zambia 08

8/9/2019 Zambia 08

1/268

ZambiaSocial Protection Expenditure and Perform ance Review and Social B udget

8/9/2019 Zambia 08

2/268

8/9/2019 Zambia 08

3/268

Social Protection Expenditureand Performance Review andSocial Budget

Zam bia

Social Security DepartmentInternational Labour Offi ce – GenevaSeptember 2008 ILO / DFID / Zambia / R.22

8/9/2019 Zambia 08

4/268

Zambia . Social Protection Expenditure and Per ormance Review and Social BudgetInternational Labour Offi ce, Social Security Department – Geneva: ILO, 2008 viii, 265 p.

ISBN 978-92-2-121251-5 (print)ISBN 978-92-2-121252-2 (web pd )

International Labour Offi ce; Social Security Department

Social protection / social security / scope o coverage / social expenditure / social security nancing /in ormal economy / poverty / trend / Zambia

02.03.1

ILO Cataloguing in Publication Data

Copyright © International Labour Organization 2008

First published 2008

Publications o the International Labour Offi ce enjoy copyright under Protocol 2 o the Universal Copyright

Convention. Nevertheless, short excerpts rom them may be reproduced without authorization, on conditionthat the source is indicated. For rights o reproduction or translation, application should be made to ILO Publi-cations (Rights and Permissions), International Labour Offi ce, CH-1211 Geneva 22, Switzerland, or by email: [email protected]. Te International Labour Offi ce welcomes such applications.

Libraries, institutions and other users registered with reproduction rights organizations may make copies inaccordance with the licences issued to them or this purpose. Visit www.i rro.org to nd the reproduction rightsorganization in your country.

Te designations employed in ILO publications, which are in con ormity with United Nations practice, and the presentation o material therein do not imply the expression o any opinion whatsoever on the part o the Inter-national Labour Offi ce concerning the legal status o any country, area or territory or o its authorities, or con-cerning the delimitation o its rontiers.Te responsibility or opinions expressed in signed articles, studies and other contributions rests solely with theirauthors, and publication does not constitute an endorsement by the International Labour Offi ce o the opinionsexpressed in them.

Re erence to names o rms and commercial products and processes does not imply their endorsement by theInternational Labour Offi ce, and any ailure to mention a particular rm, commercial product or process is nota sign o disapproval.

ILO publications and electronic products can be obtained through major booksellers or ILO local offi ces in

many countries, or direct rom ILO Publications, International Labour Offi ce, CH-1211 Geneva 22, Switzer-land. Catalogues or lists o new publications are available ree o charge rom the above address, or by email: [email protected]

Visit our website: www.ilo.org/publns

Printed in Switzerland PAP / WEI / S A

8/9/2019 Zambia 08

5/268

This report is the Social Protection Expenditure and Per ormance Review (SPER)and Social Budget (SB) or Zambia. It ocuses on ve key issues with respect tothe objective o extending social protection coverage in the country:☐ Living conditions o households with a ocus on the overall situation o poverty and

key vulnerable groups;☐ Working conditions and prevailing patterns o in ormality in the labour market;☐ Coverage and per ormance o existing public social protection interventions;☐ Current resource allocations to social protection within the current iscal

environment;☐ Future trends in the Zambian social budget

Te analysis presented is a result o a joint effort between staff at the ILO Social Secu-rity Department in Geneva (Pauline Barrett-Reid, Florence Bonnet, Krzyszto Hage-mejer, Mirtha Muñiz, † Raphael Muturi, Amjad Rabi) and the eld offi ce in Lusaka(Adrian Shikwe, Urszula Lonc, John Angelini). Florence Bonnet not only preparedLCMS 2004 and LFS 2005 datasets or the analysis but also developed the method-ology and calculated the indicators o the degree o in ormality in the labour market.Nknadu Chilombo, Mubita Lubawelba and Felix Masiye prepared background ana-lytical reports and/or direct inputs to the main report. Christine Smith was responsible

or editing the text and Irene Brown or typing and ormatting. Special thanks go alsoto those at the Zambian Central Statistical Offi ce who originally worked on all thesurveys used, and shared and discussed the methodological questions with us.

5

8/9/2019 Zambia 08

6/268

8/9/2019 Zambia 08

7/268

7

Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1. Overview of the demographic, economic and social environment 161.1 Demographic trends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161.2 Economic situation and labour market trends . . . . . . . . . . . . . . . . 21

1.2.1 Recent economic per ormance . . . . . . . . . . . . . . . . . . . . . . 211.2.2 Labour market trends . . . . . . . . . . . . . . . . . . . . . . . . . . 24

1.3 Household incomes and inequality . . . . . . . . . . . . . . . . . . . . . . 291.3.1 Composition o incomes . . . . . . . . . . . . . . . . . . . . . . . . . . 291.3.2 Distribution o incomes and inequality . . . . . . . . . . . . . . . . 31

1.4 Selected human development indicators . . . . . . . . . . . . . . . . . . . 32

2. The poor and the poorest: Living conditions and vulnerability . . . 362.1 Contemporary poverty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382.2 rends in poverty: 1991-2006 . . . . . . . . . . . . . . . . . . . . . . . . . . 412.3 Living conditions o the poor . . . . . . . . . . . . . . . . . . . . . . . . . . 442.4 Vulnerability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56

3. Degrees of informality: Prevailing patterns in the Zambian labourmarket and social security coverage . . . . . . . . . . . . . . . . . . . . . 583.1 How people work: Te employment status o the population . . . . . . 603.2 Where people work: Employment by legal orm o establishment . . . . 613.3 Paid employees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633.4 In ormality o employment . . . . . . . . . . . . . . . . . . . . . . . . . . . 643.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Contents

8/9/2019 Zambia 08

8/268

8 Zambia –Social Protection Expenditure and P erform ance R eview and S ocial B udget

4. Social protection schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . . 684.1 Contributory and other employment related schemes . . . . . . . . . . . 68

4.1.1 Te National Pensions Scheme Authority (NAPSA) . . . . . . . . 704.1.2 Public Service Pension Fund (PSPF) . . . . . . . . . . . . . . . . . 75

4.1.3 Local Authorities Superannuation Fund (LASF) . . . . . . . . . . 774.1.4 Workers Compensation Fund Control Board . . . . . . . . . . . . 794.1.5 Comparison o selected characteristics o contributory

social security schemes . . . . . . . . . . . . . . . . . . . . . . . . . . . 814.1.6 Occupational pension schemes . . . . . . . . . . . . . . . . . . . . . . 854.1.7 Pension system regulation . . . . . . . . . . . . . . . . . . . . . . . . 854.1.8 Other employment-related bene ts . . . . . . . . . . . . . . . . . . . 874.1.9 Extending co erage by contributory schemes

to unco ered in ormal-economy workers . . . . . . . . . . . . . . . . 894.1.10 Micro-insurance in zambia . . . . . . . . . . . . . . . . . . . . . . . 904.1.11 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

4.2 Non-contributory programmes . . . . . . . . . . . . . . . . . . . . . . . . . 984.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 984.2.2 Main characteristics o individual non-contributory programmes 1014.2.3 Conclusions: Opportunities and challenges to extend

social protection co erage . . . . . . . . . . . . . . . . . . . . . . . . . 1104.3 Zambia’s health system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111

4.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1114.3.2 Current distribution o health care resources . . . . . . . . . . . . . 1134.3.3 Structure o public-sector health care nancing and expenditure . . 1184.3.4 Co erage o health services . . . . . . . . . . . . . . . . . . . . . . . . 125

4.3.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

5. Government expenditure and Zambia’s Social Budget . . . . . . . . . 1325.1 Government expenditure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335.2 Development cooperation and Government scal space . . . . . . . . . . 1395.3 Zambia’s Social Budget: 2005 and 2006 . . . . . . . . . . . . . . . . . . . 142

5.3.1 Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1445.3.2 Education . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1455.3.3 Social security schemes . . . . . . . . . . . . . . . . . . . . . . . . . . 1455.3.4 Short-term and social assistance bene ts . . . . . . . . . . . . . . . . 147

5.4 Challenges in the provision o social protection in the country . . . . . 148

6. Future trends in social expenditure and nancing . . . . . . . . . . . 1506.1 Demographic and labour orce trends . . . . . . . . . . . . . . . . . . . . . 150

7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175

8/9/2019 Zambia 08

9/268

Contents 9

Appendix A. The Zambian Social Budget model . . . . . . . . . . . . . . 1831. Te logic o the Social Budget model . . . . . . . . . . . . . . . . . . . . . 1842. Population and labour market module . . . . . . . . . . . . . . . . . . . . 1873. Economic module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190

4. Pension module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1915. Social expenditure module . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1966. Government module . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2017. Social Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202

Appendix B. Population projections . . . . . . . . . . . . . . . . . . . . . . . . 2031. Application o the ILO-POP model to Zambia . . . . . . . . . . . . . . . 2032. Comparison o model results with projections rom different sources . . 2073. Reasons or using the ILO-POP model . . . . . . . . . . . . . . . . . . . . 209

Appendix C. Data issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2111. Government accounts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2122. Labour market statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2143. In ormation rom pension schemes . . . . . . . . . . . . . . . . . . . . . . 2154. Social assistance programmes . . . . . . . . . . . . . . . . . . . . . . . . . . 218

Appendix D. Tables of the Social Budget model . . . . . . . . . . . . . . 219

8/9/2019 Zambia 08

10/268

10 Zambia –Social Protection Expenditure and P erform ance R eview and S ocial B udget

Abbreviations

BHCP Basic Health Care PackageBoZ Bank o ZambiaCBoH Central Board o HealthCSO Central Statistical Offi ceDFID Department or International DevelopmentFNDP Fifh National Development PlanGRZ Government o the Republic o ZambiaGTZ Deutsche Gesellschaf ür echnische Zusammenarbeit

(German technical cooperation agency)HICP Highly Indebted Poor CountriesIH E Integrating the Healthcare EnterpriseILO International Labour Offi ceIMF International Monetary FundLASF Local Authorities Superannuation FundLCMS Living Conditions Monitoring Survey LuSE Lusaka Stock ExchangeMCDSS Ministry o Community Development and Social ServicesMDRI Multilateral Debt Relie InitiativeMTEF Medium- erm Expenditure Framework MoH Ministry o HealthMoFNP Ministry o Finances and National Planning NAE National Average EarningsNAPSA National Pensions Scheme Authority NPS National Pension SchemePETS Public Expenditure racking Survey PIA Pensions and Insurance Authority PSPF Public Service Pension FundPWAS Public Wel are Assistance SchemeRHVP Regional Hunger and Vulnerability ProgrammeSAGs Sectoral Advisory GroupsSCTS Social Cash rans er SchemeSIDA Swedish International Development Agency UNAIDS Joint United Nations Programme on HIV/AIDSUNDP United Nations Development ProgrammeUNZA University o Zambia

WCFCB Workers’ Compensation Fund Control BoardWFP World Food ProgrammeWHO World Health OrganisationZNPF Zambian National Provident Fund

8/9/2019 Zambia 08

11/268

11

In recent years, Zambia has witnessed steady economic expansion (with growth ratesover 5 per cent in the last our years) and a all in in ation rates towards acceptablelevels (below 10 per cent in 2006). Tis situation has been accompanied by lowerexternal vulnerability: Zambia bene ted rom signi cant external debt reduction in2005 and 2006. Given this relatively avourable context, it is an appropriate timeto examine the per ormance o the Government and other public social actors (e.g.,

donors, NGOs and businesses) in improving the options or the population to makea decent living, notably by gradually extending the scope and coverage o the nationalsocial security (social protection) system.

his report is an output o the irst year o work o the ILO/DFID- unded project in Zambia: “ILO Global Campaign or Social Protection and Coverage or Allas a Means to Reducing Poverty in A rica and Asia”. Te ILO supports implementationo the Zambian Fifh National Development Plan with its Decent Work Country Pro-gramme (DWCP) which seeks to promote opportunities or women and men to obtaindecent and productive work in conditions o reedom, equity, security and human dig-nity. Te activities o the project are deeply integrated in the overall Zambian DWCPobjectives and strategies.

Te report also contributes to the ILO’s Global Campaign on Social Securityand Coverage or All; and builds on earlier global analytical and policy development work undertaken jointly by the ILO, DFID and other cooperating partners. hat work indicates that social security/protection is an indispensable actor in economicand social development and decent state-building, and that even a minimum social protection package (comprising basic bene ts or children and basic pensions or the

It is not the purpose o this report to analyse in depth the sustainability o the economic expansion or theexisting legal ramework to attract oreign investment.

ILO does not make any distinction between the terms “social security” and “social protection”. Both re-ect the same array o policy instruments and interventions undertaken usually by public (also sometimes by

private) bodies, which seek to provide affordable access to health care and certain minimum income securityand other support in case o old age, sickness and disability, death o the breadwinner (particularly when thebreadwinner suffers these contingencies as a result o employment-related accident or disease), unemployment,maternity and other amily obligations. Social security/protection aims at preventing poverty but also at al-leviating existing poverty and exclusion resulting rom these or other contingencies.

Introduction

8/9/2019 Zambia 08

12/268

12 Zambia –Social Protection Expen diture and Perform ance R eview and Social B udget

elderly and persons with disabilities) would substantially reduce poverty in low-incomecountries and be affordable in the longer run.

Social security/protection systems are usually described as consisting o contrib-utory and non-contributory schemes. It is common practice to judge such social pro-

tection systems by three criteria: numbers o the population covered, scope o coverageand adequacy o bene ts. Tese are the main outputs o social security/protection sys-tems and individual programmes. o assess the per ormance o individual programmesand the overall social protection system, this report looks closely at these outputs orboth types o scheme. A comprehensive picture o the ormal schemes (including cov-erage, scope and adequacy o bene ts) is available in subsequent chapters. Coverage bysuch schemes is generally limited to workers in ormal employment, which is a small

raction o all the employed population. Scope is limited to old age, death, survivors,invalidity, maternity and workmen’s compensation and the levels o bene ts are low.Most workers who work in the in ormal economy and their amilies are excluded romcontributory schemes.

Non-contributory schemes are provided by the Government, international andnational NGOs, donors, the church, as well as by traditional practices. Te main non-contributory social protection programme in Zambia is, o course, public health-careservices, or which user ees are gradually being eliminated starting in the poorest dis-tricts. Others are social-assistance type schemes providing mainly in-kind and somecash bene ts to the most vulnerable. However, it has proved diffi cult to obtain a comprehensive picture o numbers o the population covered, and the scope and depth o provision o these schemes. Te situation is expected to improve as the Ministry oCommunity Development and Social Services (MCDSS) designed a monitoring andevaluation system, including questionnaires and a database, intended to look in detail

at the outputs and costs o all social assistance-type programmes – both public andthose run by international donors and NGOs. We looked mainly at public schemeson which a number o analytical studies have been undertaken, commissioned mainlyby the MCDSS and unded by its cooperating partners. It is generally agreed that thecoverage o bene ts and services provided is very limited and that although there isevidence o positive impacts at the local level where they operate, signi cant up-scaling would be needed to bring a measurable effect at national level.

Although different social programmes can be more or less effective in meeting policy objectives depending on their design and on governance quality, there is nodoubt that i inputs in the orm o available nancial resources to cover the costs anddelivery o trans ers are very low, then it is also diffi cult to expect high levels o output.It is also diffi cult to obtain ar-reaching outcomes and strong impacts rom such socialtrans ers. In order to assess the per ormance o a social protection system, there ore, inaddition to looking at the joint outputs o social security programmes in the country,one should examine the nancial inputs provided: government and donor allocationsto social security/social protection, other sources o nancing (contributions paid byemployees and employers, and income rom investments o reserves accumulated bysocial insurance programmes).

See or example: ownsend, 2008. Proposed Monitoring & Evaluation System or Social Protection, MCDSS/G Z, Lusaka, June 2007 (see

also at: http://www.socialcashtrans ers-zambia.org/pageID_2466950.html). We have not yet been able to include ndings o a very recent study looking at the experience o ve pilot

cash-trans er social assistance schemes in Zambia (which also compares costs and bene ts o different tar-geting approaches): Ben Watkins: Alternative Methods or argeting Social Assistance to Highly VulnerableGroups, Kimetrica or the echnical Working Group on Social Assistance, Feb. 2008.

8/9/2019 Zambia 08

13/268

Contents 13

Whereas Zambia allocates signi cant resources to education and health, in rela-tive terms, the other orms o social protection are apparently under- unded. Te con-tributory schemes collect contributions and some o them are building reserves to help

nance the uture pensions o current contributors. However, the pension schemes o

Government employees are subsidized by the general taxpayer rom the state budget.On the other hand, allocations to social assistance are very low compared with thesesubsidies.

Social security coverage will have to be extended because o the scale o in ormal-sector employment, mainly by scaling up the non-contributory programmes. Tis willrequire signi cant increases in allocation o resources.

From that point o view, the report looks at Zambia’s economic and scal per-ormance in order to make an assessment o the scal space available now and in theuture. In this context it also looks at the Zambian Social Budget – how much is spent

in total on the various social programmes in the country, what and whose needs thisspending is supposed to meet, and how and by whom it is nanced. Te Social Budgetis also projected into the uture, under status quo assumptions. Tese baseline projec-tions can be the oundation o any uture policy option analysis supporting the processo a national policy debate, particularly a social dialogue on the uture o social protec-tion in Zambia. It is hoped that it will be a use ul input into ongoing work in Zambiaon the uture design o social security and social protection within the Fifh NationalDevelopment Plan (FNDP) implemented by different sectoral ministries, and by Sec-toral Advisory Groups (SAGs), cooperating partners and other actors involved.

Chapters 1 and 2 describe the overall social and economic situation and ocuson living conditions o the population and identi ying particularly vulnerable groups,based on the 2004 Living Conditions Measurement Survey (LCMS) by the CSO.

Zambia has a population o about 11.5 million people, 49 per cent o them agedbelow 15 years and less than 4 per cent aged over 60 years. Tis relatively young countryis acing a severe threat in HIV/AIDS, which is affecting the lives and causing thedeath o working-age persons who leave behind children, widows and elderly relatives.Around 1.1 million people are in ected (UNAIDS, 2006), with HIV/AIDS more prevalent among working-age people (17 per cent). Hence, li e expectancy has allendramatically. Besides the disastrous consequences in human terms (deaths, broken

amilies, orphan-headed households), this national epidemic highlights the responsi-bility o the Government (and donors) to avoid urther losses and later to improve live-lihood opportunities. Tere is a great need or social protection.

Zambia is also suffering as a result o climate change. About 65 per cent ohouseholds depend on agricultural activities (CSO, 2005) but only 7 per cent oavailable land is arable (BoZ, n.d.). Events such as oods, droughts or epidemics arebecoming more requent. In this context it is diffi cult even or able-bodied individualsand large households (with presumably more labour) to be sel -reliant; hence the estab-lishment o permanent ood programmes nanced by the Government and donors tohelp people living in ood-insecure areas.

Chapter 3 ocuses on what people do or a living: how they work, where they work, and how integrated they are into the ormal economy. In this chapter, we analyse

According to the UNDP, li e expectancy in Zambia has allen by 14 years since the mid-1980s (UNDP,2006, p.26).

Global warming declines average annual rain all, which causes more chronic ood emergencies (UNDP,2006, p. 164).

Tere were at least six disasters o this type in the country between 2000 and 2007. See: EM-DA :Te OFDA/CRED International Disaster Database, www.em-dat.net - Université catholique de Louvain –Brussels - Belgium.

8/9/2019 Zambia 08

14/268


the range o economic activities (not necessarily income-generating activities) per-ormed by individuals by age group, sex and location in rural or urban areas. With

respect to income-related activities, we ocus on paid employees and develop workingde nitions related to ormality in employment . Our analysis is based on the datasets o

the 2005 Labour Force Survey (LFS) prepared by the CSO.Tis is the rst study o this kind carried out with existing national survey dataand it is justi ed, given that offi cial gures do not ully capture in a single indicatorthe complexity o the labour market and the working conditions o paid employees.For instance, an unemployment rate o around 16 per cent is not a use ul indicator

or policy-making because there are no unemployment bene its that justi y peopledoing nothing to survive. Despite data constraints in the labour orce survey, we car-ried out an exercise to explore the conditions o employment. ypically, in ormality inemployment exists at different levels, depending on the existence and en orcement oa legally binding contract with an employer, entitlement to paid leave, and employer participation in social security contributions. Consequently, opportunities to rely onexisting social security provisions also differ according to the degree o in ormality oemployment.

Chapter 4 describes and analyses the scope o different social protection schemes,including employment-related schemes and social assistance schemes. Considering thatonly 3 per cent o employed persons and 22 per cent o paid employees are ‘totally

ormal’ (see chapter 3) and that around hal o the population lives in extreme pov-erty, it is necessary to analyse which population groups are being supported by socialassistance schemes. Tere is a donor-supported initiative to extend the current pilotcash-trans er schemes to reach the poorest and most vulnerable segments o the popu-lation, but there still remains the challenge o integrating the in ormal sector into well-

established social protection schemes.Chapter 5 outlines the current situation o government accounts and existingiscal space. Donors are identi ied as important partners o the Government in

expanding social expenditure, mainly in health and education sectors. Chapter 5 also presents our estimates o the Zambian Social Budget, the levels o social expenditureand sources o its nancing.

Chapter 6 presents results o status-quo simulations o the Social Budgetrevealing possible expenditure trends in the area o social policy resulting rom thecurrent legislative situation. It includes a set o initial and approximate estimates o a possible minimum package o social protection bene ts: an old age bene t, targetedsocial assistance and child bene t.

Chapter 7 presents overall conclusions.

8/9/2019 Zambia 08

15/268

8/9/2019 Zambia 08

16/268

This chapter is in three parts. he irst part examines demographic trends inZambia and their main determinants. Te second part describes recent economicdevelopments and their e ect on the labour market. he third part presentsbroad aspects o the social environment in Zambia, including poverty incidence,income inequality and selected indicators o human development, preparing the way

or an analysis in greater depth o the living conditions o the poorest and most vulner-

able population groups, in Chapter 2.

Demographic trends

Currently, Zambia has a predominantly young and steadily growing population which, on the sur ace at least, has not changed since independence. Beneath the sur-

ace, however, there have been signi cant changes in the demographic situation associ-ated with recent developments in urban population growth and the unabated spreado HIV/AIDS.Zambia’s population was about 9.8 million inhabitants in 2000 and was projected at 12.1 million or 2007 (CSO, 2006d). According to the census report, 2000 was the irst year in which the male population slightly outnumbered the emale population in Zambia (CSO, 2003a). Figure 1-1 shows the evolution o the popula-tion since 1980.

Zambia’s population is young and growing ast. During the 1990s, it grew onaverage by 2.4 per cent per year, whereas the world population increased by 1.5 per centand other developing countries by 1.7 per cent. However, as able 1-1 shows, there hasbeen a slowdown over the last decades: the population annual growth rate o Zambiadecreased rom 3.1 per cent in the 1970s to 2.4 per cent in the 1990s.

Growth rates were obtained rom UN Population Division estimates (UN, 2006). Te total population insub-Saharan A rica grew at a higher rate than Zambia’s: 2.7 per cent during the 1990s.

1.1

1

8/9/2019 Zambia 08

17/268

Te main reason or this was the sharp decline in the growth rate o the urban population ( rom 6.0 per cent during the 1970s to 1.5 per cent during the 1990s) thatoffset the modest increase in the population growth rate o the rural areas. Tese differentgrowth rates are mirrored in a change in the population structure: the ratio o urban tototal population ell rom 39.9 per cent in 1990 to 34.7 per cent in 2000 (CSO, 2003).

wo actors explain this change: internal migration behaviour and the differentertility rates in rural and urban areas. Te rst actor, internal migration, is driven by

O verview of the dem ographic,

econom ic andsocial environm ent

M ale

Fem ale

Total

1980 1990 2000 2005 (p) 2007 (p)0

2

4

6

8

10

12

14

Figure 1-1. Population by sex, 19 8 0 -20 07 (millions)

(p) m eansprojec tion s

Source: C SO(200 3a; 20 06d )

Table 1-1. Population annual growth rate, 19 70 -20 0 0

Population growth 1969 -198 0 19 8 0 -19 9 0 19 90 -20 0 0

R ural 1.6 2.8 3.0

U rban 6.0 2.6 1.5

Total 3.1 2.7 2.4

Source: C SO (20 03 a)

8/9/2019 Zambia 08

18/268


economic conditions. During the period 1969-90, there was marked rural-to-urbanmigration, in line with the pro-urban economic policies applied during those years (seeChapter 2). Te opposite phenomenon occurred during the 1990s: migration was romurban to rural areas, led by the crisis in the mining and manu acturing sectors.

It is worth mentioning here that while migration is more likely to occur betweendistricts within provinces ( rom rural to urban areas or vice versa in the same province),there are some provinces that are predominantly areas o immigration (such as Lusakaand Central) and others that are a net source o migrants (CSO, 2003d). In addition,as Figure 1-2 shows, migration behaviour in the Copperbelt was notoriously differentduring the 1990s rom previous decades as a consequence o the copper-mining crisis. While in 1980 net migration represented 22.2 per cent o the population, in 2000the province was actually a source o migrants. In that year, the number o emigrantsminus immigrants represented 6.1 per cent o the population. Non-economic reasons

uelling migration are environmental, e.g. droughts that affect the livelihoods o ruralhouseholds (as in Southern Province).

Te second actor, ertility rates, has a more stable pattern than migration. Tetotal ertility rate ( FR) in rural areas is historically higher than that in urban areas

CSO (2003d, p. 28) shows that inter-district migration is an important phenomenon with positive net-migration rates in all districts o the country.

Zam bia

U rban

R ural

Children p er w om an

1980 1990 1992 1996 2000 2002 2005 (p)

0

1

2

3

4

5

6

7

8

1980

1990

2000

C entral C opp erb elt E astern Luapula Lusaka N o rth ern N o rthw estern S ou thern W estern-30

-20

-10

0

10

20

30

40

Percentage of total provincial pop ulation

Figure 1-2 . Contribution of net migrants to population by census year, 19 80, 19 90 and 20 0 0

N ote: ‘N etm igrants’is equalto im m igrantsm inu s em igrants.

Source: C SO(2003a), Figure3.3, p. 26

Figure 1-3. Total fertility rates, 19 80 -20 05

Sou rce: C SO (20 03 a;20 03 c; 20 03 e).

8/9/2019 Zambia 08

19/268

1. O verview of the dem ographic, econom ic and social environm ent 19

(see Figure 1-3). Te in uence o different ertility rates on the population structureis that the rural population grows much aster (with a FR o 6.7 in 2000 versus 4.9in urban areas) and that even without migration to rural areas, its share in the total population will increase.

Figure 1-3 shows that there is a decreasing trend in ertility rates: overall FRor Zambia ell rom 7.2 children per woman in 1980 to 5.7 children per woman in2005. Much o the decline in FR between 1980 and 2000 occurred in urban areas.

FR in rural areas remained almost constant over the same period and consequentlymoderated the rate o decrease o the overall FR.

Nevertheless, better economic conditions during the 2000s than during the previous decade – with annual economic growth around 5 per cent between 2000and 2006 – could explain a new phase o migration into urban areas, which couldchange the distribution o the population and also the path o the overall FR, as mostmigrants are o reproductive age and urban areas are more likely to attract migrants with the lowest ertility rates ( or instance, FR in Lusaka was 4.3 in 2002).

Te rise in the mortality rates is another increasingly important actor. Whileinhabitants o urban areas have easier access to health care services than those in ruralareas, they are also more at risk o being in ected with HIV than those in rural areas.

It is extremely di icult to oresee how this pandemic will evolve, becauseseveral actors must be taken into account. First, it is linked to risky sexual behav-iour, and behavioural changes take time. Second, in addition to the prevalence rates,there are variations in the progression rom HIV in ection to AIDS and rom AIDSto death, access to treatment being a key determinant (UN, 2006). Tird, levelso mother-to-child transmission also vary. However, the impact o HIV/AIDS on

uture population growth and composition is important because its prevalence is

higher among people in their 20s and 30s (the prevalence rate is over 20 per cent inthese age groups). Without this terrible health hazard and the high incidence o other diseases that

are still pandemic in the country (such as malaria), li e expectancy would improve overtime. Offi cial estimates o li e expectancy based on the 2000 census report are shownin able 1-2. Li e expectancy at birth was estimated at 50 years (48 years in rural areasand 54 in urban areas), showing an improvement in relation to the 1990 indicator that was 47 years (CSO, 2003a). However, there is no consensus about this estimate and itseems that li e expectancy could be much lower (see Appendix B).

Nevertheless, Zambia’s population is still young. Its average age is 20.9 years andthe median age is around 17 years. Figure 1-4 depicts the age distribution o the popu-lation oreseen or 2007 (CSO, 2006a).

Table 1-2. Life expectancy at birth, 19 80, 1990 and 20 0 0

Life expectancy 198 0 19 90 20 0 0

Fem ale 53 48 52

M ale 52 46 48

Zam bia 52 47 50

Source: C SO (20 03 a)

8/9/2019 Zambia 08

20/268


Te population pyramid reveals that the emale population o reproductive age(between 15 and 49 years old) is nearly hal o the total emale population. Tis broad-based shape o the pyramid has changed little over the past 20 years: the proportion

o the age group 0-14 years in the population was around 45 per cent in 1990 and in2000. Similarly, child dependency ratios are high, although a decreasing trend has beenobserved, as Figure 1-5 shows.

In line with different ertility patterns, rural areas show higher child depend-ency ratios than urban areas. Furthermore, the urban population is acing an ageing process shown by the decrease in the child dependency ratio rom 104.3 in 1980 to74.0 in 2005 (see able 1-3) and slight increase in old-age dependency ratios. However,the evolution o the prevalence o HIV in the working-age population could affectthese demographic ratios, thus increasing the burden on healthy people aged 15-64supporting dependants and the actual labour orce participation rates o older personsand young people just out o school.

Figure 1-4 . Population pyramid, 2 00 7 Figure 1-5 . Dependency ratios, 19 80 , 19 90,200 0 and 2005

0-4

10 -14

20-24

30-34

40-44

50-54

60-64

70 -74

80-84

90-94

A ge

10%8%6%4%2%0%2%4%6%8%10%

1980 1990 2000 2005 (p)0

20

40

60

80

10 0

12 0D em ograph ic ratios

Total DR

Child DR

Old-age DRM ales Females

D R = d epen den cy ratioC hild D R is the ratio of the p opulation aged 0-14 to the p op ulationaged 15 -64. O ld-age D R is the ratio of the population aged 65 yearsor over to the p opulation aged 15 -64 .Total D R is the sum of the two previous ratios.

Source: C SO (20 03 a) for years 198 0 and 19 90 and C SO (20 06 d)for years 2000 and 20 05

10%8%6%4%2%0%2%4%6%8%10%

Table 1-3. Dependency ratios in rural and urban areas, 198 0, 19 90, 2 0 0 0 and 20 05

198 0 199 0 20 0 0 20 0 5 (p)

Rural areas Child dependency ratio 104.3 90.1 94.0 93.1O ld-age dependency ratio 8.6 7.1 6.8 6.2Total dependency ratio 112 .9 97.3 10 0 .8 9 9.3

U rban areas Child dependency ratio 104.3 82.9 77.7 74.0O ld-age dependency ratio 0.2 1.9 2.5 2.6Total dependency ratio 10 6 .3 8 4 .7 8 0 .3 76 .6

N ote: Ch ild dependency ratio is the ratio of the population aged 0-14 to the population aged 15 -64 . O ld-age dependency ratio is theratio of the population aged 65 years or over to the p opulation aged 15 -64 . Total dependency ratio is the su m of the tw o p revious ratios.

Source: C SO (20 03 a; 20 06 d)

8/9/2019 Zambia 08

21/268


Economic situation and labour market trends

Over the past ew years, Zambia’s economy has improved its per ormance with respect tothe major macroeconomic indicators. Gross Domestic Product (GDP) grew by around

5 per cent per year rom 2000 to 2006. It is important to analyse the main drivers othis growth and the implications o this on the living standards o the population.

Recent economic performance

raditionally, the Zambian economy has depended on copper exports. As Box 1explains, the all in the international price o copper caused a deep economic crisisthat later uelled liberal economic re orms and public expenditure cuts. However, theeconomy is becoming more diversi ed. Whereas in 1994 mining production repre-sented around 17 per cent o real GDP, it accounted or a little more than 9 per cento GDP in 2006.

rade and nancial services increased their participation in real GDP during1994-2006 rom almost 30 to 37 per cent. Te agriculture sector remains important,representing around 14 per cent o real GDP. However, its per ormance is linked toclimate conditions that in recent years have been adverse, with recurrent droughts or

1.2

1.2.1

Box 1. From independence to the new millennium

Follow ing the country’s independence in 1960 Zam bia w as on e of the better-off coun-tries in sub-Saharan Africa, with a stron g extractive industries sector and viab le socialprotection arrangem ents in place including key p ublic service provision s. This situationw as n ot to last, how ever, as international prices of copper declined in the 1970s andw orldw ide inflation fuelled by petroleum price hikes underm ined the econom y. TheG overnm ent nevertheless m aintained tight control over econom ic and social arrange-m ents in the country during this period and it w as not until the m id-1980s that keyreform s to liberalize the econom y w ere initiated. Alongside political liberalization thatgained m om entum in the early 19 90s, econom ic reform s led to severe cu tbacks inpublic provision and em ploym ent, w hich had previously accounted for m ore than half ofnational G D P. The 1990s w ere a p articularly lean decade in w hich the various reform scom m only referred to as Structural Adjustm ent Policies (SAPs) had an adverse im pacton all population groups, but esp ecially on the poorest segm ents. A s a result, poverty

incidence increased substantially w hile the social protection program m es that had pre-viously been in place lost their ab ility to safeguard those m ost likely to suffer from theadverse effects of rapid change.

The unexpected effects of SA Ps w ere not lost on the population and there w eredem ands to rectify the situation. The 1990s saw , therefore for the first tim e, govern-m ent efforts to exam ine m ore closely the im pact of (the liberal) econom ic and socialpolicies, including the initiation of nation al surveys to m easu re poverty and vulner-ability. At the sam e tim e the international com m unity, notably b ilateral don ors, placedpressure on the G overnm ent to act on the dow nw ard-spiralling living standards. Thusthe late 1990s saw nationw ide social protection initiatives launched by the G overn-m ent covering both the form al and the inform al sectors of the econom y. These effortsexten ded into the early years of the new m illennium w ith greater internation al sup-port follow ing recognition that som ething u rgent had to be d one abou t global povertyand failure to develop in countries like Zam bia. Although such efforts have h ad som eim pact on the dow nw ard spiral of living conditions that began in the 1990s, the situ-ation in Zam bia today is still dire.Source: W orld B ank 20 05

8/9/2019 Zambia 08

22/268


Copper exports

Current accoun t

2001 2002 2003 2004 2005 2006 2007p-10 00

-500

0

50 0

1000

1500

2000

2500

3000

3500

oods (see Chapter 2) as well as to Government policies, e.g., Fertilizer Support Pro-gramme and Food Reserve Agency encouraging maize production and a lack o invest-ment in irrigation.

During this decade, high GDP growth rates have been driven by the miningand construction sectors (the latter with growth rates over 10 per cent in recent years).As Figure 1-6 shows, agriculture has given a uctuating per ormance, with short-term

contractions.In contrast, mining production has grown strongly uelled by a avourable inter-national context. Copper prices reached maximum values in May 2006. Furthermore,2006 prices were on average over 80 per cent higher than those o 2005, so that copperexports were almost USD 3,000 million in 2006. In that year the current account othe balance o payments had a surplus that represented 1.6 per cent o nominal GDP(see Figure 1-7). Copper exports are even higher in 2007. However, the impact o risinginternational oil prices – Zambia is a net oil importer – would affect the balance o payments harder during this year than the previous one.

Figure 1-6 . Recent trends in GDP, selected sectors for 20 01-20 0 6

Total

Trade

Agricu lture

M ining

Ann ual % change

2001 2002 2003 2004 2005 2006-5

0

5

10

15

20

Source: C SO (n.d.)

Figure 1-7. Current account of the balance of payments and copper exports, 20 01-20 0 6 (USD millions)

Source:M oFN P (2007)

C opper exports for 20 07 are estim ated on the b asis of exp orts in the first three qu arters of the year. P = provisional

8/9/2019 Zambia 08

23/268


On the demand side, both public and private consumption have increasedin recent years, in a context in which almost all economic activities have expanded.

able 1-4 presents aggregate indicators that signal an improvement in the Zambianeconomy, both in external and internal balances. Chapter 5 presents an assessment ogeneral government accounts.

Tis improved per ormance has occurred in parallel with increased investors’

con dence as shown by growing levels o investment. Investment has grown steadilysince 1996 and represented almost 28 per cent o GDP in 2006 (almost 13 percentage points higher than a decade ago). In addition, there has been an unprecedented growtho short-term in ows in the stock exchange during this decade, as shown in Figure 1-8.

he purchasing power o the kwacha has improved as in lation levels havereduced over time rom a maximum o around 184 per cent in 1993 to 8.2 per cent in2006. However, the value o the Zambian kwacha with respect to the US dollar is stillnot stable, as Figure 1-9 shows.

Tis erratic behaviour o the exchange rate in the short run has implications orthe national budget that depends to a large extent on the availability o donor unds, which is reducing overtime rom 43 per cent in 2003 to 16.6 per cent in 2008, denomi-nated in oreign currency. See Chapter 5.

Table 1-4 . Main economic indicators, 20 01-2 0 0 6

20 01 20 0 2 20 0 3 2 0 0 4 2 0 05 20 0 6

N om inal G D P (U SD billion) 3.7 3.8 4.3 5.5 7.4 10.5

N om inal G D P per capita (U SD ) 367.2 368.4 412.7 508.6 668.7 929.7R eal G D P (per cent grow th) 4.9 3.3 5.1 5.4 5.2 6.2

Inflation rate (end period), per cent 21.4 22.2 32.0 17.5 15.9 8.2

B ank of Zam bia interest rate (D ecem ber) 52.5 34.0 21.3 18.3 17.1 10.7

C entral G overnm ent overall balance (per cent of G D P) a n.a. n.a. -6.6 -3.3 -3.4 -1.1

Current account deficit ( per cent of G D P) b -19.1 -16.3 -15.0 -6.7 -8.4 1.6

External deb t stock (U SD billion ) a n.a. n.a. 6.5 7.1 4.5 0.7

a M oFN P (20 07a). b It includ es grants, ILO calculations based on B oZ (20 07b) and n om inal G D P figures in U SD from U N (20 07b)

Source: B oZ (20 07a), C SO (20 07f), M oFN P (20 07a)

Source:LuSE (20 07)

LuSE all share index

M arket cap italisation

2001 2002 2003 2004 2005 20060

30 0

60 0

90 0

1200

1500

1800

2100

0

50 0

1000

1500

2000

2500

3000

3500

Figure 1-8 . Recent trends in the stock exchange market, 20 01-20 0 6 (USD millions)

8/9/2019 Zambia 08

24/268


Labour market trends

o understand the current reality o the labour market it is necessary to recall whathappened during the 1990s with the structural adjustment packages and liberalization policies. Without the protection o import tariffs, the national industry collapsed and publicly managed companies were privatized. Public expenditure decreased rom 25.7to 7.3 per cent o real GDP during that decade. Consequently, ormal employment elland the labour orce had to migrate to in ormal activities.

In absolute numbers, ormal-sector employment, de ned here as employment

in ormal businesses (a discussion about this and other related concepts is ound inChapter 3) reached its maximum level in 1991: around 545,000. Since then, ormal-sector employment ell rom about 12 to 7 per cent o the working-age populationbetween 1991 and 2005. Tis trend is depicted in Figure 1-10. Te working-age popu-lation is presented as an index, so that the slope o the curve shows the rate o growth.

Looking at the sectoral composition o the ormal employment, it seems that:1. Tere are high uctuations in ormal-sector employment or reasons other than eco-

nomic per ormance. For instance, ormal jobs in construction ell strongly in 2002(by over 80 per cent) despite the act that its production increased by 17 per cent,so that ormal jobs in this sector in 2005 were about 58 per cent o the numberexisting in 2001 (calculations based on CSO, 2007e; n.d.).

2. Tere is no direct link between economic per ormance and ormal employment inthe long run, especially in those sectors that are more labour-intensive, such as tradeand general services. Figure 1-11 shows the evolution o total GDP (in real terms)and ormal employment in these two sectors over the past decade.

3. In the short run, it seems that tertiary sectors are absorbing more employment thanin previous years. In 2005, ormal employment in these sectors accounted or over65 per cent o total employment, while it was 58 per cent in 2004. Tere ore, in 2005,total employment in the ormal sector grew the rst time in more than a decade.

Tese gures were calculated based on UN (2007b). Te CSO (2006e) states that public expenditure was13.1 per cent o real GDP in 1994 and 8 per cent in 2000.

Formal employment is expressed as a percentage o the working age population because there are no avail-able labour orce gures or that period. Still, the trend is clearly negative and the absolute numbers reportedby the CSO (2007e) declined lower and lower afer 1991 (except in 2005).

1.2.2

Source: B oZ (2007 )

Figure 1-9 . Evolution of nominal exchange rate (kwachas per US dollar), 20 05 -20 07

2006

Average 200 6

2007

2005

Jan. Feb. M ar. A pr. M ay Jun. Jul. Aug. Sep. O ct. N ov. D ec.2500

3000

3500

4000

4500

5000

8/9/2019 Zambia 08

25/268


It was not possible to obtain long-term gures or total employment; however,the 2005 labour orce survey (CSO, 2007a) provides in ormation about total employ-ment by type o economic activity. Te de nition o ormal sector used in the labour

orce survey is a little different rom the one used or the long-term series (CSO, 2007e;see Appendix C) so that we do not compare the gures o ormal-sector employment

rom these two different data sources.able 1-5 compares employment in the ormal sector and total employment. In

2005, over 72 per cent o total workers were employed in agriculture, while only 15 percent o ormal-sector workers were registered in this sector. By contrast, workers inservices represented only 7 per cent o total employment.

A special eature o the Zambian labour market rom CSO (2007a) is the highunemployment rate: around 16 per cent o the labour orce (the working-age popula-tion actively looking or a job).

One reason could be the increase in the labour orce participation rate. Indeed,in the 2005 labour orce survey this rate is the highest by comparison with the ones

Sou rce: C SO (20 07e)and ILO calculations basedon U N population projections.

Figure 1-10 . Formal employment and working age population, 1 991-2 0 05

‘Services’include com m unity,social and personal services.

Source: C SO (20 07e ; n.d.)

Form alem p loym ent

W orking age popu lation

1991 1993 1995 1997 1999 2001 2003 2005

0

2

4

6

8

10

12

14

0

20

40

60

80

10 0

12 0

140Index (base 1991 = 100 )% of w orking-age population

Figure 1-11. Economic activity and formal employment, 1996 -20 05

Em ploym entin services

Total G D P

Em ploym entin trade

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

7.6

7.7

7.8

7.9

8.0

8.1Ln of total G D P Ln of form al em ploym ent

10.0

10.5

11.0

11.5

12.0

12.5

8/9/2019 Zambia 08

26/268


recorded in the 2000 census report and in living conditions measurement surveys car-ried out over the past ten years. For instance, or those aged 20-24 the participationrate was 82, 73, 76, 67, and 86 per cent in years 1996, 1998, 2002-3, 2004, and 2005,respectively. Given the different methodologies used over time in such household surveys, we did not consider it use ul to analyse trends in total employment. Moreover,according to the 2000 census the participation rate was 61 per cent or the same cohort(aged 20-24).

However, it is relevant to identi y the different patterns in the distribution olabour orce and employment based on 2005 gures. A detailed analysis about condi-tions o employment is carried out in Chapter 3.

First, urban and rural populations exhibit different eatures in terms o theirlabour orce participation (see Figure 1-12):

(a) Urban dwellers delay their entry into the labour market compared with those inrural areas.

(b) Te gap between male and emale rates during the productive years is signi cantlynarrower in urban areas, compared with that o the rural population.

Table 1-5 . Distribution of employment by economic activity, formal and total employment

Economic activity Formal sector employment Total employment

19 95 20 05 2 0 05

Prim ary sectors 25.0 22.4 73.6of w hich: A griculture 14.2 15.0 72.3

Secondary sectors 14.7 12.5 5.8of w hich: M anufacturing 11.5 9.2 4.0

Tertiary sectors 60.3 65.2 20.6of w hich: Trade 8.5 15.4 10.5of w hich: Services 35.6 39.9 7.0

Figures are exp ressed as percentage of em ploym ent in all econ om ic activities (prim ary, secondary and tertiary sec-tors). Prim ary sectors include agricu lture an d m ining. Secondary sectors include m anufacturing, electricity andconstruction . Tertiary sectors include trad e, transport, finan ces an d services.

Source: C SO (2007a ; 20 07e ; n.d.)

Source:CSO (2007)

Figure 1-12. Labour force participation rates, 2005 (percentage)

Rural

U rban

15 -1920-24

25-2930-34

35-3940-44

45-4950-54

55-5960-64

65 +

0

20

40

60

80

100

15 -1920-24

25-2930-34

35-3940-44

45-4950-54

55-5960-64

A ge group A ge group65 +

0

20

40

60

80

100

M ales Females

8/9/2019 Zambia 08

27/268


While eature (a) can be attributed to the act that school attendance in urbanareas is higher than in rural areas, eature (b) is mainly the result o the different er-tility rates o the two population groups.

Second, younger people have the highest unemployment rates. About 25 and

22 per cent o people in the labour orce aged 15-19 and 20-24, respectively, were unem- ployed in 2005. However, they represent a higher percentage o total employment thanother cohorts, given the young pro le o the Zambian population.

Tird, younger people are more likely to be employed in rural than in urbanareas. Tis is linked to the higher employability o young people in the in ormal agri-cultural sector, as Figure 1-13 shows.

Te light grey shadowed area depicts the age structure o total employment. Telines show that, in effect, younger people are employed in agriculture and older peopleare employed in mining.

Fourth, men are more likely to be employed than women in both rural andurban areas. otal employment rates were 91 per cent or men and 90 per cent or women in rural areas, while these were around 78 per cent or men and 64 per cent or women in urban areas.

Source:CSO (2007a)

Figure 1-13. Distribution of employment by age group in selected economic activities, 20 05

Agricu lture

M ining

M anufacturing

Trade

% of em ployed persons in each activity

15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65+A ge group

TOTAL

0

5

10

15

20

25

Table 1-6 . Percentage distribution of employment by four main groups, 20 05

Female Male U rban Rural Urban Rural

A griculture, forestry and fisheries 26.7 95.2 14.8 89.8M ining and quarrying 0.6 0.1 7.2 0.2M anufacturing 7.0 1.3 13.0 1.7Electricity, gas and w ater 0.4 0.0 1.5 0.3C onstruction 0.6 0.1 6.0 0.7Trade, w holesale and retail distribution 31.9 2.2 23.9 3.9H otels and restaurants 3.1 0.1 2.2 0.2Transport and com m unication 1.6 0.1 10.5 0.5

Finance, insurance and real estate 1.4 0.0 4.4 0.3C om m unity, social and personal services 26.7 1.0 16.6 2.5

Source: C SO (20 07a)

8/9/2019 Zambia 08

28/268


Fifh, employment in rural areas is concentrated in agricultural activities (95 percent o emale workers and 90 per cent o male workers) while employment in urbanareas is more evenly distributed among three economic activities: trade, services andagriculture. able 1-6 shows the distribution o employment by economic activity or

emales and males in urban and rural areas.At this point, it is clear that recurrent shocks in agriculture indeed affect a large part o the population engaged in arming, especially in the in ormal sector.

In terms o employment by type o business, most employed persons were in the private sector, which is also a signal o the less important role that the public sector hasin the overall economy by comparison with the situation in the 1980s. able 1-7 showssome aggregate gures by sex that are the basis or the Social Budget to be presentedin later chapters.

Table 1-7. Indicators of employment by sex, 20 05 (in percentages)

Female Male Total

B asic classificationPaid w orkers (as per cent of em ployed persons) a 46.2 71.8 59.4Paid em ployees (as per cent of paid w orkers) 20.1 30.6 26.6Em ployed in the form al sector(as p er cent of paid w orkers)

12.8 23.3 19.3

Paid w orkers in the form al sector by type of businessC entral G overnm ent 33.3 18.4 22.1Local G overnm ent 6.3 6.9 6.8

Parastatal 4.7 9.1 8.0Private b 55.7 65.6 63.1Total 100.00 100.00 100.00

a P aid w orkers’includ e self-em ployed, em ployers and paid em ployees. b ‘Private’includes private sector, N G O s,churches, international organ ization s and househ olds.

Source: Cross-tabulations based on labour force survey 2005

Paid workers are de ned here as all employed persons excepting unpaid amily workers.able 1-7 shows the limited coverage o the ormal sector in terms o this group: less

than 20 per cent o paid workers are employed in ormal-sector employment and thesituation is more diffi cult or emale workers. In absolute terms, around 25 per cent othese ormal-sector paid workers were emale and 75 per cent were male.

Looking at different businesses, the private sector employs around 63 per cento paid workers in the ormal sector. In the public sector, central Government employs22 per cent; local Government, 7 per cent; and the parastatal sector, 8 per cent.

8/9/2019 Zambia 08

29/268


Household incomes and inequality

Detailed analysis o households’ living conditions is based throughout the reportmostly on the results o 2004 LCMS as LCMS 2006 was not yet available when the

analysis was conducted. Te 2004 dataset which was made available by CSO requiredcleaning procedures which – in some cases – may lead to slightly different results thanthose presented in the report on 2004 LCMS published by CSO.

One can also nd in the report re erences to preliminary results o 2006 LCMS, which have as source aggregate results published by the CSO by the end o 2007 andbeginning o 2008. Te LCMS 2006 dataset was made available to the project teamonly in the rst quarter o 2008 and thus we were not able to make use o it or the purpose o this report.

Composition of incomes

An analysis o the composition o total mean monthly incomes in 2004 revealed thatthe greatest portion o household incomes in Zambia came rom salaried and wageemployment, ollowed in second place by non- arm business activities and in third place by agriculture and livestock-rearing activities. Figure 1-14 presents the averageshare in household incomes or eight main income sources.

Examining the composition o incomes in rural areas, where 60 per cent o theZambian population was to be ound in 2004, reveals that the largest share (44 percent) o household incomes came rom agriculture and livestock rearing. Other

important sources o income or rural households were non- arm business activi-ties, which accounted or 26 per cent o total mean monthly incomes, and to a lesserextent salaried and wage employment, which accounted or 17 per cent o total meanmonthly incomes.

On the other hand, salaried and wage employment was the most importantincome source or urban households, accounting or 46 per cent o total mean monthlyincomes. Non- arm business activities and remittances were also important incomesources or urban households, accounting or 22 per cent and 17 per cent o total meanmonthly incomes, respectively.

1.3

1.3.1

Figure 1-14 . Composition of average total monthly household incomesper adult equivalent: all Zambia (2 0 0 4 )

Source: LCM S 200 4

Crops and self-consum ption (11% )

Livestock an d self-consum ption (4% )

Poultry and self-con sum ption (3% )

N on farm business incom e (23 % )

Salaries and w agesand all other in cashand kind incom e frompaid em ploym ent (36% )

Pensionsand grants (3% )

R em ittances (12% )

O ther incom e (8% )

8/9/2019 Zambia 08

30/268


Figure 1-15 presents the average composition o monthly household incomes per adult equivalent or the whole o Zambia and separately or urban and rural populations.

Figure 1-16 presents the average composition o monthly household incomes ordifferent population groups ranked by average monthly consumption in 2004.

Source: C SO , 20 05

Figure 1-16 . Average share per month of different sources in total household incomeper equivalent adult by expenditure groups

Figure 1-15. Main sources of household income (2004)

Agricu ltural production(including ow n consum ption)

N on-farm ing business

R egular salary

O ther

% of total houshold incom e

0 20 40 60 80 100%

All Zam bia

R ural

U rban

Share of Incom e from crops and self-consum ption

Share of Incom e from livestock and self-consum ption

Share of Incom e from Poultry and self-consum ption

Share of N on Farm business incom e

Share of salaries and w ages and all other in cashand kind incom e from paid em ploym ent

Share of Pensions and grants

Share of Rem ittances incom e

Share of O ther incom e

D ecile 1

D ecile 2

D eciles 3 -5

D eciles 6 and over

Total

D ecile 1

D ecile 2

D eciles 3 -5

D eciles 6 and over

Total

Rural

Urban

0 20 40 60 80 9010 30 50 70 100

8/9/2019 Zambia 08

31/268


Distribution of incomes and inequality

Te mean monthly income or a Zambian household in 2004 was K502,030 whereasthe modal income ranged rom K150,001-K300,000, representing 24 per cent o the

population (CSO 2005, p. 85).he majority o Zambian households (65 per cent) received mean monthlyincomes at or below K450,000, which was below the threshold estimated as necessaryto meet the cost o basic needs.

Findings rom the LCMS 2004 point to high income inequality in the gen-eral population, with a bias towards urban areas. Urban households reported average per capita incomes over twice as large as those o rural households. O 39 per cent othe population in 2004, urban households accounted or 60 per cent o total nationalincomes.

Te Gini coeffi cient or the entire population in 2004 was 0.57. Individuals inthe bottom two national income deciles accounted or only 4 per cent o national percapita incomes, while individuals in the top two income deciles accounted or 45 percent o national per capita incomes.

Te Gini coeffi cients or the urban population were 0.50 and or the rural popu-lation 0.55, indicating greater inequality within the rural population than within theurban population.

However, comparing the income shares o the 20 per cent poorest individuals within urban and rural areas shows the relative income situation o the poorest individuals to have been less avourable within urban areas, where the bottom two incomedeciles accounted or only 1 per cent o average per capita incomes, compared with 8 per cent or the bottom two rural income deciles (CSO 2005, p. 89).

Figure 1-17 shows the number o households in various income groups in2004.

1.3.2

Source:CS O, 2005

Figure 1-17. Distribution of households by income group (2 0 0 4 )

< 50,000 50,001-150 ,00 0

150 ,001-30 0,00 0

30 0,001-450 ,00 0

45 0,001-60 0,00 0

60 0,001-80 0,00 0

> 800,000

Percen t of all househ olds

Total incom e in the last m onth (Kw ach a)

0

5

10

15

20

25

8/9/2019 Zambia 08

32/268

8/9/2019 Zambia 08

33/268


While GPD per capita has increased since the second hal o the 1990s, the HDIhas decreased or has not improved. Tis highlights the need to take additional actionbecause economic growth by itsel does not promote the well-being o all Zambians.In addition, as seen in section 1.3, income inequality is high (0.57) and there ore, the

Government’s role in redistributing wealth and opportunities is paramount.he Government’s task to improve the well-being o its citizens is di icult,given the catastrophic effect o HIV/AIDS on li e expectancy. In addition to havingintrinsic importance or people to promote their goals in li e (e.g., supporting practicalreasoning and thinking skills), education has an instrumental value because it creates possibilities or people to integrate into the labour market with more chances o ndingdecent work and o taking better care o their health by participating in preventioncampaigns, keeping good hygiene habits, and so on.

As discussed in section 1.1, the low li e expectancy in Zambia is largely attribut-able to a high disease burden. Among the leading causes o death in 2006 were malaria,non-pneumonia respiratory in ections (including tuberculosis), diarrhoea, traumas, pneumonia and other HIV/AIDS-related complications (Figure 1-19).

Although malaria is still the leading cause o death in Zambia, there has beena substantial decline in its prevalence in recent years, with incidence rates per 1000 people alling rom 400 in 2000, to just above 200 in 2004 (UNDP 2007, p. 72), andmost recently estimated at 76 (MoFNP 2006, p.98).

However, the same cannot be said or HIV/AIDS, o which the prevalence isstill very high and showing little recent improvement. HIV prevalence was last esti-mated in 2002 at 15.6 per cent o the population in the age group 15-49 (ZDHS 2002).Te concentration o death among those aged 30-44 is due particularly to the highHIV prevalence rate affecting this age group (see Figure 1-20).

Figure 1-21 clearly shows the disproportionate impact o HIV/AIDS on theemale population. Te emale prevalence rate was estimated at 17.8 per cent, compared with 12.9 per cent in the male population. Te urban population was ound to be athigher risk, with a prevalence rate o 23 per cent, compared with about 11 per cent inrural areas.

HIV/AIDS is eroding human capacity on a broad ront. Zambia now loses two-thirds o its trained teachers to HIV/AIDS, and in 2000 two in three agricultural exten-sion workers in the country reported having lost a co-worker in the past year. Te spreado AIDS is a consequence as well as a cause o vulnerability. (UNDP, 2005, p. 22)

Positive trends include declining, but still high, under-5 mortality and in antmortality rates estimated in 2002 at 168 per 1000 live births and 95 per 1000 live

Figure 1-20. Percentage distribution of deaths by age, 2004

Source:CSO (2005)

U rban

Rural

Zam bia

% of population

0

5

10

15

20

25

30

35

< 1 1-4 5-14 15-24 25-29 30-44 45-64 > 64A ge group

8/9/2019 Zambia 08

34/268


births, respectively. Tat is a decline by 15 and 13 per cent, respectively, rom those o1996. Te high rates are mainly caused by the high HIV/AIDS prevalence rates among women in their reproductive years and the consequent impact on high rates o HIV/AIDS mother-to-child transmission. On the other hand, national immunization cov-erage rebounded in 2004 and was estimated at 80 per cent o children under one yearold, a slight improvement rom the rate o 73.5 per cent in 2003. However, coverage isstill considerably lower than a peak record o 86 per cent, achieved in 2001.

Source: C SO (20 03 e)

Figure 1-21. HIV prevalence rates by sex and age group, 2002

M ale

Fem ale

% of people in each age group

A ge group15-19 20-24 25-29 30-34 35-39 40-44 45-49

0

5

10

15

20

25

30

Table 1-8 . Trends in health indicators

2 0 01 20 02 20 0 3 20 0 4 20 05 20 0 6

Supervised deliveries(per cent) 44 49 55 61 62 48

Fully im m unized children under 1 year(per cent) 86 76 74 80 90 96

U nderw eight prevalence(per cent w eight) 23 22 21 17 16 16

M alaria (incidence per 100 0 cases) n.a. n.a. 337.5 214.4 200.1 76.1

Sou rce: M oFN P (A nnu al econom ic reports)

8/9/2019 Zambia 08

35/268

8/9/2019 Zambia 08

36/268

Following the overview o the demographic, social and economic environments presented earlier, we examine poverty and vulnerability as two related and widely prevalent phenomena that characterize living conditions in Zambia, pointingout the need or greater social protection coverage to address the needs not only o themajority o the population that is offi cially classi ed as poor, but also o the poorestand most vulnerable population groups.

Poverty in Zambia is longstanding and its eradication has been a major ocus onational development efforts in the post-independence era. As a low-income country,Zambia aces numerous development challenges, e.g., historically low levels o humanand material capital, inconsistent periods o economic growth and severe povertyimplying an inability o the majority o the population to sustain basic needs. While poverty was known to be extensive, it is only since the early 1990s that systematicefforts have been made to measure it. As a result o a series o national household surveys carried out since 1991, it is now possible to report on the extent o poverty in the population across the years. Such an examination reveals that poverty has remainedextensive, with most o the population subsisting below what is considered to be theminimum standard o living.

An important aspect o contemporary poverty measurements in Zambia is thatthe actual living conditions o the population can be compared against objectivelyde ned minimum living standards in the orm o offi cial poverty lines. wo povertylines are currently in use: the ood poverty line, which marks a minimum level o

ood consumption and below which level households are characterized as extremely poor; and the basic needs poverty line, which marks the minimum level o consump-tion necessary to meet all basic needs and which demarcates the overall poor. akinginto account these distinctions, the Government estimated in 2006 that 51 per cent oZambians were extremely poor and 64 per cent were overall poor.

By contrast, assessing vulnerability implies taking a more dynamic view o well-being, i.e., one which takes into account not only current livelihood outcomes butalso the key risks aced by individuals and households in their daily lives, and theirability to deal with them when they occur. Like poverty, vulnerability is no strangerto Zambia. A large body o evidence points to the act that most o the populationis vulnerable to one or another poverty-threatening event. Such events are likely to

2

8/9/2019 Zambia 08

37/268

affect any household, or example, the death o a productive member o the householdbecause o HIV/AIDS; or may affect only certain population groups, or example theloss o livestock to disease which is a risk aced particularly by rural livestock-rearinghouseholds. Poverty and vulnerability in Zambia are closely related. On the one hand, poverty itsel re ects a orm o vulnerability, with the poor presumed to be especiallylikely to suffer in the event o ‘shocks’ due to their presumed limited ability to protect

themselves. On the other hand, vulnerable groups such as orphans, elderly widows and persons living with HIV/AIDS are more likely than others to become poor, as theirstatus almost always has some negative impact on their livelihoods and their ability tomeet basic needs.

Because o their real and potential impact on wel are, poverty and vulnerabilityboth imply states against which some orm o social protection is required. Tis couldbe in terms o prevention, i negative outcomes are avoidable; reducing risks; or mitiga-tion, i negative outcomes are entrenched. As most o the population is poor, ensuringa minimum o social protection or all could play a major role in poverty reduction. Atthe same time, there is no doubt that even in the poor majority some are poorer andmore vulnerable and especially in need o some orm o public social assistance. Teonus in this case is to identi y such needy groups and determine how they can mosteffectively be targeted.

Although this is certainly not the rst attempt to examine poverty and vulner-ability in Zambia, the ollowing analysis is distinguished in two ways. First, it makesuse o the most up-to-date empirical data on living conditions in the country, notablydata arising rom the Living Conditions Monitoring Surveys series, complemented byevidence rom a number o secondary sources. Second, the analysis responds to thechallenge o how to extend coverage o social protection in a context o widespread poverty and vulnerability by identi ying some o the characteristics which distinguishthe neediest members o society rom the erstwhile majority poor, and which could

acilitate targeting o the ormer.

Although systematic investigations into conditions o poverty and vulnerability have not been the norm inZambia, a recent and growing preoccupation with such analyses is evident, especially with the advent o thenational Poverty Reduction Strategy Papers (PRSPs). Te World Bank’s Poverty and Vulnerability Assess-ment (2005) marks one particularly recent and comprehensive attempt.

The poor and the poorest:

Living conditionsand vulnerability

8/9/2019 Zambia 08

38/268


Te rest o the chapter is organized as ollows:

☐ Part 1 discusses current estimates o the extent o poverty in Zambia and looks at past trends.

☐ Part 2 examines the living conditions experienced by the poor, re erring to variousindicators o well-being such as the ability to meet basic ood needs, education andhealth status, and access to housing and sa e drinking water.

☐ Part 3 examines vulnerability in Zambia rom the perspective o major livelihoodthreats and the individual and household characteristics identi ed with a greatersusceptibility to poverty.

☐ Part 4 identi es areas where public social protection interventions could be mosteffective in protecting the livelihoods o the poorest and most vulnerable.

Contemporary poverty

Te Government o Zambia estimated that 64 per cent o the population were poorin 2006, including 51 per cent o the population classi ed as extremely poor (livingbelow the ood poverty line), and a urther 13 per cent o the population classi ed asmoderately poor living between the core ( ood) and overall (basic needs) poverty line(CSO 2007)). Although poverty headcount ratios such as those indicated above illus-trate overall poverty incidence, they only scratch the sur ace when it comes to showing

the extent and nature o poverty in a country. Tis holds true in Zambia, where theincidence o poverty depends to a large extent on which particular area or populationgroup is taken into account or the measurement, and where the poor themselves re ecta wide range o wel are outcomes. In this section, we examine urther aspects o pov-erty in Zambia, in particular the depth and severity o poverty alongside its incidencein the population, and compare poverty measurements or different population groups. We also examine poverty estimates dating rom the early 1990s until the present, high-lighting patterns that have emerged over this period.

Poverty headcounts or the provinces in 2006 were calculated with re erenceto the same poverty lines and data sources used to calculate the national headcounts.Examining the headcounts in each province, it can be seen that in all but two provinces poverty rates exceeded 50 per cent, implying that in the majority o the provinces the populations were predominantly poor. However, even among such provinces there weresome who were poorer than others: while the poor constituted over three-quarters o the population in Western, Northern and Eastern Provinces, they made up just above two-thirds o the population in Central, North-Western, Luapula and Southern Provinces.

Te remaining two provinces, on the other hand, reported ar lower povertyheadcounts. In Lusaka Province just below a third o the population were ound to be poor and around one-sixth o the population extremely poor. Likewise poverty rates

Regarding the decomposable poverty measures pioneered by Foster, Greer and Torbecke (1984): deptho poverty re ers to the ‘poverty gap’: the average distance, measured in monetary terms, between the meanconsumption o the poor rom the poverty line. Severity o poverty re ers to the square o the poverty gap; which, in contrast to the other two measures, is sensitive to changes in income distribution (or inequality)among the poor and gives greater weight to an increase (or decrease) in wel are among those at the lowest endo the income distribution scale in re ecting a reduction (or augmentation) o poverty. For more on the useand interpretation o these measures see chapter 3 o the UN Handbook on Po erty Statistics (2006).

2.1

8/9/2019 Zambia 08

39/268

2. The poor and the poorest: Living conditions and vulnerability 39

were comparatively low in Copperbelt Province, where less than hal o the population were ound to be poor and just above a quarter extremely poor. Figure 2-1 presents thesize and distribution o the poor population in each o the country’s nine provinces. Ascan be seen in the gure, in the provinces where the poor were in the majority, extremely poor persons accounted or the majority o those who were poor, representing close to80 per cent o the overall poverty headcount in the seven poorest provinces.

Te rural/urban character o an area was ound to have a signi cant bearing on poverty. O those identi ed as poor in 2006, the vast majority (82 per cent) were esti-mated to be living in rural areas, where poverty was both more prevalent and intense by

comparison with urban areas. Eighty per cent o the rural population were poor in 2006compared with just around a third o the urban population. In addition the character o poverty in urban areas was such that it was much less intense than in the rural areas; in2006 the urban poor experienced an average short all or poverty ‘gap’ o 13 per cent othe overall poverty line, compared with an average poverty gap o 45 per cent in the caseo the rural poor. As re ected in the ratio o their squared poverty gaps, which take intoaccount not only the average o consumption across the respective poor groups but alsothe distribution o incomes within them, poverty was also ound to have been more than

our times as ‘severe’ in the rural areas than in the urban areas. Comparing the extent oextreme poverty across rural and urban areas based on the lower or ood poverty line sim-ilarly reveals that the rural population were particularly disadvantaged compared withtheir urban counterparts: those who were extremely poor made up around two-thirds othe rural-based population compared with just one- fh o the urban population.

Te large differences in poverty rates between rural and urban areas are alsoobservable in the provincial poverty headcounts presented earlier. All seven o the pre-dominantly poor provinces reviewed are at the same time predominantly rural, withthe share o their rural populations ranging rom 78 per cent in Central and SouthernProvinces to 92 per cent in Eastern Province in 2006 (CSO 2007). Te two least-poor provinces, Lusaka and Copperbelt, on the other hand were the most urbanized, with

Te squared poverty gap is de ned as:

Pα = 2 where Pα =1

/ N ( Z – Yi

/ Z)α

Where N = the total population in a group o interest; Z = the poverty line (Moderate); n = the number oindividuals below the poverty line; Yi = the adult equivalent expenditure; a = the poverty aversion parameter which takes on values o 0,1,2; and Z – Yi = the poverty gap. Pα = 2 was ound to have been 0.30 or the rural population and 0.07 or the urban population. See also ootnote 2.

Figure 2-1. Distribution of the poor within Zambia’s provinces (2 0 0 6 ) (thousands)

Sources:CSO 2007

M oderately poor

Extrem ely poor

N on poor

C entral C opperbelt Eastern Luapula Lusaka N orthern N orth-W estern

S ou th ern W estern0

20 0

40 0

60 0

80 0

1000

12001400

1600

1800

2000

8/9/2019 Zambia 08

40/268


the share o their urban populations at 85 per cent and 79 per cent, respectively. Figure2-2 shows the distribution o major urban centres (towns) in Zambia.

Rural and urban areas are distinguished not only by their different poverty ratesbut also by their different population densities, which has potential implications or the

relative prevalence o poverty in one or other area. As might be expected, urban areasre ect ar higher population densities compared with rural areas. Provincial populationdensities calculated on the basis o 2006 LCMS data indicate that the more urbanized provinces in Zambia have much higher population densities than less urbanized prov-inces. Whereas the average number o persons per square kilometre was 75 in Lusakaand 57 in Copperbelt, the averages were much lower in the remaining seven provinces,ranging rom 23 persons per square kilometre in Eastern Province to 6 persons persquare kilometre in North-Western Province.

It might thus be questioned whether the higher population densities experiencedin the relatively urbanized provinces compensated or their lower rates o poverty suchthat, other things being constant, one’s likelihood o being poor would be the same inany one o Zambia’s nine provinces. We set out to answer this question by comparingthe distribution o the total poor across the provinces against the corresponding distri-bution o the total population.

able 2-1 shows both the distribution o the poor in Zambia by province andthe distribution o the population by province. Comparing the third and fh columnsin the table, which show the percentage shares o the whole population and o all the poor respectively in each province, it can be seen that in all but two provinces the percentage shares o the poor exceeded the percentage shares o the population. Notsurprisingly, the two provinces that were the exception, and where the population per-centage shares o the poor proved to be less than the corresponding percentage shares

o the population by rather large margins, were the same ones that were predominantlyurban: Lusaka and Copperbelt.Te above ndings imply that even though the high population densities seen

in Lusaka and Copperbelt Provinces may have resulted in higher concentrations o the poor within them, their poor populations were still proportionately smaller than wasto be expected given the distribution o the overall population. Consequently, mostZambians based in the other seven, predominantly rural provinces had a signi cantlyhigher chance o being poor.

Table 2-1. Distribution of the population and of the poor by province (2 0 06 )

Province N umber of persons Percentage share N umber of poor Percentage share

C entral 1 221 667 10 879 255 12

C opperbelt 1 782 799 15 748 481 10

Eastern 1 604 257 14 1 267 363 17

Luapula 929 310 8 678 396 9

Lusaka 1 640 853 14 475 476 6

N orthern 1 482 946 13 1 156 674 15

N orth-W estern 709 095 6 507 595 7

Southern 1 453 112 12 1 058 262 14

W estern 887 183 8 740 858 10Total 11 711 22 3 10 0 7 512 3 62 10 0

Sources: C SO 20 01; C SO 20 07

8/9/2019 Zambia 08

41/268

2. The poor and the poorest: Living conditions and vulnerability 41

Trends in poverty: 1991-2006

We now turn to examine how poverty in Zambia has developed over the last feen years based on a set o nationally representative household sample surveys conducted

between 1991 and 2006. In principle, such trend analysis o poverty requires that onlyndings rom surveys similar in design be compared, so that any differences in theobserved incidence o poverty between particular points in time ollow as a result oactual changes in the wel are o the population rather than rom changes in the surveymethodology. aking the series o surveys carried out in Zambia since 1991, six o theseven meet this criterion and are consequently used to examine trends in poverty inZambia. In particular, two o the surveys the LCMS II conducted in 1998, and theLCMS V conducted in 2006, have been identi ed as providing the best current basis

or assessing poverty trends over a short-term period o eight years. Box 2 presents anoverview o all the surveys.

Figure 2-2 shows the respective shares o the national, rural and urban populationsliving below the overall poverty line between 1991 and 2006. Examining the middlegraph, which traces changes in the national poverty headcount ratio, it appears that

Poverty estimates based on the Integrated Household Budget Survey 2002/2003 (or LCMS III) are notexamined here because o the different way in which that survey was conducted compared with the othersurveys. In particular, the 2002-2003 survey was based on a panel design, involving continuous data collec-tion over a twelve-month period with households recording their expenditures and other consumption datain monthly diaries, in contrast to the oth

zambia 08

Documents