Social Relations of Poverty:

A Case-Study from Owambo,
Namibia

Inge Tvedten
Selma Nangulah

R 1999: 5

Recent Reports

R 1998: 3

R 1998: 4

R 1998: 5

R 1998: 6

R 1998: 7

R 1999: L

R 1999: 2

R 1999: 3

R 1999: 4

TJORE, Gro

Protective strategi es in the 1990s: A review of the policy discourses in UNCR and the
executive commttee. Bergen, 1998. (Price NOK 90 + postage)
DIA, Mona
Boikott til hvilken pris? Barns situasjon i Irak og Burundi. Rapport for Redd Barna. Bergen, 1998

(Price NOK 90 + postage)
NORDAS, Hildegunn Kyvik
Macmood, a macroeconomic model for the Tanzanian economy. Bergen, 1998 (Price NOK 50 +
postage )
NORDAS, Hildegunn Kyvik
Criteria for general budget support and general sector support. Report commssioned by the
Norwegian Ministr of Foreign Affairs. Bergen, 1998 (Price NOK 50 + postage)
BJERKLAND, Krstin M.
Komipsjon. En studie av skatteadministrasjonen i Tanzania. Bergen, 1998, 102

pp. (Price NOK 90 + postage)
NORDAS, Hildegunn Kyvik
The impact of the financial and economic crisis in Asia on Norway's major
development partners. Bergen, 1999,29 pp. (Price NOK 50 + postage)
NORDAS, Hildegunn Kyvik med bidrag av Inge Tvedten og Ame Wiig
Effekter i mottakerlandene av norske petroleumsinvesteringer med hovedvekt på
Angola. Bergen, 1999,39 pp. (Price NOK 50 + postage)
FJELDSTAD, Odd-Helge and Joseph Semboja
Local governent taxation and tax administration in Tanzania. Bergen, 1999, 79 pp.
(Price NOK 90 + postage)

SUMAILA, U ssif Rashid et aL.
Marine protected araeas and managing fished ecosystems. Bergen, 1999, 53

A complete list of publications and Anual Report available free of charge

For priced publications:

Surfacemail (B-economique) free with prepaid orders. For airmai1 (A-prioritaire) outside the Nordiccountries add
20%

Four easy ways to pay:

Cheque, issued in Norwegian kroner
Post offce giro, paid by International Giro: 0808 5352661
SWIFT: DNBANOBB, Den norske Bank no: 5201.05.42308
Credit card: VISA only

Order from:

Chr. Michelsen Institute
P.O. Box 6033 Postterrnalen, N-5892 Bergen, Norway
Fax: + 4755 5741 66 Phone: + 4755574000 E-mail:
cmi(qamadeus.cmi .no

Summary
Evidence from a number of quantitative studies on poverty in

Namibia suggests that poverty is widespread, albeit unevenly

distributed. The relatively high GDP per capita conceals wide

income differentials by geographicallocation, occupation, age

and gender. This study addresses urban paverty and the
importance of social relations and networks ('social capita!')

in the coping strategies of the poor - a phenomenon which
has received little attention in poverty research. On the basis

of case studies of four shantytowns in Oshakati and two rural

villages with extensive links to urban areas, it is argued that

the poorest sections of the urban shantytowns are

systematically marginalised and excluded from political,
economic and social processes. Horizontal transfers and social

capital play a raJe in spreading risk and reducing

vulnerability - though more sa for the somewhat better off
than for the poorest in a setting where social relations
increasingly have to be filled with material content to be
sustainable. People apparently trapped in poverty and
vulnerability are those for whom the shantytown encapsulates

their univers e both in socio-economic and cultural terms.
While interventions with a view to developing social capital

in the form of networks, associations, etc. are helpful, any
serious attempt to reduce urban poverty in Namibia requires

vertical transfers through an active public policy commitment

by the government.

The study has been conducted under the Agreement of Co-

operation between the Social Sciences Division of the
University of Namibia (SSDIUNAM), the Namibia Economic

Policy Research Unit (NEPRU) and the Chr. Michelsen

Institute.

Inge Tvedten, Chr. Michelsen Institute, Bergen, Norway

Selrna Nangulah, Social Sciences Division, University of

Namibia

ISSN 0805-505X
ISBN 82-90584-46-6

Social Relations of Poverty:

A Case-Study from Owambo, Namibia

Inge Tvedten and Selma Nangulah

R 1999: 5

lIi

Chr. Michelsen Institute Development Stud/es and Human R/ghts

CMI Reports

This series can be ordered from:

Chr. Michelsen Institute

P.O. Box 6033 Postterminalen,

N-5892 Bergen, Norway

Tel: + 47 55 57 40 00

Fax: + 47 55 57 41 66

E-mail: cmi(¡amadeus.cmi.no

WebjURL:httpj jwww.cmi.no

Price: NOK 50 + postage

ISSN 0805-505X

ISBN 82-90584-46-6

Indexing terms

Poverty
Urban areas

Rural areas
Social aspects
Namibia

(g Chr. Michelsen Institute 1999

Table of Contents

1. INTRODUCTION ....... ............................ ...... ................... ................. ......1

2. POVERTY IN NAMIBIA ......................................................................... 4

3. URBAN RELATIONS OF POVERTY .........................................10

3.1 URBAN POVERTY.............................................................................. .10
3.2 SOCIAL RELATIONS OF POVERTY ............................................................12

3.3 MARGINALlZATION AND SOCIAL EXCLUSION ...............................................13

4. POVERTY IN OSHAKATI .....................................................................15

4.1 HISTORICAL AND STRUCTURAL CAUSES OF POVERTY ....................................16

4.2 INCIDENCE AND CHARACTERISTICS OF POVERTY ..........................................18

4.2.1 The Population ............... .......... ........ ..................... ............ .....20
4.2.2 Socio-economic Characteristics ...............................................22
4.2.3 Health and Nutrition ...............................................................25

4.3 SOCIAL RELATIONS OF POVERTY ............................................................ 26

4.3.1 The Household .......... ..... ...................................... .................. 29
4.3.2 The Extended Family............................................................... 31
4.3.3 Neighbors and Friends ............................................................34
4.3.4 Urban Relations .... ................. ............ ........................... .........36
4.3.5 Associations ............. ......................... ...................... ......... .....38
4.3.6 Marginalization and Social Exclusion ........................................ 41

5. POVERTY IN RURAL OWAMBO ........................................................... 45

5.1 THE VILLAGES OF OMPUNDJA AND ONIIHENDE............................................ 46
5.2 RELATIONS AND NETWORKS .................................................................49

5.2.1 Rural-Urban Relations............................................................. 49
5.2.2 Rural-Rural Relations ..............................................................51

6. CONCLUSIONS .................................. .............. ............ ...................... 55

REFERENCES .... ... ..... .......... ..... ........ .... ............ ..... .... .... ............... .... .....57

C M I

1. Introduction

This study has been conducted under the Agreement of Cooperation between
the Social Sciences Division at the University of Namibia (SSD/UNAM), the
Namibia Economic Policy Research Unit (NEPRU) and the Chr. Michelsen
Institute (CMI) in Bergen, Norway. The Agreement was established in 1996,
and is currently in its second phase (1998-1999). The main objectives of the
Agreement are:

. To promote the development of professional competence at NEPRU,

SSD/UNAM and CMI in the fields of development and economic policy
research.

. To improve the quality of library services and skils at the three
institutions.

. To generate and disseminate reputable research in both published and

consultancy reports on topics of national and regional significance.

As partial fulfillment of these objectives, three joint research projects have
been carried out. T opics for research were selected through a process where
research needs identified by government, NGOs and donors, as well as the
interests and qualifications of the research institutions themselves, were taken
into consideration.1

The topics singled out were poverty and migration. Project 1 ("National
Migration Study") is a national survey of the causes, effects and trends of
migration, and Project 2 ("Macro-Economic Policies and Poverty in

Namibia") is a study of the relation between trade policies and poverty. The
current Project 3 ("SocIal Relations of Poverty") analyzes the role of socIal
relations and networks in the survival strategies of the poor through a case-
study from Owambo in Namibia. The study gives special attention to urban
poverty and the role of urban-rural links.

While there is a relatively limited number of studies on migration in Namibia
(Melber 1996; Pendleton and Frayne 1998; Tvedten and Pomuti 1998), there
has been a large number of studies on various aspects of poverty to which we
will return below. The studies reveal that poverty is pervasive in the country,
and that income discrepancies are considerable. They tend to concentrate on
issues of distribution and characteristics of poverty, using quantitative data on
income and consumption and households and individuals as units of analysis.

The selection was made in a seminar held in Windhoek in February 1997, with
participation from government, NGOs and donors. As point of departure for discussions
at the seminar, five jointly written papefS on topics of common interest were presented
and subsequently published (NEPRU 1998).

1

CM I

The current study discusses and analyzes qualitative aspects of poverty and
urban-rural links. Our hypothesis is that social capital in the form of social
relations and networks generally plays an important ro le in the survival

strategies of the poor, but that the poorest and most vulnerable in urban and
rural areas are characterized by limited sets of such networks both in time and
space. This further exacerbates their political, economic and socio-cultural
marginalization in Namibian society.

The characteristics of social relations and networks among the poor have
important implications for our understanding of urban as well as rural
poverty, and hence for policies of poverty alleviation. On the one hand
support to the development of social capital (in the form of networks or
associations) is important to alleviate the most extreme expressions of poverty
in Namibia by "helping the poor to help themselves". On the other hand, the
marginalization of the very poorest highlights the limitations of privately
negotiated informal transfers (horizontal transfers) and underlines the need for
publidy funded formal transfer programs (vertical transfers) to do something
substantial about poverty in the country.

The study was carried out in the four largest shanty-areas areas in the town of
Oshakati in Owambo, with a special focus on the richest and poorest areas
respectively in terms of income. For comparative purposes, and to ascertain
the perceptions of rural-urban relations from the rural point of view,

fieldwork was als o carried out in one vilage dose to Oshakati (Ompundja in
the Oshana region) and one vilage further away from urban areas (Oniihende
in the Ohangwena region).

The study has been carried out through a combination of quantitative and
qualitative methodologies. A baseline survey from the four informal settlement
areas from 1994 has been used as a point of departure for the urban part of
the study (Tvedten and Pomuti 1994). A more limited survey was carried out
in the two rural vilages, using the relevant questions from the urban survey.
Qualitative methodologies such as structured interviews, group discussions,
case-studies and participant observation were used both in the urban and rural
areas of study.

In the following we will first brie fly present relevant data on poverty and
migration in Namibia as these appear in existing studies, and give an outline
of our own theoretical point of departure (Section 2). We wil then move on to
analyze poverty, vulnerability and the role of relations and networks in the
urban setting (Section 3) and the rural settings (Section 4) respectively. In the
conc1uding section (Section 5), the potential and limitations of support to the

development of social capital as a means to alleviate poverty will be discussed.

We would like to thank Akiser Pomuti from SSD for his invaluable
contribution during the fieldwork in Oshakati, and Martha Naanda, Frieda
Iigonda and Gabriel Daniel from SSD for their contributions during the
fieldwork in the rural areas. We would also like to thank Dr. Stephen
Devereux from the Institute of Development Studies (IDS) for valuable

2

C M I

comments to a draft version of this report.2 Finally, om thanks go to thank the
Norwegian Agency for Development Cooperation (NORAD) for their
financIal support to the Agreement of Cooperation, and Mr. Olav Myklebust
and Dr. Gustav Klem from the Norwegian Embassy/ NORAD Representation
in Windhoek for their consistent support and encouragement.

This report and the two other reports mentioned above were presented at the seminar

"Poverty and Migration in Namibia" in Bergen, Norway 19-20 August 1999.

3

C M I

2. Poverty in Namibia

There is a large number of studies on poverty in Namibia, and evidence of the
distribution and characteristics of poverty is persuasive. The evidence points in
the direction of a relatively high GDP per capita, concealing severe income
differentials along lines of geographicallocation, occupation, age and gender.
The general political context of the following analysis is thus one of a country
with deep and widespread poverty, but also with a great deal of wealth
concentrated on relatively few hands. The combination of poverty and
inequality presents an almost unique opportunity in the African con text for

redistribution to alleviate poverty. As the following outline of poverty in
Namibia will show, however, real redistribution has so far been limited.

There are three main types of definitions of poverty used in studies from
Namibia. Absolute poverty defines a cut-off point for poverty (based on
income, consumption, expenditure or some other proxy) below which people
are considered to be poor. The justification is that income is highly correlated
with other causes of poverty and is a predictor of future problems of

deprivation. Relative poverty describes an individual's or group's level of
wealth in relation to other individual or groups in society. Relative poverty is

closely related to levels of inequality, and hence highly relevant in the case of
Namibia. Finally, definitions based on social indicators include non-income
determinants of welfare, such as domestic production, potable water,

adequate housing and basic social services including health care, sanitation,
primary education and public services.3

The most comprehensive study on poverty in Namibia is Devereux et aL. 1996
(Namibia Poverty Profile. A Report for Sida. Windhoek: SSD/University of
Namibia). The study is base d on secondary sources and includes discussions of
i) conceptual issues, ii) the incidence of poverty in Namibia, iii) origins and
causes of poverty in Namibia, and iii) anti-poverty measures in Namibia since
independence. In addition to the definitions outlined above, it also relates to
the issue of vulnerability, arguing that although poverty is typically discussed
in terms of static levels of living, variability of income is as serious for the
poor as a low level of income.

The most comprehensive source of primary data is The National Household
and Expenditure Survey (CSO 1996) which builds on the 1991 Census (CSO
1994). Households are classified into three groups according to their level of
economic resources, which is estimated from weighted data on reported total
household consumption giving a Standardized Consumption Level (SCL) of

Composite poverty indices combine several weighted indicators, such as per capita GDP,
life expectancy at birth, and literacy rates (Human Development Index), public spending
on social services, immunization and fertility rates (Social Indicators of Development
Index) and life expectancy at ane year, adult literacy and infant mortality ( Physical
Quality of Life Index).

4

C M I

N$ 7,200. According to the Survey, 53 percent or 129,758 of all Namibian
households are classified as poor by having a CSL of less than that amount.

The CSO study also measures poverty according to "food consumption ratio"
or food consumption as proportion of total household consumption.

According to this definition, a household that spends 60-79 percent of its total
consumption on food is "poor", and a household that spends 80-100 percent
of its total consumption on food is "very poor". As seen from Table 1, 10
percent of the households in Namibia are classified as very poor, 30 percent
as poor and 60 percent as not poor on the basis of this definition. 4

Table 1. Poverty by rood consumption ratio (Percentagesl 1995)

Food consumption to Classifica tion Proportion of households
total consumption
O-59 Not poor 60

60 - 79 Poor 30
80 - 100 Very poor 10

Source: Devereux et aL. 1996

Regarding relative poverty and income distribution, available data show that
Namibia is one of the most unequal societies in the world (CSO 1996; UNDP
1998). This is illustrated by a Genie Coefficient of 0.70, measuring the
inequality of income distribution among the Namibian population. The richest
10 percent of the society receive 65 percent of income, and the remaining 90
percent share among themselves only 35 percent of the national income. Put in
other words, half of Namibia's population survives on approximately 10

percent of the average income, while approximately 5 percent enjoys incomes
that are five times the average.

The extreme inc om e inequalities imply considerable differences in human
development as measured by UNDP's Human Development Index (combining
income with life expectancy at birth and level of education). Table 2 shows the
Namibian HDI for the last four years, together with Namibia's global ranking
by income and by development. Whereas Namibia is defined as a middle-
income country in terms of GDP per capita, there is a considerable
discrepancy between average income and HDI index.5 Key human
development indicators for Namibia are given in Table 3.

The improvement in HDI ranking in 1998 is solely attributable to updating on
information of literacy (adjusted from 40% to 76%). There is no improvement in average
incame, and estimated life expectancy at birth has been reduced from 58.8 years in 1995 to
55.8 years in 1998 due to the prevalence of HIV/AIDS (UNDP 1998).

5

C M I

Table 2. Comparison ofGDP and HDI for Namibia (1995-1998)

Ranking 1995 1996 1997 1998

Global HDI 0.611 0.573 0.570 0.644
GDP per capita ranking 77 79 83 85
HD I ranking 108 116 118 107
GDP rank minus HDI rank -31 -37 -35 -22

Source: UNDP 1998.

Table 3. Human Development Indicators, Namibia (1998)

Indicator Value

Life expectancy at birth (years) 61.0
Ad ult literacy (%) 81
School enrolment (%) 85
Income (N$) 3.608

Source: UNDP 1998.

UNDP has als o developed a Human Poverty Index (HPI) for Namibia (UNDP
1998). The HPI is a measure of deprivation, excluding inc om e and combining
dimensions of :

· Longevity (the proportion of the population which is expected to die
before reaching the age of 40 years);

· Knowledge (the proportion of the population above 15 years of age unable
to read or write in any language);

· Standard of living (percentage of population without access to safe water
and health services, and the percentage of malnourished children);

· Proportion of households which uses more than 80 percent of their income

on food (i.e. the definition used in the 1991 Census).

The HPI should be interpreted intuitively as the proportion of the population
suffering from poverty and deprivation.

Figures on average income and poverty in Namibia show significant
differences between regions, urban and rural areas, gender and language
groups, which imply a clustering of poverty in the country. There is a strong
correlation between average income and the poverty as measured by the HPI,
even though there are exceptions. Tables 4-7 below sum up these differences.

The figures reveal a strong geographical concentration of low income and
poverty in the northern regions (Ohangwena, Okavango, Omusati, Oshikoto,
Caprivi, Oshana), with the central regions (Omaheke, Kunene, Erongo,

Khomas and Otjozondjupa) and the southern regions (Hardap and Karas)
generally being better off both in terms of average income and poverty.

6

C M I

Table 4. Average Income and Poverty by Region (1998)

Region Income HPI Region Income HPI
(N$) (%) (N$) (%)

O hangwena 1.070 31,8 Oshana 1.922 22,0
Omaheke 3.944 30,5 Otjozondjupa 3.659 21,0
Okavango 1. 763 27,2 Hardap 5.945 19,1
Omusati 1.452 26,6 Karas 6.655 16,0
Kunene 2.203 24,8 Erongo 5.423 11,2
Oshikoto 1.6 80 24,9 Khomas 11.359 9,6
Caprivi 1.598 25,0

Source: UNDP 1998

Table 5. Average income and poverty by language-group

Language- Income HPI Region Income HPI
group (N$) (%) (N$) (%)
Afrikaans 13.995 9,3 Oshiwambo 1. 707 29,4
Caprivi/Lozi 1.692 23,1 Otjiherero 3.077 24.6
English 21.708 7.0 Rukavango 1.652 31,4
German 30.459 9,2 San 1.315 58,1
N ama/D amara 2.404 23.7 T swana 5.326 17,2

Source: UNDP 1998

The same clustering is evident when focusing on ethno-linguistic groups.
Groups from the north (Caprivi/Lozi, Rukavango and Oshiwambo) generally
have lower average income and are poorer than groups from the central and
southern parts of the country. Having said this, the San stand out as having by
far the highest incidence of poverty and the lowest average income. At the
other end of the scale all three groups of European descent (Afrikaans, English
and German) have high average income and low incidents of poverty, but
there are also differences between them particularly in terms of average

income.

Another distinction in inc om e and poverty is that between rural and urban
areas. As shown in Table 6, average income is considerably higher, and the
incidence of poverty lower, in urban than in rural areas. As seen from Table 7,
differences in income and consumption between rural and urban areas were
also captured in the National Census (CSO 1994). The differences in both
inc om e and consumption revealed by the data is a preliminary indication of
the complexity of urban poverty. Moreover, urban areas reveal considerably
high er income and consumption differentials than rural areas.

7

C M I

Table 6. Average Income and Poverty by Urban and Rural Areas

Rural/urban Average income (N$) Human Poverty Index (%) 6

Rural 1.875 24.7
Urban 7.651 16.7

Source: UNDP 1998

Table 7: Income and Consumption in Rural and Urban Areas (N$).

Income and consumption Rural Urban

Average household income
Average per capita income
Average household consumption
A vera e er ca ita consum tion

9.450
1.550
7.600
1.250

32.300
6.650

23.000
4.700

Source: Tvedten and Pomuti 1998

A num ber of studies als o reveal that there are differences in income and
incidence of poverty between social categories based on gender and age.
Women and female headed households generally have lower income and are
poorer than male-headed households (GON 1993; Girvan 1994; Næraa and
Solomon 1994). As shown in Table 8, figures from UNDP show that the
average income for women is less than half of the income for men, and women
have a high er incidence of poverty. Also here, of course, aggregated data at
this lev el conceal differences in income and consumption within each

category.

Table 8. Average income and poverty by gender

Female/male Average income (N$) Human Poverty Index (%)

Female 2.188 25.4
Male 4.454 22.6

Source: UNDP 1998

Studies als o show that the young and the elderly are particularly susceptible to
poverty (CSO 1996; Devereux, Fuller et aL. 1996): The former are vulnerable
because of the problems related to health and education in poor families, and
the latter because of the considerable pressure from other family members on
the resources they possess (including land, catde and pensions).

The poverty index on rural-urban areas is incomplete due to insuffcient data on
underweight children. However, according to UNDP (1998:21) "a complete HDI is likely
to portray an even more dramatic rural/urban divide".

8

C M I

There is little doubt that quantitative definitions of poverty like the ones

referred to above are useful to policy-makers. They provide a uniform scale
against which comparisons can be made of the incidence of poverty in
different sub-sections of the population, or of the same population over time.
Comparative data are essential in order to target resources to the poorest
groups, and measure the implications of anti-poverty interventions on
different social categories.

Having said this, the current study attempts to go beyond the quantitative
measures of poverty and vulnerability by focussing on survival strategies of the
poor themselves through a focus on relations and networks that poor people
relate to in their daily lives and mobilize in times of crises.7 We will argue that
these factors determine individual and household well-being as strongly as
levels of income and ownership of tangible assets, and hence that a
combination of qualitative and quantitative data is necessary in order to get a
realistic picture of poverty in Namibia.

As emphasized in the introduction to this report, however, although

horizontal transfers play an important ro le in spreading risks and reducing
vulnerability transfers between poor households almost by definition draw on
a smaller pool of "surplus" income than vertical transfers. Therefore, any
serious attempt to reduce poverty in Namibia requires an active public policy
commitment to redistribute not only income and consumption goods (such as
pensions and food aid), but also productive assets (such as land, agricultural
inputs, and skills) from Namibia's wealthy elite to its poor majority.

A focus on survival strategies, relations and networks has a long tradition in social
anthropology, but has recently been "reinvented" through the concept of "social capital"
currently advocated by the World Bank (Portes 1998).

9

C M I

3. Urban Relations of Poverty

3.1 Urban Poverty

We indicated above that one of the main problems with poverty data based on
income, consumption or other quantitative measures is to capture the
differences between rural and urban areas. This is becoming increasingly
important as the developing world is becoming rapidly urbanized. On the

African continent, Southern Africa stands out as the most urbanized sub-

region with urbanization rates ranging from 50.8 percent in South Africa to
24.4 per cent in Tanzania. 66.4 percent of the population in Southern Africa

are expected to live in urban areas by year 2010 (Tannerfeldt 1995; UN
1998).

An estimated 50 percent of the poorest sections of the population in Southern
Africa wil live in sprawling urban shantytowns8 within the same period of
time until 2010 (Wratten 1995; UNCHS/HABITAT 1996). Having said this,
statistics on inc om e and consumption still show that there generally is a high er
incidence of poverty in rural than in urban areas. There are, however, a

number of problems related to comparing standard measurements of rural and
urban poverty respectively.

Very of ten a single income level or "poverty line" is set for both rural and
urban households. This assumes that the costs for basic necessities are the
same for the two types of areas, which is normally not the case. A number of
essential commodities (such as housing, food, transportation, health-services
and education) are more expensive in cities and towns than in rural villages. In
most urban areas there are also fewer opportunities for reducing costs through
subsistence production (e.g. growing food) or through access to free resources
(such as wood for building and fuel).

Where poverty lines are based on a "minimum consumption basket",

moreover, the costs of food items usually take up the bulk of the basket. In
real life, the urban poor have to spend a considerable proportion of their
income on necessary non-food items such as housing, water, fuel,
transportation, education and health-services. The costs of these "other items"
is of course the very reason why such a high proportion of the urban

Poor urban neighborhoods in Southern Africa are known under different names with
different connotations. A "location" normally implies a poor African township formally
established by the South African apartheid regime; a "squatter area" implies a poor and
illegally established urban settlement; and a "shantytown" implies a poor urban area
with a mixture of illegal and legal settlements. "Slum" is normally not used in the
literature due to the words strong derogatory connotation, but is close to local terms
aften used for such areas. In this report, "shantytown" and the equivalent "informal
settlement area" will be used interchangeably.

10

C M I

population in developing countries live in poor quality, overcrowded and
insecure accommodation with inadequate provision of basic services.

Despite the difficulty of measuring urban poverty, there is a growing
realization that urban areas reveal special characteristics that make the
population in cities and towns particularly vulnerable. The renewed interest in
urban poverty was first prompted by the wave of structural adjustment
programs in the 1980s. The programs were intended to remove some of the
"urban bias" (Lipton 1980) by removing anti-agricultural price-distortions,
with the result that the urban poor suffered from price increases, contraction

of industrial and public sector employment and reductions in public
expenditure. More recently, the UN Conference on Human Settlements
(Habitat Il) in 1996 drew further attention to urbanization and urban poverty
(UNCHS/HABIT A T 1996).

The most important determinant of urban poverty (as opposed to rural
poverty) is the lahor market. People with formal stable jobs, even if salaries
are low, are much less likely to be poor than the unemployed or people with
unstable casual jobs. Earnings in the urban informal sector have generally

become increasingly low and irregular, as an increasing number of the urban
poor have come to rely on this option.

Urban areas are also characterized by a greater degree of commercialization
than rural areas. Urban households require money in order to pay for housing,
pay rent and buy basic necessities such as food, water and fue!. A number of
interpersonal social services, such as domestic work and child care, are also
increasingly commercialized. People without savings or salable capital assets
are extremely vulnerable to changes in the demand for labor and the prices for
basic goods.

Social diversity and fragmentation is typically also seen as a special aspect of
urban poverty. Urban areas attract people with different ethnic, cultural and
linguistic origins. Social diversity is likely to create new tensions and survival
strategies. Lifestyles, kinship and neighborhood support networks are seen as
different from those in rural areas, and crime and social insecurity is another
aspect of urban poverty that is given increasing attention.

Finally the urban poor are particularly susceptible to health risks, resulting
from the doseness of industrial and residential functions, competition for
land, high population densities and overcrowded housing, and the inadequate
pace at which dean water supply, sanitation and solid waste services are
expanded. In addition, though health services of ten are easier accessible in

urban than in rural areas they are also of ten more expensive.

It should be underlined that the issue at hand is not whether poverty is more
serious in urban than in rural areas. Many argue against making such a
distinction at all and for treating the urban-rural divide as a continuum rather

11

C M I

than a rigid dichotomy (de Haan 1997) 9. Given the considerable income

differentials in urban areas one may argue that it is more relevant to talk
about a "trichotomy" with shanty-areas and small semi-urbanized
communities constituting an intermediate category between the rural and the
urban. In any case, however, it is necessary for policies as well as for theories
to gain more insight into the determinants of urban poverty.

3.2 Social Relations of Poverty
With the growing awareness of the different conditions under which the rural
and urban poor live, increasing attention has been given to the importance of
alternative ways to measure and understand poverty. Two concepts have
become particularly important.

The concept of vulnerability is not synonymous with poverty, but means
defenselessness, insecurity and exposure to risk, shocks and stress (Chambers
1989). It is link ed to tangible assets such as wages, savings, housing and
domestic equipment, but also to intangible assets such as claims on other
households, patrons, the government and the international community for
resources. The related concept of entitlement refers to the complex ways in
which individuals or households command resources (Baulch 1996). These
vary between people and over time, in response to shocks and long-term

trends. They may include wage la bor, sale of assets, own production and
public provision of goods and services.

Less attention has been given to the importance of social relations and
networks between the poor, and between the poor and the less poor, more
specifically. The role of social relations and networks have been a central issue
in social anthropology for years (Mitchell 1969), but has only recently been
given more widespread attention through the notion of "social capital"
advocated by the World Bank.

Social capital is defined as "the ability of actors to secure benefits by virtue of
membership in social networks or other social structures" (Portes 1998).
Studies show that people are involved in relations and networks that are
important not only for their daily survival strategies, but also for their access
to employment, housing and social services, and their ability to mobilize
support in times of crisis. The nature of relations and networks tend to differ
with income, gender, age and other social variables (Moser 1996; Macharia
1997).

First, it is argued, concentrating on whether urban poverty is more extensive than rural
poverty diverts attention away from structural determinants which affect the life-chances
of the poor in both areas. These include distribution of land, constraints to opportunity

based on dass, gen der, race and age, government social and macro-oriented policies and
external relationships which shape exchange rates, terms of trade etc. And secondly, there
are linkages between the functions of urban and rural areas which implies that poverty in
ane area cannot be treated in isolation from the other. Interdependence between towns
and countryside exists in areas such as rural-urban migration, casuallabor, markets for
food, industrial goods and services, water supply and demand, and flows of remittances.

12

C M I

Networks are of ten given structure through the creation of voluntary

associations. Such associations is the main focus of the Wodd Bank's notion
of "socIal capital" .10 Many of them are formed in direct response to the
State's inability to provide a modicum of social and public services, and to
alleviate poverty (Tostensen, Tvedten and Vaa forthcoming 1999).

Analyses of relations and networks of ten draw distinctions between
commodity exchange (impersonalized exchange between independent
individuals); rational models of recIprocity (emphasizing material motivations
for exchange); and cultural principles for social interaction that create
enduring bonds between mutually dependent persons of ten influenced by

Mauss' notion of gift practices in archaic societies (Mauss 1967).

Generally the expansion of the market economy and urbanization have been
seen as a threat to the reproduction of value systems based on solidarity and
what has been called a "moral economy" (Scott 1976). Instead, it is argued,
urban relations are based on individual strategies and self-interested agents
acting to maximize their returns. In the process, reciprocal relations tend to
giv e way to asymmetrical (patron-client) relations based on variations in
access to economic resources. One implication of this is seen to be that people
differ in their ability to forward claims towards other households, patrons,
informal associations, the government and other institutions. Tangible and
intangible assets are created by households during periods of economic
surplus, and the ability to activate community support is thus related to the
capacity to generate such a surplus.

Urban-rural links represent special types of social relation and networks.
Existing evidence suggests that relations between rural and urban dwellers
involve exchange of food and material assets, as well as services ranging from
identification of job opportunities to child care. The extent and material
content of such relations in Namibia have been relatively well documented
(Pendleton and Frayne 1998; Tvedten and Pomuti 1998), but less is known
about the extent to which they involve the poorest sections of rural and urban
populations.

3.3 Marginalization and Social Exclusion

Until recently towns and cities have largely been seen as expressions of global
political and economic processes of modernization, with less focus on the
effects these processes have had for the marginalization of urban groups
(UNDP 1991; WB 1991). Large urban population groups are, as we have
argued, susceptible to poverty and deprivation through changes in income and
expenditure patterns, changes in social organization and traditional support
networks, and health hazards through congested settlements, sub-standard

10 If ane takes abroad view of "other social structures", then social capital is a relevant

concept also at the macro leve!. There is, according the World Bank, overwhelming
evidence that such macro-level social capital has a measurable impact on national
economic performance (World Bank 1999).

13

C M I

housing, high crime rates and inadequate infrastructure and services (de Haan
1997) .

Recent advances in the analysis of poverty have implied options for bridging
the gap between macro-orientated quantitative analyses and micro-orientated
analyses of qualitative processes in the study of urban poverty and
marginalization (de Haan and Maxwell 1998). The concept of 'social
exclusion' is currently high on the agenda, and has been defined as "the
process through which individuals or groups are wholly or partially excluded
from full participation in the society in which they live" (de Haan and
Maxwell 1998). Social exclusion takes the form of income markedly lower
than average in the society at large, failure or inability to participate in social
or political activities, or otherwise a life in the margins.

The concepts of "marginalization" and "social exclusion" have other
implications than the alternative concepts of "exploitation" and "unequal
development". Being largely de-linked from political and economic processes
in society means that general economic development will not affect the
relevant social groups, neither directly nor indirectly through "trickling down"
effects as argued by liberal economists. Re-integration and socio-economic

development among marginalized groups must either come from direct
targeted interventions, or from initiativ es by the socially excluded groups

themselves through development of their social capitaL.

The issue of marginalization and social exclusion ra is es another central issue
in social anthropology, viz. the notion of a "culture of poverty" contributing
to the maintenance of a poor section of urban populations. The notion of
culture in anthropology has moved away from a premise of cultural sharing,
to a distributive model where culture as systems of interests, values, beliefs
and knowledge is seen as differently read and construed by different
individuals and social groups (Brumann 1999). The focus is on the work done
by people in the cultural construction of their realities, and on the connections
and interdependencies that generate a degree of order and shape (Keesing

1990; Barth 1994). People in urban shantytowns will, in other words,
experience and relate to slum life through different perspectives and in
different ways from where they are in the social structure (Hannerz 1992).

The critique against the concept of a "culture of poverty" has been based on
the argument that it implied that poverty was somehow the poors' own fault,
with the counter-argument that poverty is generated and reproduced by

political and economic factors outside the control of the poor themselves

(Melhuus 1997). However, as Hannerz (1992:76) argues, "a much more
realistic view, in light of evidence, would have been to regard a culture of
poverty as resting () on the level where collective understandings are tied to a
general type of context. Shared poverty generates cultures of poverty, rather
than vice versa" (Bourgois 1995).11

11 A thorough analysis of the relation between socio-economic conditions and cultural

perceptions for the survival strategies pursued by the poorest sections of urban

14

C M I

4. Poverty in Oshakati

In the following section, we wil look at the issues raised above concerning

poverty and social relations with the town of Oshakati as our point of
departure. We will start with a brief summary of the historical and structural
proeesses behind the socio-economic differentiation existing. We will then
proceed with an outline of the socio-economic differentiation in the informal
settlements, based on data from the socIo-economic baseline study conducted
in the areas in 1994 (Tvedten and Pomuti 1994) 12. The main part of this
section will be a qualitative analysis of survival strategies in four informal
settlements, with a particular focus on social relations and networks.

Case studies like the present inevitably raise the issue of representativeness.

Our point of departure wil be that former Owambo (i.e. the regions of
Oshana, Ohangwena, Omusati and part of Oshikoto) share a number of
characteristics with the other regions in the northern part of Namibia. Hence,
the main points made in the analysis are also relevant for the regions of
Kunene, Kavango and Caprivi. The similarities are born out in statistics on
income, consumption and incidence of poverty presented in Section 2.

Each region also has one princIpal urban center (i.e. Opuwo, Rundu and
Katima Mulilo for Kunene, Kavango and Caprivi respectively), which shares
many similarities with Oshakati (Devereux, Melaku-Tjirongo et aL. 1993;
Graefe, Oherien et aL. 1994). Pinally each town has clearly separated formal
and informal settlement areas, that represent important economic, political
and social divisions. T able 9 shows the size of the informal settlements in
relation to the total population of each of the four towns.

Table 9. Informal Settlements as Proportion of Town Population

Town Total Population ,~ Informal settlements (%)

Opuwo 5,000 70
Oshakati 25.800 60
Rundu 23,000 75
Katima Mulilo 16,000 75

Source: Tvedten and Moputola 1995 * Estimates are from 1995

populations is beyond the scope of this report, but wil be pursued in forthcoming
publications.

12 The survey is based on a sample of 7% of the total shanty population, which is large

enough to draw generalized conclusions. The sample is stratified proportionally by
settlement area and not based on an overall random sample, in order to secure an equal
representation of the smaller settlement areas.

15

C M I

4.1 Historical and Structural Causes of Poverty
The current socIo-economic landscape in former Owambo and in Oshakati is
the outcome of political, ecological, demographic and economic factors that
individually and collectively shape patterns of socioeconomic welfare found in
the area today.

In brief and general terms, pre-colonial Owambo was characterized by a low
level of material production and consumption, of ten negatively affected by

multi-year droughts, epidemic stock disease or other causes related to the dry
and arid natural environment in the region (Devereux, Fuller et aL. 1996). At
the same time, Owambo society had strong kinship mechanisms of wealth
redistribution and social security. Structural impoverishment perpetuated over
lifetimes or generations was thus mainly restricted to small numbers of war-
captives and slaves and to dependent ethnic minorities attached to more
powerful elan families.

The arrival of European travelers, traders and missionaries in the 19th century
began to destabilize the political and social integration of the Owambo
communities. German colonization of south and central Namibia (1884-1904)
introduced wage labor and forced migration, and missionaries had a
considerable impact on social relations and family organization (Siìskonen

1990). However, it was the onset of South African rule from 1920 that had
the most dramatic impact on the socio-economic conditions in Owambo. They
soon established effective controlover Owambo largely by utilizing the
traditional political structure of kings, chiefs and headmen, and through the
"homeland policy". Their direct influence was particularly related to the
system of contract labor migration, and what came to be known as "Bantu
education" (Banghart 1969; Hishongwa 1992).

Dntil the mid -1960s, changes in political and economic conditions and socIo-
cultural adaptation took place within a rural setting based on small extended
vilages, agro-pastoral production, and a matrilineal system of kinship. There

were no urban areas in Owambo, and the urban experience was reserve d for
labor migrants returning to their home region for short periods of once every
two years and, later, once a year.

Oshakati was established in 1966, partly as a base for South Africa's
economic interventions in the northern "homelands" and partly as a base to
fight the SW APO liberation movement that had been established in 1959 and
enjoyed increasing support among Owambos (Hangula 1993). The town grew
rapidly throughout the 1970s and 1980s. Large military structures, a hospital,
schools, an organized open market, a meat processing plant and several
smaller factories were established. Most of these investments were made by
the para-statal Bantu Investment Corporation. A few local businessmen were
also given loans to enable them to establish their own business units.

The development of Oshakati first and foremost implied employment

opportunities for those migrating to the town. Many became employed with
the South African Defense Force (SADF), but there were also formal

16

C M I

employment opportunities in construction work, in factories, in trade, as
maids and watchmen etc. In addition, the informal sector grew rapidly. The
increasing num ber of Owambos in wage work opened opportunities for
informal tradesmen and women selling goods ranging from traditional
foodstuffs to second hand dothing, and for lodging places, tailors, barbers,
hairdressers, and prostitutes.

Most of these establishments were located in the expanding informal

settlement areas on each side of the formal town. The formal town was
divided into a white town (Oshakati East) and a black township (Oshakati
West), with the white section being separated from the rest by an intricate
system of roads, fences and barbed wire. Due to the apartheid policy and war
situation there was a strict separation between urban Oshakati and its rural
Owambo hinterland, and Oshakati for all practical purposes became a
fortified town. By 1985 Oshakati had a population of 15.000 people, and at
independence in 1990 the population stood at 25.000.

The withdrawal of South Africa at independence had immediate consequences
for employment, income and social conditions. People associated with the
South African Defense Force left or lost their jobs, and thousands of

unemployed SW APO freedom fighters moved in from the regions or from
exile. In the informal settlement areas, the transitional period was

characterized by poverty and social instability. People had no jobs, and the
informal sector had no customers.

Despite its importance in the north, Oshakati in many ways remained an
"urban backyard " in relation to urban centers in the south until around 1995
(Tvedten and Mupotola 1995). Both public and private investments were slow
to appear, and a large number of Owambos bypassed Oshakati to become
urbanized in towns like Windhoek, Walvis Bay and Grootfontein where the
lights of urban life were considered to be brighter.

At the same time new legislation vested increasing responsibility for social
services with towns and municipalities, and poor towns with a weak tax-base
such as Oshakati had problems delivering services to its population (Tvedten
and Mupotola 1995; Frayne and Pomuti 1997). Democratic participation and
accountability was limited, with low turnouts for elections and voters casting
their ballots for political parties and party lists rather than individual ward-
based candidates that would have brought politics doser to the shanty-
towns.13

Though stil severely affected by a number of structural economic and political
problems, in the past five years Oshakati has slowly regained its role as the
main urban center in the north. The basis for this expansion is primarily
related to enhanced economic investments from South African capital
(establishing shopping centers as well as manufacturing industries), and from
Angolan economic interests (involved in both formal and informal trade with

13 A change towards individual ward-based candidates was anticipated for the 1998 election,

but has been postponed until 2003.

17

C M I

goods ranging from canned fish to diamonds). The increasing commercial

importance of Oshakati also seems to have led to stronger government

involvement in the town.

The political and economic context within which the shanty-dwellers find
themselves is thus currently characterized by improved economic
opportunities, but also by a considerable increase in the population and hence
more competition for employment, housing and services. Moreover, the status
of the informal settlement areas is still unresolved. While the Oshakati Town
Council accepts them as a Il "fact of life", they stil describe the population in
the areas as "squatters" and are reluctant to plan and develop physical

structures and urban services.

4.2 Incidenee and Characteristics of Poverty
Of the total population in Oshakati of 50.000, approximately 60 percent live
in informal shantytowns and 40 percent live in formal settlement. The two
formal areas are Oshakati East (the former white area) and Oshakati West
(the former black township). There are altogether seven shantytowns, of
which Uupindi, Oneshila, Evululuko and Oshoopala are the largest (see Pigure
1 ).

As in other urban areas in Namibia, levels of income, housing and access to

social services differ significantly between the two main types of settlement
areas.14 The differences are immediately evident by the type of houses, shape of
roads and proximity to commercial centers and key public institutions like
schools and hospitals. Moreover, public services like water, electricity and
waste disposal imply that the formal settlements are deaner, greener, quieter
and brighter. The majority of people living there are government employees
with fixed monthly salaries, but private houses revealing considerable wealth
have also recently been built.

The contrasts between the formal and informal town are not only evident in
the form of physical characteristics and differences in income and

consumption, but are also born out in terms of people's perceptions about

differences in standards of living, levels of education, ways of dressing,

behavior, safety and trustworthiness. Many people in Oshakati who live
outside the shantytowns still regard them with gre at suspicion, and there are
people who argue that they will never enter some of the are as with the

historically most negative reputation. The perception about these areas among
the shanty-dwellers themselves naturally differ, but also people there tend to
perceive the shanty-areas as inferior to other areas. 15

14 The picture has been complicated by the recent increase in population, with a number of

people living in overcrowded formal houses with relatives, friends or as tenants.
15 The differences in perceptions are also evident from the names used to denote the formal

and informal towns respectively. While the neutral term odoropa is used for Oshakati East
and olukanda for Oshakati West, the diminutive okalukanda or the derogative uupereki
are normally used for the informal settlement areas. Recently the term uurnbashu (derived
from an Oshiwambo express ion for a "blow-out of diarrhea") has also come ¡nto usage,

by people from both types of settlements.

18

0-..
~
i:..

è3

'õ
~
~

Q)

~
.l"

G)..
~...~

C M I

The four major informal settlement areas are located in different parts of
Oshakati, and vary in size of the location as well as population (Tab le 11).

Table 11. Estimated population in informal settlements (1999)

Settlement Population estimate

Uupindi 9373
Evululuko 4644
Oshoopala 2877
Oneshila 4302

Source: Re-calculated from Frayne and Pomuti 1997.

The differences in the size of the population is related to historical
developments of migration and settlement, as well as current structural
conditions. Uupindi has the strongest roots in traditional Owambo society, as
the settlement was established within the traditional boundaries of the
Uukwambi ethnic group. Oneshila became the main informal commercial
center early on, at least partly due to its proximity to the main road passing
through Oshakati. Evululuko and Oshoopala were Ofiginally established as
areas for SADF and Koefoet personnel, and have had a reputation for being
more tense and violent than the two other areas. 16

The current growth of the shanty areas is influenced by urban land policies,
establishment of public and commercial enterprises, and the annual

inundation (oshana) of large areas surrounding Oshakati.

With the proclamation of Oshakati as a town in 1995, land formally belongs

to the government through the Oshakati Town Council (OTC). Urban
headmen have lost their right to allocate land within the town boundaries
(Tvedten and Mupotola 1995; Frayne and Pomuti 1997). Though the system
of freehold land is not fully implemented, it is becoming increasingly difficult
to obtain plots in the shanty-areas.

One reason for the Town Council's reluctance to allocate land are plans to
develop large commercial complexes in the vicinity of the informal settlement
areas (Hamata, Hangula et aL. 1996). Two shopping complexes ("Yetu" and
"Game") have already been opened. Additional roads, buildings and other
physical infrastructure are also likely to come into conflict with the informal
settlement areas.

Finally, the oshana surrounds all four informal settlement areas, and at times
of high floods theyenter roads, gardens and houses and represent a

16 The stigma has led both settlements to change names after independence, from Omashaka

(derived from the verb okushakena (converge) and denoting the convergence of different
languages and ethnic groups in Koefoet and SADF to EvuluJuko ("a place where you can
rest"), and from Amunkambya (derived from the term for bitter fruits from the ornwandi
tree) to Oshoopala ("a place that looks nice") respectively.

19

C M I

considerable health risk. Uupindi and Oshoopala are particularly affected
(Marsh and Seely 1992). The cheapest and most vulnerable dwellings tend to
be built by the poorest in the lowest and most vulnerable areas of the shanty-
towns.

4.2.1 The Population

The population in the informal settlement areas have a variety of
backgrounds, but some general traits can be discerned. First of all there is a
clustering of people from the same geographical areas and ethno-linguistic
groups, which implies that people ten d to settle with others with similar

background (Table 12). However, we have also shown that the majority of the
households (59 percent) do not have members of their own extended family
liv ing in the informal settlement areas. Moreover, with the increasing

overcrowding of the shanty-areas the choice of where to stay has become
much more limited in the past four-five years.

Table 12. Ethno-linguistic Affiliation (Percentages)

Income Uupindi Evululuko Oshoopala Oneshila Total

Oshikwanyama 26 45 73 46 43
Oshindonga 17 16 3 9 13
Oshikwambi 32 27 10 24 26
Oshimbadja 6 O 3 2 3
Oshimbalantu 3 2 O 2 2
Oshikolonkhadi 3 7 3 O 3
Oshikwaludi 4 O 3 7 4
Others 10 2 3 11 7

With regard to distribution in terms of gender and age, slightly more women
than men live in the shanty-areas (Table 13). This, together with the high
proportion of de facto female headed households, reflects socioeconomic and
cultural changes related to family structures, unstable household units, the
difficult situation for women in rural areas and (conversely) their economic
opportunities in towns. The slightly high er proportion of men in Oshoopala
and Evululuko has to do with the history of these areas as locations for
soldiers.

Table 13. Sex Structure (Percentages)

Sex Uupindi Evululuko Oshoopala Oneshila Total

Male 46 49 52 48 48
Female 54 51 48 52 52

With regard to age structure, the population in the informal settlement areas is
young (Table 14). 77 percent are 34 years or younger, and only 5 percent are

20

C M I

50 years or older. The age structure is only marginally gen der differentiated,
except for a higher proportion of women in the prime working age group
between 20-29 years. Here women account for 53 percent of the population.
A distinct feature of the informal settlement areas is the relatively low number
of youngsters between 5-19 years (22 percent of total). These, as we shall
explain later, are typically staying in rural areas because "the shanty is no
place for a child to grow up".

Table 14. Age Structure (Percentages)

Age Uupindi Evululuko Oshoopala Oneshila Total
Structure
0-4 12 15 19 14 14
5-9 8 11 9 8 8
10-14 7 5 7 5 6
15-19 8 6 11 7 8
20-24 16 12 14 17 15
25-29 14 18 15 20 16
30-34 10 11 9 12 10
35-39 10 11 6 5 8
40-49 10 7 7 7 8
50-59 3 1 4 3 3
60+ 3 3 1 2 2

The educationallevel in the informal settlement areas is generally low (Table
15). 14 percent of the population over school age have no education at all and
47 percent have grade 4 or less, implying a functional illiteracy rate of more
than 60 percent. 36 percent have obtained grade five to 10, while only three
percent have Grade 11 or higher. The level of education is high er among
women than men (with a lower proportion having no education and a high er
proportion having grade five or more).

Table 15. Levet of Education (Percentages)

Income Uupindi Evululuko Oshoopala Oneshila Total

No education 10 14 28 20 16
Grade 1- 4 25 21 29 22 24
Grade 5 -7 28 34 26 16 26
Grade 8 -10 27 27 13 27 25
Grade 11 - 12 9 4 4 11 7
Other 1 1 O 4 2

Pinally the majority of the population in the shanty-areas are first generation
urbanites. 58 percent of the household heads moved to Oshakati after 1980.
At the same time, 90 percent of the households believe that they wil continue
to live in the informal settlement area at least for the next ten years (Table 16).

21

C M I

Table 16. Expected Place of Residence in Ten years (Percentages)

Place of Uupindi Evulu- Oshoo- Oneshila Total
Residence luko pala
Present shanty area 100 70 93 88 90
Other areas in Oshakati O 8 O O 2
Rural areas O 19 7 10 7
Another urban area O 3 O O 1
Other O O O 2 1

4.2.2 Socio-economic Characteristics
Looking at the socIo-economic characteristics of the shantytowns, practically
all households settling in these areas are relatively poor in terms of income and
other assets compared to people settling in the formal town. Nevertheless, they
show variation with respect to a number of important traits including
employment and income, housing, ownership of other assets, and patterns of
expenditure. Table 17 reveals that the general income in the areas is low, but
also that there are considerable differences both between and within the four
areas.

Table 17. Income Distribution (Percentages)

Monthly Uupindi Evululuko Oshoopala Oneshila Total
Income (N$)

o: 50 6 5 7 2 5
51 - 100 4 5 7 O 4
101 - 250 25 24 50 19 28
251 - 500 26 24 17 21 23
501 - 1000 18 30 20 21 21
1001 - 3000 18 11 O 33 17

~ 3000 3 O O 5 2

The majority of households have a monthly income of less than N$ 500 per
month. The Standardized Consumption Level (CSL) of N$ 7,200 per year

discussed above implies a monthly income of N$ 600, meaning that
approximately 60 per cent of the shanty population is defined as poor. The
differences in income between the shanty-areas is also significant, with
Oshoopala being the poorest and Oneshila the richest. There are also
significant differences between female headed and male headed households:
Among the female headed households 69 percent earn N$ 500 per mo nth or

less, while 56 per cent of the male headed households have a similarly low
income.

The differences in levels of income is to a large extent reflected in main source
of income for the households. As seen from Table 18, 44 percent of the

22

C M I

households have formal employment as their main source of income, and 40
percent depend on informal employment. The proportion of formally
employed is smallest in Oshoopala. The majority of those having pensions as
their main source of income rely on older relatives of ten living in the rural

areas. The category "other" primarily consists of people depending on
handouts or other forms of redistribution. The proportion of female headed
households with formal employment (20 percent) is considerably lower than
for male-headed households (53 percent).

Table 18. Main Source of Household Income (Percentages)

Source of Uupindi Evulu- Oshoo- Oneshila Total
Incame luko pala
Formal employment 39 62 27 49 44
Informal employment 41 30 60 35 40
Agricultural income O O O 2 1
Pensions 12 5 3 9 8
Other 9 3 10 5 7

Income and standard of living is als o reflected in the type of dwelling a

household possesses. Table 19 shows that 64 percent of the shanty population
live in iron-shacks, 34 percent in brick houses and only 2 percent in traditional
dwellings.17 Living in brick houses is considered superior because it fulfills
urban cultural expectations and increases security of tenure.18 In addition, a
brick house represents better options for informal businesses such as food
production (okapana), tailoring or production and sale of alcoholic beverages.
As many as 61 percent of the dwellings in the informal settlement areas are
used for other than domestic purposes. Also in this respect female headed

households are poorer than their male counterparts, with 69 percent living in
iron shacks and 29 percent in brick houses, as against 62 percent and 36
percent respectively for male headed households.

Data on physical infrastructure is another indication of the standard of living
in the shanty areas, closely link ed to health conditions. Supply of urban

services is the responsibility of the local government (Oshakati Town
Council), and there are here no significant differences between the areas. 74
percent of the shanty population dep end on water from communal taps, 65
percent do not have access to proper toilet facilities, 88 percent depend on
wood for cooking, and 83 percent depend on candles for lighting. There are
no significant differences between male headed and female headed households
as regards access to such facilities.

17 Traditional dwellings (egurnba lya rniti, lit. "stick house") are relatively cheap, healthy

and easy to build. The Iimited num ber of such dwellings is an indication of the strength of
an urban culture emphasizing that traditional dwellings belang to the rural and not the
urban settings.

18 Egurnbo (i.e. the term used for traditional homesteads in the rural areas) is only used for

brick houses.

23

C M I

Table 19. Type of Main Urban Dwelling (Percentages)

Dwelling type Uupindi Evululuko Oshoopala Oneshila Total

Brick house 32 38 10 51 34
Iron shack 68 59 80 49 64
Traditional O 3 7 O 2
Other O O 3 O 1

The 1993/1994 Namibia Household Income and Expenditure Survey (CSO

1996) revealed that urban households spend 32 percent of their income on
housing, 23 percent on food, and 24 percent on consumption in kind. Rural
households, on the other hand, spend 47 percent on food, 15 percent on

housing, and 38 percent on consumption in kind. The expenditure pattern is
largely confirmed by our data from Oshakati. First of all people spend
considerable amounts of money for their dwelling, as indicated by T able 20 on
the mark et value of the main dwelling possessed. The value of dwellings in
Oneshila is high er than in Oshoopala, which reflects both the different types
of dwellings and the higher mark et value of Oneshila as a place of residence.

Table 20 Estimated Market Value of Main Dwellings (Percentages)

Market value Uupindi Evululuko Oshoopala Oneshila Total
(N$)
c: 250 1 3 3 O 2
250 - 1000 10 14 30 2 12
1001 - 2500 17 16 20 26 20
2501 - 5000 36 38 33 30 35
~ 5000 35 30 13 42 32

Table 21 reve als an expenditure pattern where food only represents about 1/3
of total consumption. There are no significant differences in the importanee of
food as part of total expenditure between the four shanty areas. According to
the standard measurements of poverty discussed in Section 2, limited
expenditure on food implies that people are "not poor". However, as we have
already argued non-food expenditures like housing (see Table 20), fuel (22
percent) and transport (20 percent) represent an additional and necessary

burden for the poor in urban areas making a direct comparison difficult.

24

C M I

Table 21. Daily Expenditure Pattern (Percentages)'~

hem Uupindi Evululuko Oshoopala Oneshila Total

Food 32 32 33 33 33
Clothing 18 19 17 17 18
Transport 17 18 16 20 18
Fuel 25 22 25 22 23
Other 8 8 9 8 8

.f Informants were asked to rank their expenditures in order of importance. A point system
has been used, giving the expenditure drawing most from the household budget the
highest value. The category "Other" includes alcohol, and is most likely under-
represented.

4.2.3 Health and Nutrition
In addition to income, consumption and access to basic social services, the
socio-economic conditions in the informal settlement areas are revealed by

data on health and nutrition.

Available information from surveys and information supplied by the
community (UNICEF 1990) (Tvedten and Pomuti 1994) indicate that there
are serious health problems in the shanty areas particularly among children
(see Table 22). Malaria, diarrhea and various respiratory diseases are the most
common. At the same time immunity rates have improved since independence
for ilnesses like polio, measles and tuberculoses.

Aids is becoming a serious problem and a great concern for the population.
Evidence suggests that more than 20 percent of the population is HIV -affected
(UNDP 1998), and death from aids is becoming an every-day affair. The
epidemic will, of course, have significant implications for economic adaptation
as well as social organization in the shanty areas.

Table 22 Health indicators

Health indicator Va lue

Infant mortality rate 70/1000
Child mortality rate 100/1000
Life expectancy at birth 60 yrs
Maternal mortality rate 552/1 00,000
Sick children 68 percent~
Women under 20 yrs with children 39 %
Use of contraceptives among women 13 %

.f Percentage of children being sick (cough, measles, diarrhea) within the past 14 days of the

interview.

25

C M I

Also under-nutrition is a serious problem, and the number of meals and type
of food eaten is closely linked to the socio-economic conditions of the

households. Undernourished children are most commonly found in poorer
households, in households where both parents are absent during large parts of
the day, and in households where the adults frequently substitute alcohol for
food. Table 23 indicates the degree of malnutrition in the informal settlement
areas.

Table 23 Nutritional indicators (Percentages)

Nutritional indicators

Under-nutrition 1) 35
Wasting 2) 9
Stunting 3) 32

1) Under-nutrition among children 6-60 months, measured in weight for age (Wt/A). A

low eight for age reflects both recent and lang term malnutrition.
2) Wasting among children 6-60 months, measured in weight-for-height (Wt/Ht). A low

weight -for-heights indicate recent illness or inadequate food-intake.
3) Stunting among children 6-60 months, measured in height-for-age (Ht/A). A low

height-for-age is a reflection of long-term or chronic malnutrition.

4.3 Social Relations of Poverty
The quantItative data presented above point in the direction of a shanty
population characterized by problems of employment, low income, high
expenditures on food as well as non-food items, limited access to basic services
and poor housing conditions. The existing situation has resulted in poor
health conditions, and nutrItional problems particularly for children.

At the same time, the data reveal that there are significant differences in the
socio-economic sItuation of individuals and households both between and
within the informal settlements. We wil argue in the following that access to
employment and income is the key determinant, with implications for type of
dwelling, access to land and urban services, savings and expenditure patterns.
In contrast to conventional perceptions in social anthropology (Keesing and
Strathern 1998), we will also argue that employment and income have
significant implications for the size and organization of households as well as
the type and nature of social relations and networks in which households are
involved.

The poorest sections of the population experience difficulties in making ends
meet both in their daily lives, and in times of special need. They are normally
not in a position to buy sufficient food, clothing, fuel, schoolbooks and other
basic necessities, or pay for expenses related to medical care, indemnities and
important socio-cultural events such as births, weddings and funerals. In fact,
comparing income, number of household members and prices for goods and

26

C M I

services, many people in the shanty-areas II Oshakati should not be II a
position to survive as a social unit at all.19

The central topic in this part of the study is how people actually cope with the
difficult situation described, and whether there are differences in the survival
strategies of the poorest sections of the shanty population (omuthigona/
ohepele) and those who are less poor (omunamake/omuyamba). As argued in
the introduction, survival strategies in poor urban areas in Namibia are to a
significant extent a question of social relations and networks. No households
can survive in a social vacuum, and relations are important not only for access
to material resources but also for social security and fulfilment of socio-

cultural obligations.20 At the same time social networks are not only about
"moral economy" altruism and community support systems. They also involve
obligations and social pressures, and may break down under stress precisely
when they are needed the most.

The importance of social relations is immediately evident in the informal
settlements. In the morning many people leave the shanty area to meet people
at work-places, in the informal market of Omata/a, in hospitals, in schools, or
in the Oshakati Town CounciL. Others leave the areas simply to seek
occasional work and see if other "good things" may happen from incidental
encounters. Among those who remain men sit in groups in front of shacks
talking and drinking tombo. Women cook food together for own consumption
or okapana, and watch each others children. Poor old men walk from
dweiiing to dwelling asking for food or other items. And people from other
places, revealed by their behavior, clothing or way of speaking, indicate

relations not only with rural areas but als o with regions and countries further
away.

All these relations carry material prestations, as well as cultural meaning. The
following case-studies indicate the importance and complexity of thse social
relations of poverty.

19 A number of quantitative studies on poverty based on housheolds as isolated units of

analysis Jeave the same question unanswered: How do people survive when recorded
inc ame is insufficient to feed all household members and pay for other basic necessities?

20 The latter is exemplified by the strong emphasis on ultimately being buried in one's rural

area of origin. Being buried in town is considered a disgrace, and proof of extreme
poverty, vulnerability and isolation.

27

C M I

BOX 1
Old man from a foreign country. Came to Oshakati to work in the construction
of the Oshakati HospitaL. Later work for SADF. Married a Namibian. Three
adult children. Has largely been unemployed since independence. Divorced ("She
found a man with more money than me"). Lives alone in a brick house, and
manages with small savings and income from a small garden. Has little contact
with his own children, and no contact with his former extended family. He does
not have close neighbors ("Most of them are women who stick to themselves"),
and spends most of the day in other parts of the shanty trying to get meals and
drinks. Has tried to become involved in the Angolan community by going to
services in the Tocoist Church, but laments that they only include their own. He
complains about the deterioration of conditions since independence, and argues
that things were better when the whites ruled. His main preoccupation is that he
will not be able to go back to his home-country befare he dies.

BOX2
Married man in the mid-thirties. Employed as a security guard for nearly ten
years. His wife by traditional marriage stays in his vilage in Ohangwena with
their three children. Stays in Oshakati with another woman, with whom he has
two children. The woman has two other children staying with her mother in her
village. The two women know about each other ("They accept that they are
two"). He manages to save money from his low-paid job, because people in the
security-firm "help each other out" and his girlfriend makes same money. Has
invested in cattle, taken care of by a younger brother in the village. Argues that
the most important is to have a permanent job, so you "know that you will have
a paycheck every month". He also prefers Oshakati to towns in the south,
because it is easier to get to now people and you will always find somebody who
can help you.

BOX3
y oung single mother with three children from 1 to 10 years. Lives in a recently
built shack. Makes and sells okapana in cooperation with another woman in the
neighborhood. Things were very difficult when her cohabitant die d in 1997. She

did not get any real help from his extended family, and her own family in the
rural had little to support her with. They now take care of ane of the children.
She received most of the help from neighbors in the neighborhood. Argues that
many men want to stay with her, but she doesnot want them. "Men are nice
sometimes, but you don't have to live with them". She is relatively successful in
the informal economy, because "I have a good place at Ornatala, and neighbors
who take care of my children when I am there" .

Our data indicate that survival strategies differ between three main categories
of households, based on type of income and sex of household head. These are
i) employed men and their household units, ii) unemployed men and their
household units, and iii) formally or informally employed single or cohabiting
women and their household units.21 22 It is important to note that identifying

21 We are aware of the methodological problems with the postulated correlation between the
quantitative data on different levels of poverty presented above, and the three categories
of households defined for aur qualitative analysis. We are in the process of generating
statistical data for the three categories to verify the correlation, but this has not been
possible within the timeframe of this study.

28

C M I

three main social categories implies that the survival strategies of people
within one category have common traits, and not that people in the same
category always follow the same strategy.23 As we shall return to, social
mobility (from poor to less poor, or the other way around) is accomplished

exactly by pursuing alternative survival strategies and establishing alternative
relations and networks.

Analyzing the survival strategies of the three main categories identified, we
will take four main types of social networks (in the sense of patterns of
connection between sociallinkages) as our point of departure. For people in
the shanty-areas the primary social networks are found in the household and
the extended family, among neighbors and friends, and in the urban and
modern setting outside the shanty areas. Formalized networks (associations)
are also part of socio-economic life in urban settlements like Oshakati and will
be treated separately.

4.3.1 The Household

The most immediate social entity for people in the shanty-areas is the
household (egumbo). The "classical" definition of a household is a social and
economic unit whose members are related by marriage or blood, living under
the same roof, pooling economic resources and "eating from the same pot".

However, very few households in the shanty-areas in Oshakati adhere to this
definition, and they differ significantly both in size, composition and social
stability.

The average household size is 5,1, varying from 1 to sixteen members.
Moreover, households typically consist of a small nuclear core, with extended
family members, more distant relatives and non-relatives on the average
making up as many as 43 percent of a household group. Many of these stay in
the household for limited periods of time, and distant- and non-relatives are
normally expected to contribute to the household economy or eat elsewhere.

The instability of the households is indicated by the fact that 58 percent of the
adults in the shanty-areas are single (ovo aveke), 23 percent are cohabitants
(oveli pamwe) without any legal status, and only 14 percent are married
traditionally, in church or at the magistrate (ova hokana hombola). 27 percent
of the households are de jure female headed, with the actual rate being higher
if counting households where men are away for long periods of time as de
facto female headed. Households thus differ considerably as regards the extent
to which they function as a coherent social and economic unit.

22 Female headed households have been grouped into one category as very few of them (20

percent) are formally employed, and the difference in income between formally and
informally employed women is negligible.

23 Each category does of course als
o include people in different socio-economic opportun it y

situations. While deeming the defined categories the most useful for aur anaJytical

purposes, non-typical cases are important for understanding the role of social relations
and networks and will be discussed in the analysis.

29

C M I

Comparing the three social categories, employed men tend to have relatively
large dwellings and large household units. Formal employment implies higher
and steady income, and is found both in government and in the private sector.
The size of the household is not solely a result of having a stable nuclear

family and many children. There is also a considerable pressure on such
households to take in relatives as well as non-relatives who need a place to
stay. At the same time better off male household heads of ten increase their
household by taking in a cohabitant in addition to the wife as an "urban form
of polygamy" (oohngoma). The wives in such households can expect to be
supported with basic necessities such as food, clothing, school uniforms and
medical expenses. Cohabitants or second wives will also normally get support,
but only for themselves and joint children. Such women will aften have
children with other men who the y have to support themselves.24

Unemployed men will normally be in a very different situation. Without a
fixed income to build larger dwellings and support a family, their household
units tend to be smaller and more unstable. There are fewer members outside
the co re nuclear unit, both due to limited means and space and because
unemployed men are less likely to have additional cohabitants. The main
income-earning option available for unemployed men is occasional and poorly
paid work such as cleaning yards for people in the formal town, unloading
trucks from South Africa and loading trucks going to Angola, getting hold of
and resell cheap imported items such as watches and perfumes, or do poorly
paid agricultural work for farmers in the vicinity of Oshakati. Most of the
key activities in the informal economy, such as making and selling tombo and
okapana, is reserve d for women. The ensuing economic importance of women
in such households has wide-ranging consequences for the position and role of
men as household heads. They aften depend on their wives or cohabitants,
and are in a vulnerable position due to their limited contribution to the
household economy.25

Female headed households generally have lower family income than male
headed households, and tend to live in small dwellings (shacks) and have
smaller household units. They als o have fewer relativ es or non-relatives living
with them than male-headed households do. At the same time, however, the
urban setting and the informal economy represent opportunities that women
do not have in the rural areas. Women in urban areas can (and of ten do) own
their own dwellings and plots, which (despite the general insecurity of ten ur e
in urban shanty-towns) distinguishes them significantly from rural women.
Furthermore, participation in the informal economy implies an income that
may be small, but nevertheless puts women in a relatively independent
position. Women are most commonly involved in the production and sale of

24 The possibility for getting formally married decreases with children out of wedlock.

People argue that with one child it is still possible, with two children it is difficult, and
with three children it is impossible. As shown in Section 4.2.4, approximately 40 per cent
of women under 20 years in the informal settlement areas have children.

25 One implication of this is that the problem of domestic violence seems less prevalent

among the poorest male headed households than among those better off. Beside their
economic situation implying limited access to alcohol, the poorest men know they can
easily be thrown out of the house they live in by their wives or cohabitants.

30

C M I

local brew (tombo or epwaka) or "fast food" (okapana) from fish and meat,
either in the shanty town itself or markets in the formal town. Some are also
involved in trade with second hand clothing, baskets, traditional medicine,
fresh fish, wild nuts and berries or agricultural products. Being married is stil
considered important, but there are also many women who argue that men are
primarilya burden and that "liking them do es not mean you have to live with
them" . Whereas some women move in with better-off men as cohabitants as
part of their survival strategy, many men seek to move in with urban women
lea ving their wives in the rural area. However, there is evidence that

economically independent women increasingly avoid longer-term co-
habitantship with men. ("Our fathers were real men (oshikopo ndora). The
men here are not real but artificIal (kopi kopi lera)").

BOX4
The size of poor households is not only kept down by economic
circumstances and the limited number of non-core members, but
also by smaller nuclear units due to excessive death rates resulting
from inadequate food and health conditions. (There is no evidence
that HIV/AIDS is more prevalent among the poorest. In fact, the
popular nation is that the disease is most common among better-
off people with a wider social network and possibility to travel to
areas where it is believed the illness is endemic, such as Angola and
urban centers in the south). The size of the core household unit
among the poor is also kept down by the tradition of poor
household units "giving away" children to extended family

members or others in a better position to take care of them on a
permanent basis (okutekulifa).

In sum, the household remains an important social unit, but is under
considerable pressure. It tends to be smallest and most unstable among the
poorest sections of the shanty population. However, while poor male-headed
households tend to break up and leave men single for periods before theyenter
new co-habitantships, poor female headed households maintain a co re unit of
mother and children. For the richer and larger household units, the main
problem is to relate to the pressure for sharing resources with relatives and
non-relatives mo ving in to the dwelling unit.

4.3.2 The Extended Family

Few if any households are, as already emphasized, self-contained as socIal and
economic units. People depend on relations and networks not only for food
and material goods, but also for socio-cultural reasons. Traditionally the most
important social entity in addition to the household both for economic and
social security is the extended family (aakwanezimo).

The Owambos are matrilineal, meaning that ego's family is traced through the
mother's kin and with the mother's brother being the most important socIal
person. There are traditional cultural rules for rights and responsibilities
related to the aakwanezimo, but these are undergoing continuous changes and
are subject to negotiations in each individual case. Authority and guidance,
responsibility for indemnities as well as inheritance are stil mainly vested in

31

C M I

matrilineal kino Egos father's extended family has rights and responsibilities in
cases where the extended family member is married, but rarely in the (more
common) situation where the biological father is not formally married with
the mother of his child. In general terms, the ro le of the matrilineal kin is
being reduced to primarily include direct descendents (i.e. consanguines), and
the role of the patrilineage has become more important with the changes in
residence pattem from uxorilocal to neolocal urban residence.26

In addition, there are changes towards an increasing "commercialization" of
extended family relations in the sense that transactions increasingly involve
money. Urban households are in a special situation as their extended family
normally does not live in the same area as they do, implying that relations
have to be maintained with people living in vilages that may be difficult and
expensive to reach. While rural family members visiting town are expected to
bring traditional foodstuffs such as milet (mahangu), spinach (evanda/ekaka),
meat (onyama) and traditional brew (omakadu/omalovu), urban family
members visiting vilages are expected to bring money or capital items such as
clothing, canned food, cooking oil, beverages, and detergents. In addition,
participation in agricultural tasks such asplowing, sowing, weeding,
harve sting and herding is of ten done by paying a laborer or relative to fulfil
responsibilities. And finally, important socio-cultural events such as birth,
weddings and funerals demand money as well as physical presence.

Urban male-headed households with employment and income are in the outset
in the best position to maintain relations with their extended family and rural
areas of origin, but they are also most susceptible to claims from the extended
family.27 With the large expenses related to both traditional and modem
marriage, employed men are most likely to be formally married which unites
extended families and defines rights and obligations.28 Maintaining relations is
pardy a question of visiting and being present on important social occasions,
and pardy of contributing with money or goods. With the commercialization

of agricultural production, male headed households with employment and
income are also most likely to have catde and own land in their vill age of
origin which further strengthens relations both culturally and by employing
people to take care of their assets.

Unemployed men with limited and fluctuating income are at a double
disadvantage. On the one hand, their options of maintaining relationships are
limited as they wil have problems visiting the rural are as and maintain the

relationship with money and other assets. Of ten they wil also lack catde and

land, which are important parts of extended family relations in their own
right. At the same time, being unemployed and poor is a violation of the very

26 Women tend to argue or fight against this developing, having vested interests (for

themselves and their children) in a system where they will inherit both from matrilineal
brothers/uncles and from husbandslfathers.

27 It is, in fact, nearly impossible for better off households to cut ties with the rural areas

even if they want to without getting into serious socio-cultural problems with the

extended family .
28 Traditional marriage transfers of cattle or other commodities have in the urban context

been substituted by lavish marriage celebrations that only the better-off can afford.

32

C M I

rationale for becoming urban and leaving on es cultural roots in the first place.
Extended family members are less likely to vis it a relative who is poor with
inferior housing, not only because there is limited space and less to gain but
also because having children or other relatives in town who have not made it
is considered a disgrace. In the same vein, many poor urban men do not return
to their vilages of origin even though socio-economic conditions would be
better there, because they cannot face their extended family members as
"fail ures" . 29

W omen and female headed households are again in a special situation. Their
ties with the extended family is to a large ex tent based on children as extended
family property. Children are of ten sent to the rural areas to be taken care of
by the mother's matrilineal family, which is als o expected to support the

urban based women with mahangu or other agricultural products. Women do
traditionally not own cattle and agricultural land in rural areas themselves,

even though the increasing commercialization of land has opened up options
for land ownership. There also seems to be lower expectations for money and
other contributions from women in town, based on the traditional perception
of men as principle breadwinners. However, there are increasing concerns in
the rural areas about the economic burden of having a large number of
children from urban-based relationships. Urban women acknowledge the
strain on relations with the rural areas this represents: "We put them there,
and if we do not contribute what can they do?"

Traditionally the elan (ezimo/epata) also had responsibilities to support people
finding themselves in difficult situations. The elan defined totemic categories
within one and the same ethno-linguistic group, tracing a common root
through the animal or plant which provided the elan name. The elans inelude
ovakwanangobe ("ox"), ovakwanime "lion", and ovakwanekamba ("hyena").
Each individual belongs to two elans, of which the mother's elan was

considered most important. Clan responsibilities were particularly important
when people traveled away from their own area into other territories, where
fellow elan members were expected to treat the visitor as a brother (or sister).
The importance of the elan was gradually reduced in rural areas with the
impact of Christianity, apartheid policies, increasing commercialization and
changing residence patters. Having said this, there are indications that the elan
is becoming an important part of people's survival strategies in poor urban
areas. Well-off households try to suppress the notion of ezimo-responsibilities
in order to avoid further burdens, whereas poor households try to evoke them
in order to be able to put elaims on a larger number of people for support.

Extended family relations are increasingly important for social and economic
security as the urban household units become less stable. They are particularly
important for maintaining involvement in the rural economy, for fulfilling key

29 Preliminary data indicate that people migrating to town normally come from the better-off

households (Pendleton and Frayne 1998), which will make the problem of failure even
more serious. It is quite common among the poorest household to complain that the only
people who will help them are their brothers and sisters. ("If you do not have brothers
and sisters from the same mother, you are in real trouble").

33

C M I

socio-cultural obligations, and for solving problems involving indemnities or
other larger one-time payments. It is als o increasingly important for obtaining
further education, which is seen as an extended family responsibility. Again
the poorest households find themselves in the most disadvantaged position for
maintaining extended family relations, primarily because the relations have to
be filed with material content to be sustainable. Having said this, it is easier
for poor female-headed than poor male-headed households to maintain such

relations because of the importance of children in the extended family.

BOXS
The changing perceptions of the role and responsibilities of the
matrilineal and patrilineal extended family respectively is perhaps

most clearly demonstrated in ca ses of inheritance. Traditionally the
principal heirs were members of the matrilineage (uterine brother or
oldest son of the oldest sister in cases of men, and oldest son or
uterine brother in cases of women with property). In both cases the
principal heir had an obligation to distribute part of the property to
all the nearest matrilineal relatives. There has been a gradual change
towards favoring the direct descendants of the male household head
and the patrilineage, at the expense of wives and the matrilineage
(partly but not solelyas a result of changing residence patterns).

Complicating this has been recent government laws enacted to secure
rights of inheritance for wives and women. Cases of inheritance are
increasingly taken to modern (as opposed to traditional) courts,
which is beyond reach of the poorest sections of the population.

4.3.3 Neighbors and Friends

In the daily life of households in the urban shanty-towns, neighborhoods and
friends are most important for survival strategies. Each shanty area is clearly
demarcated and under the traditional jurisdiction of a headman (mwene
wolukanda). Each shanty is again demarcated in to three to seven sub-areas,

consisting of approximately 150-200 households under the jurisdiction of a
sub-headman appointed by the principal mwene. A neighborhood (uushiinda)
is a more flexible term, defined on the basis of relations and networks. A
neighborhood will normally center around a street or a square, and include
15-20 household units. Most people in a neighborhood will relate to each
other in some way or another during a day, but a smaller number of

households (three to six) will consider themselves to be dose neighbors

(omushiinda shiinda, literally "neck-neck-neighbors").

Previously migrants moved in and established their own dwelling in a
neighborhood where they had some type of relationship with people already
living there. These could be relatives, people from their own village, or people
with whom one shared a common experience such as labor migration to the
South, life in exile or employment with SADF. With the current overcrowding,
however, many settle in areas where they do not have pre-established
relations. People emphasize that establishing neighbor-like relations is
becoming more difficult, and that they do not knoweverybody outside the
immediate neighborhood even by face.

34

C M I

For male-headed households with employment and income, neighbors tend to
be considered a threat to their own position. The household head is away at
work large parts of the day, and when retufling after working-hours there is a
strong pressure for sharing resources with poorer neighbors. Relations easily
become non-reciprocal, and the immediate need for establishing lasting bonds
or patron-client like relations with poor urban neighbors is limited. In fact,
when special needs arise, people in this category ten d to enter relations with

people outside the shanty-town such as colleagues or friends to avoid
situations of outstanding claims. Exceptions are small-scale businessmen based
in the shanty area who need customers, and local traditional or party-based
politicians who need followers.

As opposed to employed male-headed households, unemployed and poor men
tend to depend heavily on immediate neighbors and friends. We have shown
that they normally cannot rely on their extended family, and they are also
physically present in the shanty areas during large parts of the day. Exchange
relations involve money for food and beverages, medical expenses, cloths, fuel,
transportation etc. for the household. Depending on limited and fluetuating
income from casual work, people in this category wil have access to money at
different times and are generally only able to establish reciprocal relations
with others in similar situations. It is also common for poor men to circulate
around the shanty area in search for food and local brew, through a practice
called okunyanganyanga. As practically all food and drinks are consumed
outside (the shacks are small and extremely hot during the day), the people

doing this simply sit down in the group expecting or hoping to be served as
food and drinks are passed around. In doing this they try to evoke traditional
cultural obligations of sharing with people in need. The practice is frowned
upon by an increasing number of shanty-dwellers, who argue that people
doing this "have no culture" (i.e. do not behave in accordance with the
dominant urban cultural rules).

Women and female headed households are at the outset depending on
neighbors and friends both socially (to "watch the gate", take care of children,
etc.) and economically for loans or cooperation related to informal economic
activities. Women rarely go to cuca-shops or bars, and the neighborhood
streets, yards, waterposts or shading trees function as key meeting places. The
importanee and closeness of neighborhood relation between women is evident
by the fact that the notion of omushiinda-shiinda is primarily used by and

about women. Good "neck-neck-neighbors" wil always help you, can be
trusted ("a dose neighbor can breastfeed your child"), and will not gossip. At
the same time, women tend to depend heavily on their most immediate

neighbors and are less likely to establish contacts outside their uushinda. Some
argue that the practice of okunyanganyanga is not possible for women ("If we
do it, who are going to co ok the food to eat?"), and others argue that it is
dangerous to leave the neighborhood particularly at night. High incidents of
violence in the shanty-areas indicate that this fear is well-founded.

Neighbors and friends represent important social capital for the urban poor.
However, the ability and utility of entering relations of this type differ with

35

C M I

employment, income and gender. The poorest are largely confined to relations
with people in similar situations and with a limited "surplus" to redistribute.
Moreover, while poor men and male-headed households tend to depend on
the unpredictable practice of yangayanga, poor women and female-headed
households tend to enter more stable relations with other women in the
immediate neighborhood. Better off men and male-headed households avoid
entering relations of this type as asymmetrical relations may yield limited or
negative returns.

BOX6
The difference in the ro le and importance of the shanty-area for the two
categories of male headed households is exemplified by their relationship to
drinking places. These function as perhaps the most important meeting-
places or arenas for men. Cuca-shops (shebeens) are small shacks located in
the shanty-areas and selling traditional brew (tombo or epwaka). People sit
in groups and drink, and the brew is paid for and consumed under the
principle of "I pay today and you can pay tomorrow" (tornbo ontaku).
Cuca-shops are primarily used by poorer men in the shanty-population.
Bars are normally located in larger brick-buildings in the formal town or in
the outskirts of shanty-areas. Here beer in cans and wine in bottles are

bought individually, and tables and chairs make the setting more formaL.

Bars tend to be used by employed, better-off men who can afford the beer
and wine and who do not want to be involved in the system of outstanding
claims typical for the cuca-shovs.

4.3.4 Urban Relations

As opposed to the situation in many rural villages, people in urban
shantytowns meet a large number of people who they in the outset do not
know. On the one hand many people enter the shanty-town during the day or
at night. These may be relatives or friends of other residents, people passing
through and needing a cheap place to stay, aid workers, representatives from
the Oshakati Town Council, local SW APO or DT A30 politicians, youngsters in
search of quick and non-committing sexual relationships, and refugees from
Angola who need to find a place to hide. On the other hand, the shanty-
dwellers als o enter into relations with a number of people outside the shanty
area. These include officials from local government or the police, people from
banks or informal lending institutions, women selling goods at markets,
employees in shopping centers or bars, Chinese traders, teachers of ones
children, taxi drivers, etc.

All these people represent potential relationships for the shanty population.
They differ from the other types of relations and networks in that the people
in the outset are unfamiliar with each other, and they link the shanty-dwellers
to the modern and urban part of Oshakati. As parts of survival strategies
relationships to persons in the modern and urban sector represent important
potential assets. However, relating to them constructively also demands

JO SW APO (South West Africa People's Organization) was the most important Namibian

independence movement until 1990, and has since then been the main political party in
the country. DTA (Democratic Turnhalle Alliance) was established during South African
rule and has been the main opposition party since independence.

36

C M I

resources and the command of cultural codes and knowledge that is unequally
distributed among the shanty population. Among the most important of these
are education and the ability to read and write.

BOX 7
Owambos first modern shopping complex opened in 1996, with the fast-
food restaurant "Kentucky Fried Chicken" soon developing into the "hot-
spat" for the Oshakati middle dass. Very few of the Oshakati shanty

population have ever been there. This is not only a question of

affordability. Entering the premises als o requires knowledge about

cultural codes that are different from what people are used to: Buying fast
food (okapana) on the street implies that you can select and pay for the
piece you want, that you can eat the same way you do at home, that you
can complain if there is something you do not like, and perhaps even that
you can pay at a later stage. At KFC you have to be able to read and
understand the menu, it is a "take-it-or-leave-it" situation, and you are
normally in for a lonely meaL.

The type of external relationships discussed here are particularly important as
potential sources of employment. Both the formal and the informal labor
market largely recruits people by way of relationsships and networks, except
in the case of qualified occupations demanding special skills (teachers, nurses,
policemen etc.). Having indirect or direct access to potential employers such as
local government officials, shop-owners, patrons of informal markets, farmers
and cattle-owners in the vicinity of Oshakati, Angolan traders or people
involved in theft or smuggling increases the options for employment and
income. Conversely, not establishing external relations reduces the options for
finding employment substantially.

Again employed men with regular incomes are in the best position to relate
constructively to incidental encounters of this type. They wil be in contact
with the largest number of people during a day, and have the means to engage
in reciprocal relations ranging from obtaining formalloans to easing the
reaction of the traffic police. Employed men are also most likely to be active in
political parties, business-associations, labor-unions or other inter es t groups

where people with different backgrounds and resources meet. Ability to read
and write, which we have seen is unevenly distributed among the shanty-
population, also opens up a number of job-opportunities out of reach for the
illiterate.

Poorer unemployed men have, as we have already shown, at the outset a more
encapsulated social network primarily confined to the shanty-area. Their

limited command of urban cultural codes (including speech and dress) and
qualifications (including language skills and literacy) also inhibits them from
utiliÚng the urban modern context of Oshakati constructively. In addition,
coming from informal settlement areas carries with it a negative stigma and
practical problems related to the lack of an official address, inadequate

personal documents etc. At the same time, poor unemployed men have most
to gain by trying to exploit a wider set of relations. Finding patrons (in

business, in politics or in other spheres of life) can be vital for access to

employment, money and other assets. When established, however, relations

37

C M I

tend to be asymmetrical with low returns in terms of payment or other

prestations. Many unemployed men also become involved in ilicit activities
through patrons, ranging from petty-crime to large-scale smuggling.

Women and female headed households als o have a restricted external network
at the outset, mainly related to informal economic activities. They are less
inclined to seek occasional work or other alternative sources of income, as
many types of occasionallabor are not accessible to women. Women from the
shanty-areas also have a more restricted access to other urban arenas such as
government institutions, supermarkets, pubs and discotheques, hospitals and
schools, both because they are women and because poor women have
difficulties fulfilling the criteria (dress-co de, language, education etc.) for
crossing male dominated cultural boundaries. Establishing more permanent
relations with people in the modern urban con text is also difficult due to the
stigmas associated with living in shanty-towns. Relations easily become

exploitative: Though the traditionally liberal view of sexual relations outside
marriage in Owambo society makes the concept of prostitution ambiguous,
poor women from the shanty-towns easily find themselves in relations where
sexual services are given in return for food or other necessities.

BOX 8
As with all the general survival strategies described, there are exceptions
to the rule. Two single women in the informal settlements came to

Os haka ti ten years ago, and soon became involved in selling okapana and
other traditional female modes of subsistence. They slowly expanded their
business by establishing contacts with basket producers in the rural areas,
informal money lenders in Oshakati, truck-drivers from South Africa and
a woman from their village who has made it through secondary school
and studies in Cape Town. On the basis of these relations they have
developed a flourishing business selling Owambo baskets in Cape Town,
and importing second hand cloths for sale in Owambo. They have built a
good brick house, but still live in the shanty area. "This is where we
belang, and have aur friends and neighbors".

For the poor in urban shanty-towns, relations and networks in the formal

town are particularly important as potential sources of employment and
income. Again the poorest find themselves in a disadvantaged position due to
lack of resources and inadequate command of cultural codes and knowledge.
Poor men have no natural arenas or points of contact in the formal town.
Poor women are generally even more restricted in terms of accessible arenas,
with the important exception of informal markets.

4.3.5 Associations

Relations and networks in urban informal settlement areas are of ten given

structure through the creation of voluntary associations. These may be related
to politics (such as political parties, ethnic groups, or professions), urban
service provision (including water, electricity, garbage collection), alleviation
of poverty (churehes, local branches of the Red Cross etc.) or other types of
interest groups (such as women's organizations and associations of traditional
healers). As opposed to urban settings in e.g. South Africa, however,

38

C M I

associations do not seem to be prominent in the informal settlement areas in
Oshakati. The few existing are either based on traditional rural institutions, or
established through external initiatives.31

The most visible associations in the shanty-areas are Community Development
Committees (CDCs). They were established on the initiative of a Danish
development organization in early 1993, with the immediate task of tak ing
part in the planning and implementation of an urban upgrading project and
the longer term objective of being a self-sustainable community based

development organization (CBDO). The CDC in each shanty area was
democratically elected, and selected its own officers. The CDCs functioned
well as long as the project was present with finance and advice, but largely
collapsed when the project withdrew in 1995.

The immediate reason for the collapse were irregularities related to funds, but
the more profound reasons were lack of political legitimacy both in the
shanty-areas and in relation to external political structures. The Oshakati
Town Council came to see the CDCs as a threat to its own position as a local
authority. Internally the committees suffered from the lack of participation by
the most resourceful households (who have their main identity outside the
shanty-areas) and the least resourceful households (who were primarily
concerned with immediate returns in the form of employment and not with
the idea of community development) (Frayne and Pomuti 1997).

Other externally initiated associations include women's groups (mainly related
to SW APO), housing associations (linked to the Namibia Housing Action
Group) and various credit socIeties (such as the Namibian Development Trust,
the Rural Institute for Community Empowerment etc.).32 Common for all of
these is that they are considered external entities, and relate to a limited
number of people. The people involved ten d to be the ones with resources in

terms of money as well as education and knowledge. The poorer sections of
the population tend not to be involved, partly due to lack of necessary

resources but als o because they do not think that formal associations will do
anything for them. ("All they do is talk. They never help us").

The two institutions with the strongest impact on the shanty population are
political parties (mainly SW APO and DT A, but also the recently established
Congress of Democrats, CoD) and urban based traditional authorities. They
are not primarily important as organized structures delivering goods and

services, but as symbols of political processes and contradictions in the shanty
areas.

31 More research is needed to understand the reasons for the apparent lack of associations or

interest groups in Oshakati. Relevant factors may be to the lack of tradition for

democratic associationallife in traditional Owambo society, the policies of the apartheid
state, and the recency of the urban crisis in Namibia.

32 As opposed to rural Owambo, the church does not have a strong position in the shanty-

areas with the exception of the charismatic Kimbangistas (Simão Toco Church) which
solely recruits Angolans.

39

C M I

Only a small number of people in the shanty areas are active members of
political parties, and participation in national and local elections has been very
low. SW APO is the dominant political force in Owambo as a whole with more
than 90 percent of the votes in national and local elections since independence.
However, DTA (and CoD) has a relatively strong position in the urban shanty
areas, primarilyas a symbol of opposition to the current national and local
government. This is partly due to the large number of former soldiers with
South African sympathies in the areas, but increasingly also a sign of
frustration with SW APO due to what many perceive as exclusion from socio-
economic developments in the country.

All four shanty-areas have a headman (mwene wolukanda), and sub-headmen.
The headmen are formally linked to wider traditional authority structures, but
their status and role differ. 33 The most important tasks of the headmen have
traditionally been related to land allocation and adjudication of criminal cases
in their area of jurisdiction. Both tasks are now formally the responsibility of
state bodies (i.e. the Oshakati Town Council, the police and the court system
respectively), but due to the unresolved legal status and marginal position of
the shanty areas the headmen still deal with both types of issues albeit under
new conditions. They are not allowed to charge for allocating land, and
serious offenses within the areas of traditional jurisdiction (rape, murder etc.)
are treated both in traditional and formal courts.

BOX 9
"People in (my community) come to me the whole time, both in
connection with land and criminal cases. At the same time the councilors
are afraid of us. They know we have power, and can get them out of
office if we organize aur people. But it is also difficult to be a headman
in a town. We have a terrible problem with AIDS, and many people here
show no respect for each other. They also quarrel and fight anout a lot
of things they would not fight about in the village. I recently had a case
where ane man wanted to take his neighbor to court because the
neighbor's cock was making toa much noise in the morning. Imagine!"

Two additional institutions, albeit more in the anthropological sense of the
word and without formal organization, are in the process of becoming
increasingly important in the shanty areas. Football clubs and shanty-based
gangs (botsotsos) primarily involve young unemployed men, who we have
seen represent the most vulnerable and frustrated social gro up in the shanties.

Oshakati-based football-clubs (including the "Almighty Stars") are perhaps
the only modem urban institution where socio-economic background and
shanty-residency do es not determine accessibility. Skils are decisive, and
players from poor backgrounds are supported either by people in the
neighborhood or local businesses. Games at the Oshakati Stadium is one of
the few arenas where people from the shanty-areas and the formal town meet
as equals (provided the former can pay for the ticket, of course), and the

33 Two of the headmen in the four shanty-areas are historically part of traditional structures,

while two were appointed by the South African administration. The status and ro le of the
headmen is further complicated by their affiliation to different political parties.

40

C M I

recent introduction of TV in public places showing Namibian who have made
it in professional football has made football a career objective for many
shanty-dwellers.

Botsotsos tend to be organized in informal gangs with a patron inside or
outside the shanty-area. They normally carry out petty crime, but are also
involved in more serious offenses (robbery, rape and even murder). Most of
these activities take place outside the shanty-area where the y belong, but may
als o be carried out in the shanty itself. The population there know who the
botstsotsos are, but cannot do much due to their intern al codes of conduct and
the limited capacity and efficiency of the traditional and modern court
system.34 The botsotsos tend to defend their activities with reference to the
difficult situation they are in, and of ten blame the new government for their
fate (" (President) Sam (Nujoma) brought us hunger", "Sema okwa eta
ondjala"). With deteriorating socio-economic conditions in the shanty-areas,
botsotsos are likely to have a considerable impact on the shanty-areas in the
future.

The limited number of organized associations in the Oshakati shanty areas
seems to be related to the differences in interests and resources between the
better off households (with their primary identity outside the shanty areas)
and the poorest households (whose strategies are determined by more

immediate concerns for survival). As such, it further substantiates the
differences already described related to individually based relations and
networks of the poor and less poor respectively. Traditional and modern
political institutions seem to represent the strongest potential for community
organization and initiatives. They have clear structures of authority (i.e. in line
with traditional Owambo culture), and are less demanding on the poor
themselves. Also here, however, people actively involved tend to come from
the most resourceful households.

4.3.6 Marginalization and Social Exclusion

We have argued above that it is possible to dis cern general traits regarding the
survival strategies of the three main social categories in the shantytowns in
Oshakati. Employment and gender have been singled out as the most
important determinant for household organization, as well as for the types of
social relations and networks entered.

Our arguments are summed up in Figure 2. Employed male headed
households generally have access to all the main types of social networks

identified, and are in a position to fil them with necessary material content to
be sustainable. Poar male-headed households are much more restricted in their
relations and networks. They do not have the necessary resources to establish
or maintain them, and hence have limited access to the type of social

34 As poor young men Botsatsas of ten have weak ties with their extended family which
makes a traditional punishment less effective, or they are let out on bail by the police or
public court system due to lack of capacity to pursue cases of this type.

41

C M I

capital needed to alleviate poverty. Poor female-headed households are also
restricted in the types and extent of relations they can enter, but have doser
and more sustainable networks built on gender and on children as extended
family property. Even though they are equally poor as their male counterparts
in terms of household income, they are thus in a better position to secure
benefits through social networks.

Figure 2. Types of relations and Networks

HOUSEHOLD TYPES OF RELATIONS / NETWORKS
CA TEGORIES

The Extended Neighbors Urban V oluntary
Household Familv Friends Relations Associations

Employed Large & Extensive & Limited Extensive & Extensive &
Male-headed Stable Intensive Intensive Intensive
Households
Unemployed Small & Limited Extensive Limited Limited
Male-headed Unstable
Households
Female- headed Small & Intensive Intensive Intensive Limited
Households Stable

Care

Returning to our theoretical outline of socIal relations of poverty, the relations
entered by people in the shanty-areas indude both exchange relations based
on more or less shared cultural values, reciprocal relations emphasizing
material motivations, and impersonal commercial exchange relations. The
basis for these relations undergo continuous changes, but there has been a
general development towards an increasing commercialization and ensuing
marginalization of the poorest sections of the shanty population.

Commercial relations of exchange are in the process of changing from small
localized urban markets with options for bartering, price negotiations and
credit, to market-places with high er prices, impersonal relations and high er
socio-cultural thresholds. People without employment and income find it
increasingly difficult to be involved in normal commercial exchange relations.
Money is also entering recIprocal relations between people familiar with each
other in the urban context. Obtaining money as well as food, dothing or other
items increasingly requires payments of an equal amount either in cash or
kind. This is not only a result of "modern" and "urban" perceptions about
what exchange relations should be like, but equally much that most people
cannot afford to have outstanding daims for longer periods of time.

For the poorest, the main urban option that remains is to convert traditional
rural values of support to an urban context. However, a "moral economy"
where relatives, neighbors and friends are responsible for the well being of
people in the neighborhood is increasingly difficult to allude to. Culture in the
sense of interests, values, beliefs and knowledge is not shared in an urban
environment like Oshakati to the same extent as in a vilage setting, and

42

C M I

cultural perceptions will depend on where people find themselves in the socio-
economic structure.

What remains as the most relevant survival strategy for the urban poor is to
maintain relations with their rural area of origin. However, also in this case
the increasing commercialization has changed the nature of relations.
Maintaining relations through visits and exchange of goods demand access to
money, as do es participation in socio-cultural events related to birth, marriage
and death. The poorest in the shanty-towns do not only have to overcome

these obstades, but they also have to relate to the marginal status ensuing

from being poor and not having made it in an urban context.

The poorest sections of the population in the shanty-areas are thus

increasingly marginalized both in relation to their own immediate social

environment, and dominant urban political and economic processes. The
marginalization and socIal exdusion is exacerbated by an emerging "culture of
poverty", giving itself expressions ranging from apathy (as with unemployed
men), via redefinitions of dominant socio-cultural principles (as with women
disregarding relationships with men and emphasizing relationships with each
other), to anger and destructive behavior (as with the botsotsos). The "culture
of poverty" is strengthened by the increasing discrepancy between the actual

living conditions of the poorest sections of the population and the cultural
"implosion" through mass media, education etc. emphasizing what a
successful modem and urban life should be all about.

While the majority of the shanty-population seem trapped in poverty and

deprivation, social mobility both from being poor to less poor and from being
less poor to poor do es take place. This is normally not, as we have argued, the
result of "trickle down" effects of the general economic development taking
place in Oshakati, Owambo and Namibia for that matter. New employment
opportunities, enhanced access to capital goods, improved social services etc.
largely bypass the poorest sections of the shanty population.

Social mobility is, rather, of ten the outcome of initiatives of individuals and
groups to establish new relations either with the traditional and rural or with
the modem and urban. Preliminary case-studies indicate that upward social
mobility among poor women and female headed households of ten is based on
dose cooperation between women, pooling labor and resources related to
infomal economic activities. Among poor men and male-headed households, it
seems more common to enter patron-dient like relations with people in the
urban and modem setting that eventually lead to improved access to
employment and income. SocIal capital in the World Bank sense of voluntary
associations seems less relevant as a basis for mobility, because associations

are normally not accessible to the poorest sections of the shanty population.

Having said this, the vulnerability of the entire shanty-population is

demonstrated by preliminary case-studies on downward mobility. The less-
poor of ten depend on a single source of income in the form of employment.
Losing employment has immediate implications for income, which again has

43

C M I

implications for the ability to maintain other relations and networks.

Households with the best options for improving their situation and eventually
move out of the shanty-areas seem to be those who manage to diversify their
sources of income and relations by relating constructively both to the urban
and modem and the traditional and rural.

44

C M I

5. Poverty in Rural Owamb035

Urbanization in Owambo is a relatively re cent phenomenon, and the nature of
social relations and networks in towns is both directly and indirectly
influenced by similar relations in rural areas. The majority of the population
in Oshakati are first generation urbanites and maintain relations with rural
areas, and traditional rural culture (values, beliefs, knowledge) is one of
several cultural flows that influence the survival strategies of the urban pOOL
A proper understanding of urban relations of poverty thus necessItates an
analysis of rural areas and their links with urban populations.

We argued in the introduction to this report that Owambo traditionally had
strong kinship mechanisms of wealth redistribution and social security.
Structural impoverishment perpetuated over lifetimes or generations was
mainly restricted to small numbers of war-captives and slaves, and to
dependent ethnic minorities attached to more powerful dan families.
However, the arrival of European travelers, traders and missionaries in the
19th century and the onset of South African rule from 1920, began to
destabilize the political and social integration of Owambo communities.

As in urban areas, community support structures have continued to be under
pressure. The traditional redistribution at the community level carried out by
chiefs and headmen largely seems to have disappeared, even though people in
some communities continue to give parts of their yield to the traditional
authority. Moreover, traditional systems of redistribution through productive
activities in arable agriculture (ondjambi) and cattle production (oufita) stil
exist, but take place on a much more limited scale (Hakulinen 1992; Girvan
1994; Næraa and Solomon 1994).

In general terms redistribution of economic resources have become

individualized at the level of households, extended families and

neighborhoods. Households and extended families stil ten d to be more stable
and function as more coherent socIal units than in urban areas, but non-
family relations (neighbors, friends, workmates etc.) have increased in
importance as economic life has become more diversified. Having said this,
little is still known regarding the extent to which these general trends affect
the poor and the less poor households respectively.

In the following, we wil take a doser look at the villages of Ompundja and
Oniihende in order to obtain a better understanding of the basis for relations
of poverty in Owambo. The two vilages were selected in order to have one
rural setting dose to Oshakati and urban influences, and one further away
from urban centers. Our assumption has been that social relations and

35 Selma Nangulah has been respansible for planning, implementing and writing the rural

part of this study, and functioned as team leader for Martha Naanda, Frieda Iigonda and
Gabriel Daniel from SSD.

45

C M I

networks will differ within the two vilages, as well as between each village

and urban areas.

The rural part of this study is based on a combination of literature reviews,
structured interviews, group discussions and a small survey. The survey covers
10 percent of the 200 households in Ompundja, and 30 percent of the 51
households in Oniihende.

5.1 The Vilages of Ompundja and Oniihende
Ompundja is situated in Oshana region, only 20 km from Oshakati. Most
people from this vill age speak Oshikwambi. Ompundja is a large vilage with
approximately 200 households, and is divided into seven sections of which six
are led by sub-headmen. The seventh section is regarded as the main section
and led by Ompundja headman. The vilage has six schools, a church, two
dinics and a few small shops. The headmen are responsible for traditional
matters, while the councilors for the constituency are responsible for

government activities. A gravel road between Oshakati and Ompundja makes
it easy for people to go to Oshakati on a daily basis. Although there are

facilities in the village, people tend to go to Oshakati for shopping, milling of
omahangu, medical treatment, and leisure.

Oniihende is located between the Oshikoto and Ohangwena regions, with a
combination of Oshindonga and Oshikwanyama speaking people. The latter
are the majority, although the vill age falls under the jurisdiction of the
Ndongas. This village is very remote, situated about 30 km from the nearest
shopping center and about 105 km from Oshakati. Oniihende is a small
village compared to Ompundja with only 51 households, and is led by a
headman. Due to the poor roads and lack of transport facilities, people from
this village hardly visit urban centers. People walk over long distances in order
to get dean water, and the same applies for shops, schools, dinics etc. The
headman daims he is not familiar with the councilor of the constituency, and
only relates to the Ndonga Kingdom as a traditional authority.

People in both vilages normally live with their nudear family and dose
relatives, and not with non-relatives as is the case in towns. Households
normally consist of older household heads staying with their young children

and grandchildren. 43 percent of the households are female-headed, and 57
percent are male-headed. The household members who are absent are mostly
older children of the household head, belonging to the economically active

group between the age of 24 - 49 (27 percent). Many of these go to urban
areas to look for employment in order to support their families. Those present
in the village are mostly pensioners, unemployed, students and youngsters.

The average number of dwellings in each homestead in Ompundja and
Oniihende do es not differ significantly (12 and 11 respectively). What makes
the difference are the types of dwellings. The homesteads of Oniihende are
predominantly traditionaL. The homesteads of Ompundja have a combination
of traditional dwellings and brick houses and/or shacks. Homesteads normally
have a large number of dwelling units with specific functions. Assessing the

46

C M I

condition of the homesteads of the two vilages, rural poverty is particularly
visible in the case of Oniihende.

Household members in Ompundja have reached high er educational levels,
with 57 percent having completed Grade 5 or more (and thereby being
functionally literate). Some people from this village have als o completed Grade
12 and obtained diplomas and teaching certificates. Oniihende shows a very
high percentage (22 percent) of adults who do not have any education at all,
and 30 percentof those who have gone to school have only completed Grade

4 or less. Nobody in Oniihende has education above Grade 10. One reason for
this is the inaccessibility of schools. Another is that Oniihende is aremote
village and less influenced by modern life. Many argue that there is no need
for a young boy to go to school, as he wil look after catde. Ompundja is
situated next to Oshakati, and because of the urban influence education is
considered very important.

40 percent of the people in Oniihende and Ompundja also have access to
dwellings in other areas. These are mosdy located in towns like Oshakati,
Windhoek, Ruacana and Walvis Bay (78 percent), although some (23 percent)
are located in other vilages. Most of these dwellings (57 percent) are owned
by other family members. The survey shows that households from Ompundja
have access to dwellings in all the towns mentioned, while those from

Oniihende only have access to dwellings in Windhoek and Walvis Bay and not
urban areas in Owambo. This indicates weaker links with northern towns like
Oshakati.

Unemployment is a smaller problem in Ompundja than in Oniihende. More
people (15 percent) are formally employed in the former than in the latter (6
percent). Income from informal employment is not very important in the two
villages. Ompundja als o has more pensioners than Oniihende (10 as opposed
to 6 percent). The larger proportion of formally employed people in

Ompundja is related to higher levels of education, as well as the proximity to
town. Pens ion and formal employment are the most important sources of
household income in both villages, although agricultural produce (16 percent)
is also important (Table 23).

Table 23: Most Important Sources of Household Income (Percentages)

Sources of Income Ompundja Oniihende

Formal employment 30 21
Informal employment 5 7
Agricultural products 15 14
Pensions 50 57

The majority of the households in the two villages fall well below the poverty-
line (Table 24). There is also here a strong correlation between formal

47

C M I

employment and income, even though some households have considerable
savings in the form of cattle (see below).

Table 24: Distribution of Income (Percentages)

Monthly Income (N$) Ompundja Oniihende

-: 50 O 21
51 - 100 10 22
101 - 250 35 57
251 - 500 15 O
501 - 1000 20 O
1001 - 3000 5 O

;: 3000 15 O

The two rural communities belong to the communal area, which is governed
by state. Land is allocated to the villagers by the headmen. The size of plots
differs for the two vilages, ranging from 2.1 - 5.1 hectare in Ompundja and
1.1 - 4.0 hectare in Oniihende. Although there are differences in the size of
land, most people from the two villages did harvest last season. Households in
Ompundja harvested far more oratas of omahangu (mean:140) than
Oniihende (mean:34). The differences seem to be the outcome of various

factors. People from Ompundja harvest more because they own more land,
and make use of tractors and apply fertilizers because they are close to town.
Oniihende is a poor village and the household members cannot afford to make
use of these facilities. Poverty also lea ds to physical weakness, and people are
not able to work in the fields for longer hours.

Livestock and small stock are very important for households in the rural
vilages. As many as 85 percent of the households in Ompundja own cattle
and 90 percent own goats, while 47 percent of the households in Oniihende
own cattle and 73 percent own goats. None of the households in Oniihende
own more than 30 heads of cattle. Ownership of livestock is, as already
mentioned, an important measurement of rural poverty.

Finally, and further substantiating the differences between the villages, is the
fact that Ompundja households possess more capita L items than households in
Oniihende. Oniihende households tend to posses farm equipment and radios,
while Ompundja households als o possess modern items like stoves, cars,
refrigerators and even TVs. The scores of households expenditures indicate
that households from both villages spend most of the income on education
(Table 25). The second item for Ompundja households is transport.
Households from Oniihende spend more on food than on medicines and

transport. Fuel and alcohol are not very important items of expenditure in
either of the vilages.

48

C M I

Table 25: Household Expenditiure(Percentages) l,

Monthly Income (N$) Ompundja Oniihende

Education 21 22
Transportation 17 15
Clothing 14 16
Food 14 16
Medicines 14 14
Fuel 11 10
Alcohol 5 6
Other 3 1

l, Informants were asked to rank their expenditures in order of importance. A point system

has been used, giving the expenditure drawing most from the household budget the
highest value.

5.2 Relations and Networks
The two vilages of Ompundja and Oniihende are typical Owambo vilages,
with the former representing villages with a strong influence from urban areas
and the latter being representative of more remote rural settings. It is difficult
on the basis of our data to determine the extent to which the proximity to

urban areas explains the differences in levels of poverty, but it is likely that
there is a strong correlation primarily because of easier access to formal

employment and markets. Below we will take a doser look at the implications
of differences in levels of poverty and proximity to urban areas for the type of
social relations and networks existing.

5.2.1 Rural-Urban Relations

Today most households (67 percent) have household members living in urban
areas. Those who are liv ing in urban areas leave the villages primarily in order
to seek employment, so that they can support their families. The household
members who are liv ing in urban areas are mainly household heads or older
children of household heads. 35 percent of all absent household members
from Ompundja are formally employed and 44 percent are students, while 56
percent of the absent household members from Oniihende are unemployed
and 29 percent are students.

In Ompundja 50 percent of household members visit urban areas very of ten
(i.e. at least once a week), with very few members visiting urban areas seldom
(once a year or less). About 27 percent of the household members of
Oniihende visit urban areas sometimes (i.e. less than once a month but at least
once a year), and as many as 53 percent visit seldom. A relatively large
number of household members (20 percent) in Oniihende have never visited
urban areas. In both villages the most common reason for going to urban
areas is to visit relatives and friends (37 percent), and to do shopping (33
percent).

49

C M I

BOX 11
"I have family members who live in Oshakati, but we cannot afford
transport fees to visit them. Sometimes they came to visit us, and bring
things like sugar which is not available here. In towns ane has the
advantage of being closer to markets, unlike in rural areas. People in
towns normally do not pay for transport when they want to ga
shopping. There are many advantages in towns because everything is
there. One has access to most of the things (infrastructure). In rural
areas one has an advantage of having own fields to cultivate, while in
towns ane does not have access to this. Friendship in rural areas is better
than in towns. There a person cannot eat with a neighbour if he is
hungry. Here neighbours who have ploughs can plough your field for
free, in town nothing is done without any charge. Friends who help each
other in rural areas do not expect any kind of refund."

During the process of visiting, various items are exchanged which strengthens
the socIal relations between the rural and urban household units. When
members of the household visit the urban areas they take with them
agricultural products (82 percent). Agricultural products include omahangu,
omahangu flour, sorghum flour, beans, peanuts, and ekaka (traditional dried
spinach). In Ompundja, household members of ten bring back with them food
(68 percent), but als o money, clothing, artifacts and cosmetics. Household
members from Oniihende normally bring back money (46 percent) and food
(36 percent).

About 73 percent of the present household heads from Ompundja have lived
in towns, and very few have never lived in urban areas. In Oniihende, on the
other hand, most household heads (53 percent) have never lived in urban
areas. The household heads from Ompundja have also stayed in urban areas
longer (mean:13 years) than those of Oniihende (mean:8 years). Most of the
household heads from Ompundja left the town because of pension/retirement
(46 percent), or completion of studies (18 percent). In Oniihende, 50 percent
left the urban area because of inability to find jobs or retrenchment from jobs
they had.

BOX 12
We do not give support to those who have no contact with us and we
do not feel we have any responsibility towards them. The only time
we have it (i.e. responsibility) is when they die. We have to bury
them. I can on ly advice same one to ga to Oshakati because it is near
and ane can visit relatives there. Transport is not a big issue there,
but if one is in Windhoek it will be. We do not have any relatives in
the village, but aur friends are the most important people in aur daily
lives. In towns neighbours are not sa friendly, while in rural areas
friendship is highly valued. But even in rural areas same traditional
patterns are changing. Like in the past neighbours used to cook and
caU up other neighbours to eat and celebrate, but now it is becoming
rare. Now the rich do not like to associate with the poor, and the
poor envy the rich. People nowadays are even jealous of each other,
especiaUy if you have someone supporting you from urban centres.

50

C M I

The perceptions of urban areas in the two rural villages are similar. According
to most respondents, it is a good idea for people to go to urban areas in order
to look for employment. The three advantages of living in a town that people
found to be the most important are: i) there are more job opportunities in
towns, ii) there is good infrastrueture in towns, and iii) there is no hard work
in towns the way it is in rural areas. Other advantages of living in towns
indude availability of mark et places, doseness to hospitals, better educational
facilities, well-built houses, and options for meeting many different people.

There are als o advantages living in rural areas. The households from

Ompundja emphasize the options of getting food from the field, having access
to own fields for cultivation, and not needing money all the time. The
advantages emphasized by people in Oniihende are similar to those mentioned
in Ompundja. In rural areas there is enough space, and people are not as
concentrated as in towns. They also live doser to their extended family

members, which makes it easier to borrow from each other. Households
regard the villages as very safe areas with low crime rates, and a suitable place
to bring up children.

Respondents feel that people can move to places of preference, but that it is
better for them to go to Oshakati than Windhoek. Oshakati is near and people
in the vilages will atleast be aware of the situation of their family members.
There is als o easier access to employment in Oshakati than in Windhoek, at
the same time as going to Oshakati prevents one from spending too much
money.

5.2.2 Rural-Rural Relations

Turning to relations and networks in the two villages themselves, the extended
family as well as neighbors and friends play an important role in people's

daily life in both Oniihende and Ompundja. Most people stil have many
family members in their village, despite large out-migration. The
neighborhood in the villages is big, and in most cases consists of more than 10
households. In any neighborhood (amushiinda) a households has selected a
few households as dose neighbors (amushiinda shiinda). The dose neighbors
assist each other in many ways, meaning that they borrow from each other
with ease, help each other in connection with fields and livestock, etc.

Respondents from both Ompundja and Oniihende indicate that urban and
rural neighborhoods di£fer. The neighborhood in the rural areas is very strong,
and people are veryeager to assist each Other. In the past there was shortage
of fire/matches, and children were sent from house to house to borrow
fire/matches. Due to this neighbors started to build strong relations with each
other. In urban areas neighbors do not help each other as much as in rural
areas. Communication is limited between neighbors in town, and there is no
trust because money is such a big problem there.

51

C M I

BOX 13
"People from urban areas expect us who live in rural areas to accept them, but
when we visit them they don't accept us. They think that we are intruders
intruding in their family lives and they expect us to pay our expenses when we ga
there, while they do not pay anything when they come here. In 1990 I went to visit
my brother who has a house in Windhoek. I went there to get some financial
assistance from him. I used to work at Aus and Ondjoini, but by the time I went
there I was unemployed. My brother told me that he cannot assist me, because he
is paying for his house, electricity, water and food for his family and thus did not
have any money. He bought my mother a sec ond-hand dress and he bought me a
pair of trousers. He paid for my ticket to come back and never gave a cent. I came
back and gave the dress to my mother. I think it is better to live in rural are as than
in towns."

BOX 14
"My neighbourhood consists of four households, but only ane of them is my close
neighbor. She is the only one who I get along well with and she helps in many of
my difficulties. She sends her children to pound my rnahangu for me and even to
fetch water, you know I am pregnant and at the time I cannot do much. People do
not help ane another in town. The attitude is not good and cost of living in urban
areas is very high. My partner seeks help for money, while I am respansible to
seek help for water or fire wood when they are not there. I don't think my
neighbour is helping me because she is a woman. I think she is just a nice person
and she would do it even if I was a man. I also think that it has a lot to do with
personality rather than with her gen der. Same women are very bad people because
they gossip a lot, and if you attempt to seek help from them you will be the talk of
the village. I consider the woman who helps me as my relative, she is the most
important person to me in this vilage."

In rural vilages, seeking assistance is the responsibility of both men and
women. The difference appears when help pertains to the elugo (the place
where food is prepared), in which case the women do it. The men take
responsibility for things which are field and/or livestock related. Men in the
rural areas do not visit the elugo, and will not know how serious the food
problem is.

People have different perception of poverty, leading to different descriptions of
poor and rich households. The respondents from the two villages describe
poor households as those who do not have enough food to eat. This can be
seen by looking at the size and num ber of the omashisha. Poor households

have none or only a few omashisha and they are usually very small in size.
Rich households in the vilages are associated with large livestock herds, large
pieces of land, big homesteads with brick houses, and many omashisha.

BOX 15
"Now you are siting in a poar household, as you can see my house is not like the
others there is no chairs, nothing. I only have ane eshisha, at least I am supposed
to have ane eshisha for beans and another sorghum. I do not have enough food,
nobody else is taking care of me. My sticks (building materials) are falling apart
and I do not know who will help me to replace them. I do have relatives in urban
areas, but it is sa difficult to get in contact with them sa that they can help me.
They do not visit ffe at alL."

52

C M I

BOX 16
A rich household to me is one with enough rnahangu, a large herd of
cattle and big buildings. Lifestyles, clothing and intelligence all play major
ro les in rich families. Poor people can be recognised because they are
shameful. They are less intelligent, because they are afraid to explore
things. Rich people socialise easily because they are free. The pa or are

hungry most of the time. L think it is Gods will to have poor and rich,
because it is only then that the poor can work for the rich. It is just like
the fingers on aur hands, same are shorter same are langer. The poor are
there to serve the riches and riches to help the poor.

BOX 17
I think same people are poor and same are rich because there is a
mismanagement and misuse of aur economy in this country. Money is
improperly distributed among different groups of society and as aresult
same people suffer and become poorer and pa orer. The Namibian
government needs to set up policies sa that the allocation of the economic
policy targets both the rich and the poor. It should invest heavily in

projects that create jobs to all the unemployed. The poor need to have an
employed person for them to afford basic needs and the rich may play a
role by employing them. The big question is who will protect them from
being exploited.".

There are no institutions that assist old and poor people in the vilages any
more, apart from the family and neighbors. The headmen of the two villages
do not assist poor households, because they are also poor people. Rather,

they give advice or try to link up with local business people so that these can
help the poor. The church do es normally not help poor people any more, and
if they do it is mainly people belonging to the congregation. The only means
through which the poorest are helped is the drought relief program of the
government, but people in Oniihende complain that this do es not reach them.

A large number of households have experienced economic difficulties.
Respondents in Ompundja indicate that the main reason for their problems is
related to payments for education (40 percent). Others mention pregnancy,
drought, catde that get lost, death of family members and fire eruptions. In
Oniihende, most respondents experienced economic difficulties because of
poor harvests (43 percent), while others experienced difficulties because of
death, diseases and accidents.

Among those who had economic difficulties in Ompundja most turned for
help to the family in towns (40 percent). A smaller proportion (20 percent)

turned to family in the vilages (Table 26). In the case of Oniihende most
households seek help from friends in the vilage (86 percent), with a smaller
proportion (14 percent) going to family members in the vilage for help. This
is most likely the outcome of poverty being clustered around certain extended
family groups. No one from Oniihende approached persons in urban areas.

The most common type of help received is money (40 percent) in Ompundja,
and food (43 percent) in Oniihende (Table 27). Those who received money in

53

C M I

Oniihende were normally expected to pay back, even if the help came from
family members. Other households received help in the form of labor.

Table 26: Main Sources of Support in Times of Economic Difficulties
(P ercentages)

Source of support Ompundja Oniihende

Family in village 20 14
Family in town 40 O
Friends in vilage 10 86
Traditional leader 10 O
Other 20 O

Table 27: Main Types of Support Received (Percentages)

Type of support Ompundja Oniihende

Money (given as gift) 40 O
Loan (to be paid back) 20 29
Food O 43
Labor 20 14
Other 20 14

In sum, there is a considerable difference in the extent and nature of rural-

urban relations between Ompundja and Oniihende. Rural-urban relations are
more frequent and intense in Ompundja, which is located dosest to an urban
center. However, vill age and household wealth is an equally important factor.
Better-off households are in a position to maintain relationships with urban
relatives and friends, which is necessary for their sustainability. Very few
households in Oniihende have dose relations with urban areas.

Regarding internal relations and networks, there do es not seem to be any

significant differences between the two vilages. Traditional systems of
redistribution (through traditional authorities and agricultural production)

have largely deteriorated, and relations and networks of support have become
individualized to the level of extended families and neighborhoods. The very
poorest have the largest problems entering and maintaining relations of
support, partly because they have little to contribute and partly because of the
widespread perception that poverty is the poors' own fault.

54

C M I

6. Conclusions

This report has emphasized the differences between urban and rural poverty,
both as regards structural constraints and survival strategies of the poor
themselves. At the same time it has shown that there are dose links and

interdependencies between people in the two types of settings.

The poor in urban areas live under difficult socio-economic conditions created
by historical and structural processes, and their alternative survival strategies
are limited by different types of relations and networks. The poorest primarily
depend on relations with people in the same social category as themselves, and
options for maintaining and establishing relations with people in the
traditional and rural and modem and urban sector is limited.

The poor in rural areas live under equally difficult socio-economic conditions,
and primarily rely on relations with relatives, neighbors and friends in their
own village. At the same time traditional forms of redistribution have largely
disappeared, and structural constraints (access to land, access to employment)
severely limit the options for breaking out of poverty in the rural setting.

Urban-rural relations remain an important source of socio-economic security

both in urban and rural areas, but we have shown that these relations are least
relevant for the poorest. Lack of resources makes it difficult to fill the relations
with material content, and poverty in an increasingly commercialized context

is becoming an individual rather than collective concern.

Employment and inc om e is a key variable for under stand ing urban and rural
poverty, with implications for access to economic resources as well as the
social relations and networks in which people are involved.

The key condusions related to our main topic of the ro le of social relations of
poverty in urban areas are:

· The poorest sections of the population in urban areas are systematically
marginalized and exduded from political, economic and social processes.

· Horizontal transfers and social capital play an important role in spreading
risks and reducing vulnerability, but more so for the less poor than the
pOOL

. Transfers between households and individuals through horizontal relations

by definition draw on a smaller pool of surplus than vertical relations.

. Any serious attempt to reduce urban poverty in Namibia requires vertical

transfers through an active public policy commitment to redistribute not
only income and consumption goods, but also productive assets.

55

C M I

Following from this is the need for targeted interventions in order to reach the
poorest sections of the urban population. This has implications for the type as

well as for the nature of development initiatives. A detailed outline of
alternative interventions is beyond the scope of this report, and the following
suggestions should be further elaborated in other contexts:

Alternative interventions related to structural causes of urban poverty:

. Enhance the status and role of towns like Oshakati as regional growth

centers, by allocating more financial and human resources to local
government and other public institutions.

. Improve the basis for the establishment of private enterprises and

employment by developing proper town plans for physical and human
infrastrueture.

. Enhance the status and role of informal shanty areas by formalizing their
existence and including them in policy schemes for urban development.

Alternative interventions targeting the social capital and empowerment of the
urban poor themselves:

. Support the establishment of representative shanty-based institutions with

a particular objective to relate to local government.

. Support to the establishment of housing schemes, in order to enhance

security of tenure and improve options for alternative income-earning

options.

. Support the establishment of shanty-based urban services (water,

electricity, road maintenance, garbage collection), in order to improve
physical conditions and create jobs.

. Support to the establishment of women groups with a particular objective

to further develop productive activities in the informal sector.

. Support to the establishment of social activities (movie-theater, sports
clubs, daneing halls etc.), in order to create employment and improve the
social milieu particularly among the young and unemployed.

56

C M I

References

Banghart, Peter D. (1969). Migrant Labour in South West Africa and it Effects
on Ovambo Tribal Life. Stellenbosh: University of Stellenbosh.

Barth, Fredrik (1994). "Towards Greater Naturalism in Conceptualizing

Societies." In: A. Kuper (ed.) Conceptualizing Society. London: Routledge
pp. 17-33.

Baulch, Bob (ed.) (1996). Paverty, Policy and Aid. IDS Bulletin. Sussex:
Institute of Development Studies.

Bourgois, Philippe (1995). In Search of Respect. SeIling Crack in El Barrio.
Cambridge: Cambridge University Press.

Brumann, Christoph (1999). "Writing for Culture. Why a Successful Concept
Should not be Discarded." In: Current Anthropology VoL. 40 (February)

pp.1-25.
Chambers, Robert (ed.) (1989). Vulnerability. How the Poor Cape. IDS

Bulletin. Sussex: Institute of Development Studies.
CSO (1994). Population and Housing Census 1991: Basic Analysis with

Highlights. Windhoek: National Planning Commission.
CSO (1996). The 1993/94 Namibia Household Income and Expenditure

Survey - Living Conditions in Namibia. Windhoek: National Planning

Commission.
de Haan, Arjan (ed.) (1997). Urban Paverty: A New Research Agenda. IDS

Bulletin. Sussex: Institute of Development Studies.
de Haan, Arjan and Simon Maxwell (eds.) (1998). Poverty and Social

Exclusion in North and South. IDS Bulletin. Sussex: Institute of
Development Studies.

Devereux, Stephen et aL. (1996). Namibia Poverty Profile. A Report for SIDA.
Windhoek: SSD/University of Namibia.

Devereux, Stephen et aL. (1993). Urban Situation Analysis. A Study of Five
Namibian Towns. Windhoek: Social Sciences Division, University of
Namibia.

Frayne, bruce and Akiser Pomuti (1997). Oshakti Status Quo Report.

Discussion Document for the Development of a Community Support
Programme by Ibis for Oshakati, Oshaana Region, Namibia. Windhoek:
SSD/University of Namibia.

Girvan, LoriAnn (1994). National Report on Women, Agriculture and Rural
Development.Windhoek: Ministry of Agriculture, Water and Rural
Development.

GON (1993). Demographic and Health survey 1992. Windhoek: Ministry of
Health and Social Services.

Graefe, Olivier et aL. (1994). Informal Settlement and Institutional Survey in
Rundu Town. Windhoek: Department of Geography, Univerity of Namibia
and CRIAA.

Hakulinen, Maija (1992). Means of Livelihood in Northern Namibia.

Helsinki: University of Helsinki, Institute of Development Studies.

57

C M I

Hamata, Sylia et aL. (1996). A socio-economic Assessment of the Enclosure of
Communal Land within the Townland Boundaries of Oshakati and
Ongwediva, and the Relocation of the Ndama Settlement in Rundu.
Windhoek: SSD/UNAM.

Hangula, Lazarus (1993). The Oshakati Human Settlement Improvement

Project. Windhoek: Social Sciences Division, University of Namibia.
Hannerz, Ulf (1992). Cultural Complexity. Studies in the Social Organization

of Meaning. New York: Columbia University Press.
Hishongwa, Ndeutala (1992). The Contract Labour System and its Effects on

Family and Social Life in Namibia. A Historical Perspective. Windhoek:
Gamsberg Macmillan.

Keesing, Roger M. (1990). "Theories of Culture Revisited." In: Canberra
Anthropology VoL. 13 (2) pp. 46-60.

Keesing, Roger M. and Andrew J. Strathern (1998). Cultural Anthropology. A
Contemporary Perspective. Fort W orth: Harcourt Brace College

Publishers.
Lipton, Michael (1980). "Migration from Rural Areas of Poor Countries: The

Impact on Rural Productivity and Income Distribution." In: World
Development VoL. 8 pp. 1.24.

Macharia, Kinuthia (1997). Social and Political Dynamics of the Informal
Economy in African Cities. Maryland: University Press of America.

Marsh, Allan and Mary Seely (eds.) (1992). Oshanas. Sustaining People,
Environment and Development in Central Owambo, Namibia. Windhoek:
National Planning Commsion.

Mauss, Marcel (1967). The Gift. New York: WW Norton & Co.
Melber, Henning (1996). Urbanisation and Internal Migration: Regional

Dimensions in Post-Colonial Namibia. Windhoek: The Namibian
Economic Policy Research Unit (NEPRU).

Melhuus, Marit (1997). "Exploring the Work of a Compassionate
Ethnographer. The Case of Oscar Lewis." In: Social Anthropology VoL. 5
(1) pp. 35-54.

Mitchell, J.c. (ed.) (1969). Social Networks in Urban Situations. Analysis of
Personal Relationships in Central African Towns. Manchester: Manchester
University Press.

Moser, Caroline (1996). Confronting Crisis: A Comparative Study of
Household Responses to Poverty and Vulnerability in Four Poor Urban
Communities. Washington: Wodd Bank.

Næraa, Trine and Co lette Solomon (1994). A Socio-economic Baseline-Survey
of Uukwambi. Windhoek: SSD/UNAM.

Pendleton, Wade and Bruce Frayne (1998). The Namibian Migration Project.
Windhoek: Social Sciences Division, University of Namibia.

Portes, A. (1998). "Social Capital: Its Origins and Applications in Modem
Sociology." In: Annual Review of Sociology VoL. 24 pp. 1-24.

Scott, James (1976). The Moral Economy of the Peasant. New Haven: Yale
University Press.

Tannerfeldt, Göran (1995). Towards an Urbanized World. Urbanization and
Development Assistance. Stockholm: Swedish International development
Agency.

58

C M I

Tostensen, Arne et aL. (eds.) (1999 forthcoming). Associational Life in African
Cities: Urban Governance in an Era of Change. Uppsala: Nordic Africa
Institute.

Tvedten, Inge and Moono Mupotola (1995). Urbanization and Urban Policies
in Namibia. A Discussion Paper Prepared for the Annual Meeting of the
Association of Local Authorities in Namibia luly 1995. Windhoek: Social
Sciences Division, University of Namibia.

Tvedten, Inge and Akiser Pomuti (1994). The Oshakati Human Settlement
Improvement Project: A Socio-Economic Baseline Study. Windhoek: Social
Sciences Division, University of Namibia.

Tvedten, Inge and Akiser Pomuti (1998). "Namibia: Urbanization in the
1990s." In Search of Research. Approaches to Socio-economic Issues in

Contemporary Namibia. Windhoek: The Namibian Economic Policy
Research Unit pp. 103-133.

UN (1998). Population Distribution and Migration. New York: United
Nations.

UNCHS/HABITAT (1996). An Urbanizing World. Global Report on Humna
Settlements 1996. Oxford: Oford University Press.

UNDP (1991). Cities, People and Poverty. Urban Development Cooperation
for the 1990s. New York: United Nations development Programme.

UNDP (1998). Namibia. Human Development Report 1998. Windhoek:
United Nations Development Programme.

UNICEF (1990). Household Health and Nutrition in Namibia. Windhoek:
UNICEF.

WB (1991). Urban Policy and Economic Development. An Agenda for the
1990s. Washington D.C: World Bank.

Wratten, Ellen (1995). "Conceptualizing Urban Poverty." In: Environment
and Urbanization VoL. 7 (1) pp. 11-36.

59