isb foreignexchange controls. The exchangerate equipment. was unified and a managedfloat was introduced. 1.23 The growth in urban employment and 1.20 A second stageof reformbegan in March wages is a major determinantof the pace of pov- 1991 and was based on a series of broad-based erty reductionthrough both its direct influenceon structural measures.These reforms aimedat pro- the existingurban poor and throughthe opportuni- moting competitiveness and deregulating eco- ties it creates for migrationfrom rural areas. The nomic activity The tax regimewas simplifiedand growth of urban employmentis especiallyimpor- administrationis beingstrengthened. The trade re- tant in orderto alleviatepoverty in middle-income gime was liberalized.The ngid labor stabilitylaws countriessuch as Peru. Also, Peru's resourceen- were amended to afford provisionsfor layoffs. dowments and climatic conditionsare relatively Land propertyrights were broadened and strength- unfavorablefor agriculture.Despite the heteroge- ened.The Governmentalso has taken actionto re- neity of ecologicalzones and a variety of crops, ducethe size of the publicsector and has continued only5.9 percentof the territoryof the countrycan to adjust public sector prices, while strictly con- be countedas arableland. The per capita availabil- trollingpublic sector wages. ity of cultivableland (0.14 hectares)is even lower than that of Asia and one-third the averge for 1.21 Becausethe successfulimplementation of SouthAmerica. this economicreform program would lay the base for sustainedand opengrowth, it is the heart of the 1.24 The demand for urban labor depends poverty alleviationstrategy for Peru. Most of the partly on governmentpolicy toward the markets poor in developing countries depend on income for goods and capital as well as on policy toward from labor - from wages, fromwork on their land, the labor marketitself. Oftenindustial protection or from other self-employment.The countriesthat reducesboth the level and growthof labor and the have succeededin reducingpoverty over the long formal sector.As a rule, the greaterthe degree of term haveencouraged urban employment and rual protection,the greater the capital intensityof pro- development, thereby increasing the returns to duction. Greaterneutrality in the trade regimecan wagelabor and smallfarm production.The follow- supporta more labor-intensivepattern of industial ing sectionhighlights the majorcomponents of the expansionin import-competingas well as export- Peruvian adjustmentprogram, focusing on those ing sectors. The Governmenthas moved quickly aspectswhich could make the largestcontribution and boldly to free the trade regime.By September to poverty reduction. 1990, the number of import tariffs had been re- duced to 3 - 15 percentfor basic foodstuffs,medi- 1.22 Peru's ability to reverse the sharp decline cine, some interm--hate and capital goods, and in tax revenues is central for the restoration of some foods, 25 percentfor other producergoods,

10 and 50 percentfor other conumer goods.By then bargainingnegotiation would improvelabor-man- vitually all the preferentialregimes also had been agement relations, to the ultimate benefit of all eliminated. In December 1990, the Govemment parties. The Fujimori Administrationhas recog- amnounceda timetable to cut tariffs every six nized the importanceof reducing labor market ri- monthsto arrive at a flat 15 percenttariff by Janu- gidities which would be likely to hinder the pro- ary 1995.In March 1991 the Govemmentreduced cess of structural adjustment.Thus the Govern- the 50 and 25 percen rates to 25 and 15 percent.A ment has begunto take someactions in this area. It variable-surchargescheme designed to lead to a has extendedthe possibleprobationary period after landed price no higherthan the previousfive-year which labor stability regulationsare in force. The average c.i.f. price plus a 15 percent tarff covers guidelines on the use of temporary employment some agriculturalproducts (dried milk,wheat and also havebeen broadened,as well as the numberof wheat products, corn and sorghum, rice, and reasonsthat wouldjustify dismissals.The Govern- sugar). The Governmentalso has taken strong ac- ment has improved occupationalmobility in the tion to reduce non-tariff importbarriers. By Sep- labor marketby further increasingthe tenn for la- tember 1990, aU importprohibitions, historically bor contracts,eliminating the legal constraintsto the major importbarier, had been suspended.Lo- part-time employment and extending the proba- cal content regulationshave been eliminatedfor tionary period to one year. assembly industriesand several state monopolies have been abolished.The Goverment has begun 1.27 In most countries incomes are generally to allow some second-handimports includingve- lower in rual areas and thus rural development hicle and machineryimports. Clothing, shoes and playsan importantrole in povertyalleviation. Peru laundry items are still banned. Many quality con- is no exception.Policies to promote growthin ru- trols on importsalso have been eliminated. ral areas will be an important componentof the poverty alleviationstrategy. The new administra- 1.25 In general, based on the experience of tion has taken several measures in this area. The other countries, trade reform can be expected to newlyestablished unified exchange rate eliminated lead Peru to a more efficient use of its resources the biasagainst agricultural exports and importsof through specializationin productiveareas where foodstuffsno longer receive exchangerate subsi- its comparativeadvantage is greatest and through dies. The trade reform also eliminatedthe wide- the competitivepressures that freely availableim- spread system of exemptionsthat discriminated ports will exert on domestic producers. Labor againstagriculture. Most significant have been the would tend to be reallocatedwithin broad sectors liberalizationand deregulationof all domesticand and overallemployment opportunities would ben- foreigntrade of agriculturalproducts. The national efit from an expected accelerationin output, first rice marketingcompany is beingabolished and the in agriculture and later in manufacturing.How- national input marketing company restructured. ever, trade reform is most effective if factors of Banco Agrario credit subsidies were elimin- productionare free to respondto marketforces. In ated. Land and propertyrights were strengthened, the case of Peru, state policies since the 1960s makingland transferableand allowingit to be used have severely restrictedfactor movements.In or- as collateral. The Government also has begun der to promote more opportuities for urban em- a long-termplan ensuringland-titling and registra- ploymentand thus contributeto the alleviationof tion. poverty,a key area for fmutherreform is labor mar- ket regulation. 1.28 Faced with poor performing state enter- prises, the Governmenthas decidedto reduce the 1.26 Changing the current laws to permit role of the public sector in the economy through longer labor contracts(not subject to labor stabil- the sale of public companiesto the privatesector. ity), longer probationaryperiods, and judicial re- The size of the central Governmentis also being formswould improveemployment prospects. Less cut In the first four monthsof 1991, 50,000 civil detailed regulations and a less paternalistic ap- servantsresigned under a programoffering incen- proach tc govemment interventionin collective tives for voluntarywithdrawal.

11 1nta Respowseof the Economyand Expected Conclusion Denefits 1.31 The major reason for the persistenceof a 1.29 It is too early to judge the success of the high incidenceof povertyin Peru is the dismalper- reforms which have been implementedand the formanceof the economy.The legal and adminis- progam continuesto be subject to high risks, par- trativestructure discouraged growth. The economy ticularlybecause of the residualimpact of the pre- became closed and not enough productivejobs vious hyperinflationand the difficultiesit creates were created. The pubic sector became overex- for current monetaryand exchange rate manage- tended and wasted resources.The poor economic ment. Nevertheless,there are some encouraging record is reflected in generallypoor social condi- signs.Inflation has moderatedsignificantly to just tions, although school enrollments continued to over 4 percent monthlybetween September1991 expand in spite of the difficulties.An attempt to and May 1992 as comparedto 60 percent during implementan unorthodoxeconomic program in 1990.After tree consecutiveyears of decline,the the mid-1980's compoundedthese problemsand economy grew by nearly three percent in 1991. caused output to plummetand inflation to surge. Althoughthe process of adjustmentresulted in a The Goverunent was forcedto reduce spendingto recessionbeginning the last quarter of 1991 and extremely low levels, although social services continuing during the first half of 1992, the were protectedsomewhat Nevertheless,the over- economybegan to recover duringthe secondhalf all scarcity of resourcesmeans that Peru spends of that year. In September1991, Peru successfuly very little on healthand education.In 1990 a new tescheduled part of its obligations to the Paris administrationembarked on an ambitiousstabili- aub. It awarded the Government exceptional zation and adjustmentprogram. If this programis teatment in the form of unprecedentedcashflow successfullyimplemented it shouldlay the basefor relief betweenOctober 1, 1991 and December31, economicgrowth which would broadenthe oppor- 1992 (the consolidationperiod). tunities for people to improve their living stan- dards. The achievementof broad-basedgrowth is 1.30 Stabilizationis necessaryfor the success the first prong of a povertyalleviation strategy for of the structural reforms and to lay the-base for Peru. Lacking that, no meaningfulreduction in growth. The revampingof fiscal administration poverty can be sustained. However, the poor al- and fiscal reformsshould increase revenues. That ready have sufferedfrom the economicstagnation and the Government'sactions to reduce the state's and recent crisis. They were least able to protect role in the economyshould permit more resources themselvesfrom high inflationand likely to have to be devotedto basic servicessuch as healthand sufferedmost from reductionsin key public ser- education.The trade reform should mobilize re- vices.They havenot been able to benefitfrom pro- sources toward Peru's competitiveadvantage and tected market positions.Their incomes were al- lead to an outward-oriented,and thus sustainable, ready low and thus the recentsharp economicde- growth. Labor market reforms will improve the cline has affectedthe accessof some householdsto flow of labor inputs among sectors so that the even an adequatediet. In order to protectthe most economy can adapt to changes in relative prices vulnerablegroups in the country the poverty alle- caused by trade and fiscal reforms and will pro- viation strategy must also address the particular mote a broaderdistibution of employmentoppor- needs of poor householdsin Peru. tnities.

12 2

Poverd Proftk

2.1 The extent of poverty in Pr wsend s level weneclassified as poor.1 Tne data pre- during the stagnationof the economy and recent setd in Table21 shows thatduring the decadeof crisis. Applyingtwo nutrition-baed povery lines the 1970s the share of households classified as to a householdsuvey covering 75 pecent of the poor or extemely poor declinedin Pe, although population,results in about half of the populon the country's recordlagged that of othercountries failing below the poverty line and a fiflh being in Latin America.In addition,while the extent of consideredextemely poor. In orderto desg pro- rural povertydiminished during this period(in line gams and policies to improve the lving condi- with overall trendsfor Latin America) the extent tons of the poor somebasic informion is needed. of urban povertyrose substantially(in contrast to We need to know about their occupations,con- overalltrends for Latin America).Between 1980 sumptionpatterns, economic activities and acoess and 1986 these gains in reducing poverty were to public services and assets. In order to provide completly wipedout and the trend of a poorerper- answers to these questions a povery profle for fomnce in uba relaive to rur areascontinued. Peru is constructed. 2.3 Two studieshave examined the changein Trends in Poverty 1970-199 the incidence of poverty between 1985/86 and 1990 for IlJma2 TheestiMatS are not comparable 2.2 The United Nations Economic Commis- with those shown in Table 2.1 because of the use sion for Latin America anlyzed the extent and of differentmethodologies, including a different trend of povertybetween 1970and 1986in several poverty line, and the narrower coverage. Both countries, includingPer They estima0tda pov- studes used data from two LivingStandards Mea- erty line based on the cost of a country-specific surement Surveys, one which was camed out in food basketwhich met interationally-set require- 1985/86 and the other in 1990. Special mention ments for energy and protein. Householdswhose needs to be made of the exceptionalcircumstances expendituresfell below this level were classified under whichthe 1990survey was done. It was un- as extremly poor. Thisamount then was scaledup derlaken at the height of hyperinflationin Peru. by a factor to take into accountnon-food expendi- tures. Householdswhose expendituresfell below 2. -Povetyand bequalty DuwigUnorthodox Adjuslment 'he Cue of Peru, 19&5-1990"by Paul Glewweand Gillete Hall, 1. Fbr addidl informationon themethdol nd sources UvlngStadard Measreent StudyWoking Paper, No. 86 and used for this analysissee CBPAL1990. A*ay Eceam& Faeliar, 19851990,Instituto Cuanto.

13 TALE 2.1. LATINAMERICA: POVERtY ESTIMATESAROUND 1970, 190 AND 1986 (%OFHOUSEHOLDS BELOW THE POVWTY IN

URBAN RURAL TOTAL

1970 1980 1986 1970 1980 1986 1E70 1980 1986

LatnAca b 26 25 30 62 54 53 40 35 37

LatinAmerica 10 9 11 34 28 30 19 1S 17

(for Perudata above): Povertyline: d (1988Prices) Natl Currency 123,985 140,892 140,892 83,530 94,920 94,920 US$ 552 627 627 372 422 422

Exptm povertyline: Id (1988Prices) Nlat'curncy 61,987 70,440 70,440 47,731 54,240 54,2A0 USS 276 313 313 231 241 241

Per capita GNP (1980prices) US$ 1,062 1,081 1,023

a. Povertyline based on cost of food basketneeded to meetenergy and proteinrequirements and an estimatefor non-foodexpenditures. b. In additionto Peru, includesArgentina, Brazil, Colombia, Costa Rica, Mexico,Urguay and Veneuela. c. Extremepoverty line based on costof food basketneeded to meetenergy and proteinrequirements. d. Percapita.

Source: CEPAL(1990).

Duing the oiO1tweek suvey penod the increase PortY ProfIle in price level rangedbetween 1-5 percentper day. Derving a series of values in constant prces for 2.4 PERU LIVING STANDARDS SURVEY expendituresover the survey period alone is not (PLSS).The poverty profile for Peru is based on only very difficult,but the resultsare unavoidably informationfrom a Living StandardsSurvey car- subjectto more than the usual marginof error.Ad- ried out during October/November19913. It cov- justing the data to make comparisons between ered households in four regions, Lima, Urban 1985/86and 1990 is even more problematic.Nev- Coast, and the Rural and Urban Sierra, which ac- ertheless, although the precise magnitude of the count for about three-quarters of the popula- changeand the poverty incidencerates are subject tion. The areas selected within the sierra region to a large margin of error, the directionof change were subjectto securityconsiderations. The survey is unambiguous. The studies estimate that poverty rose by between 3-4 times between 1985/86and 3. mis analysishas benefitedfrom an examinationof several 1990 and the inaease in extremepoverty was even studieson povet (andrelated topics) which were based on the Survey. higher.Thes findingsconfirm what would be ex- Severalinformation Worldfrom Bankthe 1985/86studies appearedPeru Living as StandardsLiving Standards pected in lightof the trends in maconomic in- MeasurementStudyWorking Papers. An annotated bibilography dicators,wages, etc. discussedin Chapter 1. ofsome of thework done in Peruappears in AnnexB.

14 TABLE2.2. PERU: DISUUlRIIION OF TOTAL EXPENDITUREBY DECILEs ALLSAMPLEAREAS

Meanpreapita Disttibutionof Ratio of FoodExpendituwes Decile AnnualExpenditumes 1/ Expenditures to TotalExpenditures (%) 4/ (NewSoles) by Decile(%)

Unadjusted21 Adjusted3/ Unadjusted2/ Adjusted3/ Unadjusted2/ Adjusted3/

1 145 216 2 2 71 71 11 254 369 3 4 69 68 III 340 477 4 5 66 65 IV 424 578 6 6 62 62 V 505 688 7 7 61 60 VI 595 797 8 8 58 59 VI 709 941 10 10 56 56 Vill 879 1,146 12 12 53 53 IX 1,178 1.509 16 16 47 48 X 2,472 3,018 .3 31 35 38

Total 750 974 100 100 S8 S8

1. At first weekOctober tAna prices 1991. 2. Householdexpenditures divided by householdsize. 3. Householdexpenditures divided by adult equivalents. 4. Withineach decile.

Source: 1991PLSS genetally followed a format based on the World auveys. The processing time is also much shorter Bank's Living Standards Measurement Study than for othersurveys making it particularlyuseful (LSMS)methodology. The measureof household in countries undergoing significant economic welfare is total consumptionexpenditures. 4 There changes. However, three limitationsof the data are several advantagesof consumptionas com- need to be kept in mind. One, expendituremea- pared to incomedata. One, it better reptesentsthe suresof welfareare inherentlypartial because they welfare leveioof cettain groups whose incomes do not take into accountfactors such as healthsta- are difficult to estimate,for example,the self-em- tus and accessto publicservices. Thus expenditure ployed, includingfarmers and those working in measuresneed to be supplementedwith a discus- the informalsector and thosewho benefitfrom in.- sion of basic needs.This alreadyhas been donefor kind remittancessuch as food assistance. Two, the populationas a wholeand will be coveredspe- since individualsare more sensitive to questions cificallyfor poorerhouseholds later in the report. about incomethan to those about expenditure,the Two, although the 1991 survey covers a signifi- latter data shouldbe more accurate. cant portion of the population,the compositefig- ures are not representativeof the entire country 2.5 The data from the LivingStandards Sur- because certain regionswere not included.In pre- vey includes a wealth of infornation that is not viousanalysis, the incidenceof poverty in two of generally availableeven in extensive household the regions not included in the 1991 survey, the coastal and selva rural areas, was similar to that in 4. For a discussionof thisapproach and theLSMS methidology the sierra urban area. (See Table 2.3 for current se 'The Distributiotof Welfarein Peru in 1985/86" by Paul estimates of the incidence of poverty in the sierr Olewwe,tving Stadards MeasurementStudy Wofking Paper urbanarea the incidence of poverty in the ura No.42. urban area.) The incidence of poverty in the utban

15 27 POVERTY MEASURES.Two poverty TAsL 23. lines have been calculated.Both are based on the PERu 1991 POVERTYHEADCOUNT capacity of a householdto consume a basic food No. basket. Households are classified as extremely poor if their total expendituresare less than the UpperPovery Line (poor) 8,601 cost of a basic food basket.Households are classi- Lowerpovert LinM (extmly poor) 3.476 fied as poor if their total expendituresare less than PopulatioCovered by Survey 16,024 the costof a basic food basket plus an estimateof (TTAL POPULATONOF PERU) (zooo) non-foodexpenditures. In classifyinghouseholds as extremelypoor or poor, regionaldifferences in 1. Headnountcovers only sampledregions the cost of living and in the compositionof the food basket have been taken into account.6 Ac- Source:1991 PLSS cording to these definitions,out of the 16 million people in Peru covered by the sample, about 3.5 million people or about a fifth fall below the ex- selva, the other region excludedfrom the surve;', trme poverty line and an additiona 5.1 million was very low, and it accounts for a very small fall between the extreme poverty line and the shareof the total population.The areas of the rual higher poverty line. Adding the two groups, sierra excludedfrom the survey because of secu- roughlyhalf of the sample populationfalls below rity problems are some of the poorest areas of the higherpoverty lineSable 2.3). Peru.Thus, the figuresprobably understate the ex- tent and depth of povertyin the rmal sierra.Tbrem 2.8 REGIONALDISTIMBUTION. Where do in the contextof a countryin which many fear for the poor live? Among the four regions in the their personal security, it was very difficult to sample, the incidenceof poverty is highest by far implementthe surveyfor high-incomehouseholds. in the rual sierra (Table 2.4) Two-thirdsof the The informationon the upper decilesof expendi- householdsliving there are poor and 47 percentof tures is subject to more than the usual errors and the total fall into the extremely poor category. may be significantlyunderestimated. Only 10 percent of the householdsin Lima are classifiedas extremelypoor, compared to about 20 2.6 Beforelooding at poverty,Table 2.2 pre- percentin both the urban coast and sierra regions. semtsthe disecfehtron of total expenditureby In each of these three regions the share of poor population decile, where decile I contis the householdsis about half. The rural sierra is esti- ponrest 10 percentand decile X the wealthiest10 matedto containthe largest numberof extremely percentao thepopulatiio.The daaareprovidedbo poor (1.6 million). Because of its size Lima con- both an adjusted (consumptiondivided by adult tains the highest numberof poor (3.2 million). equivalents)Sand an unadjusted(consumption di- vided by number of household members) basis. 2.9 Table 2.5 providesinformation on house- The data show that householdsin the top decile hold size and per capita consumptionfor the ex- accountfor a third of total expenditurescarped tremelypoor, the poor and the rest of the popula- to two percent, four percent and five percent,re- tion. The typical poor person lives in a larger spectively,for the bottomthree deciles.For the top householdthan the non-poor,on averagelarger by decilehousehold, food represents40 percentof to 1.5 individuals.There is little differencebetween tal expenditurewhereas for the lowestthree deciles the poor and non-poorhousehold on the numberof the share is about 70 percent. wage eamers, but because of the larger household each worker in a poor householdsupports 4 family

6. Ihe e-tremepovaq yine (adjusted,per capita)ranges from 5. Themethodology used is thesame as Glewwe (1988). Adlts 430new soks/yearin the urbansiera to 500 newsoles/year in overthe age of 18are given a weigt ofone, whie childrenaged Uima.Thepoertyne(adjusted,percapita)rangesfrom 700new 13-17,7-12 and 0.6 are given weightsof 0.5, 03 and 0.2, sols/year in the ra siera to 980 newsoles/year in Lima. See respectively. AnnexB for moredetails on thecaklulation. 16 Emembers wherea for the non-poor the number PERU: REGIONALDfSTRIBUTION F fills to about 3. On average,the valueof per capita POPU:ATION BY DIOVERTIYGROU food consumptionin an exlremelypoor household is only 30 pecent (on an adjustedbasis) what it is Region AU Non-Poor Poor in a non-poorhousehold. n the case of total con- 40 44 37 sunption,the fracdon is20percenof the levelfor lima non-poorhouseholds In poor householdsthe cor- 19 18 19 respondingf for food and to lconsmption Urba Coast are about45 percentand 30 percent,respectively, 20 23 18 of the non-poorlevel. UrbnSierra_ RuralSierra 2114 7210 As shown in Table 2.6, femade headed 100 100 100 housholds makeup a smallershare of the ex- Total ____ A tremelypoor and the poor comparedto their share in the total popiuation. However, the mean per Region ADn Et mly Pow Ote capita annual expenditures(unadjusted) of a fe- male-headed household are only 15 percent greater than for a male-headedhousehold. Fe- 40 19 46 male-headedhouseholds are found to a greater Lima 19 9 uextent in Lima,accounting for a fifthof house- UrbanCA holdsin the capitalcity. Contraryto the pattem in 20 17 21 other areas of the country, here female-headed Urban Sierram householdsare found disproportionatelyamong 21 45 14 the extremely poor, accountingfor 26 percentof RuralSierra WFNINWNER M the total. 100 100 160 Totall 2.11 Data on languageand expenditurelevels Source: 1991 PLSS indicate the disadvantagedposition of the Indig- enous popution in Peru. Slightlyover 60 per- cent of Quechuaspeakers are poor and a third of them are extremely poor. In the case of Aymara speakers, 86 percent are poor and 70 percent of

TABLE 2.5. PERU: HOUSEHOLDSIZE AND PER CAPITA CONSUMPTIONBY POVERTYGROUP Extremely Poor Poor Non-Poor All HouseholdSize 6.0 5.8 4.3 5.0 AdjustedHousehold Size 3.9 4.0 3.4 3.7 No.of WageEamers per Houehold' 1.4 1.6 1.6 1.6

Food Consumption($/annum) 132.1 205.9 509.5 369.3 AdjustedFood Consumpton ($/annum) 221.1 324.2 735.7 488.8

TotalConsumption($/anum) 192.4 329.2 1,111.2 750.0 AdjustedTotal Consumption($/annum) 327.4 528.8 1,642.3 937.7

1. Includesonly paid workers Notes:(Limg prices October1991), new soles

Source: 1991 PLS 17 . TABLE27. TABE- BY PERU:SEX OF HOUSEHOLDHEAD PERU: DISTRIBUTIONOF THEPOOR AMONGTHE POOR LANGUAG1:OF INTERVIEW Non- sex remely Poor Poor All Edmic Poor (ofwhich: .roup Extemdy Poor Total Male 88 8S 83 Poor) Female 12 1S 17 Spai 42 (13) 58 100 38 100 Source: 1991PLSS Quechua 62 (34) ______Aymaa _ 86 (70) 14 100

Souc: 1991 PLSS them are extremely poor. Taken together these households account for 40 percent of the ex- for tremelypoor and abouta quarterof the poor, com- food by the poor. The commoditiesaccounting bread, pared to their share in the sample populationof the largest share of expendituresare rice, own slightlyunder a fifth. sugar, tubers and meat. Consumptionfrom production is substantialonly among the poor in 2.12 The age of the head of householddoes not the rural sierra (among the regions sampled) appear to be strongly correlated with poverty. (Fable 2.10). For both the poor and the extremely food con- Based on the total sample, those aged 26-45 are poor, slightlyover half of the value of on somewhatoverrepresented among the poor. Older sumption is accountedfor by own production for which heads of households are consistently averagein the rural Sierra The products share of unrderrepresentedamong the poor. In Lina, mi- own production accountsfor the largest and dairy grants are only slightlymore likelythan non-mi- consumptionare chicken (90 percent) grantsto be found amongthe extremelypoor and products,tubers, meat, vegetables and eggs (70-80 in the poor. About 44 percentof the residentsof the percent). There are only marginal differences these capital are migrants. the pattern among the extremely poor. For products, the poor would be insulated from out 2.13 EMPLOYER.The poor are found largely changesin marketprices. It shouldbe pointed areas in among the self-employed.The unemployedhave that the samplesurvey does not includethe The the highest incidenceof poverty followedby the Peru where rice and sugar are produced. most self-employedand privateworkers. Reflecting the coastalrural and the selvaregions account for have extent of the Peruvianeconomic crisis, about half of this production.However, previous studies of all households headed by a self-employedor shown that these farmen are generallybetter-off com- private worker are poor. The self-employedare than sierra farmers because they are more have better off than private workers in the sierra The mercialized and these products generally on the opposite is true in Lima and the urban coast. In been protectedby the Government.Based from spite of the sharp decline in public sector wages, survey data, the poor would stand to benefit as a the incidence of poverty is lower among these declinesin the price of wheat, rice and sugar these workers. Thus the poor would not be dispropor- resultof a decreasein effectiveprotection for tionately hurt by reductions in public sector em- crops. ployment.Public sector workers are better off in in- Lima than in the other regionssampled. Although 2.15 Althoughthe Governmenthas eliminated it would be expectedthat the recenthyperinflation put subsidiesand guaranteedprices for agricultural of would have had an adverseimpact on retirees,the output, for someagricultural products a scheme sur- poor are found to the least extent among this surchargesis applied.As of June 5, 1992,the between group. chargeson wheat, flour and pasta ranged 30-38 percent;rice and sugar receiveda 13 and 16 2. 14 EXPENDITUREPATFERNS. Table 2.9 percent surcharge; and yellow corn and sorghum liststhe most importantexpenditures on purchased received a 17 percent surcharge.The equivalent

18 TABLE 24L PERU: Dbibulon of the Poor by Employer of Head of HOU.isbI|

Employer All Extemely Poor Other

Public

Private

Self-Employed

Unemployed

Non-Active 7 1 Other

Total

Employer All Poor Non-Poor

J ~~~~~~~~~~~4 11 16 Public 22 22 22 Private

Self-Employed

Unemployed _ 1 _

Non-Active 785 Olher _g* 100 10010 Total__

Soce: 1991 PLSS percentagesurcharges on dairy productswere 34 dairy products, the poor may benefit from the percentfor powderedmilk and for anhydrousmilk scheme. Purchased dairy products account for a fat.7 What is the impact on the poor of this lowerbudget share for the poor relativeto the non- scheme?For severalproducts- sugar, wheat flour, poor and poor farming households consume pasta and rice - the policywould appearto hmt the largely their own production. The impact of the poor. These items account for a higher budget surchargeon yeiow maizeand sorghumis consid- share for poor households relative to non-poor ered in paragraph2.21. householdsand even poor fiuminghouseholds are net purchasers of these products In the case of 2.16 Table 2.11 lists the nine most important non-foodexpenditures for the sample population. 7. For folr tament of the surge scheme see Peru- For the poor the most importantexpenditues are Agiult Poles for Enomic Effciency (No. 10605-PB, detergents,public transportaflon,soap, and cloth- September1992)2 ing. in Lina, the poor spend more on public trans-

19 portation(slightly over 20 percentof non-foodex- PRRU-. D TO LE r.o9m penditure) than in other regions. In the sierra re- GtoUP Vgion,on the other hand, expenditure shares are higherfor consumergoods such as soap, detergent, and clothing.Although the poor stand to benefit Eicrmely Poor Non- from any price decreases in these commodities re- po"r Poor sulting from deregulationand trade liberalization,

Rice2 13 11 8 a subsidywould be well-targetedto the poor only Bread2 11 11 8 if tbey consume more than the rich in terms of ab- Sugar2 9 7 4 solutequantities. None of these itemsmeet this test Tubers 8 7 6 because for all the products the proportion con- Meat 7 a I sumed by the non-pooris greaterthan the propor- Oil 6 5 3 Beans 6 6 6 tion of the non-poor in the population. Chicken 5 6 7 Noodles2 4 2 .17 Informationon the characteristics of the F dibSS 4s housing of the poor gives some indicationof the DairyProucs 4 6 8 Vegebles 3 3 3 distibution of public servicessuch as water, sew- Others 19 21 31 erage and electricity. As well as being important measures of welfare, some, such as source of Total 100 100 100 drinkingwater and sewerage,have important im- 1. P_Shadufod only,exudes _-omp plications for health status. Only slightly over half ownproduction of extremely poor households have access to a 2. Coveredby the erumWesurcarge sdcm public water systern, comparedto 80 percent for Soul : 1991P1.S the total sample.The rest of the extremelypoor are dependent on either wells or rivers/springs.The extremely poor who do have access to a public water systemhave beenmore affectedby cutbacks in water supplythan the rest of the population.In TAKLE2iQ the rural sierra less than half of the populationhas PERU:FOOD CONSUMPTON OFPOOR access to a public water system, comparedto 90 HOUSEHOLDSDYTYE OFACQUISTON percentin the other areas coveredby the survey. INRURAL SIERRA Purchased ownProduction 2.18 Less than 30 percentof the extremelypoor havepublic sewerageservice, compared to slightly Rice' 86 14 over 60 percent for the total sample. The rest are Breadw 84 16 either dependent on latrines (40 percent) or lack Suar 93 7 facilities (30 percent). In the rural sierra the low Meat 33 67 accessof the population(8 percent)to public sew- oil 94 6 age service is even more striking. The poor use BMW 71 29 gas/oi/kerosenefor lightingmore frequentlythan Chico3 11 89 wealthier households where electricity dominates. Noodba 92 S For cooking, the use of kerosene is common for DairyProducts 33 68 households of all income levels, but among the Vegetables 19 51 poor wood is even more widespread. Others 28 72 All 49 51 2.19 Educadon levels are strongly related to 1. Covrd bythe Govneasurdargemsceme poverty. The percentage of heads of households with less than a primary education is relatively Sor: 1991PLSS small (7 percentof the total sample).It is not sur- prising that two-thirds of these households are

20 TABE 2-1 PEWR%NO-FOOD EXPENDITUREPATTERN BY POVERTYGROUP

AU emely Poo Othe

10 14 10 PublicTmnsportadon ------7 8 7 Clotbing(Adlt) _S_ _ 6 11 7

Soup,Shampoo,oo¢ 6 1S S~~~~~~~~~~ 5 Detegnmts FumininrHousehoM -upn 3 3 3 Shoes 3 03 Fuel and LubricaX

Clothing(Children) _ 52 38 ... 5 Ohes 100 100 100 All

All Poor Non-Poor

to 179 Publc Transportation 7 7 8 Ckbdas(Adul) _ 6 ~~~~~~~~6 7

6 9 5 Soup, Shampoo,etc.l ___ | _ 6 10 S 4 1 ..... t IuehodEquipmentU IIIIUIIEUIU 3 ------3 ------4-- Shoes II_ 3 14 Fuel ad Lubcat 3 4 2 Clodt(Chiden 52- 42 54 Others ------100 ------100 100 AB

Source: 1991PLSS

21 TARLE 2.12 PERU: PUBUC SERVICESFOR THE POOR (% OF HOUSEHOLDS)

Extemely Poor Poor Total Rwal Siens

Sourceof DrinkingWater Public 55 70 80 Well 26 14 8 River/Sprig 14 9 7 WaterTruck 1 4 3 Other 4 3 2

Hours of Public Water Suppy in Housebold 1-3 hours/day 21 18 15 46 hours/day 1S 14 12 7-12hours/day 17 21 20 13 or more hours/day 47 47 52

Sewerage Public Service 28 49 63 Well-Septic 4 5 4 Latrie 40 23 16 None 29 23 17

Sourceof Ughting Electricity 49 71 81 OiaslO/Kerosene 44 24 16 Candle/None/Other 7 6 4

CookingFuel Electricity 1 1 3 Gas 4 14 28 Kersene 39 51 45 Coal/None/Olher 7 4 2 Wood 49 30 22

Sourc::1991 PLSS

classifiedas poor. In the ual siea, however,the the percentagesof poor and non-poorhouseholds share is double that for the total sample and the who own land. However,there is a significantdif- proportion classified as poor is about the same. ference in the amount of land owned. The poor About two-thirds of extremely poor households have roughly 50 percent less land than the non- are headed by someonewith no more than a pn- poor. Irrigatedland is more valuablethan non-irri- mary education,compaed to about a third of non- gated land. About a third of poor householdsown poor households.Those who have not completed irigated land, comparedto abouthalf for the non- secondary educationare more likely to be found poor. The poor householdalso owns on average amongthe poor than those who have finishedsec- less irrigatedland than the non-poor.Also, poorer ondaryschool. households in the rural sierra tend to have a greaterproportion of workersengaged in agricul- 2.20 The incomes eaned by the rral poor de- ture than is the case for better-.ff hoseholds. pend in part on the land they own. Table 2.14 pro- vides data on land ownershipamong the poor in 2.21 The crops produced in the rual sierra by the rural siem ere is virtly no differencein the largestshare of poor farm householdsare pota- 22 TABLE 2.13. PERU: EDUCATIONLEVEL OF HOUSEHOLDHEAD

HighestGrade Completed by Head BxtremelyPoor Poor Non-Poor Mean per Capita Anna Expenditure1/

None/initial 12 10 4 430 Primary 55 47 30 543 IncompleteSeconda 14 14 12 633 CompleteSecondary 16 21 27 808 Non-UniversityHigher 2 3 7 969 IncompleteUnive-sity 0 2 4 1160 CompleteUniversity 1 3 16 1429 Total 100 100 100 874

TotalSample RuralSiera

None/Initial 7 16 Primary 38 58 IncompleteSecondary 13 11 CompleteSecondary 24 12 Non-UniversityHigher 5 2 IncompleteUniversity 3 1 CompleteUnivesity 10 1 Total 100 100

1/ All firstweek October1991 LimaPrices, new soles

Source: 1991PISS toes, barley and wheat (Table 2.15). The crops which are produced by a higher share of poor relative to non-poorfarmets ae barley,potato, oca PERU:AGRICUTURAL L OWNER and olluco (Andeantubers). Yelow maie and rice Pi THEIRURALSIERRA are grown by a larger share of non-poor farmers, while wheat and white maize are grown about Poor Non-poorTOt equally among poor and non-poor farmers. Thus Total poor fumers would benefit from icaes in the priceof potatoes or badey (or gainsin pfitability). AllFam LAnd: As already mentioned,the suivey did not encom- Householdsowning land (%) 81.6 84.1 82.5 pass the area wherence gros However,acoding dh s o sn0i.9 . to the 1985/86PLSS, the poor were not impor Ladwning householdonly 0.34 069 045 producersof this cropand it is unlikelythat thishas IrrigaedLand: changed.8 The Govermment'ssurchge on yellow Householdsowningland(%) 36 48.7 40.2 maize (see para 2.15) benefits largely non-poor Capita: farming households.Also, the productionof sor- w ghum (also covered by the surcharge) by poor mmo item: farmersappears to be negligi'ble. Distributionof RurwalSiera 67.8 32.2 100.0 Popuation s*ne poorinLain AmaicaduingAdjostment -A CaseStudy of Source: 1991PLSS Peru,' by Paul Okwweand Dens de Tray, lng Standards MeasurementStudy Woring Paper,No. S6. 23 thirds are poor. TABL 2.1S PERU:CROPS GROWN IN * Becauseof its size Limacontains the high- RURALSIER8 A est number of poor. Poor Poor Non-Poor Crop Exreey more mem- Cm Ex_____y_Poof Pow No__ _ The7* poorhouseholdhasslightly bers than the non-poorhousehold. As a w Maize mresult,each workerin a poor household % whofam 13 14 22 supportsmore family members. kg/capita 32 47 221 * Povertyis widespreadamong the indig- White Maize enous population. % whofam 9 10 9 kgcapita S4 73 42 Most of the poor are eitherself-employed

Wheat' or workin the privatesector. The incidence %who fam 50 5s 50 of poverty is highest among the self- kg/capita 27 29 203 employedand the unemployed. Rice' %wbo fam 3 3 S * Themostimnportantfoodcommoditiescon- kg/capita 656 787 2956 sumed by the poor are rice, bread, sugar, tubersand meat. Barley %whofarm 60 54 42 The most important non-food commodi- kgtcapita 37 40 180 ties consumedby the poor are detergents, Potato publichtansportion, soap, and clothing. % whofarn 69 69 63 kg/capita 111 139 444 * Onlyslightly over half of extremelypoor Oca householdshave accessto a public water % whofarm 11 12 7 syskm. Less than 30 percenthave public kg/capita 50 48 126 sewerageservice.

%who farn 8 10 7 * About two-thirdsof extremelypoor kg/capia 47 40 141 householdsare headedby someonewith no more than a primaryeducation. Those 1. Coveredby the Goves surchrgesdCe. 1 who have not completedsecondary educa- thecase of what and ice venfamg ouseholdsw tionare more likely to be foundamong the neAtpsdmm Crabk2-10) poor than those who have finishedsec- Soure:1991 PLSS ondaryschool. * Among landowninghouseholds in the ru- Sumun ral siem, the poor have roughly50 per cent less land than the non-poor.They are 2.22 TMepreceding sections have descrbed the also less likelyto own irrigatedland and characteristicsof the poor. In summary,the most even if they do own irrigatedland, on av- important conclusionsof the poverty profle for erage,they own less than a non-poor Peru are: household.

The incidenceof povertyis highestin the * Poor householdsin the rural sierra have rumalsierra; nearly half of the households more workersengaged in agriculturethan living there are extremelypoor and two- the non-poor.

24 ing the rehabilitationor constructionof The most importantcrops producedin the small-scaleirrigation works. Measures to rural sierra by the poor are potatoes,bar. increaseproductivity of the crops most ley and wheat. The crops whichare pro- importantto the poor (barley,potato, oca ducedby a highershare of poor relativeto and olluco) would also be beneficial. non-poorfarmers are barley,potato, oca Othermeasures are unrelatedto agricul- and olluco. tural policies.For example,given the higlherdependency ratio of the rural poor Inipilcatlonsfor Poverty AUeviationStrategy and the low productivityof agricultural employment,these householdswould ben- 2.23 The broadguidance provided by the pov- efit fromopportunities for non-fannwork. erty profile for the design and components of a povertyalleviation strategy for Peru is summarized * In relationto the poor (as opposedto the below. extremelypoor), the povertyprofile indi- cates that most of these householdshave The extentof povertyin Peru is substan relativelygood access to public services, tial. About half of the populationsurveyed including education. These households fall belowthe higher of the two poverty are poor because they are either in lines and a fifth below the lowercutoff. low-paying,low productivityjobs in the Givenresource constraints and the need privateor informalsector or unemployed. for fiscal disciplinetargeted programs What is needed to raise their expenditures shouldfocus on the extremelypoor. above the povertylevel are moreproduc- tive jobs. And they are relativelywell-

4 The incidenceof extremepoverty in the equippedto take advantageof more em- rural sierra and amongthe indigenous ploymentopportunities. For this group populationlargely accounts for the inci- what are most importantare the execution dence of extremepoverty at the national of the economicreform program and lib- level. Thus, progams to alleviatepoverty eralizationof the labor market. amongthe extremelypoor shouldtarget the rural sierraand the indigenouspopula- * Basedon consumption(and production) tion. These groupssuffer not only from patterns,the poor would benefitfrom de- low expenditurelevels, but also from low cdinesin the price of wheat,rice and sugar accessto public services,such as educa- as a result of a decreasein effectivepro- tion and water and sanitation.Improve- tection for these crops. Sinceclothing is mentsin access,particularly for education, an importantnon-food expenditure, the are neededto reduce poverty.These poor would gain from the eliminationof householdsare also at greaterrisk for poor the ban on second-handclothing imports. health becausethey lack clean water and Given the importanceof expenditureson sanitationfacilities. detergentand soap, the poor also would benefitfrom increasedcompetition in the * The fact that workingin agricultureis manufacturingor importingof these prod- positivelyassociated with povertypoints ucts, leadingto a declinein price. to the need to look for ways to raise in- comesof fam households.Some possible * The surchargescheme should be recon- measuresrelate to agriculturalpolicies. sideredby the Govemmentbecause, with Poorer farm householdshave less land the possibleexception of dairy products, than better off householdsand the land is the poor are not benefitingfrom it. For less likely to be irrigated Reformsrelated someproducts (sugar, wheat flour, pasta to land titling shouldimprove access to and rice) the poor are adverselyaffected land and there may be scopefor promot- because even poor farminghouseholds

25 are not purchasersof the commoditiesand their livingstandards. Although less documented, becausethe itemsaccount for a larger good health status also has a role to play in en- budget sharefor them than for the non- abling people to improve their lot. The poverty poor. The surchage on sorghumand yel- profile has shown that the extremely poor in the low maize benefitsbetter-off fams be- rual sierra are at greaterrisk for healthproblems causethey producethese crops to a greater because of low access to public water and sewer- extentthan the poor. age systems. Thus, improvementsin social ser- vices - education,health and nutrition- are critical 2.24 The povertyprofile showsthe importance for the poor. of education in enabling householdsto increase

26 3

Social Services

3.1 Social servicescan play a dual role in al- gaps in coverage,and the intensityand quality of leviating poverty. First, improvementsin health, services is inadequate.There is also worrisome nutritionand education,direcy addressthe worst evidenceof a decline in the demand for services. consequences of being poor. Second, there is In nurtion, food securitygenerally has declined ample evidence that investing in human capital, for the poorest.Peru suffers from relativelyhigh especially in education,attacks some of the most rates of child malnutritionand low birth weight. importantcauses of poverty.The programswhich There are also micronutrientdeficiencies in some have the greatest impacton the poor are primary areas of the country.The response to malnutrition education and health care. This section reviews has excessivelyrelied on food assistanceand given the status of servicesin education,health and nu- insufficient attention to complementary health trition,focusing on thoseactivities most important measures,micronutrient supplementation, and the for the poor, and highlightsthe major issuescon- promotionof better healthand nutritionpractices, frontingthese progmms. includingbreast-feeding and chfldgrowth and de- velopmentmonitoring. There are also problemsof 3.2 In education, there has been an impres- poor coordinationamong the agencies active in sive expansionin primaryenrollments. However, nutrition,a lack of an overallframework for nutri- this has been coupled with a decline in expendi- tion interventions,an absenceof norms andguide- huresper studentand an exodusof trainedteachers lines and weak nutritionsurveillance. from the system. Repetitionrates are high, par- ticularlyin the early grades.In addition,there are Education wide disparitiesin educationalattainment and ef- ficiency within the country. Another concern is Status of Services the persistenceof a relatively high rate of Wit- eracy as little progresshas been made in rual ar- 3.3 Peru expandededucational opportunities eas and among women. In health, in spite of an substantially during the last two decades. The expansionin primarycare infrastructure,Peru has share of 6 to 14-year-oldsattending school in- some of the worstindicators on the continent,par- creasedsteadily from 78 percentin 1972 to 89 per- ticularly in rural areas. Many of these conditions cent in 1991according to Governmentfigures. The could be improvedwith more concertedefforts in high proportion (97 percent) of each age cohort primary health care. Also, the full benefitsof the who enroll in first grade on time (between ages 6 network of facilities are not being realized be- and 7) suggeststhat parentsare committedto send- cause of the scarcityof equipmentand the poor ing their childrento school.However, during this distribution of personnel. In general, there are time the real level of expenditureson education 27 but the high repetitionrate is an obstacleto univer- TABLE3A. sal completion of primary school. It also creates PERU:WAGE AND BUDGET INDICES incentivesfor early dropoutand is responsiblefor 1980 198S 1989 1990 a large degree of age heterogeneityin each grade which makes teaching very difficult. Students TeacbersSalares 100 66 24 - spend an averageof 13 years in each grade so that BudgetIndex 100 86 39 26 by grade 6 almost three-quartersare overage. On Mm Wage 10 s 31 27 average, for each graduate the system spent re- c-uCollan 100 61 so 29 sources equivalentto 8.4 years of schooling,40 White-Collar 100 80 60 31 percentmore than whatwould be necessaryin the absenceof repetitionand prematureleaving. Source: Peruen Numeros,Cuanto S.A. 3.7 The high level of repetitionin first grade seen in Peru is commonin LatinAmerica. To ad- generally declined. The decline in real expendi- dress this problemmany countrieshave expanded tureson education,coupled with expandingenroll- the coverageof pre-schools.Peru has significant ment, resulted in a shatp drop in per student experiencewith pre-schooleducation. There is an spendingafter 1985, from $175/studentto about extensivenetwork of formal pre-schoolsand non- $80/studentby 1989-90. formal pre-schoolcenters with strong community involvemen,(PRONOEI). About 70 percentof the 3.4 The impact of fewer resources on the populationof 4 to 5 years old is enrolled in pre- quantityof educationalservices was ameliorated school.The proportionof childrenserved by non- by the even sharper drop in the real level of formal centers increasedfrom 15 percentin 1980 teachersalaries. Salarylevels fell for most groups to 39 percent in 1989. However, evaluationsof in Peru during this time, but by 1989 the decline non-formal programs suggest that, while they eperienced by teacherswas greaterthan the drop work relativelywell in the area of early childhood in the minimumwage or in the averagewages for stimulation,they do not do as well preparingchil- blue and white-collarworkers. In response,many dren to enter primary school. These centers are trainedpersonnel left the teachingprofession. As a mostly located in poor areas and are organized result, between 1982 and 1990 the proportionof more as day-care centers, staffed by volunteers uncertifiedteachers increased from 25 to 50 per- who havereceived litle trainingand lack the basic cent. teachingmaterials or facilitiesrequired to prepare childrenfor school. Issues 3.8 Nationalaverages mask sharp differences 3.5 Accessto prmaty educationin Peru is al- in educationalattainment and internal efficiency most universal.The main educationissues related betweenregions and between urban and rural ar- to povertyare: (a) the high level of internalinef- eas. Departmentswith the lowest educationalat- fclcency,in part explainedby the poor quamtyof tainment (percentage of students completing teaching; and (b) the signficant disparies tn Grade 6) tend to have the highest proportionsof educationalattainment and Internal effiiency uncertifiedteachers and schoolswith no electric- withinthe country. Table 3.2 summarizessome ity, water or sewerage.Based on 1981-88data, the indicatorsof internalefficiency in primaryeduca- highestpercentage of studentscompleting Grade 6 tion. The impressiveincrease in enrollmentshas was the 85 percentrecorded in Lima. The capital not been matchedby improvementsin educational also has one of the lowest percentages of attainment uncertifiedteachers (47 percent).At the other ex- treme, the lowest share of students (28 percent) 3.6 A comparisonof grossand net enrollment completedGrade 6 in Ayacuchowhich has one of ratios over time suggests that there are enough the highestproportions of uncertifiedteachers (65 places to providefor all children6 to 11 yearsold, percent). In Ayacuchoit takes 15 years of school

28 TANA 3.2 PERU:INDICATORS OF INTERNAL EMCIENCY OF PRIMARYEDUCAlnON, 198 Evolut of GOssand Not Ewolment Ratio, 1960-1990

Eumflut RatIos 1960 1970 1980 1990 GrEssEollmnt Ratio 83% 105% 114% 118% NetEnrollment Rstio 77% 86% 87% Grades One Two Thino Four Fve Skn Total

RepetitionRates 32% 21% 20% 16% 16% 6% 20% Over-AgeStudents by Grade Oneyearormore 32% 47% 57% 64% 70% 72% Twoyeamsormome 10% 18% 26% 32% 38% 40% ThreyYosmormom 4% 8% 11% 14% 18% 19%

Avenage#YearsperGrade 1.6 1.4 13 12 1.2 1.1 13 SummryResults Percentagewho Graduate 77% P _cetageof Entering Ohort whoGrauate without Repatig 22% Percentagewho Graduate without Repeaing 29% StudentYears Per Gradat 8.4 AverageNumber of Yearsin PimaryEucaion 6.4 AverageNumber of GradesCompleted 4.9 ibputOutput Ratio(Student Years per GraduaWYeas in Cycle) 1.4

Soures: UNESCO,Statistical Yearbook; Schiefelbein, E. and Helkkin, S. Peru Acoess,Repetitibo and Efficicy in Primy Education.OREALC, October 1991. resources per graduatefrom Grade 6, 25 times teachingpractices there do naotdiffer substantially more than what would be required in the absence from practices in regular schools. Since most of of repetition and dropout. On the other hand, in these teachers have received very little training Lima it takes 7 years to produce a graduate,only they tend to use the objectives,contents, method- 17 percent more than if no repetition or dropout ologiesand materials that havebeen developedfor existed. a setting of one teacher per gamde,without adapt- ing them to the reality of multigradeteaching. As 3.9 The lack of textbooksand teachingmate- a result, teaching is of poor quality and much of rials is a serious constraintin multigradeschoos, the time spent by students is irrelevant.Most of most of which are located in rural areas. Few pri- theseschools also lack furniture,have not been de- mary students have textbooks and schools have signed to meet the needs of multigradeteaching, little or no teachingmaterials to supportteachers' have not been properlymaintained and lack even activities.In the absenceof textbooks,students at- basic sanitaryfacilities. tendingmultigrade classes are left idle or assigned irrelevant tasks while the teacher attends to stu- 3.10 The poor qualityof primaryeducation and dents in a differentgrade. Studiesof conditionsin signiicant intra-countrydisparities are reflectedin rural areas and multigrade schools suggest that sudent perfomnce. Whilemo recentinformation

29 is not available,a 1981 studyrevealed that a large percentageof students reachingGrade 6 had not TABLEM3. mastered basic skills in mathematics,social sci- MORTATY RATE Y GION ences and naturalsciences (See StatisticalAppen- ( tOlvebhts) dix Table 1.13).Students living in the siea region and other ural areas performed the worst and Range those in Limathe best Since teachingconditions Avrage by Department have deterioratedsubstanutially in Peru durig the 1980s,both in terms of the amount of resources Amazona as allocatedto educaton and the numberof certified ACroqpa 73 teachers, performancetoday is likely to be even Chavin 87 lower. Gtau 94 72 96 Inca 123 88 126 wae 101 62 131 3.11 A study of the academicperformance of Maniategn 105 73 114 studentsattending secondary schools in Limapre- MaoaIs 87 73 95 sents a similarpicture. A large percentageof sec- S M-LaLibrtad 70 66 81 ondary students taking the test at the end of the Ucayali 85 cycle scoredbelow a minimumstandard in math- Lma-ca 57 55 58 ematics, nuae, social sciences, and natur sci- tOW 83 55 131 nces. In spite of the low levels of academicper- formance25 percent of the relevant age group is 'The 1991/92Demograpc and HealtnHrvey ress enrolledin institutionsof higher education.Gov- forPeru bndicate a nationalIMR of 64/1,000,with raes emrnentspending on higher education represents forrions fm 30,0Lma)to 103/1,000(Inca). about a third of total expenditures on education Source:Mist&Y ofHeakh, INFJ (1984 data), surpassed on the continent only by Venezuela.This indicates a need to re-examine budget priorities, and to explore whether public spendingon higher educationresults primarilyin increasedproducdvity or postoned unemployment Healthand Nutition

3.12 Anotherissue is the perstence of a rda- Health Conditions tively hih rate of iteracy. Between 1961 and 1981the illiteracyrate deceased by slightlymore 3.13 Peru faces serious health problems, in than half, from40 to 18 percent.However, in 1986 terms of both mortalitYand morbidity.Infant and illiteratesstill represented13 percentof the popu- maternal mortalityrates are among the highest in lation.Census and householdsurveys show that il- LatinAmerica. There are wide disparitiesin health literacyis primarilyconcentrated in rurwalareas and conditionswithin the country. The risk of infant affects disproportionatelywomen, the Quechua death in the Inca, Wari or Mariateguiregions is speaking populationand those living in the geo- twice that in Lima(Table 3.3). At the departmental graphic area called"Trapecio Andino". More than or provincial levels, differences are even more a third of the populationliving in rural areas was striking.A major share of morbidityresults from illiterate of which 70 percentwere women.Ciov- causeswhich would be preventablethrough vacci- emnmentefforts to eradicate illiteracy have been nation or improvementsin sanitation.In 1991, for more effective among males and urban dwellers example,Peru faced the most serious outbreakof than among womenor rural dwellers.During the cholera in the Americas in this century. Typhoid eightiesprogress to reduce illiteracyslowed con- fever and hepatitis are both endemic. In some ar- siderably comparedto the gains achieved during eas, leishmaniasis,yellow fever and dengue are the 1970s.As a consequence,the numberof illiter- importantsources of morbidity.Added to already ates increased by 1.5 percent between 1980 and serious morbiditypatterns is evidence that some 1986. important diseases are on the rise. Tuberculosis

30 cases have increased 20 percent over the last 10 PERU: TABLE35. years and malaria cases have doubled over the PERUTV4SMTOF CONTAC!S 1989 same petiod. N THEU 1 IRNAL-CHI) PROGRAM

Status of Services Numberof GrOup Contactsper 3.14 The primary health care network has ex- hIdividual panded over the last tel. yeats. The number of health centers has increasedby 60 prcent and the Pgnant Women 1.9 number of health posts has doubled. Curently JAmting Women 0.4 posts Familyplanning 1.7 about 1,000 health centers and 3,000 health UnderOne Yea 1.6 exist nationwide(about 1.9/10,000inhabitants) 14 Yeuars 1.6 potentiallyan acceptablenumber depending on the 5-14Years 15 geographic dispersion of the populationand the Vacinations< 4 Years 1.8 personnel and equipment available at eachsa- O amongthose raivin carefm lishment- Ministryof Healhi

3.15 Table 3.4 shows the coverage of some ac- Source:Ministry of Health tivities under the Ministry of Health's matenal and child healthprogram. The coverae of infants underage one is probablyoverstated becase par- ents mustvisit a healthfacility in orderto receivea for all activities.For example,on average,lactating birth certificate. Often the visit occurs when the mothersreceiving care have only 0.4 healthcon- child is a coupleof days old and is not followedby tacts per year and pre-school and school-age later visits.About 60 percentfor prenatalattention groupsshow similarlylow numbersof contacts. is relatively high. However, coverage of other equally important services is extremelylow. For 3.16 The availability of pharmaceuticalprod- example, family planning progwamsserve only 6 ucts is a critical factor in the provision of high percentof womenin the feraileage group,well be- quality health services, both to achieveeffective low regionalnorms. Examiningthe qualityof care treatmentand to create the confidenceof benefi- for vulnerablegroups, Table 3.5 shows that it is ciariesand healthworkers. Pbarmaceutical supply significantlybelow the targetlvels of the Ministry to the public sector has sufferedfrom: (a) a disor- _anized logistics system; (b) budget constraints; TABLEX3 and (c) the failure to place high priority on drug PERU:COVERAGE OF THE provision.In 1984,about $145 million was spent MATERNAL-CLWPRO.RAM 198 in Peru on medicinefrom privatepharmacies, in- dicatingthat the commercialsector is an important Total PercentRecewmg source.However, in the face of decliningincomes Population Population Minisy of Health and higher prices for importedmedicines, many Subgroup ('000) % people are relyingmore on the publicsector, exac- Pregnat women 645 57 erbating the Ministry of Health's supply problems. Lactaing women 630 13 FamilyPlannng 1 4,950 6 Issues Less than one year 630 86 1.4years 2,221 2 3.17 While the distributionof primary health care infrastructureis good, the poor distibuton 1. Womenof fertle age of professionalresources is a major issue. There are three dimensions to the problem. First, al- Somrce:Mhrisfrh of Health thoughthe totalnumber is adequatein somefields, such as doctors, health professionals are ten times

31 T,,3 recent economic crisis has been the marked NURSESERDOCTOR physical detedoration of the buildiW and shortages of equipment As a result the signifi- 1975 1980 1985 1990 cant cons;tcton program for health centers has not had the intendedresults. As of 1990only thr- Peru 0.5 0.70 0.66 0.66 quartersof the centerswhich had been constucted werefunctioning because of the lack of equipment ainAmedca 2.01' 1.95 200 - a. 1974 3.19 One effect of the recent economic crisis a 1974 seems to be a very low utilization of primary healthcare servicesas seen in a small sample of Sour: Minisy ofHealth and Orosb, Sodal health centers in Limaand Piura (a city in the apending Lan Ameica, WorldBank Discussion costal region).In severalcenters, the averagepro- PaperNo. 106. ductivityof the professionalwas 2 consultations/ hour, whereas an acceptable norm would be 41 hour. Other examples of low production are the more availablein Limathan in the least developed few emergencypatients seen on the eveningshift, regions.In the sierra region,physician availability the smallnumber of laboratoryanalyses done per is onlyabout 1/34,000people. Second, the compo- day and the low occupancyrate in maternityhospi- sition of skills is not optimal mainly because the tals. In the South LJimaUDES, an analysis of ser- ratio of nurses/doctorand nurses/populationis too vice production demonstrates a significant de- low. Throughout the country the availability of crease in most activitiesbetween 1988 and 1990. nurses is clearly insufficient.PAHO recommends The average numberof consultationsper hour for a ratio of betweentwo to four nursesper doctor.If all personnelin healthcenters was 2.44 for doctors the ratio is lower,it impliesthat doctors,who cost and 2.04 for midwives,relatively low considering moreto train and to pay, will have to performtasks the highpopulation density. The poor utilizationof for whichnurses are adequatelyskilled. 'The ratio health services is likely to be due in large part to of nurses to doctorsin Peru is belowone, substan- the lack of effectivesupply. Part of the explanation tially lower than the average for Latin America. may be user perceptionsof limitedvalue because Third, as alreadymentioned, the aspect of service of lack of equipment,supplies, etc. in healthfacili- most imPorant to the poor is primary healthcm. ties and poorlytrained staff. The resourcesfrom the Ministryof Health in this area are 1/16,000for physiciansand 1t20,000for Nutitlon nurses, extremelylow ratios.Health professionals who work for the public sector have been ad- ChildNutri0onal Status versely affected by the decline in real salaries in the civil service.This has causedmorale problems 3.20 Chronicchild malnutritionis a seriousand and work stoppages.A majority of Ministry of widespreadproblem. A nationwidesurvey carried Healthworkers were on stike for four monthsdur- out in 1984 found that six percentof childrensuf- ing 1991 in the midst of the cholera epidemic. fered from some degree of acute malnutrition Nevertheless, the total supply of doctors and (wasting)and over51 percentsuffered from some nurses in Peru rose at an annual average rate of degree of chronic malnutrition (stunting). Both nearly 5 percent between 1985-1990,exceeding acute and chronic malnutritionwere concentrated the rate of populationgrowth. among childrenin the sierra regions.Over half of sierrahouseholds had malnourishedchildren, com- 3.18 Althoughthe numberof primaryhealth pared to 13 perc- r. of Lima households, 16 per- facilities appears adequate,one of the effects ofthe cent of selva haus,holds and 19 percent of other coastalhousehold. While overall, 6 percent of 1. "SocialSpendinginLatinAmerica:TheStoryoftel980st"by childrenwere severelystunted, 11 percentof chil- Marr rosh,Wodd BankDiscuon PaperNo. 106. dren in the sierra were severely stunted. Only 1

32 percent of children were moderatelyor severely TAKE3.7. wasted.Child nutritional status had improvedlittle PEU:sDsEA TRENDS between 1972 and 1984. The 1991/1992Demo- (C pMe,Mpopuaien) graphic and Health Survey results suggest that little has changedsince 1984. This recent survey 1980 1989 again found 50 percentof childrenunder age five sufrn_ g frm somedegree of chronicmaltiton. Tios 93 124 AcuteRespity Iness 470 3,146 3.21 A regionalcomparison of child nutritional AcuteDisihe Dease 444 1,379 status in Peru with Bolivia, Colombia, Ecuador S and Northeast Brazil based on UNICEFdata for Suveillan the mostrecent yearsavailable in each country in- dicatesthat in 1984 Peru had the highest propor- tion of moderately-malnourishedchildren (11 per- period. cent) and a relatively high share of severely mal- nourished(2 percent)based on weightfor age cri- 3.24 Data suggest that food security has de- tera. The data also suggestthat higheroverall food clined for the poorest quintile of households in availability is no guarantee of better nutrition Lima as comparedto 1985/86.This is in spite of among children,since Peru's food availabilityis the fact that the shae whichLima residents spend better than countries with better child nutrition. on food decreasedacross all expenditurelevels be- Other factors which influence child nutrition in- tweenJune/July 1990 and October1991. The most clude the lack of adequate householdpurchasing likelyexplanation is a decline in the "lative price power,inappropriate feeding beliefs andpractices, of food (Table 3.9). For other parts of Peru, poor and child infection. householdsin the uralsierra continueto spend the highestproportion on food, followedby the poor 3.22 Patternsof growth falteringamong Peru- in the urbancoast and the urbansierra (Table 3.8). vian childrendunng the weaningperod (between 6 and 24 months)and later improvementssuggest Micronutres that child growth is jeopardizedmore by inappro- priate weaning practices than by lack of food. 3.25 A national smuveyin 1986/87 estimated Comparedto the U. S. NationalCenter for Health that goiter is endemicin 88 percentof communi- Stafisticsreference population, the growthof Peru- ties in the sierra and selva and that the average vianchildren tends to slowby 6 monthsof age and prevalenceamong school children is 34 percent. to resumeexpected rates by 24 monthsof age. The problemis most commonin areas where diets have no naturalsource of iodine. Since 1940, Peru 3.23 In lightof the close link betweenmalnutri- has had laws governingthe productionand distri- ton and infection,information on morbiditytrends bution of iodizedsalt. In practice, however,only providesadditional insight into the nutritionsitua- 60 percent of the population consumes iodized tion. Surveillance data from the Ministry of salt. Reasonsfor this low coverageinclude higher Health's EpidemiologyOffice highlight significant cost, irregulardistribution from productionpoints increases in the incidence of acute diarrheal dis- along the coast to endemic zones and the large ease and acute respiratoryillnesses. As Table 3.7 numberof naturalsources of non-iodizedsalt. Na- illustrates,from 1980 to 1989, the rate for acute tional programshave successfully overcome simi- respiratory illnessesincreased six-fold and that of lar obstaclesin Brazil and Bolivia by subsidizing acute diarrheal diseases doubled. This suggests the increasein productioncosts and by involving greater precariousness in living conditions, de- small productionfacilities which marketsalt from creased effectivenessof preventive health mea- these naturalsources. sures and the increasinglyvulnerable nutritional statusof the population.It is unclearto what extent 3.26 No comprehensivestudy has been under- increases also reflect better reporting in the later taken to determinethe prevalenceand concentra-

33

ka TABLEM3.L TABLE3.9 PERU:CHANGES IN HOUSEOLD PERU:FOOD CONSUMiON AS FOODSHARE PERCENTAGEOF TOTALCONSUMPllON BYREGION AND ENDITlE LEVELS FORLur J ouSEHOLS FROM15686 TO1991 Quintilo 1985/86 Junewuy October Poo0oCPoo" AU 1990 1991 10% 30% 1 53 64 59 Urban 1985/86 65% 62% 49% 2 55 61 55 Coast 1991 69% 67% 55% 3 51 55 52 4 48 51 49 cdan 7% S% 6% 5 38 42 33 Urban 1985/86 64% 65% 48% Average 45 49 43 Sioera 1991 61% 63% 50% Source:Glewwe and Hall, 1991; Cuanto 1992 chae -3% -2% 2% Rual 1985/86 72% 77% 72% DevelopmentProgram is a nationwidepromotion Siera 1991 76% 76% 68% program In additionthe Ministryoperates the Na- tional Program to Control Goiter and Endemic dcue 3% -2% 4% Cretinism(PRONABCE). The Child Growth and Source:Glwwe, 1987; to92 DevelopmentProgram aims to promoteintegrated child health, improvechild survival and improve the qualityof life. Four of the program'seight ob- tion of iron deficiencyanenia. Smallscale studies jectives relate to growth monitoring:(a) monitor- amongpopulation subgroups indicate that anemia ing growth as a way to target children for nutri- is a public health problem, particularly among tionalsupplementation; (b) promotingsupplemen- pregnantand lactatingwomen and children.With tary feeding for at-risk mothersand children; (c) increasedeconomic hardship, rates of anemia are identifyingand rehabilitatingmalnourished chil- likely to have risen as people reduce consumption dren throughcommunity food distribution;and (d) of animal and other iron-richfoods, relativelyex- conductingcommunity outreach to stimulatede- pensive sourcesof calories. mand for growthmonitoring. The other four relate to improvedprenatal nutrition, breastfeeding and Nutritink Services approprate weaning; (a) improvingbirthweight throughmaternal care, includingfood, iron and io- 3.27 Programsto address malnutritioninclude dine supplementation;(b) encouragingjoint rooms food assistance, health programs and education in hospitalsfor mothersand newborns;(c) provid- campaigns.Peru's extensiveexperience with food ing trining at healthservices on techniquesfor ex- asance is coveredin the next chapter.However, clusive, long-termbreastfeeding; and (d) promot- feeding programsare not the complete answer to ing appropriate weaning. While program objec- improvingnutritional status. Malnutritionis not tives are comprehensive,its budget and activities onlythe result of a lack of food. Parasites,diseases are inadequateto achievethem. In 1991, the oper- and other health problemsare often to blame. To ating budgetwas under$300,000. The low funding be effective,therefore, nutrition programs need to for the programcontributes to a high concent,ation be combinedwith health programsincluding ma- of actividesin Lima.The programalso centralizes ternal and child health, immunization,sanitation, supervisoryresponsibility, making supervisionof dewomting,oral rehydrationtherapy and micronu- health centerpersonnel an expensive undertaking trient supplementation. Education in primary with limitedinvolvement of the subregionallevel. health and nutrition is also crucial where local practicesare inappropriate. Issues

3.28 The Minis of Health'sChild Growth and 3.29 Nutritionin Peru is characterizedby the

34 participationof a widerange of actors- the central sis for a nutritionsurveillance system. The Minis- government, regional govemments, community try of Health and regionalgovernments have been groups, donor agencies,private firms and NGOs. collaboratingwith the NGO PRISMAto establish Considerable gaps and duplications occur In regionalsurveillance systems. This work has be- the hnctions of these actors. Withinthe central gun in Arequipa,Lima and Iquitos. The Govern- government,the Ministryof Health is formallyre- ment soughtto promote greater interagencycoor- sponsiblefor problemdiagnosis. However, t does dinationby organizingthe SISVANNetwork, a not always guide actions by other central gov. loose associationof governmental and non-gov- ernmentagencies. PRONAA operates on the ba- enmmentalagencies active in nutrition.Within tight sis of its own problem identificationas does the resource constraints,a nutritionsurveillance sys- Ministry of Education (School Feeding: "La tem to monitorsmall sentinel populationsrepre- EscuelaDefiende La Vida").No clear mechanism sentativeof importanttarget groups would be ap- exists for prioritizingproblems or for channelling propriate.Selected local health authoritiescould governmentefforts to resolve them. Multilateral periodically monitor nutrition status under the and bilateral donor agencies, non-governmental guidanceof regionaland Ministryof Health offi- agenciesand communitygroups base their activi- cials. This would complementcurrently available ties on their own perceptionsof Peru's nutrition data dawn largelyfrom programbeneficiaries for problems. programmonitoring and evaluationpurposes.

3.30 Thefunction of establshig norms and 3.33 Peru does not have a nutiton policy. standards largely nglcted. The Ministry of The absenceof a frameworkhas led to duplication Health only establishesstandards for food quality of functions among state and private agencies, a and growth monitoringprotocols. Multilateral and lack of atteaion to achievingmajor improvements bilateralagencies establish their own standardsin in nutritionalstatus and less than optimalresource other importantareas. The Govenmmentshould de- use. Peru needs a national nutrition policy to velop standards in three key areas: (a) unit costs achievefour majoraims: (a) to establsh consensus for interventions;(b) food assistanceprogmming on the priority ranking of problems and target practices; and (c) informationcollection and im- groups; (b) to identify key functions in nutrition pact evaluation. and establish institutionalresponsibilities; (c) to set programstandards for impact, informationre- 3.31 No agency of the central government Is quirements, unit costs and quality, and (d) to responsible for mobilizing resources Instead, strengthen its dialogue with extemal agencies. donors base contnbutionslargely on their own di- While a fundamentalconsensus exists that preg- agnoses, funding constaints, dialoguewith gov- nant and lactatingwomen and childrenunder age ernmentand non-governmentalagencies, and spe- five are the most vulnerable group the nutrition cificproect proposas. Outsidethe centralgovem- policy will need to establish Peru's priority rank- ment, comm ty groups,NMOs and regionalgov- ing among other important groups. How will ernments mobilize resources based on specific choicesbe madebetween the maternal-chiidgroup proposals.The centralgovernment could probably and school children? Will rral populationsbe mobilizegreater resourcesfor nutritionif it took a given preferenceove urban ones? Will resources more active role in its dialoguewith donors. The be directed proportionallyto all regions or will Government should assign this function to the criticalregions receive the resourcesrequired for Ministryof Health and it shouldseek to mobilize significantadvances before funds are channelledto resourceson behalf of bothgovermental and non- other areas? A substantialproportion of nutrtion governmental agencies on the basis of priorities resources come from external donors and major and implementationcapabilities. progams are implementedby NGOs.Thus, it will be importantto seek donor and NGO feedbackon 3.32 Peru does not have a formal system for the policy as it is being developed.As part of its nutrition swrveillnce on a nal leveLHow- strucua adjustmentprogram, Peru prepareda nu- ever, severalongoing initiatives could form thi ba- tritionplan in 1991.This plan proposesreforms in

35 foodassistance and a strongerrole for the Ministry illiteracy. Health statlusis poor is Peru. Needed of Health.Under the plan, donatedfood use would equipmentis lacking and personnelis poorly dis- be rationalized to achieve nutrition objectives, tnbuted. The intensityand quality of services are with priority atteationto poor familieswith preg- not adequateand the latter may havecontnbuted to nant and lactating women and chDdren under a decline in the demand for health care. Peru suf- three. The plan reflects valid objectives and the fers from relatively high rates of child malnutri- Govemeit owneds to takestop implementto it. tion. The responseto malnutritionhas excessively reliedon food assistanceand given insufficientat- Coucluslos tentionto complementaryhealth measures, mnicro- nutrient supplementation,and the promotion of 3.34 This chapterhas reviewedthe status of the better health and nutrition practices, including education,health and nutrition servicesmost im- breast-feedingand child growthand development portant for the poor. The major issues confronting monitoring.The next chapterwill carry this analy- these prgrams also havebeen highlighted In edu- sis fiuther by looking in detail at the use of these cationthe main concernsare higb repetitionrates, servicesby the poor and identifyingactivities for a wide disparitiesin educationantainment and effi- povertyalleviation program. ciency within the country and the persistenceof

36 4

Towards a PovertyAlleviation Strategy

fttroduction broad-basedgrowth and improved,more equitably distrbuted social services. A discssion of two 4.1 In the longrun povertycan be reducedonly suchprogams concludesthis chapter. Mechanisms if broad-basedgrowth is achieved.However, the to generate employmentare the subject of much poor must have sufficient access to health and interestin Penr Food assistanceis a major public education services. Otherwise they may not be programspending substantial resources. educatedor healthyenough to take advantageof the opportunitiesavailable in an expandingeconomy. Sodal Services Also,it will take timeto dismantlebad policiesand respondto new incentives.The benefits of growth Education may not flow immediatelyto some of the poor, pariculalyinPeruwheremacroeconomicinstabil- 43 EDUCATIONAND THEPOOR. Poverty ity and economicdeterioration have been so severe influencesschool attendancein two ways. First, and where extremeregional disparities exist in the childrenof poor familiesbegin attendingschool at welfareof the population.Vulnerable groups such a later age; second,a larger proportionof children as mothersand youngchildren, the old and the sick of the poorestfamilies never attend school or drop- are not able to participatefully in the economy.For out very early. Data showthat attendancerates for these reasons,a strategy to alleviatepoverty needs chidren betweenthe ages of 6 and 10 are lowerfor to incorporatespecial efforts in social services.The thepoor.(rable4.1)Amongpoorhouseholdsinthe issues relatingto basic education,primary health rural sierra,8 percentof childrenaged 6-10 are not and nutritionservices, which werediscussed in the attendingschool, a high figure in the context of previouschapter, suggest cerain linesof action.In neary universalprimay educationin otherpars of this chapterthat informationis supplementedwith the country.' Two possible explanationsfor the dataon the use of socialservicesby the poorinorder higherattendancerates for olderchildren from poor to Wdentifyspecific p amsfrpoverty alleviation householdscomparedtothose for youngerchildren are thatthey start schoolat a later age and that they 4.2 Even with improvedsocial services,in the short-runsome groups may remainvulnerable be- 1. Anendacemm fmmte PLSSare higher than Govenment cause they are alreadyliving on the edge. For these admat (ppg.aph 3.3).The Govenmnt figures ane deived peoplewelldgnedsociacompensationproams fromenrome informadonfrwm schools wad esimates of the could complement the two main elements in pouaonfiee dh oemagecategorks.n pthecaseoftheSSb the poverty alleviation strategy - promotionof hndef anbd on rpopdlmation

3- TABLE" . PERU:SCHOOLATTENDANCEDY 4.6 EDUCATIONPROGRAMS TO ALLE AGEGROUP AND POVERTYLEVEL VIATEPOVERTY. The firstpriority is to increase the internalefficiency of schoolsin poor areas.The ExtremelyPoor Poor Non-Poor poor academicperformance of primary students ADAa andlow completionrates among those in ruraland poor uban areas, demonstratethe importanceof 6.10yxeasf improvingthe capacityto retainstudents, to move AntrtdsSchool 94.1 95.6 989 them fasterthrough the systemby reducingrepeti- DoesNot AttendSchool 5.9 4A 1.1 tion,and to increase the knowledge they get through

1114 yars schooling.Improving conditions in rural multi- Atends School 98.8 993 99A grade schools would particularlybenefit the poor. DoesNot AttendSchool 1.2 0.7 0.6 This can be done by developingspecialized text- Pi"al Sian books and materials and providing teacherswith Rtaininggeared specificallyto the requirementsof 6.10 years multigradeschools. AttendsSchool 92.2 92.1 97.2 DoesNot Atend School 7.8 7.9 28 4.7 Second,a programto providetextbooks 11.144ye and teaching materials and to train teachers for AttendsSchool 97.1 98.0 98.6 otherschoolsinpoormarginalurbanandruralareas DoesNot Ated School 2.9 2.0 1A can improve the quality of teaching,increase stu- Source:1991 dent time dedicated to school work and provide Sourc:1991 PLsS morerelevant learning experiences. These contib- ute to reducingrepetition and to improvingintemal repeat grades. efficiency which will load eventually to higher primary school graduationrates at fewer school 4.4 Attendancerates for female students are years per graduate. somewhatlower than for male students,but by a relativelysmall amount Thepatternisevidentinall 4.8 Third, pre-school programsthat concen- areas except Uma,but the pattern is more pro - nouncedamongtheextremelypoor,withtheexcep- TABLE4.= tionof the ruralsierra where it is presentamong all PERU:KINDERGARTEN AIF1ENDANCE BY households. POVERTYGROUP

4.5 One factor behindthe relativey high rep- Ermly Poor Non-Poor etitionrates for childrenfrom poorerhouseholds is Poor lower kindergartenattendance (Table 4.2). About Peree Atending 22 25 37 86 percentof 5-year olds in non-poorhouseholds attendkindergarten, whereas the sharefor the poor PRONOE 43 34 19 is 55 percent snd for the extremely poor only 45 Go[e K 8 4 percent.Thelowerattendancemayreflectapercep- o 1 1 o tion of low qualityservices or a lack of knowledge of the benefits of kindergartenbecause the main Kindat by Age reason given for non-attendanceis that the "chil- 2 9 7 10 dren are too small", not financial constaints or a 3 18 18 24 lack of facilities.Children from poor households 4 34 40 60 attend mainly PRONOEI or Government-run 5 45 55 86 schools,while childrenfrm better off households attend private kindergartensin substantialnum- Sore: 1991PLss bers, as well as Government-runschools.

38 trate on prepaing studentsfor primary education TABLE4.3. have been shownto be useful in other countriesin PERU:PERCENTAGE OF CILREN reducing repetition in the early primary grades. RECEIVINGALL REQURD VACCINATIONS Such programs both improve the likelihood of (Sff Reported) student promotion and reduce the risks of ay Poor* Non- 0-2 3-5 6 9-12 dropoutand shouldbe stressedmore in the poorer poor. areasof Peru. Stepsneed to be takento improvethe preparationof childrenfor schoolin the PRONOEI Al Are 75 81 52 78 86 90 and Government-runkindergartens and to promote ijma 84 86 58 88 94 96 UrbanCoast 75 80 61 68 85 93 greater awareness of the values of kindergarten UrbanSierra 77 82 53 81 91 89 among the poor. One way to attract poorer students RuralSiem 65 70 37 69 75 78 would be to offer school feeding programsat this level in some localities. 'Ages0-12

4.9 Fourth, a program to teach non-Spanish Soume:1991 PLSS speakingwomen basic functionalskldls should be part of povertyalleviation efforts. The incidenceof based or PLSS-basedrates for youngerchildren. povertyamong non-Spanishspeakers is vety high. mmunizationcoverage for childrenunder two in Since illiteracycreates obstacles to using informa- other areas ranges from 50 - 60 percent. These tion and hence to impoving living standardsfor comparetocoverage of nearly 90 percentfor older themselvesand theirchildren, poor illiterate women children.In general,vaccination rates are higherfor are trapped in a vicious cycle. Basic literacy and childrenin betteroff households. numeric skills can help to break this cycle. Pro- grams to reduce illiteracyin this group must con- 4.11 Vaccinationcoverage usually overstates siderthattheyarewidelydispersedandthushardto access to health services because immunizations reach and that their free time is minimal.Experi- are the focus of specialcampaigns. Table 4.4 pre- nce from other countriessuggests using NGOs. sents data on the share of ill people who seek Theyhave the capacityto reach deprivedareas and medicalcare. As expected,these numbersshow a to incorporateliteracy programsinto ongoingac- lowercoverage than for vaccimtions.There is also tivities,thereby reducingcosts. a wider gap between the poor and the non-poor. Overall, half of all ill people seek medical care. Health Amongthe extremelypoor, this drops to about one in three. As for vaccinations,that gap is matched 4.10 HEALTHANDTHEPOOR.Vaccinations and in someareas (ruralsierra) exceeded by differ- are important to improve the health of children. ences betweenregions. Access to healthservices is Datafrom the ExpandedProgramme on Immuniza- more equitablydistributed in Limarelative to other tion (EPI) indicate that coverage of infants under 1 with 3 dosesof DPT (a commonmeasure of vacci- T .4LE" nation programsuccess) was 71 percent in 1991. PERU:PERCENTAGE OF ILL WHO SEEK The lowest coverageregistered under EPI was 59 MEDICALAITENTION percent for the measles vaccine. Yet based on the Eeely Poor Non-Por response of mothers, recent data from the PLSS Poor suggest lower coverageof immunitions among children under age two (Table4.3). In the wral ADAreas 34 44 56 sierra, only 37 percentof childrenunder two were Lima 44 53 59 Urba Coast 49 54 63 reportedto be fully immunized.While this may be UrbanSiea 31 40 55 partially explainedby the inclusionof infants too s Rau &erra 24 27 46 young to have been fully imunized it appearsto signal a breakdown in services since rates among Sowce 1991PLSS older children are much higher than either the EPI-

39 areas.There is a particularlylarge gap betweenthe whichhave public seweragefacilities account for poor and the non-poorin the sierra. fewer than 40 percentof the cases.

4.12 Where are the poor likely to go for care 4.15 Health Propms for Poverty Alevia. when they are ill and whom are they likely to tDon.The first priority is to provide mothers and consult?In general,the poor are more likelyto use young childrenwith adequate primary health health centers, health posts, and pharmacies,than care. This wouldinclude: (a) pre-natalanti-tetanus the non-poor.Nevertheless, the highestproportion immunizations,andironsupplementation;(b)sani- of the poor (30 percent)use hospitals.On the other tary home deliveryand trainedpersonnel for high hand, clinics - most of which are private - ar. risk deliveries;(c) child immunizationand growth overwhelminglyused by wealthierhouseholds. In monitoring;(d) family planning informationand Lima, whilethe extremelypoor are more likely to supplies;(e) healtheducation; and (f) treatmentof visit a health center,the poor make greater use of commoninfectious diseases and first aid. hospitals.In the urbansierra, the poor use hospitals to a greater extent and there is little reliance on 4.16 Thegreatestbarriersthehealthsystemfaces health centers,particularly by the extremelypoor. in providing care are lack of supplies and inad- In the rural sierra,the extremelypoor makegreater equate human resources,particularly in areas of use of hospitalsand less use of health centers and extreme poverty. It is important to improve the posts than the rest of the population.The poor are supply of basic equipment, mateials and medi- more likely than others to consult paramedics, clues necessaryfor carrying out these activities. pharmacistsand traditionalhealers. But across all These would includecold chain and related equip- expenditure levels the main source of medical ment for immunizations,iron tablets, ORS packets consultations are doctors, obstetricians, and and basic drugs, familyplanning supplies,educa- nurses.InLima and the coastalurban regions, they tionalmaterials, growth monitoring equipmern, di- accountfor over 80 percentof consultationsby the agnostic equipmentand materialsfor first aid and poor.In the ruralsierra the shares are 66 percentfor deliveryprocedures. Also critical is the hmprove- the extremelypoor and 60 percentfor the poor. meat of human resources in primaryhealth care, through technical and managementtraining and 4.13 Table4.5 providesfigures on the share of improvedworking conditionsor other incentives those ill or injured who purchasedrugs. The data for staff in poor areas.The Governmentalso needs show that the extremely poor household is less to shift human resources into primary care from likely to purchasedrugs than others.However, in other areas. both Lima and the coastal urban regions there is oniy a minor difference between the extremely 4.17 Since deterioratedinfrstrcture is likely poor and the rest of the population.In the sierrathe to poseanother significant barrier to primaryhealth fraction of extremely poor householdsthat pur- care, these facilities shouldbe repaired. As the chase drugs faUsfrom the 80 percentregistered in distributionof infrastructureis generally adequate, Lima and the coastal urban area to less than 60 the Governmentneed not constructnew facilities. percent. There is only a small regional variation However,if poorliving conditions make it difficult in the fractionof non-poorhouseholdb purchasng to attract staff, a programto upgrade staff hous- dgrg. ing for primary health care facilities might be helpful. 4.14 The lack of access to a public sewerage system is clearly related to problemsof diarrhea 4.18 Cholera prevention and treatmentis a among children.While only 17 percent of house- secondpriority followingmaternal and child care. holds lack seweragefacilities, they accountfor 35 Cholerais a serioushealth problem. The fact that it percentof child diarrheacases. Some improvement affectsprimarily the poor lendsit addedurgency as is seen with use of latrines; 16 percent of house- a poverty issue. In the short run, the Government holdsrely on latrinesand they account for newly20 can strengthen and expand the educationalcam- percentof the cases.The 63 percentof households palg launchedfollowing the first outbrealkIt can

40 TABLE4.. PERU: DRUG PURCHASESBY POVERYf LEVEL

Location All Areas Lhna Urban Coast

Extremely Poor All Extremely Poor All Extremely Poor All Poor Poor Poor

% of lIU/njuted Wbo Purchased 66 74 79 82 84 84 79 81 83 Drugs

Location UrbanSierra RuralSier

Etremely Poor All Extemely Poor All Poor Poor

% of IlMlnjured Who Purchased 58 67 76 55 59 68 Drugs

Source: 1991 PLSS

aLsoensure that all health facilities have adequ- infantsthroughfourtosixmonthsofageandensure ate suppliesof medicine,rehydration solution and that weaning foods are adequate for growth. Pro- materialsto treatacte cases.In the mediumterm, gramswhich prevent or treatinfections in the wean- programsto improvewater qualityand basic sani- ing period(diarrhea, acute respiratory ilness, vac- tation conditionsthrough water system rehabili- cine-preventablediseases) will also contribute sub- tation and latrine Installation are also important stantiallyto improvingchild nutrition. The children in rural and margial urban areas. who are the most importantto reach are thosefrom poor householdsin the sierra. 4.19 The third priorityis to strengthenhospital emergency services In poor areas To do this, a 4.21 The programswhich can promote appro- programshould include supply of basic equipment, priatebreastfeeding and weaningare maternaland suppiies and medicines,introduction of manage- child healthcare and nutritioninformation, educa- ment improvements, rehabilitation of infra- tion and communicationsprograms. A compre. structure,vehiclesandcommunicationsequipment, hensive program begins at pregnancyto ensure and taining and improvedworking conditions for adequatematernal nutrition and to introduceex- staff. pectantmothers to the importanceof breastfeeding, and continuesthrough regular infant growthmoni- Nutritfon toring/nutritioneducation contacts to promote ap- propriateweaning, immunization, and prevention 4.20 NUITION PROGRAMSFOR POV- and treatment of infection, particularly oral ERTY ALLEVIATION.The most criticalperiod rehydrationfor diarrhea. As an adjunct, such a for nutritionin Peru is fromthe age of about6 to 18 program should include close monitoring and months,when infantsare weanedfrom breastmilk supplementation to rehabilitate malnoed and provided with other foods. This is when the children and more intensivenutrition training for sharpestdeclines in growth occur and lead to short mothersand caretakers.A nutrition information, stture which is not corrected later in the child's education and communications (IEC3program life. Thus the highest priority nutritionprograms ctnuseappropriatemediatodisseminatemessages arethosewhichpromoteexclusivebreastfeedingof and improve pracficesin breastfeeding,weaning

41 and the prevention and treatment of childhood among other groups and if the project improves infections. infrastructureor facilitiesused by the poor. About two-thirdsof unemployed-leadedhouseholds are 4.22 A secondprionity is the protectionof vul- poor and about 30 percentof these householdsare nerablegroups with inadequatefood intake (mal- extremelypoor. However, the unemployed-headed nourishedchildren, lactating and pregnant women, householdsaccount for a very smallproportion of thesickandinstitutionalized,pre-schoolers)through the poor (3.3 percent)and the extremelypoor (4 supplementation.This is discussedin the contextof ereent).Thus employment programs cannot be the food assistancelater in this chapter. maintool to addresspoverty. In orderto betterreach the poor, an employmentprogram should focus on 4.23 A third group of high priority nutrition theaeas wherethe incidenceof povertyis greatest interventionsaddresses deftciencies of micronutri- amongthe unemployed(area outsideof Lima)and ents, particularlyiron, iodine and vitaminA. Such wheretheunemployed account for moreof the poor programsare impoant withinapoverty alleviation (theurban coast and urban sierra). Poor farmers also programbecausethepoorarelesslikelytoconsume would benefit from other employmentopportuni- these micronutrientsbecause of the expense(iron, ties. Two issues related to cost-effectivenessof vitaminA) and becausethey live in the areaswhich these programsare discussed below - wage rates are mostdeficient (iodine). The Goverment's goi- and administrativemechanisms. ter prevention and treatment program (PRONABCE)needs to be supportedand to add 4.25 WAGE RATES.When a poverty reduc- measuresto ensurethe iodinationof salt fromsmall tion strategy includes a temporary employment producersand temporaryiodine supplementation program, the persons employedshould be poor. in endemiciodine deficient areas until iodizedsalt Thiscan mosteasily be accomplishedif the wage is use becomes more widespreadLVitamin A defi- set sufficientlylowthatpersonswith higherpaying ciencycan be addressedparialy throughthe IEC (andhigherproductivity)altemativesarenottempted programby promotingconsumption of vitamin-A to leave them in orderto gain temporaryemploy- rich vegetablesand fruits, and adequate oil con- ment with the emergency program. A low rate sumption.The maternaland child healthprogram ensuresthat an automaticself-selection will occur. can also prevent vitaminA deficiencyin children It alsomeans that resourceswil stretchfurther than by includingsemi-annual vitamin A distributionin if the wages werehigher. An analysisof the impact conjunctionwith immunizationor when children of the Bolivian Emergency Social Fund (ESF) are treated for illnesses. Iron deficiency can be showed that the magnitudeof benefits which ac- addressedbyincludingironsupplementationamong crmedto the workerswho participatedvaneo de- the activitiesof the maternaland child healthcare pending on the pte-ESF employmentstatus. The pgam (particularlyfor pregnant women) and averageESF worker experienceda 35 percentin- by fortifying staple foods with iron where cost- creaseinweeklyeamingsoverandabovewbatthey effective. could have expectedwithout the ESF. However, those workers who were either unemployedor Soda! CompensationProrm who, prior to the ESF, had salary levels well be- low average,experienced increases five timesthat 4.24 PUBLIC EMPLOYMENTSCHEMES. high.2 Employmentprograms can reducepovery by pro- viding employment(and thus income)to those in 4.26 Considerable administrative advantage need and can generateadditional benefits through accompaniesthe use of an appropriatewage rate. the rehabilitationand reation of infiasucture. No wmezatest needs to be applied to ensurethat Theseprograms canbecost-effectivebecausepoor beneficiaies are poor since only the poor will be r-oplearewillingtoworkforarelativelylowwage. The impactof an employmentprogram on poverty 2NmigeneandPrdam,HowDidWorkrsBeaefi wil be greater if unemployed-headedhouseholds hm Bolvia'sSoLialFund? DheWorldBankEconomicReew, are found to a greater extentamuag the poor than May 1991,volume 5, No. 2, pp 367-393.

42 TABLE 46 PEMU:UNEMPLYEl BY POVERTYGROUPS (%) All Etemely Poor Others

2.S 4.0 2.1

2.5 3.2 25

3.0 7.5 2.2 Urbancoast 5.5 1.2 Urban Siem 2.3 2.7 21 RuralSiera m2.1

AM Poor Non-Poor

2.S 3.3 1.7 AU _ 2.S 4.0 ~~~~~~~~~~~~1.6 Limal 3.0 3.8 2.3 Ubanstcoa. 1.9 2.6 1A4 Urban Sierra 2.3 2.6 2.0 RuralSienra /.

Source: 1991 PISS willing to work at the wage rate offered. For the tract monitoringmust leave some flexibility for reasonsoutinedabove the maximumaverage wage skilledand supavisory personnelto earnmore than rate recommendedfor any temporaryewployment the minimum,project criteria should favor contrac- programis the legal minimumrate. What propor- tors whosewage bills reflectgeneral observance of tion of the labor forcewould offer themselvesfor the minimum.The projectcriteria shouldalso en- workif the legalminimum wage were offered? One sure that projectsbe relativelylabor-intensive. The source asse.ts that nearly one-third of the labor BolivianEmergency Social Fund(ESF) included a force earned the legal m.nimum in 1990.3Yet requirementthat no less than 60 percent of any officialsources claim that only about 10 percentof project'sbudget be spent on labor. the labor force earn the minimum wage. Either figure suggests that an emergency employment 4.27 ADMINISTRATION.Adesignfeaturethat programthatofferedthelegalminimumwagewould would promote cos.-effectivenesswould be the not lack prospectiveworkers. In order to improve locally demand-drivenselection and administra- targeting,there isan argumentforofferinglessthan tion of projects. Projects should be prepared by the minimumwage. However, the incometransfer residents and representativesof the community. mightnotbesufficienttoalleviatepovertyformany Thisensuresthatprojectsrankedlocally asthe most households and using less than minimumwage needed are done first and that the overallprogram wouldmake it moredifficult to use privatecontrac- has strong support. It also avoids administrative tors. Whilethe process of bid solicitationand con- over-centralizationand the costs of a large bureau- cracy.In orderto increasethe impactof an employ- 3. lusftwo Cuanwo,Ajuste y EconomiaFamiliu 98S-990. mentprogram on the poorestgroups, some consid- Lima,1991,p. 30. eration might be given to geographicaltargeting.

43 Althougliall regionscould be eligible to receive 431 INSTITUIONS. A wide variety of gov- financing,higher-income locations (Lima, for ex- emmentalandnon-governmentalorganizationsplay ample)might be requiredto offer a greaterpropor- importantroles in food assistance.Within the Gov- tion of local counterpartcontributions. In order to emient, the National Food Assistance Office encourageefficiency in resourceuse, theparticipa- (O.NAA)was responsiblefor coordinatingthe re- tion of privatecontractors also shouldbe promoted. ception, transport and distributionof food assis- tancecommodities nationwide. In practice,ONAA Food Assisance performedthese functions foronly aportion offood assistanceprograms. The Government has recently CurrentStatus: Magtude,Agendes,Pomg s created a new agency, PRONAA (the National FoodAssistance Program) which combines ONAA 4.28 This section examines how well current and PAD, anotheragency formerly responsible for food assistanceprograms meet the dual objectives some food assistance to urban women's groups of protectingfood securityfor the poorand protect- (comedores).Within the Ministry of Health, the ing the nutritionalstatus of vulnerablegroups. It National Feeding and NutritionInstitute (INAN) begins with a review of food assistance - its coordinatesthe implementationof several direct magnitude,institutions and existingprograms. It food assistanceprojects. then discussesmajor issues in food assistanceand providesrecommendations for bothbroad rt loo is 4.32 Non-GovernmentalOrganizations possess and specificprograms. considerable capability in food assistance. CARlTASoperates a nationwidenetwork which 4.29 MAGNITUDE.Food assistance in Peru is distributesa greatervolume of food assistancethan of growing importance in relation to worldwide ONAA.CARE,PRISMAandOFASAimportfood food assistanceand Peru's social sector resources. sommodities,design and supervisefood programs ne severe needs facing the populationmake it and mobilize donor support for food assistance imperativethat theseresources be used effectively. programs. CARE also monetizes imported food While the volume of worldwidefood assistance commoditieson the commercialmarket in Peru to increasedby about 40 percent,food assistanceto generatelocal currency for food assistance progrm Peru increasedby about240 percent(1980 - 1987). In 1991, food assistanceprograms provided over 4.33 EXISTING PROGRAMS. Eight major 226,000metric tons of food to at least nine million food assistanceprograms operate in Peru.Three of beneficiaries.In the same year, while the central these programs target food assistanceto women government'sbudgets for the social sectorminis- and children; one to school children; two to the tieswere $186 millionforHealth and $245million urban poor; and two to the rural poor. Table 4.7 for Education,donor food assistanceresources to- summarizescoverage and cost information. taled at least $150 mnillion. 4.34 The three programswhich target food as- 4.30 The transfervalue of food assistanceis also sistanceto pregnantand lactatingwomen and pre- importantrelative to the overallexpenditures of the school children are PAT-PAMI, PANFAR and poor in Lima.According to resultsfrom the 1991 Vaso de Leche. The PAT-PAMIproject is funded PLSS, in Lima food assistanceand other transfers by the World Food Programmeand implemented fromnon-profit organizations represent one fourth of by ONAA,INAN and the Ministryof Education.It totalexpenditures in extremelypoor householdsand operatesit Lima-Callao,Ancash,Cajamarca,Cuzco almostten percentof expendituresfor poor house- and Puno.It aimsto protectthe nutritionalstatus of holds.Without this assistance, the extremelypoor in pregnantand lactatingwomen and pre-school chil- Limawould consume at the sameexpenditure levels dren.The PANFARprogram is fundedby AIDand as the extremelypoor in the ruralsierra. For the poor implementedby sub-regional health authorities outsideLima, this accountsfor a mninorpercentage (UDES)under the supervisionof INAN.ONAA is of household expendituresbecause the levels of responsiblefor food warehousingand the NGO food assistanceoutside Lima are relativelylow. PRISMAis responsiblefor transportingfood to the

44 TAU 4.7. ffERU:MJOR FOODASSISTA4CE PROGRAM 1991

Vssode Leich Escuela PAT- (IUa. Defende Progrm Name PAMI PANFAR Calbs) la Vida ARCA DAIREZOD Cawdor PRODIA

Poam TYpOCH MC MCii Sdcool Rual Poor Rwr Poor lgaImPoor UrbanPoor FoodSoure WFP AID Treas Treasury AID WFP WPP AID MetricTons 17,300 21,000 7,100 .. 1,20 6,600 22,20 31,200 FoodValue (in USSmilibns) $6.7 8 0 S10.8 . $0.5 $33 S4.0 S12.1 TotalProgrmValue(inUS$millons) .. $23.7 S113 $13.7 S1A - - StS3 Beneficaries 209,000 596,000 1,200,000 3,500,000 45,000 71.000 440000 635,000 FoodCohBcnflcy $32 $13 $9 .. $11 S47 $9 S19 TotalPogram Cost/Beneficiary .. $40 S9 $4 S30 .. $24 FoodVBeneficbay(kg) 83 35 6 .. 28 94 50 49 FoodCost/Kilo $039 $038 $1.52 .. $0.40 $0.50 $0.18 $0.39 Pogram CosVi/togrm Food .. $1.13 $1.59 .. $1.09 .. - $0.49 Ohr Costsas % Food Value .. 197% ,% .. 170% . 27

SouTce:Compibtion from Maguina,Yamada, MOE, AID,WFP, CARE,PRISMA, CARlTAS repom UDES level. This program operates in 22 of the PogamopeatesiAncashApunmacAyacucho. county's 25 dpartments. It also seeks to improve Cajamarca,Cuzco, Huancavelica, Junin and Puno. the nutritional status of pregnant and lactating The DAIREZODProgram operates in Ancash, women and pre-school children through supple- Apudmnac.Areqipa,Ayacuho,Cajamarca,Cuzco, metary feeding.Although Table 4.7 includesin- Huancavelica,Ica, La Libertad,Moquegua, Piuras formationonly for the Vaso de Leche Prgram in PunoandTacna lhe beneficiariesof theseprojects LiTmaand c,IIlao, the Vaso de Leche program is are relativelyfew and food assistanceis onlyone of funded by th,. Governmentof Peru and operates several components. nationwide,wherever provincial governments are willing to assume responsibilityfor its manage- 4.38 PROGRAM COVERAGE AND INCI- ment.It providesmilk to childrenunder six and to DENCE.The October1991 PLSS information pro- pregnant and lactatingwomen on a daily basis. It vides insightsinto the way some food assistance also supplementsschool children up to age 13 programs actually work. lhe survey covers the during school vacations. Its objective is to stop Vaso de Leche program, food assistance through malnutritionamong its target groups. comedoresand othersimila outlets,and municipal progms. First, it suggests that food assistance 4.35 The major school feeding program is the reaches a significantproportion of the extremely EscuelaDefende La Vidaprogram created within poor or the criticallypoor only in Lima.There 74 the Ministryof Educationin 1990under the Social percent of the extremelypoor receive some food EmergencyProgram Thisprogram seeks to reduce assistance4.Among the poor, 37 percent benefit, school desertionand absenteeismand to improve whileamong the non-poor,only 6.5 percentbenefit. academicperformance. From August 1990 through In all other areas, fewer than four percent of the July 1991, the program provided sporadic food extremelypoor or criticallypoor benefitfrom food assistance to 3.5 miUion school children. Fifty assistance. percent of resourceswere spent in Lima, which accountsfor 21 percentof schoolenrollments. 4.39 Second,in Lima most food assistanceap- pears to be relativelywell targetedto the poor, the 4.36 The two major programs which provide non-poorgenerallyhavelowparticipationrates and direct food to the urban poor are the WFP-sup- receivearelativelysmall share ofprogram benefits. ported comedor program and the AID-supported However,in other areas, 44 percentof the benefi- PRODIAprogram. Combined they targeted over 1 ciariesof all food programscombined are fromthe millionurban beneficiaries in 1991.Both programs non-poor. employthe same vehicle-the comedor,commu- nity kitchenswhich self-selected groups of women 4.40 Third, actual coverageappears to be sig- in poor urban neighborhoodsestablish to provide nificantlylower than programminglevels. For ex- low-costlunches to their membersor the general ample,VasodeLeche program documents indicate public. The WFP comedor program operates in thatittargetedl.2millionbeneficiariesinLimaand Lima and Callao through Caritas. Imported food Callao alone. At the same unit costs, the $42.5 commoditiesaremonetizedtoprovidefundsforthe million national budget was sufficient to reach purchase of local foods which are distributedto about 4.5 million. Yet, accordingto the survey comedores.ThePRODIAprogramoperatesinLima- results,only about 700,000beneficiaries were re- Callao, Cajamarca, Chimbote, Piura, Puno and ceivingmilk in October1991, and not all of them Trujillo. Its nutritional objective is to provide from the Vaso de Leche program. Of these, 88 comedoruserswith40percentoftheirdailycalorie percentwere in Lima.The actual coverageof all and 50 percentof their dailyprotein requirements. food assistancethrough women's groups (club de

4.37 The two programswhich target food assis- tancetotheruralpooraretheAID-supportedARCA 4. This assmes nooverlap of beneficiariesamong programs. It is possiblethat some people patticipate in severalprograms. in and the WFP-supportedDAIREZOD program. which case the oveall peentage who participatewould be ARCA, or the Food-AssistedRural Agroforestry lower.

46 madres,cocinapopular, comedorpopularandoga the governmentmay later be expectedto meet. comun)was about 500,000,or approximatelyhalf the number that the two urban programshave as 4.44 Food assistancedoes not reach important their combinedtarget groups. groupsamong the poor-the rnualpoor or the chroni- callyill such as tuberculosissufferers. Even among 4.41 Fourth,the municipality'srole in provid- groups reached,the lack of complementaryhealth ing food assistanceis surpisingly large. In Lima, care and nutritioneducation measures means that nearly 17 percentof the extremelypoor and nearly impacton nutritionalstats is likelyto be negligible. 8 percentof the poor receivefood assistancefrom the municipality.In both cases,these percentages 4.45 Thelack of establishedstandardsandprac- are higher than for any other sourceof food assis- tices for program implementationposes several tanceexcept for VasodeLeche. Municipalitiesare problems.First, differentagencies fulfill the same relativelyless effectivethanothersourcesatdirect- functions using different approaches (e.g., ing food assistanceexclusively to the poor. While PRONAAand Caritasboth handle food logistics). only 10 percentof all food assistancebeneficiaries Second,programsprovidefoodcommoditieswhich in all areas are among the non-poor, nearly 25 arenot cost-effective.Given pronounced variations percent of those receivingfood assistnce from in the costs of nutrients,Peru's food progmming municipalitiesare not poor. is not orientedto maximizethe nutrient value of each dollar spent. Third, programswith the same 4.42 Finally, programsare not necessarilyef- taget groups and objectivesvary in the levelsand fectiveat reachingtheir intendedtarget groups. An durationof the assistancethey provide. Fourth, unit examinationof Vaso de Leche beneficiariessug- costs of programs vary substantially, as shown gests that the progmmis less successfil at reaching earlierin Table 4.7. Finally,no mechanismensures its intendedtarget groups than at reachingothers. thatall food assistanceprograms are monitoredand While it aims to reach children under six and evaluatedsystematically. women, the largest proportionof its beneficiaries are schoolaged children. Fifty percent of beneficia- Major Recommendations riesare betweenthe agesof6 and 14,while only one in four is underage 6. Nearly20 percentof benefi- 4.46 ESTABLISH THE POLICY FRAME- ciaries are not poor. WORK.TheGovernmentshouldfocusitsattention on areas which individualagencies cannot under- Major Issues in Food Assistace take-setting priorities,identifying viable strate- gies and evaluating approachesand impact. The 4.43 The major issuesin food assistanceare the Governmentshould spearhead a collaborativeef- lack of an overallpolicy framework,the failureto fortwith donors and non-governmentalagencies to reach importanttarget groupsamong the poor, and devise a strategic plan for food assistance. This the lack of establishedstandards and practicesfor plan would spell out the objectives,target groups, program implementation.Peru does not have an institutionalarrangements and total resource re- overall policy frameworkor strategicplan which quirements.The plan wouldalso set standardsfor would set forth the objectivesof food assistance, levels and duration of assistance,entry and exit defme institutionalarrangements, assign priorities criteria, unit costs and implementationstandards. to targetgroups and identifyappropriate strategies. This plan would ensure that the food assistance In its absence,practice is largelydetermined by the strategiesof all agencies contribute to the same individual judgements and actions of particular nationalobjectives (see paragraph3.33). agencies.This leads to the use of substantialdevel- opmentassistance and Treasuryresources without 4.47 DEFINETARGET GROUPS. Special ef- central guidance,clear objectives,or measurable fortsare requied to reachhigh priority target groups impact.This is inappropriate because decision mak- not well served by food assistanceat present.The ers lack incentivesto be efficient,and becauseeven failureof food assistanceto reach the rural poor to non-governmentalprogams createdemands which date suggests that other mechanismsshould be

47 explored. Food coupons have been successfulin * timely distribution; other countries. Experience shows that success * efficientdelivery and logisticssystems; and with couponsdepends on severalkey prerequisites: * strong participationby beneficiaries. (a) infrastructureto identifybeneficiaries and dis- tribute coupons;(b) food marketingand banking 4.50 In Peru, the areaswhere standards are most networksto acceptthem; and (c) adequatesecurity necessaryare: (a) the institutionalarrangements for measures to protect against fraud. Under such a foodlogistics; (b)the selectionof commodities;(c) system,geographic targeting can be used to reach unit costs;(d) levels andduration of assistance;and poor areas. The Governmentcould encouragethe (e) monitoringand evaluation.Logistics are crucial testingof this or othermechanisms for reachingthe for effective food assistance programming,and rural poor by providing funding to agencies to dependon budgetingadequate resources and em- experimentwith new approaches. ployingefficient operating mechanisms. Food lo. gtiscs in Perucost between 7 and22 percentof food 4.48 For the urbanpoor in general,who are less value, or between $18 and $55 per metric ton of needythan otherson the basis of nutritionalrisk or food. In the past, ONAA sufferedfrom a lack of food security problems, the Governmentshould establishedperformanceandcoststandards,aweak encourage agencies to explore other potentialy informationsystem and dispersionof efforts out- more cost-effectiveforms of assistance. For ex- side its centralrole of managingfood logistics.The ample, ways could be found to facilitatethe pur- government'sdecision to maintaina state agency chase of food and equipmentby comedorgroups. such as PRONAAto managefood assistancelogis- Comedoresconstituteanimportantnetworkamong tics should be based on a determination that the urban poor and are organized and skilled at PRONAAismore efficient and less expensivethan mobilizing support Their abity to respond to privatesector alternatives.It should have to com- waves of economiccrises has been central to sus- pete with the private sector. Its funding should taming the urbanpoor in social emergencies.Any come largely via service contracts based on the attemptto restructa urbanfood assistanceshould volumeof food andthe mix of ports andbeneficiary bring comedorleaders into the processof identify- destinations.Acarefulstudyoflogisticscostswould ing and designingalternatives. enablethe Governmentto plan adequateresources to cover them. 4.49 ESTABLISH IMPLEMENTATION STANDARDS.The implementationof food assis- 4.51 Commodityselection should use the least tance shouldbe standardizedto a greater extent. expensivesourcesof nutrients acceptable to benefi- This would complementthe introductionof a clear ciaries.Milk is often used asa sourceof protein(e.g. policy framework.International experience points VasodeLeche). Yet with milkat $1,300/metricton to key factorswhich can improvethe implementa- (1991 WFPprice for dried skim milk),it would be tion of food assistance.These include: possibleto obtainfourtimes as muchprotein forthe same price with wheat or soy-fortifiedbulgur, or * targetingto high risk groups; twice as much with lentils or rolled oats. While * the carefulselection of rations; $1.00willpurchase 3,00caloriesofwheatorcor, * anticipatingand planningfor dilutionand the same dollar spent on rice purchasesless than substitutioneffects; half this amount. Efforts to optimizeration mixes * strong nutritioneducation; shouldalso consideradditional factors such as the * growthmonitoring/nutritionalstatusmoni- bulk and digestibilityof different commodities. toring; Staplefoods usedin assistanceprograms should be * stronglinks with otherprimary health care fortifiedwith iron. services; e anticipatingand planningfor opportunity 4.52 A set of unit cost guidelines should be costs of participation; developed for food assistance to determine how * clear andapplicable entry andexit criteria; much it shouldcost to reach a specifictarget group * considerationof seasonalfood consump- in a given locationwith a givenlevel of assistance. tion patterns; When transportationexpenses push the cost of

48 importedfood above its local cost and when local 4.56 The Vaso de Leche program needs to in- suppliesareadequate,importedcommoditiesshould crease its rationlevel and strengthenthe qualityof be sold to purchaselocal food. managementand controls in orderto reach a higher proportionof its targetgroup and have a significant 4.53 Oncefood assistanceobjectives and target impacton nutrition.This can be donethrough more groupsbecome more explicit,it shouldbe possible intensivemanagement training and nutritionedu- for agencies to develop standards for levels and cation to local program managers.It should also duation of assistance. Beneficiaryrations should seek ways to provide sources of nutrients which be basedon an understandingofthe nutrientgaps of cost less than milk. participants. Programs with the same objectives and target groupsshould provide similarlevels of 4.57 TheLaEscuelaDefiendeLaVida Program assistance.For example,to improvethe nutritional shouldtarget children in grades1 -4 and placefirst status of malnourishedpre-school children, chil- priority on rural sierra provinceswhere the inci- drenwho areat the samenutritional risk shod be pro- dence of povertyis high.Because it was started as videdwith similar levels of supplementaryfood for an emergencyprogram, it has relied heavily on the periodnecessary to reversethis risk. The duration resourcesalready within the educationsector. Fu- of assistancewill dependon each child's progress. ture budgetsshould incorporategreater resources for programadministration so that schoolfeeding 4.54 The Governmentshould assume the pri- does not divert resources away from educating mary responsibilityfor monitoring the perfor- children.Future implementationshould also seek mance and evaluating the Impact of food assis- to overcomesome of the program's early weak- tanceprograms.TheMinistryofHealthcouldevalu- nesses, including an excessive concentrationin ate differentproject strategies on a samplebasis to Lima and a lack of continuityin feeding. test their cost-effectiveness.This would strengthen the Government'sdialogue with donors and enable Conclusion it to allocate resourcesto the most cost-effective strategies.The Ministrycould also conductimpact 4.58 Thischapter has outlinedsome of the com- evaluationson differentprograms to determineif ponentsof a povertyalleviation strategy for Peru. the strategiesare effectiveat achievingthe desired The first plank of the program is the promotionof objectivesin nutrition. broad-basedgrowththrough implementation of the stabilizationand adjustmentprogram. The second ogram SpecificIssues and Recommendations plank comprisessocial service programsfocusing on activitiesmost critical to the poor and on the 4.55 MCH programs could be improved by rural sierra,where the incidenceof extremepoverty introducingmechanisms for bettertargeting, stron- is highest.These programsinclude improvements ger linkage to other services and better impact in the quality of primary and pre-primaryeduca- evaluation.Food assistance to the MCHpopulation tion,promotion of higherattendance at the kinder- shouldbe expanded.In ruralareas, the use of food garten level by childrenfrom poor householdsand couponsshould be exploredto protect vulnerable literacyprogramsfornon-Spanishspeakingwomen mothers and children.The PANFARprogram is in rural areas. In health,the emphasisshould be on appropriateforareaswith fairly good healthservice primaryhealth care for mothersand children.Spe- coverage,good outreachto identify beneficiaries cial effortsare neededto increasevaccination cov- and adequate infrastructureto allow warehousing erage, family planningservices and the supplyof of food commoditiesat health centers. An advan- basic drugs in the sierra region.Nutrition interven- tage of this approachis that it strengthenscontact tions need to be more closely linked to health between the health center and the maternal and improvementsandanationalnutritionpolicyshould child population.This initial contact may be used to be defined and implemented.Recommendations provideother importantservices such as pre-natal were made on the design of two social compensa- care,immunizations,treatmentofdiarrheaandacute tion programs- public employmentand food assis- respiratoryillnesses and nutritionand health pro- tance.Thenext step after defininga povertyallevia- motionactivities. tion strategyis to executeit. The followingchapter 49 disces some of the issuesthe Governmentis pwvt in Peu and how for the Governmenthas likelyto fce in implmentig pgms to reduce advancedinid ftygapovertyaUeviaonsutegy.

.

50 5

Implementinga PovertyAlleviaton Strategy

ntroduction programto lessenthe severty of povertymight focus on bringing the extremely poor up to some 5.1 Earlier sections have set forth key issues moreacceptable level of welfare.Extreme povery and recommendationsin the social sectorsas they is prevalentin Pem and the extremelypoor are the relate to poverty alleviation.This section exploes least likely to be liftedout of povertyin the short- two important issues in the implementationof a run by opporatities arisingfrom adjustmet. 11i poverty alleviation strategy. First, we examine suggests that In Per poverty alleviation hter. what available infoTmationreveals about how to ventions should be rgeted prmarly to the ex- target measures to the pooresl. Second, we con- tremely poor In order to bring them up to a siderthree facts in the instiutionalarrangements more acceptable levd of welfre for povety alleviation:(a) the role of a social fund; (b) the regional compensationfund; and (c) the 5.3 Whiletargeting m anismsare often dis- role for NGOs.Tie constaint posed by the lack of cussed in terms of transfersto the poor to increase securty in some pats of the countryis also briefly expenditurelevels, anotherapproach is to target a discussed The chapter concludeswith an exami- safety net of services to the poor to protect them nation of the Government's poverty alleviation from the most deleteriouseffects of low expendi- strategy. ture levels. This approachwould seek reductions in infant and child motality, malnutritionand in- Targetng fection; and increasesin literacyand schoolatten- dance to inree the welfare of the poor indepen- 5.2 The best targeting achieves the greatest dent of expenditurelevels. impact at the lowest cost. Targeting mechanisms depend on objectives,the interplaybetween effi- 5.4 Targetingseeks to distribute transfers or ciency concerns and political factors, and the some other benefits only to those identifiedas the strengthof institutionalcapacity. Under the broad beneficiarygroup in the interestof efficiency.The heading of "poverty alleviation" objectives, a administrativemechanism must not cost so much muber of optionsexis, each with implicationsfor to operatethat it absorbsthe savngs from exclud- targeting A programto reducethe nationalaggre- ing the non-needybeneficiaries. If targetingcosts gate poverty gap might focus on the regions with more than it saves,universal distribution is a bette the highest incidence of extreme poverty.A sec- option. Shor-term efficiencygains fiom tugeting ond program might aim to reduce the number must also be weighed against longer-term or percentageof people in povery by concentrat- sustinability. If targeting is so effectivethat only ing on those just below the poverty line. A third the poor have a vestedinterest in a prowam,then it 51 is unlikelyto sustainthe public support necesary latrines and the absence of household lighting. for it to survive politicalprocesses. This suggests Seventy-onepercent of thosewho use keroseneare that broadeningthe beneficiarygroup somewhat to poor (51 percent are extremelypoor). Sixty-eight includethose with a strongerpolitical voice may percent of households who use latrines are poor be desirable.Somewhat "inefficient" targeting of (46 percent are extremely poor). All households benefits may be the cost of ensuring long term withouta sourceof lightare poor; but they repre- sustainability. sent only one percent of the poor. Within regions (Lima,urban coast, urbansierra, rural sierra), few 5.5 Targeting uses indicatorsclosely associ- characteristicsserve as usefultargeting indscators. ated with need, direct testing to assess need, or The successful targeting of food assistance pro- programdesign which makes the opportunitycosts gramsin Limasuggests that a combinationof self- of participationso high as to exclude those with targetingand geographictargeting can be used ef- betteralternatives (self-targeting). Indicator target- fectively. ing and direct testingboth dependon selectingin- dicatorswhich potential beneficiaries cannot easily InsttutionalArrangements manipulate.Self-targeting is often achievedby re- quiringbeneficiaries to work to receivebenefits or 5.9 THE ROLE OF A SOCIAL INVEST- by providingpublic goods of lower qualitywhich MENT FUND. To determinethe role of a social the non-poorwould prefer to obtain privatelyat a investmentfund in carrying out a poverty allevia- higherquality. tion strategy, two questions must be addressed. one, is a social fund appropriate given country 5.6 In Peru,geography is the indicator with conditions;and two, what are the main prerequi- the greatest value as a targeting mechanism. sitesfor its successfuloperation?' A social invest- Even with no additionalinformation, a household ment fund is a mechanism now being experi- in the rural sier.; has a 68 percent likelihoodof mentedwith in many countries.It is a financialin- being poor, and a 47 percent likelihoodof being termediarywith a streamlinedbureaucratic struc- extremely poor. By targeting benefits to rural si- ture which mobilizes and channels resourcesto a erra households,nearly one in everytwo extremely wide rangeof implementinggroups who carry out poor householdscan be reached. small projects proposed by communities. The projectsrange from labor-intensiveconstruction or 5.7 Anotherindicator which is stronglyasso- rehabi'itationworks which generate employment ciated with poverty is language.Throughout the for the poor to activitieswhich improve social ser- country, 70 percent of Aymara speakers are ex- vices for the poor. Experienceshows that a social tremely poor; among Quechua speakers,32 per- investmentfund is most appropriatewhere exist- cent are extremelypoor. Targetingall Aymaraand ing infrastructureis in need of rehabilitationand Ouechua speakers will capture 41 percent of the where existinginstitutions are weak,as in Peru. In extremelypoor. Since most non-Spanishspeakers addition,a socialfund is a suitablemechanism when: live in the sierra, geographictargeting would also reach them. Non-Spanishspeaking households in * a countryfaces an emergencyand needs other regionsare not disproportionatelypoor. quick and tangibleresults;

5.8 Some characteristcs closely associated * the emergencyis of a temporarynature, with poverty suggest the types of assistance e.g., policiesare in place that will ensure which would manly benefit the poor. Nearly all growthand redistributionin the medium (82 percent ) households which draw their water from wells are poor (26 percent are extremely 1. The foUowingdiscussion is drawn from 'Easing the Poor poor). Thus, a program to improve watersupply through EconominCrisisand Adjustment-The Story ofBolivia's for well-users would benefit primarilypoor house- Emergency Social Fund' by Jorgenson, Grosh and Schacter. Latin Americaand the CaribbeanTechnical Department, Regional holds.Other housingcharacteristics closely associ- Studies Program, Report No. 3. and the Povety Reduction ated with poverty includethe use of keroseneand Handbook, World Bank 1992

52 to long term; activitiesto benefitthe poor. It is intendedto be a * the countryneeds a social safety net be- short-te*rmresponse to worseningpoverty until the cause social servicesare weak; capacityof the Ministriesof Healthand Education * the agency can be allowedto work with as is sufficientlystrengthened. FONCODES finances broada range of institutionsas possible; small projects (thus primarily rehabilitation/up- * overallpolicies favor privatesector devel- gradingof facilities,not new construction)and re- opment;and gional governments have an advisory role in its a willingnessexists to acceptmarket based operation,including project selection. solutions,i.e. when there is not a "planner knowsbest" attitude. 5.13 Progressin makingFONCODES fully op- erational was slow and initiallythere was confu- 5.10 Peru meets these conditionsand thus it sion over whether the agency was going to be a would appear that a social fund could play a useful "demand-drive"or a "top-down"operation which and importantrole in the poverty alleviationstrat- adverselyaffected the qualityof someprojects. - ow- egy. Yet, social funds involvetradeoffs. Since so- ever,in mid-1992perfamance began to improve. cial fundsare demand-driven,projects risk not be- ing integratedwith investmentpriorities. In Peru, 5.14 REGIONALCOMPENSATION FUND. this issue may not be critical.Investment spending The RegionalCompensation Fund (FCR) has the by centralministries is almostnonexistent because potential to contribute to a poverty alleviation of the fiscal austerityprogram and the decentrali- strategy. This fund was created by the January zation of social service provision.A social fund is 1990 Law of RegionalGovernment Financing to unlikely to reach key vulnerablegroups (for ex- finance regionalcapital spending.Until now, the ample, mothers and children), although they are FCR has been financedon the basis of a formula - likely to benefit from the rehabilitationof infra- 8 percentof Treasuryresources and 25 percentof structure.This is inherentin a relianceon demand- the revenue from the generalsales tax. The distri- driven and market mechanisms.Thus, a social butionof these fundsto the regionshas been based fund would need to be complementedby interven- on a formula as well. Half of the funds are allo- tions in education,health and nutrition. cated equally to all regions and the other half in accordance with population size. No attempt is 5.11 Experience with social funds in other made to channelmore resources to poorerareas. A countriesindicates that effectivenessdepends pri- changein this formulato enablethe fund to play a manly on the capacityto select the right projects role in povertyreduction would only be feasibleif which in turn dependson the quality of staff and tax revenuesincrease and after fiscal stabilization their independencefrom political pressures. Be- is achieved. cause social funds are "wholesaletoperations, ad- equatesupervision a-id monitoing of sub-projects 5.15 THE ROLE FOR NGOs. The magnitude is essential. andseriousness of the povertyfacing Peru call for the activeparticipation of all resources,including NGOs. 5.12 THE NATIONALFUND FOR SOCIAL In 1990,NGOs mobilized between $170 millionand COMPENSATIONAND DEVELOPMENT.In $200 millionin externalfunding. The UnitedStates August1991 the PemvianGovermnent established channelled about $100 million through NGOs a social investment fund, the NationalFund for Qarly in the foamof foodass-stance); Holland - $20 Social Compensation and Development million;Cnad 2 , Germanyand Italy - $10 million (FONCODES).It is a temporarypublic institution each and Switzerland- $4 million.NGOs have a with technical, administrative and financial au- comparativeadvantage in mobilizingextenal re- tonomywhich financesprojects proposed by gov- sourcesbecase of theirfexibiliy ani theirability to ernment agencies, NGOs and community-based respondto donorpriorities. organizations.The focus is on employmentgen- eration, but FONCODESalso funds projects in bealth, food assistance,basic educationand other 2 Cada esimatei fo 1991.

53 5.16 NGOs are also able to reach underserved the Vaso de Leche. An estimated 23 percent of populationsgroups effectively.NOOs often pro- NGOswork in healthand nutrition.Some operate vide social servicesto populationswhich have in- communityhealth posts in responseto community adequateaccess to state servicesdue to terrorism demand.NMOs in healthand nutritionare flexible or emergencysituations. In 1990 NMOspartici- and able to experimentwith a varietyof modelsfor pated activelyin the NationalDrought Thnergency service provision, cost recovery and information Program,the Socal EmergencyProgram and the management.NOO activity in education is con- school feeding initiative,La EscuelaDefiende La centratedin non-formaleducation. Vida. NGO participationin the drought program demonstratedtheir abilityto operateeven in secu- 5.20 WhileNGOs operate throughout tlh coun- rity emergencyzones. The SocialEmergency Pro- try, they are concentrated in Lima and lack a gram demonstratedtheir abilityto mobilizeeffots sttong presence in the selva. DESCO(Center for rapidlybehind an emergencyinitiative. NGO par- the Study and Promotionof Development)esti- ticipationin La EscuelaDefiende La Vidaenabled mates that nearly60 percentof NGOsare based in this programto expandrapidly and providerations Lima,another 12 percentin Cusco and Puno and 5 at low cost. percentin Arequipa.Since Uma NGOsoften oper- ate in other areas of the country, concentrationin 5.17 CARITAS,the social service arm of the termsof serviceareas is not as pronounced.Nearly Catholic Church, has a particularlystrong com- half of NGOsoperate only in urban areas, another parativeadvantage in reachingunderserved popu- fourth operate only in rural areas and one-fourth lationsbecause it is the only institutionother than operatein both areas. the state which can draw upon a nationwidenet- work. This presence extends to areas where gov- 5.21 NGOs are constrainedby vulnerabilityto ernment activity is minimaland other NGOs are donor influence and weakness in formulating not active,such as isolatedregions of the selvaand policyproposals. As NGOsmobilize resources in- sierra. CARTfAS'participation has provencentral dividuallyor as small consortia,the sector lacks a to manylarge-e emergencyrehabilitadon efft mechanismto ensurethat resourcesare directedto the highest prioritytarget groups or activities.In- 5.18 NGOs are a strong constituencyfor pov- volvementat the micro levelleads to difficultiesin erty alleviationgoals and programs.In manycoun- seeing the broad picture and in formulatingfar ties, povertyalleviation initiatives have failedbe- reachingpolicy proposals. cause beneficiarieslacked a strongpolitical voice. In the last five years, NGOs in Peru have devel- 5.22 Collaboration between the Government opedcoordinating bodies which strengthen the ad- and NOOs is constrainedby real and perceived vocacy function by channellingcommunications barries. The dispersionof many Governmentpro- and providing fora for discussion.The National grams over a broad range of state institutions Associationof Centersregisters 22 such NOOnet- makes NGO entry into the process difficult. Per- works.The advocacyrole of NGOsis strengthened sonalisticmanagement of some governmentpro- by their ability to maintaincommunications with gramsmakes collaborationwith NGOsdependent the poor.While NOOsdo not "epesent" the poor, on the will of individualmanagers rather than on they often serve as a channel of communications institutionalmchanisms. NGOsare not organized enablingthe poor to expresstheir concerns. as a sector and they lack the mechanismsto plan and establish collectivepriorities for spending or 5.19 NOOs are numerousand operatein many impact evaluations.Their actions are frequently sectors and areas of the country. At least 250 not motivatedby long term analysisand planning. NGOs are active in healthand nutrition,education They differ widely in their institutional,manage- and food assistance.They also assistin the institu- ment and administrative capabilities, and their tion,aldevelopment of local governments.NGO ability auanalyze problemsand propose policies. collaboration with municipal governments has They lacka clear visionof their operationalcapacity been instrumentalin intmducingprogms such as both individuallyand collectively.They are not

54 operationalin all areas of the country with severe many of them innocent victims caught in the povertyproblems. crossfir of Latin America's most violent and bloodyguerrilla movement. Community orniza- 5.23 Perceivedbarriers are moreimporant than tions and the workns of local and centralgoven- actual anesin constainig Govemment-NGOcol- ments have beenseverely affected. In incrasingly laboration.Govemment staff do not see NOOs as large areas, the central Governmentis unable to partnersin the identificationof national priorities providebasic servicesof law and order, muchless in povertyalleviation nor as subcontactos to ex- providesocial services. pand the Government's implementationcapacity in achievingthese goals. They perceiveNGOs as Poew AllyIaton Progm of theGovenient competitorsfor a fixed levelof resourcesand con- trol. They object to competingwith agencies not 5.26 POVERTY ALLEVIATION SJTRAT- bound by the constraints they face. Likewise, EGY.The Govemmentprepared a povertyallevia- NOOsmistrust Govenmden motivation for greater tion stategy in 1992.It is generallyconsistnt with collaboration.They are concemedabout the state the findingsand recommendationsof the analysis tuning over inmpossibletasks to NMOsin order to conined in this report.The stategy sets clear and avoid accountabilityfor failure.They are reluctant appropriatepriorities for a povertyalleviation pro- to share informationwith Government agencies gram in Peru, thereby providinga frameworkfor becase they fear the Govement wantsto control poverty reductionmeasures, both for the Govern- their activities. ment and donors. Toe challenge and test of the Government'sresolve will be the extent to which 5.24 WbereverNGOs have a comparativead- both the designand executionof specificprojects vantage,the Governmentshould be willingto mo- are consistentwith the principles set out in the bilize and channelresources through them. Several povertyalleviation strategy. steps wouldimprove NGO participationin Peru's povertyalleviation strategy. The govenmmentand 5.27 The strategy recognizesthat the allevia- NG0s could work to determine the comprative ton of povety i Peru demandsessentially an eco- advantagesof NGOsby regions,sectors and popu- nomic strategy The recovery of the economy, lation groups. With NOG participation,the gov- based on a progm of stabilizationand liberaliza- ernment coud define the areas for autonomous tion measums,is crucialfor the attack on poverty. NGO activity and determine which activitiesre- The objective set forth is to improve the living quire coordination.Ihe goverment and NOOs conditionsofthe extemely poor throughprogms could also determine jointy the magnitude of pro- in the areas of food assistance, health, education grams and projects which NOOs are capable of and employmentgeneration projects. underaking. The govermuent should establish dear and efficientmechnisms for NGO colabo- 5.28 The taret group for the Government's ration. NGOs shouldavoid terial and sectoral povertyalleviation efforts are the extremelypoor. isolation which leads to duplication of efforts. The goal is to reduce the incidence of extreme EmergingNGO consortiaprovide one mechanism povertyfom the ament level of about 20 percent whichshould be suppoed to fostergreater coordi- to 15 pere by 1995.The maintargeting mecha- nation,dialogue and joint planningwith the state. nismwould be geography.More detailed goals for 1995 are also set in the areas of health,education Securiy Constaint and employmentgeneration. In order to address the problemsof extremepoverty in the countrythe 5.25 The task of improvingsocial services in strategyoutlines 5 morespecific objectives many rural areas is made much more difficultby the violentactivities of the terrorist orgization, * Generateemployment and incomein rural SenderoLuminoso (Shining Path). Since Sendero and marginalurban areas throughprojects launchedits militarycampaign in May 1980,the focusingon socialinfiasucurendsupport resulting conflict has claimed ovr 24,000 lives, to production;

55 * Providefood assistnce in order to ensure 5.31 In the area of heal, the main lines of ac- that the basic requirementsof the most tion includerestoring a system for monitoringthe vulnerablegroups are met; health status of children; promoting the partic- ipation of communitiesin preventive activities; * Improvethe coverageand qualityof pri- improvingpre- and post-natalcare and promoting mary and preventivehealth services with breastfeeding,and improvingthe qualityand effi- an emphsis on rural and margia uba ciency of services through trminingprograms for areas, with a view to contrbuting to the healthpesonel. Coverageof primaryand preven- reductionof infant morbidity-mortality tive serviceswould be increasedthrough an expan- and the protectionof the healthof poor sion of the network of health services in women; undetseved zones. Relatedto that, existinginfra- structure (hospitals, health posts and health cen- * Improvethe efficiencyand qualityof edu- ters) would be rehabilitatedwith an emphasison cation especiallyat the primary level in healthposs and centers.There would be coordina- rural and urbanmarginal areas; tion with the agenciesresponsible for sanitationin order to provideservices to poor areas. Attention * Meet the basic needs of school-agechil- to the problemswhich cholera has generatedalso dren throughschool feeding pwrgams. would be strengthened. distributionof texts and teachingmateral and the needsof adults throughliteracy 5.32 In the area of nutrion, the principallines programs;and of actioninclude rationalizing and improvingfood assistance programsso that they reach the target - Providefinancial support through population;increasing the nutritionalstandards of FONCODESto activitiesengaged in by food assistance;and implementingeducation pro- the poor. gramsand a systemto monitornutrition levels.

5.29 The Governmentenvisages a major role 533 For education, the strategy proposes: to for FONCODESin the implementationof this reduce the illiteracyrate, especially in rural area stregy. There is explicitrecognition that the strat- and amongwomen; to lower repetitionand drop- egy implies an increase in the volume and effi- out rates and to expandpre-schools in poor areas; ciencyof social spendingby the Government,the to improvequality through teacher training and up- formerto be achievedparly througha reallocation grading infrastructureand teaching methods;to of spendingtoward social investments&Social ex- integrate the community and the pnvate sector, penditures also would be oriented more to the and to orient the system more to the needs of the poor. In addtion to an effectiveFONCODES, this labor market. Educationwould also play a more would requires_trgthening the servicesprovided activerole in supportng the resolutionof other so- by the Ministries of Education and Health and cial problemssuch as malnutritionand poverty. other agenciesto the poor in Pem. Concldusion 5.30 The poverty alleviation strategy sets out principal lines of action in three areas: employ- 5.34 Thereare two sourcesof financingfor the ment generation;health, food assistanceand nutri- poverty alleviation strategy. The first is govem- tion; and education. In the area of employmet ment resourcesand the secondis donor resources. geeration, the focus is on providingincome eam- Public resourceswill ccntinue to be constrained ing opportunitiesand/or food-for-workprogams over the nextseveral yeas becauseof thestabiliza- for the poor in rural and marginal urban areas tion progam and the time it will take to restoretax thrwoughprojects in social and productive infra- revenues.However, in order to attract sufficient struct Smallenterprises would also be assisted donorresources the Govemmentmust demonstrate by deregulation,as well as taining and credit. clearlyits own commitmentto povertyalleviation.

56 In 1990 it is esftimatedthat totalgovenunent spend- larly in the shot-run. Measuresare also needed to ing on health and educationreached only 2.2 per- protect vulnerablegroups and to ensure that the cent of GDP, well below the average for Latin poor are able to take advantageof the greaterop- America. It is recommended that over the next portunitiesin the reformed economy.In order to three years the Governmenttake steps to improve addressthese latter needs, this reporthas examined the cost-effectivenessof public spendingfor social the profile of the poor in Peru and recommended servicesand review the possibilityof reallocating priority policies and progrms targeted to them resources from the tertiary level to the primary The Governmenthas made a start in this area. Its level.As tax reformand the measuresto strengthen poverty alleviationstrategy sets clear and appro- tax admiis tonaise revenue generation, it is priatepriorities and goals for povertyreduction ef- also recommendedthat the Governmentendeavor forts within a frameworkof a contindiationof eco- to increase the resources allocated to health and nomic policies that would maintain macroeco- education. Spending on education and health of nomicstability and promotegrowt The challenge aboutfour percentof GDP is the averagefor Latin and test of the Government's resolve will be to Americaand about the levelthat was spent in Peru what extent current and future policies and pro- duing the first half of the 1970's. grams are governedby that strategy. The recent improvementin FONCODESis an encouraging 535 The Governmentof Peru has taken an un- sign. But in order to ensurethat the poor reap the portantstep to reduce povertyin the country.Over benefits of adjustment, the Govenment should the mediumto long-runwhat is neededto aUleviate strengthen the social services most cntical for the high levels of poverty is broad-basedgrowth. them andprovide more effective safety nets for the That is the objectiveof the stabilizationand adjust- mostvulnerable in the country. ment prograL However,it is not enough,particu-

57 ibliography

AID. "Analisisdel SectorSalud en el Peru. Lima, Latna en losAfos Ochnta. May 1990. ManoD1991. De Wit, Ton. "ONGs en el Peru y la Planificacion Amaty laon, Cauos and Curonisy,Dante. La,4i.- Estrategica".unpublished wordng paper ptpred mmacion end Perm Oentrode Investgcion for the WorldBank. May 1991. Univetsidaddel Pacifico. Lma, Peru. May 1990ed. FAO.Food Balane Sheets.1979-81Average 1984. AramburuC. Petrua M y Tgliafro G.LosNiWles de FAO.Food Bance Sheets. 1984-86Average 1991. Vi'daen las Provias del Penu INP/ANDEP, FAO. Produtn YearbookVolume 43. 1989. Llina 1989. Fernandez,H. "Eficienca lntema del Sistema Educ- alab, Eduardo. Evaluaciondel ProyectoPLoto Bono ativo.1 In Fernandez, H. and Rosales J. Educa- Mateno hntil en Honduraseunpubished wokdng cionjUnaMrada HaciaAdenro.IPP, Lima 1990. paperprpared for the WorldBank. September 1991. GarciaH. "El SectorSalud en el Peru: Resumende los Berg.Alan. "Guidelines for Waokin Nutrion". PHN Resultadosde la MisioneFebreto 1991. TecaluicalNote 8612 WorldBank, March 1986. Glewwe, P. The Distribuion of Welfare in Peru in B3sley,lgnty and Kanbur,Ravi. ThePrinples of 198586. LSMS WorkingPaper No. 42. The World Targeing.PRE Woking Paper385. World Banlk Bank, 1988. March 1990. *3lewwe,Paul and de Tray, Denois. he Poor in Latin Ca, 1L,Andia, Y. y Leon, 0. Evauacion de los Americaduing Adjustment-A CaseStudy ofPen*. astemasde abasecmiento de agna, acantarilad LSMSWorking Paper, No. 56, WorldBank 1989. diposiclon de aguas residualesy saneamino de Glewwe,Paul and Halt Gilet Poverty andinequali as zons urbanomarginales en el During Unona=dxAdiusnent: 7seCase of Peru marcode la epidema delcolera.CEPISIOPS, Uma, 1985 to 1990. LSMSWorking Paper, No.86, World Abuil1991. Bank1991. Cbiroque,Sigfredo. -Evolucion del PoderAdqusitivo World Bank,Perw Economic and SectorReforms to delDocente Peruano: 1980-1991-" (Unpublished) SustainStabiliation ad Lay the Founda for De- nsituto de PedagogiaPopular, 1991. wvopnent White CoverEconomic Report Cobun, J, Bitran R., Griffin, C Levine, R. and Grosh, Margaret,Social Spending in Latin America: RobeR,R Aswsesn ofHealthSytems, Ihe Story of the 1980's. DiscussionPaper No. 106. Financig and Policy OpwonsinArequa Regin, TheWorld Bank, 1991. Pera HFS TechnicalReport No. 3 AID, February Gro, Marget Unpublishednotes on a taxonomyof 1991. targetingmechanisms and theircosts and outcomes. Cocwhane,Susan H. The Effes of Educationon Fer- February1992. tility and Mortality."Education and Training Gra, Mare "The JamaicaFood StampsProgra". DepartmentDiscussion Paper No. 26. The World WorldBank LATHR. June 1990. Bank, PHR1E.Washington, D. C., 1986. Jorgensen,Steen, Grosh, Marpgret and Schacter,Mark" ComisWnEconomicaParaAmericLatinoyel Caribe Easing the Poor through EconomicCrisis and Ad- (CEPAL), Magniatd de la Pobreza en America justment- TheStory of Bolivia's EmergencySocial

59 Fund". Latih American and Canibean Teecnical tific PublicationNo. 524,Washington, 1990. Deparment RegionalStudies Proam, Repot no.3. Pavone,and Petters M., "EconomicImpact of Cholera INEI. DireocionTecnica de IndicadoresEconomicos. in Peruin 1991",WHO/PAHO Study Perm Compendlo Estadistco 1989-990. UAms I=ta M. "Efectos de la crisis economica en los Septiembre1990. Servicios de Salud Matemo e Infantil y en sU INEI. Direccion Nacional de Estadisticas Basicas. poblacion. Peru 1985-1990." OPS, Lima, Mayo "IndicadoresSociales. Boletin No. 2 Uma, 1991. Septiembre1990. PNUDIJNP.La Pobreza en el Peru: Diagnostico y Instituto CuantoAjuste y Economia Familiar, 1985- PropuestasdePoiticas. Bogota,1990. 1990,Lima, 1991 Psacharopoulos,G."The Contributionof Educationto Lopez de Castilla, Marta, Oriheula,Ana Maria, and EconomicGrowth. International Comparisons." In Chiroque,Sigfredo. "Maestros Rurales." IPP, Lima Kendrick,J. W. (ed) InternationalComparisons of 1990. Productvityand Causesofthe SlowdownAmerican Lopez de Castilla,Marta, Trapnell, Lucy, and Rengifo, Enterpre Institute/BallingerPublishing Co. Cam- Moiso.. "Ser maestroen La Amazona."IPP, bridge. 1984. Lima 1989. INIDE. Quinche,Daniel et al. "Documentode Base:Situacion y MendozaI. "Proyectode Apoyo a is GestionLocal del Perspectivasde la FormacionMagisterial en el Peru: ProgramaSocial de Emergecia en Salud,Nutricion Pimer SeminarioNacional Sobre Fonnacion Magis- y SaneamientoAmbiental. Informe Final." OPS, terial",LUia, Per, 1991. Umr, Febrero1991. Schiefelbein,E. And Heikkinen,S. Peru:Access, Rep- Ministryof Education.Situacion de la Educacionen los etitionand Efficiencyin Prinauy Educain CentrosEducativos Unitarios del Departamentode OREALC,October, 1991 Arquipa. Senado de Is Republica. "Encuesta de Opinion: Ministerio de Salud. Haca un Sistema Nacional Violenciay Pacificacionen el Peru." Lima, Abril Reglonalzado e Integrado de Salud Lma, Peru, 1989. Enero 1991. Stelcner,M., Arriagada,A. and Moock,P. Wage Aeter- Ministerio de Salud I OPS / UNICEF / USAID. mnnantsand SchoolAttaimnentamong Men in Perm "Malnutricionpor microutrientesen el Peru" Lima, LSMS Working paper No. 38. The World Bank, Agosto 1991. 1988. Ministerio de Salud. "Boletin lformativo sobre la UNICEF,(Comite de Coordinaciondel Programs de Epidemiadel Coleraen Peru."lima, Mayo1991. CooperacionPERU-UNICEF). Plan deAccionpor Ministeriode Salud. OficinaGeneral de Planificacion. la Infanca: Proridades Sociales del Desarrollo Evauacionde laEstrategra deAtencionPrimaria de Hwuano, la ReconstruccionNacional y la Paz SaluduPeru 1988-1990.Lima, 1991. ULms,Junio 1991. Ministeriode Salud. La Epidemia del Cokra en el UNICEF.Estado Mundial de la Infancia 1991. New Peru. Lima,Junio 1991. York, USA, 1991. Ministerio de Salud del Peru. Plan Operativo 1991. UNICEF.ElAjuste SociaLPerum Hacia un Desarrollo Direccion General de Salud Integral. Direccion EsenciaLLima, 1991. Ejecutivade Crecimientoy Desarrollo. UNICEF."La Infancia, la Pobreza Y la Mujer en el Ministeriode Saluddel Peru y la OPS.Analisisde los Peru:Analisis de Situacion."Uima, 1991. Recursos Institcionales en Alimentacion y UNICEF. (Carlson, Beverly and Wardlaw,Tessa) A Nutriion del SectorSalud Peru, 1988. Globat Regional and CountryAssessment of Child Ministeriode Salud del Peru y la OPS. Fornacion de Mabarion StaffWording Paper No. 7.1990. Personalde Saluden el Peru. LUma,Peru van de Walle, Dominique,Poverty and Inequalt in August,1989. Latn America and tde CaribbeanDuring 1970's Moock, P. and Bellew, R. Vocationaland Technical and 1980's:An Overviewoftdie Evidence. Education n Penu WorkingPapers, WPS 87, The Webb, R. and FernandezBaca, G. Peru en Numeros WorldBank, 1988. 1991. E4tonialNavarete, ULma, Abril 1991. Newman,Jorgensen and Pradham,"How Did Workers WorldBank. Primary Educadon. A WorldBank Policy BenefitFrom Bolivia's Social Fund?", 7he World Paper. TheWorld Bank,Washington, DC., 1990. BankEconomicReview, May 1991,volume 5, No.2, World Bank. World DevelopmentReport 1991. The pp 367-393. Challengesof Development The World Bank/Ox- Ohno,Izumi, "Peru Overviewof FY 1991 Budgetand ford UniversityPress. Washington,D. C. 1991. Social Expendiures"World Bank, February1991. Yamada, Gustavo. "Peru's Poverty Profile with the OPS. Las Condicionesde Saluden lasAmericas. Scien- Available Data" unpublishedworking paper pre-

60 paredfor theWorld Batk. August1991. Yamada,Gustavo. "Peru's Glass of MilMProgram" us- Yamada,Goustavo. 'Peru's Feedingand Nutrition PFo- publishedworking paper prepaed for the World gramfor HighRisk Families(PANFAR)r unpub. Bank.July t991. Lishedworking paper prepared for theWorld Bank. July 1991. Yamada,Gustavo. -Peru's Poplr Kitchens"unpub. lishedworkig paperpepard for theWorld Bank July 1991.

61 PERU

POVERTY ASMNT AND SOCLALPOLICIES AND PROGRAMSFOR TH POOR

I Am= A

...... wow

-Zsi ...... M-,,s`*Me,-M-,--,:,:K*s:s: :3:S**AR.W ...... U -M

AN ......

O hi ...... *

VW,

iX

65 . l i!

S l |ii

s S .Eg

' E! ! | = | ~! ~ ~ ~ ~

E~~~~~'rl | i| *e= 4

66 Anex B

RECENTREPSEARCH ON POVERTYIN PERU

Bamo Central de 'Docmlet on Banco nl dLg ya, Mms of Health, Poverty (unprocessedworkig paper, 1990). St al. Proyecto Salud y Seguidad Socila Asse 1985186Peru Living Standards (1991). Idcludes3 dunt. Docum_n Survey. Methodologyfor caLculationof No.1is a Mapade Sad (1988). Comparison povertyline differs fom most other studies. is madeof differnt partsof the healthsystm The authorsconverted the physicalitem of as estmatedby prwvidersof sices and the food consumptionino their equivalentin resuts of the 1985/86PISS, Constucts a caloriesfor eachindividual. A regrion was map of the productionof health servicesin nmbetween those caleoiesand the regisred differe regionsof the couy. Con)uslon per capia foodexpendiu andt wasused to i that in general the healh system is not esimate the expnditures on food needed to producingadequate services. Estimates hauta flulfl the required daily calories of 2,900 for quart of the population Is unco d by an adult equivalent. Paper also inctudes a health seices. Doc_uet No. 2 is "Eectos basic needs assessment. Auth lculate the D ibutivosdel Gastc Publico en Salud cntrbuDt tDopoverty of various factors- Conclusioni hat the provsio of public education,occpation, area of residence,etc. health servicestends to alleviatepoverty in The document also reviews several urban areas, but not in rua areas. Poor (not employment progams in Latn America, extremepoor) bendit significantlyfrompublic including the PArT, and food assistance health services becaem they resid mostly In programs in Pen:. urban areas. Total impact of public spending in health on income dubon is not Natioal PEa6 Instituteand UNDP. La sificant. Docum No. 3 is ctos Pobreis en d Peru: Diagnostc y Dsrbutvos del Programa Enfeedad- opuests de PoNUca(1990). Analyzes Materidaddel InSto Peuano de Segurd characterisi of povety in 1986 usingthe Soclal. Thisprgm onlycovers 20 percet 1985186 Peru Liing Standard Survey and of the popation and only 6 percent in the trends since 1981. Combines use of povety lowest quintile or in rural areas. Also line and basic needs Recomme", diues the impact of cial secuity beft progams to address poverty, and contrbis on the distrbuion of income. Finding is that it slightly lessen Insdtu ,Ajuste yEconomlaFamWi incomeinequalities. 1965-190(1990). Examinesthe Impactof eoonomictrends between1985 and 1990on World dL A seriesof studiesbasd on the houseoldconumption in Lima comparing 1985/86PLSS were published as Ivg iormaton from two Peru livig Standards Standdss StudyWoringPaper. Sureys (1985/86and 1990). Also eaineseey coveredtopics such as povertyproffle, the Impact of the measwurestaken by the labor markets, and socia service. Govenm t in August 1990 on low and middle-incomehouseholds in Lima using data from anotherupdaft of the PLSS.

67 AnnexC

Calcuton of Poverty LUes

1. Two estimates of the povet line to meet t calorie and protein eqirements have been used in this report. Both ae becase conumption patterns and the based on the capacity of a householdto availabili of certai food items are quito consumea basic food baset. -Households differen from the other egions of the are classifiedas extremelvpoor if their total county. In order to select the specific expeditus are less thn the cost of a basic productsand the ditbution of osumption food basket. Householdsare classifiedas over broad categories (meat, vegetables, gM if their expenditus are less than the fiuits, et.), the consumptionpatterns of cost of a basic commoditybasket inchuding small farmersin the siera as shown in the foodand non-fooditems. In order to define ENCA (1971-72) household survey were the laKtterpoverty line, the expenditure used as a model. The resulting basket neceossayto purchase a basic food basket supplies 2241 calories and 55 was inflatedby a factorto take into account grams/person/day). The composition of non-foodiems. The factor was determined both the Liur and the nual siewa baskets by locating the decile in which the mean was checked against acual food purcas expditure on food was closest to the cost pattems of survey respondeats and both of the basic food baket. Then the cost oi were foundto reflectcumnt patternsamong the basic food basket was scaled up using low-income households in tbhes areas. the same proportion of food/non-food Deails are prvided in Table4. The basket exenditur. Region-specificfactors for the was then vahledusing the averg prices of food/non-foodratio were used to reflect four cities in the sierra, weightedby the acua expealitu patteuns. Thus the popldationof each. The broad consistency expenditure level at which householdsare of these baskets with curent consumption classifiedas poor vates by regionbecause pattens was checked with data fram the of dfferences in both prices and 1991 LivingStandards Survey and modfied conumption patters. accordingly.

2. The calculation of the basic food 3. Both the Lima (and urba coast baket and its cost also differed slihgtly and the sierra food baskets rreset depending on the region. In the case of economic mes to meet mtnal Lima, the basic food basket determinedby reqiements taldng into account the the National Isitue of Ntion (INAN) coumption patterns amonglower icome was used. It provides 2,169 caloriesand koseholds. The baskets are meat to 62.3 gams of protein per person.' These provide a reasonableguide to the level of levels are similar to the amount of calories expeiiture reqwiredby householdsto meet and proteinsin the consumptionbasket used theirmntrional rqmnents. Neverless, in the earlier CEPAL work (2155 calories. thse baskets do not necessarilyrpresent and 58 grams of protein/day.2 The INAN the le"t expensive means to meet basket reflec t consumption patterns nutitional reqiements. Thus howhlds foundamong Lima laborers in the 1971-72 with lower food expendures could ENCA (Encuesta Nacional de Consumo conceivably meet the same mnonal AHmntai) and is widely used in Lima requirebents through more e tfficen among both private and public sector pur1hases. ageies. For the urban coast, the same comosition of the basket was used, but 4. In order to facilitate oompaD products were valued at the averageprices across regions aU the baskets wero then of three cities in the region weighted by expesd in prices of Lima for the first populatiao. For the sierra it was necessary week of October 1991. to moditythe compositionof productsused

1. Thecalculation of requimnentsfor the indiviualtakes into account the age and geder compositionof fth population.

2. Magnude de la Pobrea enAted#,a Laud en losAnos Ochea, CEPAL1990 68 Ta 1: P : -. Povrt Lie for Extem Poo an Poo b Reglon

Exeom Povty Limn11 PovwtyUrno 2! "w Saba) (NewSe..) Unadjusted31 Adjuted 41 31 Adjuw 41

Uka 315 SOO 617 979 UrbanCoast 315 492 53m; 833 UrbanSirra 258 433 488 786 RuralSIera 258 488 372 704

Nowt ibpea in hnsPtiw OeO_ 1S1 11 Cgt fdboded bedmetp epkulywo R Ct dofuudl bade w epltuse a# Heusebel .xp.sdbAomidie by Iwmd I.

Tab 2: PE.: Por L_sUt Rati to Expend Olstdn MdMm by Rigion

Dade Rag 2!

Ptyt. 1I PovertyUiw Poer1 POVIh PerCapul Fll!p PA_N (ewSabd New So,) La" Lime Exeendbs Mom Em. Mm Exm. ULa 500 9799 V-V- 1276 40% 77% UbanCoast 492 833 Il-ll VI-VI 925 63% 90% UbanSr 433 786 I1II1 V-VI 900 48% 80% Rurd Sierra 488 704 114N1 VII.4X 534 91% 132%

NoW. r- we.emmIn Lb_ pd*n Osu*e IW9

21on A_S 2/ Bond on e-sind expenitm dIebutsn Tabl 3: BasIcFood Bat for Lmnd iwo Coast of the HNaonalInstte for Foodand Nutdton (INA)

Candad Producto Recommed.

Lecheen Polvo 2 tros Huevos 2 urJd. Higado 0.25 Kg. Jurel 0.50 Kg. Zapdb 1,0 Kg. Tomate 0.2 Kg. Ajos 0,015 Kg. PorejIl 0,03 Kg. Pina 0,30 Kg. Naranjs 0,25 Kg. Pbtano 3 Unid. PapaBlanca 1,0 Kg. Arroz Cornente 0,84 Kg. PanTolete 19Unid. AcetosCompuseo 0.18 Lt Margarina 0,05 Kg. Azucur Blanca 0,30 Kg. Sol 0,03 Kg. CanetaA Granel 0,01 Kg. Te(Cajax57grs)2/ 0,02 Kg. SI. 5.25S

Nate: Oe food bssketfora f coomposodof husabnd,AF* nd44den (2160ec*ike e 62.3 vee of protaI ps parasoJ.

* Limeprke hI October 1989 new soles

70 Tabe 4: Basic Food Basketfor the Silra Clanitdad CmnidadRequeuida Requwida por peona (am.) ,pr famia tars.) Productos Rubro Producto Cames 45 Cordeso 7.5 45 Vacuno 2.5 15 Polio 2.5 15 Menudencias 5 30 Otras cames 7.5 45 Huevos 20 120 Lechev Defivados 25 Leche Fresca 20 120 Quesos 5 30 Vireas 50 Tomato 8.33 50 Col y colifor 12.5 75 Zapalto 4.17 25 Cebolla 16.67 100 Zanahoria 8.33 50 Frutas 75 Naranjas 25 150 Otras frutas (manzana) 50 300 Lqgumninosas 50 ArveJas 14.29 86 Frijoies 17.86 107 Habas 17.86 107 Tuberculosv Raices 497 Camote 15.27 92 00. y Ollueo 40.71 244 Papa 400 2400 Chuno 20.35 122 Otros ubercuilos 20.35 122 5erealesv pofivado" 327 Arroz 45.21 271 Cebada 33.9 203 Mai grano 71.58 249 Maiz fresco 22.6 136 Qulnius 11.3 68 Trigo 0.14 181 Pan 100 600 Fideos 22.6 136 Azuc 60 Azucar blanea 16.8 101 Azucar rubla 43.2 259 Aggi,Usv Grasas 47 Mantecas 31.33 188 Aceite 15.67 94 Sl.3.96B AfOT,Sto the bW_tVdthf /SWrt h eh _dbh_ *em th hwaa. sVwVG4BA1171-72 Awms ,ed *4Ww. t M| t *AM ,A I4 and

_*

55g,SfSef ph ..p..w* 71 *Uhpme Gabbe 1S/1 .. Aim D

PMU LPVMIGSTANDARDS MEASUM SURVEY,1991

SELCE TA EKIVlEl Standards Measurement&ewm

E.1. Locationof the Poor by Regon E.2. EstimatedNumber of Poorby Region E.3. Distributionof Populatonby Regionand Povt Lovel E.4. Characterics of householdhead by povty lvel: AS are (Colum %) E.S. Cha of houeholdhead by povertylevel: All ares (Row%) E.6. Charactstic of householdhead by povy levd: Lima(Column %) E.7. Charactriscs of householdhead by povertylevel: Lima(Row %) E.8. Characteriscsof householdhead by povertylevel: SierraRural (Column%) E.9. Characics of householdhead by povety level: Siera Rura (Row %) E.10. Dbtribudonof total expenditr by povertylevel: All are E.11. Distributionof totalexpendits by povertylevel: Lima E.12. Ditrution of totalexpendiures by povery level: Rur Sien E.13. Housingcharacteistics by povertylevel: AlUare (Colum %) E.14. Housingcharacteristics by pove level: Al areas(Row %) E.15. Housingcharaceristics by povertylevel: Lima(Colum %) E.16. Housingcharacterstis by povertylevel: Lima(Row %) E.17. Housingcha isis by povertylevel: SierraRural (Column%) E.18. Housingcharacics by povertylevel: All areas(Row %) E.19. Occupationof householdhead by povertylevel E.21. Coverageof vaccinationschidW-. ages 0-12 years: All areas E.22. Coverageof vacciatons childrenages 0-12 years: Lima E.23. Coverageof vaccinations,children ages 0-12 years: SierraRura E.24. llnessby povertylevel: All areas E.25. Illnessby povertylevel: Lima E.26. Illnessby povertylevel: SierraRur E.27. Sourceof care by patientpoverty level: all areas E.28. Sourceof care by patientpoverty level: Lima E.29. Sourceof careby patientpoverty level: Siera Rural E.30. Incidenceof food assistance E.31. Useof food assisanceby povertylevel E.32. Effectof foodtrfers on expenditurelevels E.33. EducadonLevel and Schoolatndance by Region

75 | | 1 I I I c | | |Y;s aW |~~~~~~~~~I Wk * ; X 'I* | ".l

;t a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

-9~~~~~~~~~~~~~~~g i~~~~~ !Xs|'ki g #lgW , TALe 5-4. CHARACTISTICSOF HOUSEHOLDHED BYPRTY LEVEL:ALL MAIS (PRCENTAGE)

EXTREMELY N PER-CAPITA CHAR ERISTICS P OTER POOR OTHERS TOTAL AAL EXPENDITURES1/ (NEWSOES) UNADJUSTED2/

OEMDEROF HUSHOLDHEAD NALE Be 82 85 81 83 74 FEMALE 12 18 1S 19 17 83

AGEOF HUSEHOD HEAD 16 TO25 YEARS 5 4 5 S S 704 26 TO3S YEARS 22 17 20 16 18 720 36 TO45 YEARS 30 24 2 23 25 7To " TO5S YEARS 22 22 23 21 22 711 S6 TO65 YEARS 13 19 16 20 18 821 MORETHA 6S YEMS 8 14 10 1S 13 867

LANGsAGEOF

SPAwMSN 59 90 76 91 84 821 UIaECUA 19 8 14 7 10 449 TMARA 22 2 10 2 6 242

EMPLOYEROF sOLD HEAD PUILIC SECTOR 8 1S 11 16 14 nta PIIVATESECTOR 20 22 22 22 n/a PRIVATEHONE 1 0 1 0 1 n/a SELF-ENPLOYED 49 38 44 37 40 n/a UPAID FAMILY RWER 9 5 7 5 6 nta OTHER 1 0 0 0 0 n/a UNEMPLOYED 4 2 3 2 3 n/a RETIRED 8 17 12 18 1S n/a

HIGHESTGRADE COMPLETEDBY HEAD NOME 11 5 9 4 6 431 INITIAL 1 1 1 0 1 418 PRIMARY 55 34 47 30 38 543 INCOMPLETESECONDARY 1S 13 1S 12 13 633 COMPLETESECONDARY 16 26 21 27 24 808 NON-UNIVERSITYHIGER 2 6 3 7 5 969 INCOMPLETEUNIVERSITY 0 4 2 4 3 1160 COMPLETEUNIVERSITY 1 12 3 16 10 1429 OTHER 0 0 0 0 0 874

100 100 100 100 100 750 78- 1/ At first eek october Lim prices 1991. 2V Hosehold *xpenditures divided by household size. TABLEI-$. CNARACTERITICSOF NOUNO HNEADI POVERTLEVELS ALL AE (PERCENTAGE)

EXTELY MM P-CITA CHRCTERISTICS POOR OTER POOR OTHERSTOTAL ANNtAL EXPEDITURES1/ (NEW LES) UNADJUSTD2/

GENE OF IWUSEHtllDHEAD NALE 19 81 48 53 100 743 FEMALE 13 8 40 60 100 m

AGEOF H0SEi)D HMA 16 TO 25YEARS 19 Sl 4 54 100 704 26 TO35 EMS 22 78 52 48 100 720 36 TO 45YEARS 22 78 S0 50 100 ?00 w TOSS EARS I8 82 48 S2 100 711 S6 TO 65 TE4RS 13 87 40 60 100 021 ORE HAN65 YEARS 12 8a 36 64 100 867

LANGUAGEOF HOUSEHOLDHEAD SpANISH 13 8 a42 5S 100 821 CUECHUA 34 66 62 38 100 449 ATMARA 70 30 8S 14 100 242

EMPLOYEROF HOUSEHOLDHEAD PUSLICSECTOR 10 90 3? 63 100 nIa PRIVATESECTOR 1T 83 47 53 100 n/a PRIVATEHOE 41 60 92 8 100 n1a SELF-EMPLOYED 22 78 50 50 100 nWa UHPAIDFANILYUORXER 29 71 53 47 100 nla OTHER 56 44 100 0 100 n/a UNEMPLOYED 29 71 68 32 100 nta RETIRED 10 90 35 65 100 n/a

HIGHESTGRADE COWPLETEDBY HEAD NOME 30 70 65 35 100 431 INITIAL 33 67 70 30 100 418 PRIlPRY 26 74 58 42 100 543 INCOMPLETESECONDARY 20 80 52 49 100 633 COMPLETESECODARY 12 88 41 60 100 808 NON-UNIVERSITYHIGHER 6 94 28 72 100 969 INCOMPLETEUNIVERSITY 0 100 23 77 100 1160 COMPLETEUNIVERSITY 1 99 13 87 100 1429 OTHER 0 100 0 100 100 874

1 At f1rst week october Linaprices 1991. 2V Household expenditures divded by household size. TABIIE516. CNAACTERISTICUOFI EOUWNW1 BY FTr LEVELt LINA (KNTAGE)

M Pa-CAPITA CNARACTERISTICS EXTETLY ANNAL PoM onmi PM Grim TOTAL v wDIuns 1/ 1MEvSOLES) UNADJUSTED2/

NDEROF u"SEHODHED NALE 74 S0 81 78 79 1019 FINALl 26 20 19 22 21 92

AGEOF HWSEH NEAD UP TO1SY EARS 0 0 0 0 0 2062 16 TO 25 YEAS 2 3 3 3 3 1136 26 TO135YE 25 17 19 16 17 "6 36 TO45 YE 23 24 23 25 24 98 A TO5SYTUS 26 24 26 21 23 90 5 65YEARSO 14 20 18 21 20 1066 N110TUN 65 E 11 13 11 13 12 1187

LANUAGEOF NQUSEN01DHEAD SPANISR 89 97 93 96 96 1028 UECHUA 9 3 7 2 4 576 AYNARA 2 0 0 0 0 669

EMPLOYEROF UWUSUW HEUA PUSLICECTOR 10 14 10 15 13 n/a PRIVATESECTOR 27 29 31 28 29 nVa PIVATENNE 7 1 2 0 1 n/a SELF-EMPLOYED 37 33 35 33 34 n/a UNPAIDFANILY 1m1EQ 2 1 1 1 1 nIa OTHER 3 0 1 0 0 n/a UNEPLOED 3 3 4 2 3 n/a RETIRED 11 20 15 21 19 n/a

HIGHESTGRADE 01PLEEDBT HED NMEl. 6 3 5 1 3 605 INITIAL 0 1 1 0 1 636 PRINAY ST 29 41 24 31 715 INCPLETESECONARY 19 14 19 12 15 774 CONPLETESECONDARY19 31 27 33 30 961 NON-UNIVERSITYHIGHER 0 6 3 8 6 1238 INOMPLETEUNIVERSITY 0 4 2 5 4 1440 COI1PLETEUMivEsiTY 0 12 3 17 11 1884

TOTAL 100 100 100 100 100 1010

80 1/ At first we* ztbw Lim prim 1991. 2/ Householdexpendtum divided by housemholdsize. TABLEE-7. CHARACTERISTICSOF HOUS HEADBY POVERTYLEVEL: LIHA (PERCENTAGE)

CHARACTERISTICS EXTRENELY MEANPER-CAPITA POOR OTHERSPOOR OTHERSTOTAL ANNUAL EXPENDITURES1/ (NEW SOLES) UNADJUSTED21

GENDEROF HOUSEHOLDHEAD KALE 7 93 41 59 100 1019 FEMALE 10 90 38 63 100 972

AGE OF NOUSEHOLDHEAD P TO 15 YEARS 0 100 0 100 100 2062 16 TO 25 YEARS 4 96 39 61 100 1136 26 TO 35 YEARS 11 89 45 55 100 965 36 TO 45 YEARS 7 93 38 62 100 987 46 TO 55 YEARS 9 91 45 55 100 900 56 TO 65 YEARS S 95 36 64 100 1066 NORE THA 65 YEARS 7 93 36 64 100 1187

LANGUAGEOF HOUSEhOLDHEAD SPANISH 7 93 39 61 100 1028 QUECHUA 19 81 69 31 100 576 AYMARA SO 50 50 50 100 669

EMPLOYEROF HOUSEHOLDHEAD PUBLICSECTOR 6 95 31 69 100 nla PRIVATESECTOR 7 93 43 57 100 nWe PRIVATEHOME 44 56 89 11 100 n/a SELF-EMPLOYED 8 92 41 59 100 nta UNPAIDFAMILY WORKER 10 90 40 60 100 nta OTHER 67 33 100 0 100 n/a UNEMPLOYED 11 89 67 33 100 n/a RETIRED 5 96 31 69 100 n/a

HIGHESTGRADE COMPLETEDBY HEAD NONE 17 83 71 29 100 605 INITIAL 0 100 50 50 100 636 PRIMARY 14 86 54 47 100 715 INCOMPLETESECOMNARY 10 90 52 48 100 774 COMPLETESECOMDARY 5 95 36 64 100 961 NON-UNIVERSITYHIGHER 0 100 21 79 100 1238 INCOMPLETEUNIVERSITY 0 100 22 78 100 1440 COMPLETEUNIVERSITY 0 100 10 91 100 1884

1/ At first weekoctober Lifa prices 1991. 2/ Householdexpenditures divided by householdsize. 81 TAMLE5-. CHAACTEItICS OF NWUEIOL EM BYPOV LEVELS, SILEA RURAL (PERCENTAE)

EXTEMLY M PER-CAPITA CNAmACTERISIC PM OTERS POOR OHR TOTAL AmA EXPENDITURES1/ (NEW SOES) UNADJUSTED2/

uENuu oF OUWEHOWHA MAE 92 84 90 84 ST 384 FENALE 8 16 10 17 13 471

AUEOF NELD ED 61TO25 EiRS 5 a 6 8 7 372 261TO35YERS 28 16 2S 14 21 2m 36T 045 YEARS 31 19 29 15 24 300 i TO 55 19 19 19 20 19 436 56 TO 65YEARS 12 19 14 20 16 494 NORETHAN6S6M 5 20 7 24 14 543

LAMOLUAOF NONEHOLDNE SPAISH 37 64 44 67 53 471 UECIA 2as 28 30 26 28 377 AIOM 34 a 27 7 19 207

EPLOER OF HOWEHWDNEM PUBLICSECTOR S 10 6 11 8 nta PtIVATESECTOR 1 9 13 9 11 n/s SELF-EMPLOYED ST 52 5S 53 54 n/a UNPAIDFMILY lAVR 17 23 20 22 21 n/a OTHER 0 0 0 0 0 n/a UNWLOYER 3 2 3 2 2 n/a RETIRE 5 4 S 4 S n/a

HIGHESTGIUD COIPLETEDBY HEWD NMOE 12 16 15 14 14 360 INITIAL 2 1 3 0 2 231 PRIMARY 61 57 59 57 58 384 INCONPLETECONDARY 12 10 12 8 11 359 CWLETESECONDAY 12 11 10 14 12 489 M1ONI-UNIVERSITYHIIER 1 2 1 3 2 571 INCOIPLE UNIVERSITY 0 1 0 2 1 636 COIMPETUNIVERSITY 1 2 1 2 1 627

TOTAL 100 100 100 100 100 395

1/ At ffret week octbeor Lma prfcee 1991. 2/ Huehold e.dittures dIvided by houehold size. 82 TABLE1-9. CHARACTERISTICSOf HUIEOL HA U PNTY LEVEL: SIER WAN (PERMAE)

EXTRMLY NM P-CITA CHARACTISTICS PO OTHES POOR OTNERS TOTAL An" EXPENDITURES1/ (KM SOLES) UNADJUSTED2/

GENDEROF IOUSEHOLDED KALE 42 S8 61 39 10 384 FEMALE 23 n48 52 100 471

AGEOF HuSEmLD HMEAD 16 TO25 YEARS 31 69 52 48 100 372 26 TO35YEAS 54 46 73 2 100 29 36 TO4S YEARS 53 48 74 26 100 300 T S5YEAtS 40 60 59 41 100 436 56 T0 65 YAS 31 49 51 49 100 494 M TUTHN65 YEARS 1S 85 30 70 100 S5

WIAGE OF HWUSEHOLDHEAD S18H 28 72 49 51 100 471 GUECHIA 40 60 63 37 100 377 AYAMA 75 26 85 1S 100 207

EPLOER OF HOUSEHODWHEAD ULIC SECTOR 25 75 43 5? 100 n/a PRIVATESECTOR 52 48 68 32 100 nWa SELF-MPLOED 43 S8 60 40 100 n/a UWID FMILYVKR 34 67 5t 43 100 n/a OTHER 0 100 100 0 100 n/a UNLOY 46 54 76 24 100 n/a RETIRED 46 54 67 33 100 We

HIGHESTGRDE COMPLETEDSY HEKD NE 35 66 61 39 100 360 IIITIAL 60 40 100 0 100 231 PRIMARY 42 s8 61 40 100 384 INCOM SECONDARY 45 55 68 32 100 359 COMPLIETSECDARI 41 59 53 47 100 489 MON-UNIVERSITYHIGHER 19 81 19 81 100 5n INCOMPLETEUNIVERSITY 0 100 15 85 100 636 COlPLETUNIVERSITY 16 84 2 73 100 627

1/ At first weekoctboer Lim prife 1991. 2/ Holuse1d e3spiltures dfvided housbeoldsize.

83 TABLEE-10. OISTRI1UTSONOF TOTALEXPENDITURES BY POVERTY LEVEL: ALL AREAS

KEANPER-CAPITA DISTRIBUTIONOF RATIOOF ANNUALEXPENDITURES 1/ EXPEUDITURES FOODEXPEND1TURES/ (NEWSOLES) (1) TOTALEXPENDITURES UNADJUSTEO2/ ADJUSTED3/ UNADJUSTED2/ ADJUSTED 3/ UNADJUSTED2/ ADJUSTED3/

EXTREMELYPOOR 192 327 5 9 70 68 NONEXTREMELY POOR 873 1,194 95 91 56 55 TOTAL 750 974 100 100 58 58

POOR 329 529 20 33 65 63 IONuPOOR 1.111 1.642 80 67 52 50 TOTAL 750 974 100 100 58 58

l At firstweek October Line prices 1991 2/ Household expenditures divided by household size 3/ Household expenditures divided by adult quivalents

TABLEE-11. DISTRIBUTIONOF TOTALEXPENDITURES BY POVERTY LEVEL: LINA

NEAMPER-CAPITA DISTRIBUTIONOF RATIOOF ANNUALEXPENDITURES 1/ EXPENDITURES FOODEXPENDITURES/ (NEIWSOLES) (0) TOTALEXPENDITURES UNADJUSTED2/ ADJUSTED3/ UNADJUSTED2V ADJUSTED 3/ UNADJUSTED2/ ADJUSTED3/

EXTREMELYPOOR 239 381 2 3 62 59 NONEXTREMELY POOR 1,074 1,392 98 97 50 50

POOR 429 660 17 28 57 56 NOMPOOR 1,399 1985 83 72 46 45 TOtAL 1.010 1,276 100 100 51 51

1/ At ffirtweek October Lima prices 2/ Householdexpenditures divided by household size 3/ Nouseholdexpeditures divided by adult equvfalents

TABLEE-12. DISTRIBUTIONOF TOTALEXPENDITURES BY POVERTY LEVEL: RURALSIERRA

NEARPER-CAPITA DISTRIBUTIONOF RATIOOF AUMMLEXPENDITURES 1/ EXPENDITURES FOODEXPENDITURES/ (NEWSOLES) (2) TOTALEXPENDITURES UNADJUSTED2/ ADJUSTED3/ UIADJUSTED2/ ADJUSTED3/ UNADJUSTED2/ ADJUStED31

EXTREMELYPOOR 166 302 17 32 75 74 NON EXTREIELYPOOR 550 836 83 68 70 70

POOR 211 378 32 55 75 74 NO POOR 668 1085 68 45 68 66 TOTAL 395 534 100 100 72 72

1/ At first wek October Lime prices 84 2/ Nouseholdexpenditures divided by household size 3/ Householdexpenditures divided by adultequivalents TOLE i-13. HOUSINGCHARACMISTICS BY POTVT LEM: AU.AREA (POTAGE)

EXTREELY MA P-CAPITA CNARACTERITICS POOR OTHERSPOOR OTHR TOAL AMA EXPEN- DltURES1t (vimUS SOLES) UNJUSTED 2/

SOURCEOF DRINKINGMATER

PUBLIC INSIDEIIELLINO 44 73 sT s a 67 INSIDEBUILDING 5 a 7 7 7 624 WTSIOESI)LD1iH 6 S 6 S 5 "6 WELL 26 4 14 3 8 m RIVER 14 5 9 5 7 377 VAM TRUTU 1 3 4 2 3 S84 OTHER 4 2 3 2 2 601

HOURSOF PUBLICWATER SUPPLY

1-3 NOM/M DAY 21 15 19 15 670 4-6 NOURS/P DAY 1S 12 14 11 12 695 ?-12 HOURS/EDAY 17 21 21 20 20 m 13 ORMORE HOURS/PER DAY 47 53 47 56 52 904

SORCE Of LIGHTING

ELECTRICITY 49 88 71 89 1 S4o KE_OSENE/OIL 44 9 24 9 16 336 CANDLES 6 3 5 2 3 413 NO 0 0 0 0 0 327 OTER 1 0 1 0 0 551

ERA SERVIE

PULIC SERVICE 2 71 49 76 63 924 MULL-SE"IC 4 4 5 3 4 5S5 LATRINE 40 10 23 9 16 423 NMI 29 1S 23 12 17 448

00OKINGFURL

ELECRICITY 1 3 1 4 3 1595 GAS 4 34 14 41 29 116l KEROSENE 39 46 51 39 45 621 COAL 1 0 1 0 0 404 wow 49 16 30 15 22 37 Oar 6 1 3 0 2 215 DONOT COK 0 1 1 0 1 762

TOTAL 100 100 100 100 100 750

1/ At first sdk october Lm prsi 1991. 85 2/ Houeold exip ftur.s dividd by househotd size. TABLEE-14. HUSING CRACEISTICS BY POVERTYLEEL: ALLARS (PERETAGE)

EXTRENLY NEA PER-CAPITA CHARACTISTICS POOR OTlERS PODR OTHERS TOTAL ANUPALEXPEI- DITURES1/ (NUEVOSSOLES) uNADJT 2/

SOURCEOf DRINKINGWATER

PUBLIC INSIDE DWELLING 12 88 39 62 100 Sy5 INSIDEWILDING 13 a 6 54 100 624 OUTSIDEBUILDIN 20 60 54 46 100 646 WELL 61 39 82 l 100 265 RIWElt 36 64 62 38 100 377 WATEITRUC 9 91 67 33 100 58 OTHE 32 68 66 34 100 601

OURSOF PIUBLICATER SUPPLY

1-3NO RIPERDAY 17 83 48 52 100 670 4-6 HOURS/PEDAY 1S 85 46 54 100 695 7-12 HOURJPER DAY 10 90 41 S9 100 88O 13 OR MOREHOURS/PER DAY 11 89 36 64 100 904

SOURCEOF LIGHTING

ELECTRICITY 11 89 40 60 100 845 KEROSENE/OIL 51 49 71 30 100 336 CANLES 32 68 68 32 100 413 NM 50 50 100 a 100 327 OTHER 45 55 73 2 100 551

SEWERAGESERVICE

PU3LICSERVICE a 92 36 64 100 924 VELL-SEPTIC 18 82 59 41 100 555 LATRINE 4 54 68 32 100 423 NMUE 30 70 62 38 100 48

COlllNl FUEL

ELECTRICITY 5 95 21 79 100 158S Gas 3 98 22 78 100 1187 KEROSENE 16 84 53 47 100 621 COAL 30 70 76 24 100 404 mODm 41 59 63 37 100 387 OTNHER 74 26 90 11 100 215 DO OTICOK 8 92 60 40 100 762

1/ At first weokoctoer Lie prices 1991. 2/ Household spmnditures divided by household sfie. 86 TABLEE-15. HUING CMAAACTERISICSBYPOTY LEVELt LINA (ERCENTAGE)

EXTRMELY M PER-CAPITA CHIRACTERISTICS POOR OTHERS PamR OTHERS TOTAL ANNUIAL EXPENDITURES1/ (NUIWES80LES) UNADJUSTE

SORCEOF DRIKINGVATER

PUBLIC INSIDEDWELLING 60 80 67 86 78 1112 INSIDEBUILDING 9 7 8 7 7 n3 OUTSIDEUILDINiG 19 7 12 4 7 620 WELL 5 1 4 0 2 444 WATEtTRUCK a S 9 2 5 592 OTHER 0 1 1 0 1 670

HOURSOF PUBLIC VATER SULY

1-3 RSIMERDAY 33 14 22 12 16 769 4-6 URSIMR DAY 14 10 12 9 10 810 7-12 H0URS/PERDAY 25 25 25 25 25 1004 13 ORME OURS/MRDAY 28 51 41 55 50 1198

SOCE OF LIGHTING

ELECTRICITY 8S 97 92 98 96 1032 KEROSENZ/OIL 8 2 4 1 2 544 CANDLES 2 1 2 0 1 480 NONE 2 0 1 0 0 327 OTHER S 0 1 0 1 374

SEVERAGESERVICE

PUBLICSERVICE 62 86 7n 92 86 1009 WELL-SEPTIC 11 5 10 2 5 S0S LATRINIE 12 5 9 4 6 670 NONE 15 4 8 2 5 551

COOKINGFUEL

ELECTRICITY 3 3 1 5 3 1909 ¢aS 9 50 25 62 1303 KEROSENE aS 45 72 32 48 668 WOOD a 0 1 0 0 29 DO1T COOK 2 1 2 1 1 785

TOTAL 100 100 100 100 100 1010

1/ At firt weekoteter Lim pries 1991. 2f ouwsholdexpegtum divided by hbwuhtd siz. 87 TABLE5-16. NOUSINGCHARACEIICS SY POETY LEVELS:LIA (PENTAGE)

EXTRENELY NEANP-CAPITA CHARUCTERISTICS P0 OTHERS NONt OTES TOTAL ANA UPWI UES 1/ ( SOLES) IIWJTE 2{

SOURCEOF DRINKINGWATM

PUBLIC INSIDEDMELLING 6 94 34 60 100 1112 INS1E BUILDING 10 90 44 56 100 733 WTSIDEUILDING 19 81 65 35 100 420 WELL 21 79 93 7 100 444 VA TERCK 12 a8 71 29 100 52 OTNER 0 100 60 40 100 6R0

NOURSOf PtIBLICWATE SUPPLY

1-3 NOURS/EDAY 15 85 53 47 100 769 4-6 NOURS/PERDAY 10 90 46 54 100 S10 ?-12 NURS/PERDAY 7 93 38 62 100 1004 I3 ORNOREMRS/PER UY 4 96 31 69 100 1198

SOURCEOf LIOHTING

ELECTltICITY 7 9 39 61 100 1032 KEROSENE/OIL 25 75 65 35 100 544 CNLES 13 S8 as 13 100 480 NONE SO SO 100 0 100 32? OTNER 60 40 80 20 100 374

SEWERAGESERVICE

REULICSERVICE 6 94 35 65 100 1009 VELL-SEIC 15 85 74 26 100 50S LATPINE 17 83 63 38 100 670 NONIE 25 75 70 30 100 551

COOKINGfUEL

ELECTRICITY 7 93 7 93 100 1909 GMS 2 99 22 78 100 1303 KEROSENE 14 8? 60 40 100 668 WOm 50 SO 100 0 100 2 DONOt COOK 8 92 58 42 100 785

1/ At first uk oct.bw Liim price 1991. 2/ Hosehold experditurs divided by houehold sie. 88 TABLEE-17. IOUSINGCHARACTERISTICS BYPCNERTY LEVELS: SIERRA RURAL (PERCENTAGE)

EXtREEILY NEANPER-CAPITA CHARACTERISTICS poOR OTHERSPOOR OTHERSTOTAL AUAL EXPENDItURES1/ (NUEVO$SOLES) JUSDTI2/

SOURCEOF DRINKING WATER

PUBLIC INSIDEDWELLING 23 38 24 44 32 S0 INSIDEWILDING 5 8 7 6 7 403 OUTSIDEUWILDING 4 4 4 5 4 406 WELL 37 14 39 14 23 269 RIVER 28 34 32 31 32 378 WATETRUCK 0 0 0 0 0 284 OTHER 2 2 3 1 2 329

HOURSOF PUBLIC WAtER SWLY

1-3HUS/PER DAY 9 17 7 21 14 560 4-6 OURSPERDAY 6 13 8 14 11 494 7-12 HOURS/MPERDAY 6 9 6 10 8 m 13OR NOREURS/PER DAY 79 61 79 55 67 426

SOURCEOF LIGHTING

ELECTRICITY 19 39 22 44 31 Su8 KEROSENEjOIL 74 S0 68 47 59 332 CANDLES 7 11 10 9 9 423 OTHER 0 0 0 0 0 3%9

SEUESASESVICE

PUBLICSERVICE 3 12 4 14 a 606 HELL-SEPTIC 3 8 3 10 6 637 LATRINE 51 29 42 31 38 346 NONE 44 52 51 45 49 370

COOKINGPUEL

ELECTRICITY 0 I 0 1 0 2016 GAS 0 1 0 2 1 764 KEROSENE 3 10 4 12 7 470 COAL 0 0 1 0 0 252 XIWm 84 84 s8 84 84 393 OTHER 12 3 10 2 7 217 00 lOT ClK 0 0 0 0 0 282

TOTAL 100 100 100 100 100 395

1/ At first weekostobw LU prtes 1991. 2/ HouseholdexpendItures divided by hcuseholdsfze. 89 TABU1-18. 151SINGCNAACERISTICS BY PETY LELIs 81ER OUAL (PERCENTARE)

EIREL NEARPER-CAPITA CHARACTERISTICS POOR OTHERSPOOR OTHERSTOTAL ANNUAL EXPENIIDURES1f (NUEVOSSOLES) ADJUSTED2/

SOURCEOF DRINKING WATER

PRULIC INSIDEDWELLIIM 29 71 45 SS 100 506 INSIDE3UILJDIN 31 69 2 38 100 403 OUTSIDEBUILDING 41 59 43 100 406 NELL 64 36 7 24 100 269 RIVER 36 64 61 39 100 378 WMTERIRUCK 0 100 100 0 100 284 O1HER 51 49 81 19 100 329

HOURSOF PUBLIC WATER SUPPLY

1-3HOUSIM PDAY 20 80 25 i5 100 560 4-6 SIPER AY 16 83 36 66 100 494 T-i2HRS/M DrA 21 79 38 62 100 m 13OR NORE HURS/PER DAY 36 64 S8 42 100 426

SOURCEOF LIGHTING

ElECTRICITy 24 76 43 s5 100 S08 KOSENEIL SO 50 68 32 100 332 CANDLES 32 69 61 39 100 423 tHER 0 100 100 0 100 349

SEWERAGESERVICE

PUHLICSERVICE 14 86 29 71 100 606 WELL-SEPTIC 20 80 30 70 100 637 LATRINE 54 6 67 33 100 346 NOME 37 64 63 37 100 370

COOMINGFUEL

ELECtRtCITY 0 100 29 71 100 2016 GaS 0 100 1 83 100 764 KEROSENE 17 83 33 67 100 470 COAL 36 64 100 0 100 252 WOOD 40 60 60 40 100 393 OTHER 73 27 89 11 100 217 DO IT Col 0 100 100 0 100 282

I/ At first weuk october Lifs prices 199S. 2/ Nouhold explnitum divided by heubold size.

90 TABLe -19. OCCUPATIONOF HOUHD HAD BY PERTY LEVEL(PERCENTAGE)

EXTRELY POOR OTERS POOR OTERS TOTAL

ALL AREAS: PUBLIC 8 15 11 16 14 PRIVATE 20 22 22 22 22 PRIVATEtME 1 0 1 0 1 SELFPEMPLOD 49 3o 44 3 40 UPAID FAMILYTRE 9 5 7 5 6 OTHERS 1 0 0 0 0 UIEIWLOYENIT 4 2 3 2 3 MUIACTIVE 6 17 12 18 IS LIMA PUBLIC 10 14 10 1S 13 PRIVATE 27 29 31 28 29 PRIVATEHWE 7 1 2 0 1 SELFEMPLOYED 37 33 35 33 34 UNPAIOFANILYVOK 2 1 1 1 1 OTHERS 3 0 1 0 0 UNEMPLOYMENT 3 3 4 2 3 MH ACTIVE 11 20 15 21 19 uSD1 coAST PUBLIC 10 12 10 13 11 PRIVATE 31 23 22 26 24 PRIVATEHIE 0 0 1 0 0 SELFEMPLOYED 40 40 46 34 40 UNPAIDFAMILY OKE 0 2 1 2 2 OTHERS 0 0 0 0 0 UNEMPLOYMENT 8 2 4 2 3 1101ACTIVE 12 21 17 23 20 lIA SIER PUBLIC 13 26 23 25 24 PRIVATE 22 17 20 16 18 PRIVATEVWE 1 0 1 0 0 SELFENPLOTED 47 38 42 37 39 UNPAIDFANILY VWXER 2 2 2 2 2 0 OTHERS 0 0 0 0 UNEMPLOMNENT 6 1 3 1 2 NONACTIVE 9 i7 10 19 15 SIERRARUR PUBLIC 5 10 6 11 8 PRIVATE 14 9 13 9 11 SELFEPLOYED ST S2 5S 53 54 MIPAIDFANILYUE 17 23 20 22 21 OTHERS 0 0 0 0 0 UNEMPLOYMENT 3 2 3 2 2 MONACTIVE S 4 5 4 5 URWA PUBLIC 11 16 13 17 16 PRIVATE 27 25 26 24 25 PRIVATENHOE 3 1 2 0 1 SELFEMPLOYED 41 36 39 34 36 UMIPAIDFAMILY WR(lR 1 2 2 2 2 OTHERS 1 0 0 0 0 UNEPLOYENT 5 2 4 2 3 MONACTIVE 11 19 14 21 18 100 100 100 100 100

91 TABLE*-21. CoEAG OF VACCINATION14 CHILDRENAMES 0-12 YEARS: ALL ARM (PCNTWAUE)

All gm *ne Total

ulE 17 22 1 loS fRLXE 78 20 1 100

0-2 TMS 52 45 3 100 3-5 tEAS T8 22 0 100 6-S EARS as 13 1 100 8-12YEAS 90 10 0 100

BY POVERTYLEVELS EXTREMELYPOP 73 25 1 100 OTHERS 79 19 1 100 CRITICALLYPOOR 75 23 1 100 aTHERS 81 18 1 100

BY JINTILES I (POEST) 72 26 1 100 II 77 21 1 100 III 79 20 1 100 IV 80 17 1 - 100 v (ULTHIEST) 88 11 0 100

/ Vaccinationae am, Po.11, Tripte aN Meles

92 TAI I-2U. CWhRAIOF CIUATIU Ul CNiLOU AM 0-12 VIMO LINA (PERCHIAU)

All Seem Toul

By GlEDER ALIE S 15 0 1W FEMALE 05 13 1 100

YAGE GRUP 0-2 m s 0 2 100 3-Tam SO 11 0 100 6-STES 94 6 0 140 S-1lTun 96 4 0 100

By POVE LILS EXTREMELYPOOR as 16 0 100 OTHM SS 13 1 100 CRITICALLYPMOR 84 14 1 100 OTIERS S6 12 0 100

ST UINTILES I (POCEST) aS 13 1 100 it 82 1? 0 100 III 85 14 0 100 IV so 10 1 100 V (EALTIEST) 86 12 0 100

If VYcclntotns Ul ;, Pltoo, TrIple aN Nesest

93 TA3LEE-D. COVE"RAOF VACCINATIONV. CHILOENAGES, 0-12 WARs SIER RUL (PERCENTAGES)

All So= total

By erder NIb 66 32 2 100 Fme 46 31 2 100

0-2 YES 37 5S S 100 3-5 YEAS 69 30 1 100 6-STmS 75 23 2 100 8-12 EAR 78 21 1 100

BY POETY LEVELS EXTREMELY 63 33 2 100 OTHERS 69 29 2 100

CRITICALL 65 32 3 100 OTHERS 70 29 1 100

BYQUINTILES 1 63 35 1 100 !1 66 30 2 100 III 67 29 5 100 IV 6 33 1 100 V 77 23 0 100

I/ Vaecnottionsam BCG,Polio, triple ad N tesi

94 TABLEE-24. ILUIESS BY PWERThLEVELt ALL AUM (MECENTAGE)

EXTRIELY POOR OTERS POOR OTHERS TOTAL

X PEOPLERETING ILLNESSOR1t I Y 24 33 31 34 32

X PEOE ILL SEEKINGMEDICAL CAE 34 54 4 56 50

DAYSILL 9 9 9 9 9

DAYSICTIVIE 2 2 2 2 2

Vote: Ibm dys Itt an day Inactiverefer to pest 4 weeks

TAIILE -25. ILLNESSBY POETY LEL: LINA (PERENTAOE)

EXTRMLY PO OTHIERS POOR OTERS TOTAL

X PEOLE RETING ILLNESSOR INJY 35 36 35 36 36

% PEOPLEIUL SEEKIN MEDICALCARE 44 sT 53 59 56

DAYSILL 9 9 9 9 9

DAYSINACTIVE 2 2 2 2 2 note: Neon das ftt and d"s lnwtive reterto post 4 weeks

TABLEE-26. ILLNESS Y PMERTYLEVEL: SIERRARURAL (PERCENTAGE)

EXTREMELY POOR OTHERS POOR OTHERS TOTAL

X PEOPLEREPORTING ILLNESSOR INIURY 26 36 29 3? 32

X PEOPLEILL SEEKINGNEDICAL CARE 24 41 27 46 34

DAYSILL 9 8 9 a 9

DAYSINATIIE 2 3 3 2 3 note: Nen dap ilttnd dys nactive referto pest 4 weks 95 TABLEE3-2. SOURCEOF CARE31 PATIENTPOVERTY LEVEL ALL AREAS

Extrmoly PoW others PoW others Total

Hosp1tal 30 3? 37 35 36 Center 24 1 18 12 14 HeatthPost a 6 7 5 6 ComaMityPost 2 1 2 0 1 Ctinic 11 29 19 34 2? Pharmeacy 14 9 11 9 10 ome of Practitioner I I 1 1 1 Home of Patient 7 3 4 2 3 Other 3 1 1 2 2 Total 100 100 100 100 100

IABLEE-28. SOURCEOF CAREn PATIENTPOVERTY LEVEL: LIMA

Extretely poor Other Poor Others Total

Hoepital 24 3S 39 35 3 Center 42 13 20 11 15 HealthPost 6 2 3 2 2 CaunmityPost 1 1 1 0 1 Clinfc 8 33 23 37 31 Pharmacy 1S 10 11 9 10 Nomeof Practitioner 0 1 1 1 1 Howe of Patfent 0 1 1 2 1 Other 4 1 2 2 2 Total 100 100 100 100 100

TABLEE-29. SOURCEOF CARM BY PATIENTPOVERTY LEVELS: SIERR RURAL

Extremely Others Poor Others Total Poor

Hospitat 2? 1S 19 23 21 Center 17 19 19 18 18 Helth Post 14 31 24 27 26 Commfty Post 6 1 3 1 2 Clinic 8 12 9 13 11 Phiamcey 13 7 12 6 9 How of Practitewr 4 2 4 2 3 How of Patient 9 6 9 4 7 Other 2 4 1 6 3 Total 100 100 100 100 100 96 Table E-30. IncNence of Food Assistance ...... *...... Extromoly Poor Poor Non-Poor All ...... All Areas: Vasa de Loch. 36.2 81.3 18.7 100 Club de Madres 32.1 88.2 11.8 100 Coetna Popular 57.6 100 0 100 ComederPopuar 32.4 94.4 5.6 100 Otlt Commn 39.5 100 0 100 Nw,ictpaltdad 33.7 75.s 24.5 100 ...... Mean Incidence: 38.6 89.9 10.1 100.0

Lam: Vaso de Leche 34.1 86 14 100 Club de Madres 33.1 89.4 10.6 100 CocinaPopular 57.6 100 0 100 ComedorPopular 32.8 95.4 4.6 100 Ottoaoun 0 100 0 100 Mknieipatidad 34.3 77.1 22.9 100 ...... _ owanIneidwnce: 32.0 91.3 8.7 100.0

OtherUrban: Ctub de adres 0 100 0 100 CocfnaPopular 0 0 100 100 ComedorPopular 7.6 80.5 19.5 100 ollatoamun 100 100 0 100 Municipalidad 0 0 100 100 ...... _ Mon Incidence: 21.5 56.1 43.9 100.0 (withoutVaso de Loche)

Rural: Vaso do Leche 66.7 90.1 9.9 100 Club de Madres 16.7 SO 50 100 CocinaPopular 0 0 100 100 Cacedr Popular 86.2 100 0 100 OlttaCm 0 0 100 100 municipalIdad 50 100 0 100 ...... Mean Inciddnce: 36.6 56.7 43.3 100.0

Vase do LecheOnly: (allareas) under 3 3.0 8.6 1.5 10.1 3 - 5 5.4 12.0 2.9 14.9 6 - 10 12.4 26.0 4.9 30.9 11 - 14 5.8 15.1 4.0 19.1 15 - 49 8.9 17.7 5.0 22.7 50+ 0.4 1.5 0.8 2.3 ...... Total: 35.9 80.9 19.1 100.0 97 TWLEE131. USEOF FOWOAISIPAS I POETY LEVEL

EXTREILY POOR OTHE POOR OTHR ALL

ALLAREAS: VASOE LE 7 4 7 2 4 CLUBDEOfSo 2 1 2 O 1 COCINAPOPALt I 0 1 0 0 COOEDPOPULAR 2 1 3 0 1 O.LACOIN 0 0 0 0 0 .MICIPAI.O 3 2 3 a2 TOTAL 16 a 1S 4 10

LINA VASODE LECNE 32 7 16 3 9 CLUI0E DRES 11 3 6 1 3 COCIIAPOPULAR 4 0 2 0 1 CONR POPULAR 10 2 6 0 3 a" COI= 0 0 0 0 0 IIMICIPALIDAD 17 4 8 2 5 TOTAL 74 16 37 7 22

OTNERU13M 1/ ;OE INADRES 0 0 0 0 0 COCIIAPOPULA 0 0 0 0 0 CONEDORUPPUL 0 0 1 0 0 LA CI 0 0 0 0 0 NULICIPALIDAO 0 0 0 0 0 TOTAL

RRL VASODE LECHE 3 1 3 1 2 CLUIDE NADRES 0 0 0 0 0 COCHAPOPULA 0 0 0 0 0 CONEDORPOPULAR 0 0 0 0 0 OLLACOU 0 0 0 0 0 IUICIPALIDAD 0 0 0 0 0 TOTAL 4 2 4 1 3

1/ Costs y Serra U#bom

98 a' 0o

s X: 8 t w s J

i j n H } i~~~~~

9~~ 3}§'; I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -@j U fJ

U! 11111-f_

mf 0a gg -_--- g~~hiIi I I"*w

3a E'0"X-oo-

, 3 |1';~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0

i o~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- PERU

POVERTY AT AND SOCIAL POLICIES AND PROGRAS FOR TE POOR

STATISICAL APPENDIX SmAN mAL

1. Education 1.1 Enrollmentby Level 1.2 Distributionof Enrollmentby Level 1.3 AnnualGrowth of Enrollmentby Level 1.4 Distributionof Enrollmentby Leveland PublicSector, 1989 1.5 Enrollmentby Age and Level, 1989 1.6 GrossEnrollment Ratios by EducationLevel in SelettedSouth American Countries, 1960- 1990 1.7 EducationBudget by Functionand Region,1991 1.8 Distributionof EducationBudget by Functionand Region,1991 1.9 EducationBudget by Line tem, 1991 1.10 Intnal Efficiencyof PrimaryEducation, 1988 1.11 Flow of a HypoteticalStudent Cohort 1.12 Indicatorsof InternalEfficiency and Reurces by Region 1.13 StudentsAchievement in Grade6 by Regionsand Subject,1981 1.14 ComparadveEducation Indicators for SdectedLatin American Countries HI. Healh

2.1 ReportedCases of SelectedCommunicable Diseases, 1985-1988 2.2 Evolutionof CholeraEpidemic, 1991 2.3 Prevalenceof Low BirthWeight, Hospital Maria Auxiliadora Lima, 1987-1990 2.4 Coverageof ImmunizationProgam for CbfldrenUnder 1, 1985-1990 2.5 Health Establishmeit and Beds, 1985-1990 2.6 Distrbutionof Hospitalsand Bedsby Provider,1988 2.7 Availabilityof SelectedHealth Resourcesby Region, 1988 2.8 HumanResources in Health, 198S-1990 2.9 HumanResources in Health(per 10,000Inhabitats), 1989-1990 2.10 Supplyof HumanResources for BasicHealth, 1990 2.11 Medicaland ObstetricalConsultation Per HourWorked, UDES, Lima, 1988-1990 2.12 HealthServices in the Southof Lima, 1988-1990

103 TABLE1.1 ENROLLMENTBY LEVEL (IN THOUSANDS)

1960 1970 1980 1986 1900

PRIMARY 1500 2500 3200 3537 3968 SECONDARY 198 550 1200 1421 1714 HIGHERNON-UNIV 4 24 S6 89 223 HIGHERUt"IV 31 109 248 354 512 TOTAL 1733 3183 4704 6410 6417

TABLE1.2 DISTRIMON OFENROLLMENT BY LEVEL

1960 1970 1980 1985 1990 PRIMARY 86.6% 78.5% 68.0% 65.4% 61.8% SECONDARY 11.4% 17.3% 25.5% 26.3% 26.7% HIGHERNON-UNIV 0.2% 0.8% 1.2% 1.8% 3.5% HIGHERUNIV 1.8% 3.4% 5.3% 6.5% 8.0% TOTAL 100.0% 100.0% 100.0% 100.0% 100.0%

TABLE1.3 ANNUA GROWTHRATE OF ENROLLMENTBY LEVEL

19070 1970-80 1980-90 1980-85 1985-90 PRIMARY 5.2% 2.5% 2.2% 2.0% 2.3% SECONDARY 10.8% 8.1% 3.6% 3.4% 3.8% HIGHERNON-UNIV 19.6% 8.8% 14.8% 11.8% 17.9% HIGHERUNIV 13.4% 8.6% 7.5% 7.4% 7.7% TOTAL 6.3% 4.0% 3.2% 2.8% 3.5%

Sourc: MInfsttyof Educaton,PAnnin Office

105 0

Tabb 1.4 PERU: Drbutln of Enrolmentby LOv and Publl Setor. 1989

TOTAL PUBLIC % PUBLIC Lewd Formd Non-Formal Totd % Fonna Non-Fonmd Totd Fonmal Non-Forna TOtW ProSchool 531.8 344.7 878.5 11.5% 415.5 337.4 7S2.9 78.1% 97.9% 85.9% Prlma, 3874.6 37.7 3912.3 51.4% 3519.9 33.8 3S53.7 90.8% 89.7% 90.8% Seconday 1872.4 20.9 1893.3 24.9% 1820.1 14.6 1634.7 86.5% 69.9% 86.3% Speial Ed 21.6 0.0 21.6 0.3% 19.4 0.0 19.4 89.8% 0.0% 89.8% Hight Ed. O.% TeacheTnding 49.1 0.0 49.1 0.6% 44.S 0.0 44.5 90.6% 0.0% 90.6% Technology 148.1 0.0 148.1 1.9% 81.6 0.0 81.6 55.1% 0.0% 55.1% Vocaonad 206.4 18.9 225.3 3.0% 103.9 17.2 121.1 SO.3% 91.0% 53.8% Arts 7.0 0.0 7.0 0.1% 7.0 0.0 7.0 100.0% 0.0% 100.0% Unlvekty 475.5 0.0 475.5 6.2% 322.9 0.0 322.9 67.9% 0.0% 67.9% TOTAL 7186.5 422.2 7608.7 100.0% 8134.8 403 6537.8 85.4% 95.S% 8S.9%

Soumee:MliOr of Edutn Ffw Office, OD tmt of Stabtti Tabe 1.S: Enonimentby Ase andLevel. 1989

Ages FORMAL NON FORMAL Higher Enroml PreSch PrimarV SecondarySpecia Ed PeSch Pimary SecondaryEducaton TOTAL Population Rate 1 3.3 0.0 0.0 0.0 4.8 0.0 0.0 0.0 8.2 623.5 1.3% 2 4.5 0.0 0.0 0.1 8.1 0.0 0.0 0.0 12.7 616.2 2.1% 3 87.5 0.0 0.0 0.2 97.5 0.0 0.0 0.0 185.2 609.9 30.4% 4 183.1 0.0 0.0 0.4 112.6 0.0 0.0 0.0 296.1 603.6 49.1% 5 241.6 28.9 0.0 0.8 117.9 0.0 0.0 0.0 389.2 600.6 64.8% 6 11.6 517.8 0.0 1.0 3.8 1.6 0.0 0.0 535.9 593.4 90.3% 7 0.2 529.2 0.0 1.2 0.0 1.3 0.0 0.0 531.9 585.9 90.8% 8 0.0 527.4 0.0 1.4 0.0 1.4 0.0 0.0 530.2 578.2 91.7% 9 0.0 525.4 0.0 1.7 0.0 1.6 0.0 0.0 528.6 570.5 92.7% 10 0.0 521.7 0.5 2.0 0.0 1.7 0.0 0.1 526.0 562.6 93.5% 11 0.0 494.1 24.1 2.1 0.0 2.2 0.0 0.2 522.7 554.4 94.3% 12 0.0 289.1 215.2 2.1 0.0 4.0 0.1 0.5 511.0 547.5 --93.3% 13 0.0 186.6 290.1 1.9 0.0 5.6 0.2 1.2 485.7 542.5 89.5% 14 0.0 132.7 322.8 1.6 0.0 7.0 0.4 3.4 467.9 538.9 86.8% 15 0.0 63.9 361.6 1.3 0.0 11.4 1.2 8.3 447.7 534.9 83.7% 16 0.0 37.5 338.8 1.1 0.0 11.5 1.7 22.0 412.6 530.7 77.7% 17 0.0 13.7 152.8 0.8 0.0 11.7 1.8 44.8 225.5 526.6 42.8% 18 0.0 5.3 86.1 0.5 0.0 13.4 2.5 79.9 187.7 522.5 35.9% 19 0.0 1.8 42.8 0.3 0.0 11.2 3.0 99.6 158.8 518.1 30.7% 20 + 0.0 1.0 37.6 0.8 0.0 46.4 9.8 644.9 740.6 2499.8 29.6% TOTAL 531.8 3876 1872.4 21.6 344.7 132.0 20.9 905 7704.4 14877.0 51.8% % 6.9% 50.3% 24.3% 0.3% 4.5% 1.7% 0.3% 11.7% 100.0%

Soure: Ministy of Education,Planning Office, Opt. of Statstics

0 Figure1: EnrollmentRates by Age, 1989

0.9Os ..

0.7 -

-0.6 ~0.5

04-

03.

0.2 ~~~~~O4~ ~ ~ ~ Oe 0.1

0 1 3 s 7 9 11 13 15 17 19 2 4 6 8 10 12 14 16 18 20+ ABUs

108 60t I If' Ij''Fit

I 8 i88 I !I;J"Ii;hi I S;g;;; S

£_ "s _a q_ _8s * O if' Bbb4bX *.b . bppI@A$ 3 t

" Qg"8s" uS a| XI .4 .4 .4. . 4bt

"l b e . ;.tx " > h x .4

= J p p @.. gr -r

.4..44.44. 34

.40 *tSt

aXXX@X$@ t8 Tale 1.7 Educatin lBudgt bYFFaon aWdR8sh9L 1691 Unli)

% nro Region Sated" Maul Son TOl 9.2% Gnu s59160 687960 444083 8691203 8.7% 459.1 223.1 4.8% Anizonas 3190942 237330 177698 606270 4.7% 736.2 15.8% Nor atd" ddNlManon 10536460 1006531 660729 12235715 16.1% 86.8 1.9% Ucyll 1365911 83430 62573 1501914 2.0% 441.5 9.5% Inca 65696053 557970 41847s 7572501 9.9% 512.0 11.0% Sn Mardtn-La Llbeid 7058609 705528 461093 8223230 10.8% 10.9% 512.3 11.0% LesULbe,adoes- Wed 7348468 544320 408241 8301029 11.9% 621.4 13.4% Andr" Aveli QCMe 7365327 828190 848144 9029661 315.9 6.8% Chovn 4840814 3s86o3 269123 5468767 7.2% 6.2% Araup 4216448 45o540 231225 4931211 6.5% 287.2 460.6 9.7% Jose Cados MWueteod 7768136 507060 380296 8646491 11.4% 100.0% TOTAL 65722321 6966689 4419982 76108992 1000% 4s6.7

Souwe.: MIEb*te cb lduecl. Abn de Ovew'So cb Sbetar Imadon, 1991

TOW. 1.0 DOt*_um of Eduhata Budgetby Funhtoned Reglon 1991

Sali MateWas Sendoe TOtal Region 82.8% 10.4% 6.7% 100.0% Gnu 88.5% 6.6% 4.8% 1000% Anmaona 86.1% 8.2% 5.6% 100.0% Nor Oiet del Mranon 90.3% 5.6% 4.2% 100.0% aely 87.1% 7.4% 5.S% 100.0% na 85.8% 8.6% 5.6% 100.0% Sn Martin- La Ubertad 88.5% 0.6% 4.9% 100.0% LosLbobrtedoms - Wed 81.5% 9.2% 9.4% 100.0% Andr AveinoCcere 88.5% 6.6% 4.9% 100.0% Chown 85.6% 9.1% 5.3% 100.0% Arequipa 8U.7% 5.6% 4.4% 100.0% Muwum - Tao - Puno 86.4% 7.8% 6.8% 100.0% TOTAL

Sowo.: Mbtatdr de Eduen.wL Pkn do Cbsw.nd del Sec*r Edwadan, 1991 Educscon, 191t

TaWb1.9 Educton Btuest byLnc hem. 1991 Onduelm Pubic Treasury % Tota % 150325806 52.5% 150325806 S2.2% 84066177 29.4% 84066177 29.2% MOEEudet (1) 517S8987 18.1% 61758987 18.0% Tnsfwrs to ODescntbInstt (3) 86259629 23.1% 66259629 23.0% Talon to Unlverdee (31 84066177 29,4% 89586938 29.8% Not BudgetMOE (1-2)- (4) 517586897 18.1% 53276848 18.S% Budgeof Oecent Inst (5) 13596S641 47.5% 135965PA1 47.2% Budget of Unwvesi"e (6) 286291447 100.0% 288184208 100.0% Reional Govnmet m Tota Educaton Budget(4+5+7)

Seoe: heJt r E _ n do Oeaa,ulo dotSetor leim_ 1991 110 Tab 1.10 Pen: bhuea Wlee of ?Edure emv 188

Tranion Rates

e1 69-2 0e3 , 4 0-5 -6 Promtoton 02.5 73.0 76.1 80.1 03.5' 04.4 Repetion 31.5 21.4 10.9 16.1 1*.1 5.6 rpout 6 a 5 3.8 0.4 0

Tabe 1.11 Flow of* HWpodhee Studnt Cohet

Sti In Numberof Year 1 2 3 4 5 6 School Graduae 1 1000 1000 2 315 625 940 3 99 331 400 890 4 31 133 335 345 844 5 10 48 14 307 277 806 6 3 16 es 173 251 231 782 218 7 1 6 26 79 18S- 290 562 241 8 0 2 9 32 03 109 30s 16O 82 e 0 1 3 12 41 87 143 10 0 0 1 4 16 39 6o 37 11 0 0 0 1 6 10 24 15 12 0 0 0 0 2 a 9 6 13 0 0 0 0 1 2 3 2 14 0 0 0 0 0 1 1 1 15 0 0 0 0 0 0 0 0

Student Yea. 14.0 116 167 955 911 800 6360 701 Promted 912 a54 801 765 761 761 4865 Rqepeate 460 248 212 154 147 45 1266 Dropouts 88 58 53 36 4 0 239

Yearsof Schooling per pmoed student 1.6 1.4 1.3 1.2 1.2 1.1 1.3 % ExtbA8e G-1 G-2 6-3 604 0.5 0-0

Oneyeror more 31.5 48.2 56.9 63.8 69.0 71.3 Two yeaso more 9.9 17.7 25.5 31.6 37.8 39.6 ThreeYoem orm 3.1 6.2 10.1 13.5 17.4 18.7

Perentagewho Graduat 76.1 Prntage of Enteig Cohortwho Grduat withoutRepeaftg 21.8 Percentagof Grduste.Graduatng who Repadng 28.7 Student Yea per Graduate 8.4 AverageNumber of YearsIn Primary Education 6.4 AverageNuber f Gtrade.Completed 4.9 InputlOut 1.4

Sow..: Sd _efe/sE andHeah,S. & Pot& Ac.., rsWs arnd Wyhr In pr&niy .d&wat1W.ORAALC, Octob 1991. N Table1.12 PERU:Indicators of IntemalEffciency and Resources by Region

Percentage % stdent Avwese Percentage % Schoolswith Repters ompleting # yearsto P tage Uncerffied no eletrkict PrmaryEd Grade6 graduate literaes Teachors water, sewere Depwbents 1988 (8188 8188) 1981 1988 1978

Ayacucho 27.8% 28.3% 15.2 45.1% 64.9% 77.1% Huancavelka 27.2% 30.3% 15.8 44.2% 71.3% 81.9% Caamarca 22.7% 33.9% 15.3 35.4% 58.6% 80.5% Huanuco 29.6% 34.1% 14.7 32.4% 61.5% 83.4% ApuTmac 33.3% 39.5% 13.9 52.3% 51.7% 78.9% Ucayll 27.6% 41.1% 12.2 11.4% 58.5% 85.7% Amazonas 26.8% 41.4% 13.8 26.8% 53.8% 81.8% Loreto 31.6% 41.5% 14.2 15.0% 70.8% 85.6% Cuzco 26.6% 44.4% 11.9 37.2% 48.9% 61.2% SanMatin 30.8% 50.3% 11.5 16.4% 61.5% 82.2% Madrede Dios 26.6% 50.8% 10.6 10.7% 72.5% 83.5% Puno 22.2% 53.8% 10.8 32.7% 60.5% 61.5% CalIlo 9.3% 56.3% 9.5 2.9% 47.4% 5.0% Pium 17.6% 57.9% 9.8 22.0% 62.9% 75.6% Junin 18.4% 58.1% 9.6 18.6% 48.1% 54.4% Ancash 25.7% 61.2% 9.9 28.4% 52.9% 72.2% La Ubertad 16.8% 61.4% 9.4 17.7% 33.6% 61.0% Pasco 22.1% 66.9% 8.9 22.4% 75.6% 70.2% Lambayeque 16.2% 67.1% 8.3 13.4% 51.7% 68.4% Moquegua 12.4% 72.4% 8.4 12.4% 34.1% 51.7% Tacna 11.9% 73.6% 7.9 8.9% 28.8% 37.4% Arequipa 12.9% 78.4% 7.7 10.8% 29.4% 38.9% Ica 13.4% 78.4% 8.2 6.8% 16.2% 37.9% Tumbes 20.1% 80.2% 7.9 8.5% 51.0% 69.0% Uma 10.1% 84.5% 7.0 4.5% 46.6% 22.5% Tota Peru 20.0% 57.0% 10.3 18.1% 51.2% 64.3%

Sotace: ,FwnarH. and Rosaes,J. EducacIn ww Mlr8doHead A dnro. Irno, IPP, 1990 Table1.13 PERU:Students Achievement In Grade6 by Regonsand Sublect 1981

N8tional Schools Areas Reoions Subject Averae Public Prva Urban Rural Lma Costa Sierra 80h Mathematics Avg 7 7 11 7 8 9 9 5 6 % Faling 76% 81% 64% 72% 85% 63% 73% 93% 93% Langage Avg 14 14 17 15 13 16 14 13 15 % Failing 15% 19% 3% 10% 27% 7% 14% 24% 20% NaturalSciences Avg 13 14 17 14 12 18 15 11 14 % Failing 23% 29% 7% 16% 38% 8% 22% 46% 26% Social Sciences Avg 12 12 13 12 10 12 12 11 14 % Faiing 38% 41% 31% 36% 4E% 34% 33% 51% 45% Average 12 12 14 12 10 13 13 10 12

Source: Mkintry of Education.Diagnostico del Rendimiento Academico de los Estudlantesde SextoGrado de EM. Lim, 1981 Table1.14 CoMphavi Educatic dbatormsfor SelectedLatin Ameica Cmutdse

t _mvate pop Edutol Exp olm AsS of16+ As pwcet Pomfad I Sea Totd Female GNP O Exp or"s Nt Grow Nt ARGENTNA 90 54 42 27.2 1975 NA NA 2.5 9.5 108 Se 21.6 1980 6.1 NA 3.6 16.1 106 94 74 40.8 1S87' 4.6 6.0 1.9 8.9 115 BOUVIA 31 21 11.2 1975 M6.8 NA 3.5 NA 85 73 16 12.0 1980 NA NA 4.4 25.3 87 72 36 37 27 10.6 1987- 25.8 34.9 8.1 20.1 97 78 BRAZIL 26 9 10.7 1976 24.3 NA 3.0 NA 88 71 34 14 11.9 1980 25.5 NA 3.5 NA 99 81 16 10.9 1987- 22.3 23.5 4.5 17.7 95 84 38 CHILE 34 15.6 1975 NA NA 4.1 12.0 112 94 48 63 13.2 1980 8.9 NA 4.6 11.9 112 98 56 17.8 1987* 5.6 NA 3.6 15.3 102 89 74 COLOMBIA 8.0 1975 NA NA 2.2 16.4 118 39 10.6 1980 14.8 NA 1.9 14.3 128 75 44 56 13.6 1987' 11.9 12.9 2.7 22.4 122 73 ECUADOR 28 26.9 1975 NA NA 3.2 25.9 101 78 39 61 36.5 1980 19.8 NA 5.6 33.3 113 29.3 1987* 17.6 20.2 2.8 21.3 114 56 PARAGUAY 19 16 6.7 1975 NA NA 1.6 14.0 100 83 87 26 8.5 1980 12.6 NA 1.5 16.4 103 88 29 24 8.8 1987* 11.8 14.6 1.5 16.7 101 PERU 46 14.6 1975 27.6 38.2 3.5 17.6 113 86 59 19.4 1980 18.1 26.1 3.2 12.8 114 87 64 25.5 19870 13.0 20.1 2.4 16.4 118 URUGUAY 60 16.0 1975 6.1 NA 2.2 10.0 107 60 17.3 1980 NA NA 2.6 9.3 106 91 77 47.8 1987' 4.6 4.1 3.1 15.1 109 VENEZUELA 81 45 35 18.1 1975 NA NA 4.5 NA 100 86 41 34 21.4 1980 16.3 NA 4.4 14.7 109 89 54 44 26.5 19870 13.1 14.5 3.5 16.6 110

I*) 1987 or late

1991. CUANTO Soures: UNESCOStetstil Yeaubook.1990 Peau: Peruen Numme.s and EfficIencyIn 100nary WwodBank. S"ol IndicatorsDate See Peru: Pem. Access, Repetion Educton

114 Table2.1 ReportedCases of SdectedCommunicable Diseases 1985-1988 (Thousands)

Ano Measles Tuborculosis Typhoid Malaria Hepatitis

1985 9.4 22.7 17.9 35 7.3 1986 6.1 24.7 17 36.9 7.3 1987 6.2 22.1 18.2 39.9 8.8 1988 5.8 28.7 20.6 35.2 6.9 1989 ^ 1 24.1 14.8 35.5 6.3

Sourw: MAs of meah, O0ke of 4WdmIkgv A'ovA,odInformatin

Table2.2 Evolutionof CholeraEpidemic March- May 1991

Epidemhologkal Cases Hospitals Deahs Week N N % N %

5 901 255 28.3 12 1.3 6 11581 2030 17.5 50 0.4 7 16751 3959 23.6 47 0.3 8 16077 4576 28.5 92 0.6 9 14988 4743 31.6 82 0.5 10 19175 7504 39.1 155 0.8 11 19325 8570 44.8 201 1 12 19533 9440 48.3 231 1.2 13 16571 7776 47 148 0.9 14 16967 7532 44.4 156 0.9 15 12953 5622 43.7 108 0.8 16 11849 5319 44.9 122 1 17 9014 3943 43.7 121 1.3 18 8116 3504 43.2 99 1.2 19 6951 3051 43.9 132 1.9 20 6513 2399 36.8 105 1.6 21 5128 1760 34.3 106 2.1 22 4354 1311 30.1 79 1.8 23 3957 1803 45.6 73 1.8 24 2024 844 41.7 33 1.6 25 480 189 39.4 9 1.9

SORce: cuwa0. Peun aSme, 1991. * PrevblonbfenneUee

115 Table2.3 Prevalenceof Low BirthWeight HospitalMaria Auxdiadora Uma1987-1990.

Y"r Total Numberof <2500 9. Births Births %

1987 2741 312 11.4 1988 4611 563 12.2 1989 4226 674 15.9 1990 4088 616 15.1

w)h0 nunmb 1991

Table2.4 Coverageof ImmunizationProgram for Chiklrenunder 1 1985-1990 ANTI- Year Polio DPT BCG SARAMPION

1985 33.3 33.8 32.4 34.6 1986 50.3 49.8 53.6 40.7 1987 44.7 42.6 61.3 35.2 1988 60.8 60.9 70.0 52.0 1989 58.8 58.3 61.8 52.1 1990 ' 60.0 59.0 71.0 53.0

1: M6

116 Table2.5 HealthEstablishmnts and Beds 1985-1990

Health Health Year Hospitals Beds Centers Posts 1985 349 30032 859 1931 1986 353 30413 920 2600 1987 354 30629 963 2851 1988 368 35715 1017 3141 1989 368 35715 1019 3162 1990 368 35715 1019 3173

Sowe.:MAs of HWh

Table2Lo Distrbutionof Hospitalsand Beds by Provider 1988

Hosptals Beds Instituton n %n PublicSector 193 50 26070 80

Mln. of Health 129 33 16867 52 SocialSecurt 36 9 5661 17 ArmedFofces Pol. 15 4 2856 9 PublicSoc.Bee. 3 1 200 1 tP Ste 10 3 486 1

Prvate SecW 193 50 6447 20

PrivateCUnics 135 35 4502 14 PrivateCompanis 41 11 914 3 Cooperatives 11 3 533 2 ChaitableInst. 5 1 483 1

TOTAL 386 100 32517 100

SON".: MktY ofHKamm

117 TabW2.7 Aviabily of Seletd HealthResource by Regon1988

Per1000 InhabIants Regon Beds Estblishments Doctors Nurses Amazonas 12 3.7 2.6 3.4 Ceceres 12.2 3.2 2.5 5.5 Arquipa 21.1 2.7 12.5 19.3 Chavin 8.8 2.5 2.5 1.4 Grau 8.8 1.7 2.6 2.2 Inca 7.5 2.4 1.5 4.4 Warl 11.9 3.3 4.1 3.1 Marategul 10.2 2.7 2.4 4.7 Maranon 7.1 2.1 0.4 4 S M-la Ubertad 11.8 3.2 7.9 5.6 Ucayal 9.8 5 1.9 0.5 Uma 24.8 1.1 23.8 13.1 callao 29.2 1.2 - 9.6 TOTAL 15.3 2.2 9.5 7.4

hwJds hosp4tasend heah contos. d post SaWee: Mfib*y - Helh

Tabl 2.8 HumanResource I Heat 1985-1990

Year Doctors Nurses Obstetrician Pharmacist 1985 18103 13951 3114 4541 1986 19635 14709 2868 5350 1987 20031 15429 3137 5567 1988 20979 16140 3338 5875 1989 21856 15796 3437 6113 1990 * 22900 17600 3900 6400

* Auvisl.nbthEwte Sowoe: Mh*try of f. WU, Gwwl Ofeo of StiSw

118 Table 2.9 Human ResourcesIn Health Iper 10.000 lnhabltants)

Year Doctors Nurses Obstetrical Pharmacists 1985 9.2 7.1 1.6 2.3 1986 9.7 7.1 1.4 2.7 1987 9.6 7.4 1.5 2.7 1988 9.9 7.6 1.6 2.8 1989 10 7.3 1.7 2.8 1990 10.2 7.9 1.7 2.8

* fotvalone Numb.,s Sowo: MtE1Ststlc Copendiwn 1989-1980

Table 2.10 Supply of Human Resourcesfor Basic Health 1990

Doctors Nurses Region Per 10,000 Inhb. Per 10,000 Inh. Coast 0.5 0.6 sierra 0.3 0.8 Jungle 0.5 0.3 Lima 1.1 0.3 TOTAL 0.6 0.5

Souwce:PETRMeA M. PAHO,May 1991

119 Table2.11: Medicaland Obstetrical Consultatfon per HourWorked UDES,Uma 1988-1990

Year MedicalConsultantionlHour ObstetricalConsultation/Hour 1988 2.66 2.01 1989 2.23 2.07 1990 2.36 2.03

Average 2.44 2.04

120 Table2.12: HealthServices In the Southof Lima 1988-1990

ActivIy 1988 1989 1990 %Chango ('000 miles) ('000 miles) ('000 miles) ('000 miles)

Cons.Extema; 878 667 654 -25 Cons.Emergency 123 99 94 -24 MedicalAttendants 709 708 491 -31 HoursWorked by Doctors 266 227 208 -22 ObstetricalAttendants 165 148 128 -22 HoursWorked by Obstetricians 82 71 63 -23 Hospitalizaton 19 17 17 -12 LaboratoryExaminations 469 325 296 -37 X-Rays 49 28 19 -61 Presciptions 569 444 345 -39

121