. t
.fQ-A.

53...":

. .L . I. .

 

1

1......533 :
: Ill}...

.5531:

6':

5.5.3} 3
. 4"...)
rt... ..

9.21.1. v (I.
A .2 I... I155!
our.) .4.

4.3.7.1.?!
. v) (I

twirl
.319...» , .k
.15).... A. .a (314:?!
:i.....1.s .ac‘... 3...!
yaw...?333\§
.tliiglt!
3.4!»... £5.22.
0.. .2. it.
\u.......::.....
I 1.3.5»... .25...
2....odilaaxi...
t‘ 3);!)

, .be; 3..ng
1.3.1.2.. lav... .v.‘:..9l$i\a
3.5:... . 1.1.3. 12:75:31.... .1):..)’. .
31...... :. 5......1] 3723:]..1 1.3.5.)...551).
)3 51:54.81 {A‘.I5.¢nl’l.o??.1\..

.I‘l.
.34 v. .a .
Sign

..
, . it'll
(5..) I): '13): fizz...
A... .31.... as: a: £253! 53:34.13)?!
:1... 1.29)....313: 5.213.}..5’1.’ . {31.035411}
.1 at... . V 1-..... :..v..|\...l..)fi. .0.§z..f..{. ti.)
)5 . la! 313‘ Villa...)
)1 .341.
Iris}!!!

t.
52......7 .13. 53.55....
$21.3...575132E .1...‘ .
9|: )3... .. . 5713.11}...
.lKfiEbiaLl.)
. .1 it!!!
2.))...Slt
.I:)3)\Ia .

:II,.‘.(tr!:
. a)r..3>~xn....
\n... . i-P.\.1...§t>:..xx.
:11... 1:23.!)
1:... i.

51):)... x i
11:31) 3. II:

I...» . .(i. $5.153; alizf.

I: .l . 31.1..-) $1513.!
.2 33.23135». . {yrlsznttr .
1... ~55|1§§i Z: )lvy‘.lttta..?lal.>
A!) 3.11:8le , .‘1Enxlllii?{r‘
.: 3.9.. I 7.. kill—3.

3;. .\
.1533... is.
.121...
5):...57: 1:
£91 I l.
1.. 1...... .. s
.15)....
I i» I. .55: v
1)....

£5.21! .3.)
. 31......11 . , .
. ...l...).l. . , (4!!!! {fr
30.... ‘83.. .99 (11.01-..- 2...: A\.
:rl)§ gift- . 31!).
r . nt'flt. 159......
ii‘ (I

tilt! I
.l’lvtil:ni I)
i.)i§§)a!..53¢¥.

. A!§1§.|Yv)1.

.tp)....!' 53.1.3)!!! £1.31... v!

)al...1$1 .«8. u I. ‘11:... .1 .

I31... 2. J .
).D.l1.l..: .1!!.). I..\
ii?.$ff..'.l. \;.\L
.EIVI§ I\ III...

rfv'lflttltl
5.31).?! 4. A

3. J. 55‘ . 9.1.5:).
.3 t. t! t.:lf.¢....\
2..)
(Insist I
7.)...142... o).l)v.\(.\rl 1....
I I»... ...rl...tl1€+ln$l€rrlr({r $9I\Ifr.ll
l . . .k : Di:)v§ltvu}x {\L

 

a!
(2....

 

of ; lllllllllll||l|||l|l|||ll|l|llllllllllllllllllllllllllllllll

31293 01706 9836

This is to certify that the

thesis entitled

COPING MECHANISMS OF RURAL HOUSEHOLDS DURING
FOOD SHORTAGES IN KWAZULU-NATAL, SOUTH AFRICA

presented by

Brent McCusker

has been accepted towards fulfillment
of the requirements for

M.A. degree in Geography

 

or profess”

Date December 16, 1997

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

hhvfi—fi—r‘fi.

' 7443— H'V—vav“
' RT T‘fi—T—fifih

      

LIBRARY
Michigan State
University

 
    

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

MAGIC 2

DATE DUE DATE DUE DATE DUE

 

 

F E‘Bfi it i ii‘ fia-‘A

 

NOV ”1 9
W

 

A A

 

 

 

A

JUL (3 31320620 3

 

 

 

 

 

 

 

1/98 CJCIRC/DaleDuep65-DJ4

COPING MECHANISMS OF RURAL HOUSEHOLDS DURING FOOD SHORTAGES
IN KWAZULU-NATAL, SOUTH AFRICA

By

Brent McCusker

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements

for the degree of

MASTER OF ARTS

Department of Geography

l997

Abstract

Food shortages are common throughout Africa. South Africa’s unique
political economy has hindered research on food security in that country. This
study examines the food security of a rural area of the KwaZqu-Natal province.
It first uncovers some basic information on the area then proceeds to analyze
relationships in the data. The use of GIS in food security studies is explored and
spatial relationships are explained.

The food security situation was found to be similar to other parts of Africa.
Food shortages occurred after droughts in the 1980’s and 1990’s. Political
institutions were found to have impact on the food security equation, as well as
the loss of hunting rights due to encroachment of game wardens from the nearby
Hluhluwe-Umfolozi Game Reserve. Basket making provides a significant source

of money for food in times of shortage.

Copyright by
Brent McCusker

1997

Acknowledgments

Many individuals made this work possible, but none more than my wife,
Eileen. Through those long winter nights alone, she persevered and encouraged
me regardless of my absence. I would like to extend my thanks to my advisor, Dr.
David Campbell whose constant support and comments gave this thesis direction
and meaning and to the Department of Geography for extending their financial
support. I owe a great debt to Dan Weiner who helped me pursue my interest in
Africa, Cathy Oelofse, Peter Schmitz, Shirley Brooks, Thokazani Xaba, Annie
Maharaj and Dr. David Hemson- without their help I never would have survived
the UDW experience, and to Rebecca Manyoni for her wonderful and endless
knowledge of Zidonini and the English language. To Ella Fitzgerald for her
wonderful voice which sustained me during the many trips up and down N-2. And
finally to my parents, who have always supported me despite the fact that they

could never understand why I would ever want to go to Africa.

 

Table of Contents

List of Figures

List of Tables

List of Terms

CHAPTERS

1 . INTRODUCTION

A. Statement of the Problem

B. General Hypotheses

C. Specific Hypotheses

D. The Setting — South Africa

E. The Setting — The Study Site

2. METHODS

3.

A. The Interview Schedule and lnfonnal Interviews

8. Quantitative Data Analysis

REVIEW OF THE LITERATURE

A. Theoretical Setting : Political Ecology

B. Questioning Scale

C. Political Ecology, Traditional Resource Management and
Global Change :Divergent Views

D. Critique

E. Food Security in Africa

F. Food Security in South Africa

vii

viii

(DUI->00

21
23
26
26
29

33
34
36
58

4. RESULTS 61

A. Summary Statistics 62
B. Analysis of Statistical Tests 71
C. Results of Testing 72
D. Qualitative Analysis 75
E. GIS Analysis 79
F. The Hypotheses and The Findings 84
5. DISCUSSION 96
A. Findings and The Study Area 96
B. Broader Questions and Issues 104
C. Problems and Opportunities 109
6. CONCLUSION 112
APPENDICES 116

BIBLIOGRAPHY 118

vi

List of Figures

Figure Page
1. South Africa, 1997 7
2. Zidonini Study Area 10
3. KwaZulu-Natal with Cities and Study Site 13
4. Hlabisa Study Region 16
5. Approximate Area of “The Corridor” 17
6. Percentage of Respondents with One or Fewer Fields 86
7. Percentage of Respondents with More than Three Fields 87
8. Respondents Reporting No Assistance from the Zulu Government 88
9. Number of Shortages in the Past Ten Years 89
10. The Impact of Hunting on the Number of Food Shortages 90

11. Respondents with Three or More Shortages & Distance from the Nduna 91
12. Respondents Reporting No Hunting Activity and No Help from the Gov’t 92
13. Sites with the Highest Levels of Vulnerability 93

14. Percentage of Respondents Who Hunt 94

vii

List of Tables
Table
1. Components of the Vulnerability Index
Typology of Early Warning Systems
Constraints to the Use of Early Warning Systems
Age Cohorts for the Sample at Zidonini
Occurrence of Food Shortages and Severe Food Shortages
Vulnerability Index

Chi Square Results for Proximity to the Nduna Test

9°N9’P‘P9’N

The Hypotheses and Findings

viii

25
46
52
63
66
71
73
84

Terms Used in this Thesis

(As read in the interview)

Dwelling — a single house.

Household- the members of the family that live in a particular dwelling.

Homestead — a group or cluster of dwellings.

The Tree — a site name given to the center of Zidonini, used for clarity given that

“Zidonini” also refers to the entire study area.

The Ridge- a given site name near “The Tree”.

The Nduna — the site where the Nduna of the area lives.

Coping mechanism — a response to a food shortage intended to increase the

family food supply.

(Food) Shortage — a time when you or your family were hungry because you

didn’t have enough food or money to buy food.

Severe (food) shortage — a shortage that was far worse than all the others. You

went for many months very hungry.

* ALL REFERENCES TO “SHORTAGES” REFER TO FOOD SHORTAGES

ix

Introduction

Human relationships with the environment range from the manipulation
of raw materials into complex machinery to production of food to an aesthetic
appreciation of landscapes. The production of food for subsistence occupies the
daily routine of millions of people throughout the world. Food security is a
contentious issue in most of Africa as it is bound up in politics, societal relations,
and relations with outside organizations. As a percent of total employment,
farming is by far the most important activity in Africa. These farmers manipulate
their environment to produce their daily subsistence, generate cash, and provide

their families with the basics of life.

, South Africa is no exception to this pattern. Although systematically
neglected by the apartheid government, subsistence farmers comprise a large
proportion of the population. Their relationships with their environments are
becoming ever more the focus of attention of researchers and the new South
African government. The new government has initiated a series of upliftment, or
poverty alleviation and empowerment, programmes aimed at providing relief to
the often-neglected rural smallholder. This upliftment, however, must come
through an understanding of the rural context. Patterns of food shortage, coping
mechanisms, and expectations are issues that need to be examined when

studying rural livelihoods.

Food security is defined by the World Bank (1986) as "access by all people at all

times to enough food for an active and healthy life". As a component

2

of environmental security (the cumulative physical and social processes which
provide livelihoods), food security can not be overlooked or marginalized because
agro-businesses provide large stocks of food. The loss of soil fertility, hunting

rights, food security, biodiversity, and rural livelihoods are all interrelated.

In addition, global climate forces are acting on southern Africa, as the
1997 El Nino promises to be one of the strongest of this century. If past El Nino
events are any indication, the upcoming two years could be disruptive of

southern African food securities, economies, and agricultural progress.

This thesis focuses on issues of food security including the history of food
shortages and drought, the coping mechanisms and sources of food during times
of stress, and expectations/preparations for future shortages. The type of
information presented in this study is significant in formulation of policies to
combat rural hunger in South Africa. Furthermore, few detailed descriptions of
food security in the rural areas of South Africa exist. Without detailed knowledge
of the processes behind food insecurity, mistakes may be made in the rural
development process. I will focus first on the setting in which the study takes
place, then describe the methods used to gather data, followed by a review of the

literature, analysis of the data, and finally a discussion of findings.

Statement of the Problem Food security in Africa is elusive for the
majority of poor, rural people. Particularly pressing is the need to identify the level
of food security in the Republic of South Africa. Access to food was
systematically and purposely understudied in the apartheid era resulting in a lack
of even rudimentary information on the food situation in the rural areas of South
Africa. The cycle of drought in southern Africa and entitlement failure often
catches relief institutions off guard leaving people vulnerable to hunger and
famine. Coping strategies need to be identified and famine indicators must be
studied to prevent disaster on a widespread scale. A strong El Nino began
developing in the summer of 1997 in the eastern Pacific. Historically, El Nino
events have led to drought in southeast Africa (both the 82-84 drought and the
91-93 drought were results of El Nino). The potential for drought and hunger is
high for late 1997, 1998, and early 1999, making the type of information

presented in this study critical for effective policy making.

The impetus of this study is two-fold. I conducted a small study in the
rural farming uplands in the KwaZulu-Natal province has collected first, basic
information on the level of food security in South Africa. General information on
recent hunger periods, coping mechanisms, and people’s expectations of and
preparations for potential drought in the near future has been gathered using a
short quantitative questionnaire and informal interviews. Second, I addressed

questions regarding spatial patterns in the data. Relations between local political

4

structures and the distribution of food during hunger, access to markets, and

proximity to a game reserve have been examined to determine relevant patterns.

In summary, this study will provide information on the level of food
security in one community representative of rural Kwa-Zulu Natal. Identification
of coping mechanisms and discussion of recent hunger and drought history will
factor heavily in this work. Spatial patterns will be examined to gain a greater
understanding of the geographic processes behind the relations of production and

distribution.

Some General Hypotheses Based on a review of the literature, several
hypotheses related to the main themes are postulated. First, because of South
Africa’s level of development and unique history, coping mechanisms should be
different from those found in the rest of sub-Saharan Africa. Methods of coping
with food shortage are set in the context of the locality and in turn the nation.
South Africa’s unique political economy should have a large affect on the coping
strategies of rural inhabitants as this political economy shapes opportunities and
challenges. Second, the cycle of drought in the study area should be significantly
similar to general patterns in the rest of southeastern Africa. Particularly, droughts
should have occurred in 1981-84 and 1992-94. Environmentally induced
shortages may follow a similar pattern, but will vary according to local and

regional conditions, however, drought does not always lead to entitlement failure.

 

 

 

5
Thirdly, rural people's expectations of the role of the government in the

next food shortage period will be high. The newly elected government of the
African National Congress will be relied on to provide services often neglected in
the past. Finally, preventative measures for the next famine should complement
coping mechanisms. Preparation for famine is not a reaction governed simply by
the onset of famine, especially in southern Africa where drought can be expected
one or twice every ten to fifteen years. The interview schedule is designed to

provide information needed to support or reject these hypotheses.

Specific Hypotheses Specific research questions include:

. What is the history of drought and famine in the Zidonini study area?
o Is drought or entitlement failure more responsible for food shortages in Zidonini?

. How do the local and provincial political structures influence the level of food
security?

. What are the coping mechanisms to entitlement failure?

. What is the role of labor migration in the food security equation in Zidonini?, Does it
hold to the convention wisdom on homeland areas (the homelands were/are
primarily labor reserves)

0 What are the people expecting from the government regarding food?

0 What do the people believe will help them secure food?

 

 

 

The Setting — South Africa South Africa is located at the very southern
tip of Africa between 22 and 35 degrees South, 17 and 33 degrees east. Climate
of the nation varies according to altitude, latitude, and proximity to the Indian and
Atlantic Oceans. The western portion of South Africa is drier and warmer than the
east most of the year. The southwest coast, with a Mediterranean climate is an
exception. The eastern and northern plateau, commonly known as the Vaal, is
slightly moister than the western semi-arid region. The east coast and slopes of
the Drakensberg Mountains are the wettest parts of South Africa, receiving
approximately 100 cm of precipitation annually compared to the drier western
average of under 25 cm per annum. The Drakensberg Mountains rise in the east

transecting the nation.

Vegetation is primarily savanna in the North and East, except along the
coast where more tropical vegetation can be found, and in the west where desert-
like shrub can be found. Agricultural production in general is highest in the East
and North. Sugar cane and silviculture predominate along the KwaZulu-Natal
coast and heavy industrial farming predominates in the North. Smallholders are

most commonly found in the Eastern Cape and the uplands in KwaZulu-Natal.

 

 

com 00? o

 

    

millrl
.5.
RED 59me -. oqmo Eofim>>
Q... 8.8 59:62 a
_emz 2&me /
new . mgmww $9.“. 6 .. A.
w. .
a.
mmcmfiEzaz “mm; 5.62
/\ 5

a . 966:0
moc_>o._n_ Emctoz moron. _mcozmz

8:305

 

32 .822 538 up 2:9“.

 

 

 

 

 

 

 

 

 

 

Drought is recurrent in southern Africa’s climate. Southern African
droughts are cyclical in nature, rather than endemic as in the Horn and Sahel,
primarily due to circulation patterns. The region can expect to experience a major
drought once about every five to ten years. The last such event was in 1992-3.
The previous dry spell in the early 1980’s was rather protracted and by some
accounts it did not end until 1988. A review of drought history shows that major
below average rainfall periods occurred in 1910-12, the late 1930’s, 1968-9,

1978, and 1982-4 (Tyson, 1987).

Historical Background South Africa emerged in 1994 from over 300
years of colonial rule. The first inhabitants were Bantu speakers from the north.
The Dutch arrived on the scene in the early 17th century and their ancestors
remained in power until the first free and democratic elections were held in 1994.
Political structures instituted in this century severely handicapped the African
smallholders. Years of dispossession culminated in the 1913 Native Lands Act,
which stripped away from Africans their few remaining rights to land. Legal
segregation began in 1948 with the election of the National Party. As part of the
segregation plan, blacks were herded onto “homelands” and “townships” in a
grand scheme entitled “apartheid”. They were relegated to the least productive
areas of South Africa and were kept there through the use of strict pass laws and
other methods of a police state. Population densities in the homelands were
extremely high. Coupled with poor soils and lack of access to modern farming

techniques, smallholders lived on the brink of famine and hunger. Migration to the

9

urban areas was the only option left for many, however, this too was strictly
controlled. Often, critics of the apartheid regime accused the government of
“dumping” urban residents into the rural homelands, exacerbating the situation.
Through this entire period, agricultural improvements and innovations centered
on the white community with very few resources allocated for indigenous

agriculture.

Current Politics The freely elected government of Nelson Mandela has
begun to address issues important to the disadvantaged African community. The
Reconstruction and Development Programme (RDP) promises corrections to the
historical injustices committed under apartheid. The ruling African National
Congress has not gone unchallenged, however. The lnkatha Freedom Party has
a very strong following in parts of KwaZulu-Natal. Political violence has plagued

the region in recent years.

Smallholders in the KZN uplands are, in summary, faced with three main
problems — recurrent drought, the legacy of apartheid, and political instability.
Any attempt to understand KwaZulu-Natal’s smallholders must address all three

issues.

10

023?. mEmO 2 . ,
_No_oE,5- 632:2: \

..2.A\,.l\\.,

AI

_, [TEmEmcx

       

-. x, 89m 0E,
X ,

£283 “.88: 9F ,
2 , .832. 92+...

4: , ocm>om
25.56.32 ,

            

.. __ ._cm_m>>_bm_>_
25225.28, > m
, .,.,,.,,.,mcm o cmw

no.2. 63w EEoEN " N 959...

 

 

   

8:2 58m

 

moE< £30m D

mogm. 9239225 Co 5ch ‘
mumom II
cozm>2w

woe—duo". ”“24. >35

 

11

The Setting — Study Site The fieldwork component of this study was
undertaken between 28 February and 30 March 1997 in the KwaZulu-Natal
province of South Africa‘. Several suitable sites existed at the time of study,
however, one was selected for its exemplary attributes. Criteria for selection
included the following: access to the site had to be open at all times, rural small
holders had to dominate the economic infrastructure, and access had to be

granted by the local traditional authorities.

The Site - Background and History After reviewing several possible
areas, | determined the site most suited for the study was an area just west of the
border of the Hluhluwe-Umfolozi Game Reserve in northern KwaZulu-Natal.
Several other sites in the same area were chosen as alternates. The area is
referred to generally as Zidonini, but in reality is a series of concentrated
settlements including Zidonini proper (referred to henceforward as “The Tree”),
Ximbakazi, Madlinjini, Sovane, Madlwaleni, and Zagomfe just to name a few. The
orthophotos of the area do not show these names, referring instead to the area
as Zidonini, however, each settlement area has a specific name. “Zidonini” will be
used throughout as a generic descriptive of the entire area. The area lies about
15 kilometers to the east of Hlabisa, the administrative center of the Hlabisa

magisterial district, of which Zidonini is a part. The border of the Hluhluwe-

 

l Fieldwork was also conducted on several weekends and holidays through 10 May 1997.

12

Umfolozi Game Reserve was in clear view from Khalweni, Zidonini proper (“the
Tree”), and Zagomfe. Access to the area was open and controlled by the local
Nkosi (chief) and Nduna (sub-chief). After speaking with the Nkosi and various
Ndunas l was allowed to conduct research in an area of about 6—8 square
kilometers described above. My study was the first major research in the area. A
researcher from the University of Durban-Westville, South Africa had interviewed
two local residents and introduced me to the area. Outside of this, no researchers
had been to the area in the time span of local knowledge. Several potential sites
were discounted after the discovery of extensive research operations. Many
areas in northern KwaZulu-Natal have been overburdened with research, as local
academics from Durban find it convenient to drive a few hours north for research

projects.2

 

2 This “insight” provided by Shirley Brooks, Cathy Oelofse, Peter Schmitz, and Ann Vaughan.

 

 

 

      

ocoywaocm tom

5850

\

939568565 0

      
               

  

c3935
_comcmaE. 0
28 £922”. _QEmconEmsx
0 6:35 .8950

rmimmfimf
o

W. 25962

<mm< >035
2mg;

 

 

e25 warez

856E D

8:6 Se: 0

acum04

 

 

95 scam Ba 8:6 5? 332.23%; a 2:9".

 

 

 

 

 

The infrastructure of the area Is typical of rural South Africa. All roads in
the area were dirt, yet passable as the rainy season was late. The KwaZulu
homeland government provided electricity to the area in 1989 and 1990. Primary
and secondary schools were established in the mid-seventies along the main
road through the area. Transportation is provided primarily by private
transportation companies using either commercial buses or “combie” mini buses.
Certain areas have been provided with windmill driven water pumps and storage

tanks. Water pumps are present in Zidonini and Sovane.

The Hluhluwe portion of the Hluhluwe-Umfolozi Game Reserve abuts the
eastern edge of the research area. The reserve was established as a protected
area for game in 1897 and was expanded in 1989. Some of the residents in the
Zidonini area were pushed off their lands in 1989 when the Hluhluwe and

Umfolozi reserves, then separate, were joined.

The sale of Zulu handwork - basket, mats, skins, etc. - contributes
somewhat to the annual income of the residents. At the time of my arrival the
residents of the Zidonini area had not engaged in basket trade in over eight
months. Residents on the main north - south road, in contrast, had several

encounters a month with outsiders seeking Zulu handicrafts. The women of the

 

 

15

area were primarily responsible for the production of grass mats and baskets,
while a few men carved wood and worked skins, even though hunting is
restricted in South Africa without permits. The residents of the area report having
been harassed by local game wardens when hunting game that escaped from the
reserve. Reports of livestock damage due to lions or other predators were rare,
the largest complaint being the loss of grazing land due to alienation for the

reserve.

Members of the Manyoni lineage from Nongoma (80 km to the north-
west) settled the Mpembeni area late in the last century. Elders report that
intense droughts in the Nongoma area forced several families south-east in an
attempt to find better grazing and farming areas. Other residents report arriving in

the 1920’s as a result of expansion of the Hluhluwe reserve.

 

Hlabisa Study Region

Figure 4

 

 

uurunm
Q\uru

\
\‘\u\\
e

’—
I
a
“II"!

Key

‘:
I.
”z

’1
’1
’nm no,
’6

[:l Nkosi's Administration

\
\‘
\\l\
“m
m“

’11:

\

D Nduna's Administrations

\

\
IV“ if

“n

3‘"

A Points of Interest

AWinn“

‘mm m"
\\\

\
\‘Il
minnmunnurul""‘/Im\\“‘ """lumun\“‘

Road Network of 283 1 BB

IIIIIIIIIII

"anqu

\
\
|\\‘\\\

a

_)\\\

a
3
e
1
I

’4
,

’1
I
6

“mini: =
5’ aluminum”

'2
$
7
,
I
«

|:| Hluhluwe Game Park

uuull‘ww’r. '
’Imnuumnl

IIIL

 

 

.NDUNA MANQELE

NununumNm”
\\

S" m‘
\‘
mnuurl"

(—

"4,
’ln\\\'

  
 
 
   
  
   
 
  

(D
Z
O
_I
I
U)
2
<
Z
I)
D
Z

4
I
’a

~\ullla,
StrumS

I;

Immrmm\\‘

4

Hluhluwe-Umfolozi Game Park

’I

I \

I], [(1 ~‘\\\\
'Im

(‘

p

1.

 

 

 

 

 

l7

 

 

 

_No_o..E:

52:00 of.

0' O
000 o
'l 0
it 0
953521 . mm_Mm_I
(mm/3‘ >DDFW

 

€52.30 «Fr. “_o 33 oqu_xo.an< "m 959...

 

18

The area was incorporated as part of the KwaZulu homeland in the early
1970’s and the KwaZulu government undertook various “betterment schemes”
throughout the 1970’s and 1980’s. Political authority was exercised through a
parallel administrative system. The KwaZulu government sited officials in the seat
of the Hlabisa magisterial district, Hlabisa, and worked with the local Nkosi. The
KwaZulu government reinforced the Nkosi’s authority, and government services
were often provided through the traditional leadership. The Nkosi of the
Mpembeni area held court along the main road about 5 kilometers south of
Zidonini. The Nkosi ruled through sub-chiefs (Nduna) in outlying areas. I had
approval of the two Ndunas in the far east of Nkosi Hlabisa’s authority zone.

Authority to conduct research was granted by the Nkosi through the Nduna.

Chapter 2

Methods

Data Collection Several methods were utilized in the collection of data
for this project. Because no primary data existed for the area under study, both a
quantitative interview schedule and a qualitative participant observation routine
were undertaken. Archival data collection was limited to gathering information on
general food issues and collection of air photographs (orthophotos). A GIS was

used to identify spatial patterns.

Sampling the area proved to be a special challenge. No complete lists
could be found detailing all of the residents in the study area, however, my
interpreter was helpful in mapping the area for use in random sampling. Each of
the research areas was driven through in order to make comparisons against the
orthophotos. The orthophotos dated from 1982, but most dwellings were
represented. The missing homes were added on a sketch map created for each
area. A number was assigned to each dwelling in the area to be sampled and
households were then drawn. The goal was to interview the male and female
heads of the household. The sketched maps from which the samples were drawn
were double-checked with local residents to make sure everyone in the area had
been included The sample was stratified by dwelling clusters or homesteads. A
total of 20 clusters were identified and sampled. The dispersal of homestead

clusters for sampling.

19

20

was fairly uniform throughout the study area. Dwellings not immediately in a

given cluster were assigned to the nearest homestead.

The direct refusal rate for the survey portion of the field research was
exceptionally low3. Only twenty-two people out of 145 refused to participate
(15.17%). Out of this number eighteen were men and four were women. A non-
response test was conducted on eight of the refusals in an attempt to identify
reasons for refusal. All eight were men and attributed their refusal to the fact that
the interpreter was female and they thought the questions were going to be about
“women’s work”. Of the remaining fourteen refusals, six were “too busy”, two
people thought I was a “spy” for the parks board and the other fourteen gave no

reason for their refusal.

The relatively low refusal rate can be attributed to the selection of an
interpreter from the area. All respondents appeared to know, or know of the
interpreter, which greatly aided in obtaining interviews. Each interview was
conducted with only one person. Usually the group selected by random sample
would gather at a tree or other common meeting site to begin interviews. Each
person would be taken aside and asked questions. Often interviews were
conducted in a home, but most respondents preferred to meet outside their

homes. Each interview took between eight and fifteen minutes. The interpreter

 

3 Interviewing men was exceptionally problematic. In some cases men were not at home, but more often men
avoided being interviewed by leaving when they saw us or got word of our imminent arrival. After several
discussions with my interpreter and the mens’ wives, it was discovered that men thought that since I had a
female interpreter I was only asking about “women’s issues” or they didn’t want to be interviewed by a
woman interpreter. I compensated for this bias by hiring a male interpreter for informal, qualitative
interviews.

21

conducted every interview in Zulu. The interpreter was trained in the first six days
after onset of the study. After translation by the interpreter, data was recorded
and some information was then gathered outside the interview schedule. Often a
group would be asked questions about the history of the area, the basket trade,
or hunting, among other topics. This informal session usually took place after the
interviews had been conducted. Questions and concerns from the respondents
were addressed either before or after interviews. ln following the Human
Subjects requirements, each respondent was informed that they could, at any
time, refuse or terminate an interview and that their names were not being
recorded in any way. Each respondent was told how to contact the author and for

whom the author was working.

The qualitative portion of the research provided critical information on the
history of the area and local conditions, attitudes, and expectations. The
interpreter and the Nduna were interviewed extensively before the onset of
interviewing. Older respondents as well as community leaders and a researcher
from the University of Durban-Westville who was working in nearby areas were

interviewed qualitatively.

The Interview Schedule and Informal Interviews The interview schedule
was designed to highlight basic food security issues in South Africa (see
Appendix A for the complete schedule). The scope of the study precluded
extensive analysis. The main objective of the research was to identify the
presence or absence of food security in a fairly “typical “ rural area of South

Africa. The interview schedule was designed to accomplish three major goals.

22

The first was to identify periods of food shortage and their causes. The second
was to discern coping mechanisms, and the third was to understand individual

expectations regarding food shortage.

The qualitative routine was incorporated to collect valuable data outside
of the rigor of a structured questionnaire. Key informants provided substantial
information on issues both in the structured interview and many that were not.
Some respondents felt the need to continue after the structured questionnaire.
Information provided outside of the questionnaire was both substantial and

relevant and was incorporated into the study.

A small pilot study was conducted before the actual sampling was
undertaken to identify problems in the interview schedule. Twenty people were
questioned with the interview schedule and then a follow up qualitative session
was conducted for back translation, identification of misunderstandings, and

differences between expected and actual responses"

 

4 This difference refers to general topics, not specific answers. For instance, when asked about government
assistance to the area, many respondents said that the government had never helped them, although power
lines hung overhead.

23

Quantitative Data Analysis Analysis was conducted on 145 cases
drawn through a stratified random sample mentioned above to provide both
summary statistics and tests for relationships. Each interview (case) was
assigned a record number and a site location for spatial analysis. After
collection, the data was coded, crosschecked, and analyzed in SPSS for
Windows version 7.5. Summary statistics and Chi-Squared tests were performed

on pertinent variables.

The purpose of this thesis to understand the causes of food insecurity
and related coping mechanisms. The indicators of food insecurity most often cited 1
were discussed in the Review of Literature, including environmental, social, and \g
political conditions. In the operationalization of the interview schedule these were
represented as distinct variables/questions. However, the vulnerability of persons
or groups in society must be viewed together, not separately. The conditions of
vulnerability (few fields, no access to political power, poor implements, etc.) are
not independent of each other. A family with little land will rarely have large farm
implements and conversely, a family with political power is more likely to have
better maintained fields and to perhaps have even diversified into other non-farm
activities. The idea of a “package” of vulnerabilities needs examination to

determine whether or not the assumptions are valid.

24

A “Vulnerability Index’ was created to test just such assumptions. Six
variables were categorized and recoded to create an approximation of the level of
vulnerability of each respondent. The six variables - the number of fields, the
number of shortages, the length of time as a resident in the study area, hunting
patterns, the number of severe shortages, and receipt of help from the KwaZulu
government, were aggregated to highlight the most vulnerable respondents. The
criteria for the index are presented in Table One. Testing of the index can be

found in the Data Analysis chapter.

25

Table 1. Components of the Vulnerability Index

 

 

 

 

 

 

 

 

, . , 7 Variable Under ’f 9 ' 3‘ =Qualifying Condition ,
Censide‘ration : ’ (Considered “Highly Vulnerable’i’I-i l ‘
Number of Fields Less than one field
Number of Shortages Greater than 3
Length of Residency in Area Less than 10 years
Hunting Non hunting respondent
Number of SEVERE Greater than 2
Shortgggs
Assistance from KwaZulu No help
Government

 

 

 

Chapter Three

Review of the Literature

Theoretical Setting: Political Ecology Complex political, economic,
social, and environmental conditions all play a role in the level of food security in
any given area. At times, one of these factors may weigh more heavily on the
food security equation, but nonetheless all are issues that need close
examination. The political ecology of land degradation is particularly useful when
examining food security. Agricultural production links the two issues firmly
together. This thesis is predicated on the political ecology approach to
understanding human interaction with the environment. The political ecology
approach emerged in the 1970’s as not only a critique of classical economics and
traditional resource management, but also as a newer formula for assessing
human-environment interaction. Several major themes are addressed under a
political ecology framework, however it has no rigid structure of analysis or
procedure and has been interpreted in many different ways. Political ecology can
generally be defined as the study of how people’s decisions regarding land use
are affected by socio-political institutions, at the local scale of analysis, and how
nature is incorporated into the wider economy at the regional and global scales
(Blaike, 1985; Blaike and Brookfield, 1987). Scale of analysis is particularly
important in political ecology and will be discussed in detail. Themes from two
major paradigms, political economy and ecology, are integrated to form the basis

of political ecology.

26

 

27

The Contributions of Marxism Classical Marxism maintains that
capitalism is inherently volatile and is unsustainable due to crises faced by capital
in relation to production. These crises are used to explain many instances of
human behavior toward the environment, especially as the environment becomes
more commodified. The first contradiction of capitalism is an internal
contradiction. As capital seeks to lower costs and increase profit it takes self-
destructive steps to achieve its ends. This often takes the form of cutting wages
or terminating workers. As the worker has less and less to spend, the market for
the product of capital decreases. The second contradiction arises as a direct
result of the first. As Martin O’Connor notes “capital defends or restores profits by
strategies that degrade or fail to maintain the material conditions of their own
production” (O’Connor, p.165, 1994). The reproduction squeeze faced by capital
forces it to use any available resources to the maximum level to maintain a
constant level of profit. No regard is placed on the fact that the resource may
become exhausted in a shorter period of time. Capital accumulation accentuates
the scarcity of resources. As resources are devoured by production hungry
capital, material limits to growth become ever more apparent; nature is not
infinite. This destruction of environment through intensification is the basis of the
second contradiction of capitalism. Those with the more rigid Marxian views
within political ecology suggest that perhaps even a triple contradiction of

capitalism exists. The first of the three, as Jean Paul Deleage (in O’Connor,

 

28

1994) notes, is that the exploitation of a collective resource benefits only the elite.
Second, the distribution of costs and benefits accrues to the present generations
disproportionately so that future generations bear more cost. Finally, the human
species enhances its own interests at the expense of other species (paraphrased
from O’Connor, 1994, p.49). These contradictions help to explain the destruction
of nature and what is termed the ‘production of nature’ in political economy. The
production of nature refers to the commodificatlon and incorporation of nature into
the production process. Nature is regarded as a productive resource (such as
labor or capital) and treated as such. Little concern is shown for the destruction of

nature that results from its incorporation into the production process.

Very little understanding of any human action can be achieved without an
understanding of the human interaction with the environment. All human activity
stems from a relationship with nature. Humans use soils, forests, minerals,
animals, and hosts of other natural resources to survive and thrive, however,
production in a capitalist economy has been de-linked from the environment. The
environment has been viewed as controllable and easy to manage. But with
global environmental change becoming an increasingly important issue, many
have been forced to reconsider their long held beliefs regarding nature.
Environmental movements are gaining increased support and recognition in their
struggles for ecological protection. The physical processes so long ignored by
many are finally being valued for their contribution to the sustenance of life on

earth. Many people have become alarmed at the change in the acidity of rainfall

29

or at the destruction of forests. The relevance of ecology to everyday life has
become more appreciated by the average individual as news reports of ecological
destruction portray a picture of gloom and doom. Ecology studies may benefit
from the heightened attention in the long term, but it will take a concerted effort to
change human behavior, especially toward production and the ensuing

materialism that results.

Questioning Scale: The Locality Political Ecology focuses heavily on
the importance of the locality in studies ranging from farming techniques to
political controversies over land. Many researchers within political ecology have a
tendency to reify local traditions, customs, epistemologies, and social
movements. Peasant farmers possess practical knowledge of the land and soil
and have much to contribute to research. Peasants have come under attack in
neo-classical and traditional Marxist interpretations of economy and production.
They are often characterized as inhibiting progress and change in farming
systems. In many cases, they refuse to adopt certain agricultural practices and in
fact destroy their own reproductive capacity. Political ecology, however, rejects
the notions of the “backward peasant” who inhibits progress and advancement.
Quite the contrary, peasants are not backward and obstructionist. They must deal
with political pressures and conflict and cope to the best of their ability with
changes in their lifestyles, customs, and farming practices. Land alienation, labor
shortages due to migration, civil strife and poverty shape the decisions made by

what Blaikie and Brookfield refer to as “ land managers”. These land managers

3O

face constraints on their daily reproduction and must manage wisely what scarce

resources they have just to survive.

Questioning Scale: the Region Richard Black incorporates a regional
scale of analysis in his article “Regional Political Ecology in Theory and Practice”.
The major components of a regional political ecology incorporate the integration
of human and physical approaches to land degradation. Both social constructs
and purely physical constructs are synthesized to form a better understanding of
the degradation of the environment. This approach also highlights the interactions
between people and their environment, noting that human conceptualization of
the environment is constantly changing. For example, in the transition from a pre-
capitalist to a primitive capitalist society many long held understandings of the
environment are altered. Nature becomes commodified and a new dialectic
between humans and the environment emerges. Regional analysis allows for
area-wide assessments of environmental degradation to better scope out the

extent of ecological problems.

Linkages between Scales Political ecology integrates the different
scales -global, regional, local, household- by analyzing the vertical and horizontal
linkages between and within each. Horizontal linkages include the social
processes and interactions between actors, conflict over resources, and

production within the given scale. Blaikie and Brookfield describe vertical linkages

31

as a “chain of explanation”. It starts with the land managers and their direct
relations with the land (crop rotations, fuel wood use, stocking densities, capital
investments and so on). The next link concerns their relations with each other,
other land users, and groups the wider society who effect them in any way, which
in turn determines management. The state and the world economy would
constitute “the last links in the chain.” (Blaike and Brookfield, 1987, p 27)’ The
chain of explanation details the dialectic between how land management
decisions made by a regional bureaucrat or global economic change may affect
local land managers and how local managers decisions impact the regional and

global decision makers.

Cindi Katz gives perhaps some of the best examples of these linkages in
her article “ Sow What You Know: The Struggle for Social Reproduction in Rural
Sudan” (Katz, 1991). She highlights vertical linkages when describing the impacts
of the Suki Development Project on the population of Howa in eastern Sudan.
Her work showed how the project shaped images of the environment, especially
children’s, and how this environmental knowledge contributed to social
reproduction. Examples of horizontal linkages include the effect of labor
withdrawal due to the project on the children’s education and their contribution to

family labor.

32

Donald Moore highlights the vertical and horizontal linkages between the
scales of analysis in his article “Contesting Terrain in Zimbabwe’s Eastern
Highlands: Political Ecology, Ethnography, and Peasant Resource Struggles “
(1993). He notes the conflict over resources between the residents of Kaerezi
and the park officials of the Nyanga National Park in eastern Zimbabwe. Land
alienation by the park created discontent in the local communities. The park
interfered with local agriculture, hunting rights, and the basic customs of the area.
Moore presents vertical relationships in Zimbabwe as in disarray. The
government agencies do not work as one between the locality, region, and
nation. Many disharmonies existed between what officials in Harare view as good
policy and what officials in the park view as good policy, and both are out of sync
with what local leaders want. This break down of social consensus allows a local

chief to play one element off the other to cause even more discontinuity.

These vertical and horizontal linkages are critical to understand political
ecology. Political ecology relies on analysis of the relationships between scale to
determine why certain decisions are made. Constraints from above (such as
government policy restrictions, donor loan requirements) and below (i.e.
unwillingness of peasants to adopt government policy) shape the decisions land
managers make on a daily basis. Political ecology stands out from many other
theories of resource use because it does address the issue of scale, power, and

finkages.

33

Political Ecology, Traditional Resource Management, and Global Change:
Divergent Views Several major differences exist between political ecology and
the traditional resource management or global studies approach. Humans and the
environment are more closely related in a political ecology approach, as they are
considered part of nature. Under a traditional resource management or global
studies approach humans are portrayed as being above nature, able to dominate
and control it. Nature becomes an object of management to be used exclusively
for human advancement. Overpopulation, poor soils, and climate are
deterministic conditions that effect development in a traditional resource
management or global change approach. Humans are in a contest or conflict with
nature and the rational management techniques of the two traditional schools will
help humans prevail. Political ecology on the other hand, deals with socio-
political constraints faced by people and how those constraints manifest
themselves over space. Global economic conditions, national government policy,

and corporations all influence the decision making process of land managers.

Finally, the divergent political outlooks shape the view of the peasantry in
each theory. Political ecology tends to portray the peasantry in a more positive
light. Peasants are constrained from above. They make decisions faced with
those constraints and only act in a way that preserves their own interests. If that
may result in resource destruction, then the structure of society is to blame not
the peasantry. The traditional approaches however, regard the peasantry as

unwilling to change and use Malthusian assumptions to advocate policy which

34

would adversely effect them. Peasants are often considered backward or
ignorant. These visions of the peasantry have a considerable impact upon what

actions are suggested taken to solve environmental (and social) problems.

Critique Some questions arise after a thorough study of political ecology.
The traditional Marxist approaches assume that ecological destruction occurs
only in capitalist economies. Little is mentioned of the destruction of the
environment in socialist economies. A political ecology of socialism would provide
a counterpart to the present theory that could attempt to explain environmental

destruction in the formerly socialist world.

A second oversight in political ecology occurs in the assumption of
vertical linkages. Clear examples of impacts on locality from above are illustrated
but not the reverse. Few examples in the literature refer to how the locality can

impact the region, nation, or global system.

Thirdly, the peasantry becomes romanticized. Political ecology does not
address problems caused by peasant behavior. The individual is capable of
making choices. The bludgeoning of the peasant with structure often blinds

political ecologists to more underlying issues such as deviant behavior.

Finally, the term is political ecology. But most often the ecology is
completed hidden by over-emphasis on the political. Many ecologists maintain
that political ecology is nothing short of basic political science with some regard

for environmental issues.

35

In summation, the theory of political ecology is an integration of politics
and economy which uses particularly Marxian methods of analysis to
conceptualize the structural constraints which effect the decisions made by land
managers on an everyday basis. Assumptions are made regarding linkages of
scale, which presume a dialectical relationship between the various scales. The
theory provides insights on the relationship between social actors in each scale of
analysis. These relationships are critical in understanding why people make
certain decisions over land. Despite critical questions regarding ecology and the
reification of the peasantry, the political ecology approach is most applicable for
analysis of food security issues. The relationships between actors in this arena,

and between actors and the environment are critical to the food security process.

36

Review of the Literature: Food Security in Africa

Several issues permeate the literature on food security. Common themes
include definitions of food security, the causes of food shortage, the progression
of food insecurity through different stages of intensity, inter- and intra -household

differentiation, coping mechanisms, and early warning systems.

Food shortage, food insecurity, hunger, and malnutrition are terms used
to define a condition where basic levels of food consumption are not met. Food
security will be used in this thesis as it describes most succinctly the processes
that lead to a lack of food, the use of coping mechanisms, and contemplation of
prevention measures. The most widely used definition of food security comes
from a 1986 World Bank report in which food security is characterized as "access
by all people at all times to enough food for an active and healthy life" (World
Bank, 1986). Several variations on this definition include precise caloric intake
standards, distribution ratios, and other variables, however most definitions of
food security are similar to the World Bank definition. This definition is the
culmination of many years of research and inputs from various disciplines. While
not universally accepted, this definition provides a starting point from which to

build a greater understanding of food security.

37

According to Hesselberg (1994), food security research has progressed
through three major phases. The first phase of research and policy planning was
one where the primary emphasis was focused on "production orientation".
Production was seen as the key to ensuring food security. Food availability, or the
supply of food, was believed to be vital to ensuring a healthy population. The
second phase shifted research and policy planning to a more market oriented
arena. Emphasis was placed on "getting the prices right", a definition well suited
to the era of structural adjustment. Finally, Hesselberg maintains that the current
research is consumption oriented. Food access has taken priority over food
supply. The notion of food access encompasses ideas such as the differentiation
of households and the effect of this differentiation on food access and gender
roles and inequalities. Rukuni and Eicher (1988) also distinguish between the
supply and demand phases within food security research. They note, like
Hesselberg, that the era of focusing on supply as a key component of the "food
security equation" has faded as the focus has shifted to issues of demand, or

access.

Buchanan- Smith (1995) observes that "famine is no longer solely- or
indeed even primarily- attributed to food availability decline (FAD), but
increasingly to food entitlement decline". Lambert (1994) also presents his
arguments in a demand-side approach. Both support the shift of paradigm within

food security research and expand it through the use of illustrating case studies.

38

Food Shortage and Coping Mechanisms The progression of food
insecurity through different stages of intensity is maintained by Corbett (1988),
Dirks (1980), Sperling (1987), and many others. As described by Corbett these
stages include an “adaptive” stage in which assets and labor are sold,
progressing with increased insecurity to the sale of “key productive assets” and
finally with the most intense food shortages, “mass migration” (Corbett, 1988).
Watts (1983) makes note of the ten most predominant strategies in his Nigerian

case study in the following order:

. Collect famine foods

. Borrow grain from kin

. Sale of labor power

. Engage in dry season taming
. Sale of small livestock

. Borrow grains

. Sale of domestic assets

. Pledge farmland

COQNCDCfl-hOON-A

. Sale of farmland
10. Migrate out

This progression is remarkably similar throughout the literature on food
security for all regions of Africa. Cutler (1986) categorizes coping strategies in
eastern Sudan into three stages, providing additional support to Corbett and
Watts. Initially, adaptive strategies are adopted. These include the sale of

livestock, labor migration, and use of credit. Next, key productive assets are sold,

39

including tools, animals, and in severe cases, land. In the final stage, migration Is
the only option left to escape hunger and death. Rahmato (1988) indicates that
peasants in Ethiopia face food shortage first with "austerity measures" such as
eating lower quality and quantities of food. As the situation worsens, temporary

migration occurs, followed by sale of assets. The final strategy is mass migration.

Campbell (1990) suggests a similar progression of coping strategies.
Quoting Cassanelli, he notes ten distinct stages through which Somali peasants
passed during a particularly severe food shortage. Webb and Reardon (1992)
document a similar progression of households in Burkina through stages of
response to shortage. They note that income diversification is the first adaptation
to changing conditions followed by asset disposal and the acquisition of loans and
in the final stage consumption is altered (Webb and Reardon, 1992). Webb
(1994) reports similar findings from Ethiopia noting a similar progression of coping
strategies from asset sales to altered consumption and migration. Lambert points
out however, that an undue focus on coping strategies can lead to
misinterpretation. He argues that coping strategies are “only one piece of
information in an overall mosaic” and further states that “focusing on coping
strategies tends to overestimate people’s resilience“ (Lambert, 1994,p.342) due
to the narrow focus often placed on the process of coping to the detriment of

other information.

40

Coping with drought has progressed, in some regions, to coping with
daily life (Lambert, 1994). Campbell (1990) notes that coping mechanisms should
be viewed in the context of the changes faced by rural dwellers. He maintains
that coping mechanisms are dynamic processes and warns of conceptualizing
them as static and unchanging. We have the ability to understand insecurity and
aid areas in peril; however, a reactive role is not enough. Campbell (1990),
Corbett (1988), Vogel (1994), and many others have suggested the development
of early warning systems. The ethos for an early warning system (really, a series
of systems) is the need to preemptively “strike” at food insecurity and it menacing
brother - famine. Early warning systems would help detect the areas in most
need and target aid and assistance to those areas. Corbett (1988) warns,
however, “the existence of these systems is far from a sufficient condition for the
avoidance of famine" (Corbett, 1988,p.1109). The role of early warning systems
in South Africa has been addressed by Vogel (1994) and will be discussed in the

following section.

Coping strategies differ from place to place and time to time. The notion
that coping mechanisms remain static through time is misleading. C0ping
strategies that were used in 1900 may or may not be utilized today. Further,
Cekan (1992) maintains that coping strategies must be monitored seasonally.
She observes that rural Malians have different coping strategies for each season.
Lambert strongly challenges the nature of coping strategies when reflecting on

rural Mali, "firstly, the term [coping strategies] is not dynamic and fails to

 

41

encompass shifts in food strategies taking place in a context of rapid social and
economic change in Sahelian environments " (Lambert, 1994, p.336). In areas of
chronic drought, coping strategies become more than short-tenn phenomena, but
are a mode of normality (Lambert, 1994). Thus, considering coping strategies as
static and unchanging can lead to false conclusions. The evolution of coping

strategies is important to the understanding of current patterns of behavior.

Southern Africa Food security at the national scale in southern Africa
has several complexities not experienced by other regions. First, the relationship
between southern African nations and South Africa impacts greatly on the food
security situation in that region. Secondly, the relationship between nations in the
Southern African Development Conference (SADC) provides an added element
to the southern African food situation. However, on the issue of coping
strategies, southern Africans follow patterns similar to those of their northern

neighbors.

Southern Africa has structures similar to those in other regions of Africa.
The legacy of colonialism is a most striking example. South Africa, Zimbabwe,
and Namibia have just emerged from years of colonial rule that lasted longer than
was the case for any other nation in Africa except Western Sahara. The land
alienation policies so familiar to black South Africans were experienced by
subjugated populations in countless other African nations in the heyday of

colonialism.

42

The relationship between SADC nations and South Africa changed
immediately with South Africa's first free elections in 1994. Apartheid South
Africa strove for a host-client relationship with its neighbors in which it could
control the most important element of life in southern Africa- food. South Africa's
apartheid government used food exports for political leverage and destabilization.
Although SADC nations imposed boycotts on South Africa, the latters influence
and economic power meant that SADC nations had few options but to trade with
the economic giant. While most nations refused to trade openly with South Africa,
clandestine trade in agricultural products nonetheless took place in the early

1980's (Lipton, 1988).

The dependency of the frontline states (those nations closest to South
Africa) on South African agricultural exports was greater than that of other SADC
states. The apartheid government in Pretoria attempted to maintain food
dependency, thus encouraging some degree of political dependency. While
thousands were going hungry within South Africa, grain exports kept countries

such as Lesotho, Swaziland, and Botswana in line with Pretoria's political agenda.

Within SADC, the need for a counterbalance to South ’Africa's influence
was recognized. Policies to encourage exports were followed by member nations
Zimbabwe most notably. SADC's most successful member in confronting food
security problems was Botswana. Diamonds, a stable political climate, and
immense international support have led to the development of a series of
programmes that help provide a basic level of nutrition during shortages.

Botswana's leading role in the development of such a system, coupled with

 

43

Zimbabwe's high agricultural output, especially in 1985, aided in counterbalancing
South Africa's influence, however, intra-regional trade has been poor at best.
(Lipton, 1988; Rikuni et al, 1988; Rikuni and Wyckoff, 1991; Prah, 1987) Trade
between SADC nations in agriculture has been one of the key policy initiatives of
the regional food security initiative. As noted above, emphasis has shifted away

from just such supply side plans, providing SADC with new challenges.

Coping Strategies in Southem Africa Coping strategies at the local scale
in southern Africa are not remarkably different from those in other regions of
Africa. Zinyama, et al. describes the various economic, social, environmental,
and political methods for COping with drought. Coping strategies progress from
prevention to modification of conditions, to sharing consequences, and finally to
bearing the consequences (Zinyama, et al, 1987). Research in Malawi (Babu and
Mthindi, 1995) and Botswana (Hesselberg, 1994; Teklu, 1994; Hay, 1988)
suggest coping strategies similar to those reported by Zinyama. Research in
Zimbabwe, however, has been almost exclusively supply—side oriented. Mudima,
et al. writes" to date there had been little empirical work in Zimbabwe to
investigate neither [sic] household coping strategies for dealing with recurring

food shortages in low rainfall areas..." (Mudima, et al, 1987).

While differences exist between southern Africa and other regions at the
national scale of analysis, similarities are common at the local scale. This is not to
say that food security is not context specific, but rather some generalities can be
found across regions. The progression of intensity of coping mechanisms is just

one example of how southern Africa is similar to other regions. However, the

44

influence of South Africa Is one of the many dissimilarities. A crucial question
regards South Africa itself. What is the level food security in South Africa? How
do people cope with drought? The next section will outline the food security

situation in South Africa.

Famine Prevention - Early Warning Systems and Entitlement Theory
After coping has been identified, the next step is prevention. The prevention,
mitigation, and alleviation of famine have been approached from many differing
perspectives. The terms entitlement, prevention, and early warning systems refer
to distinct approaches in fighting famine. The earliest famine prevention methods
date back to the early rule of the British in India when hunger codes regulated
food supply in an attempt to forgo the incredible famines so often experienced by
that country. The most common approach to combat hunger was work-for-food
schemes. Colonial governments such as the Portuguese in Cape Verde and the
British throughout Africa and Asia often relied on work—for-food to keep local

populations from death (Sen, 1980).

More recently, international relief agencies have been playing a larger
part in food relief assistance to troubled areas, especially South Asia and Africa.
As African governments become less and less able to handle chronic famine, aid
agencies and international governmental and non-govemmental organizations
have taken a more active role. These new actors in famine relief have brought
with them high-profile media coverage. This coverage has caused constituencies
at home to wonder whether or not disasters such as the mid-80's Ethiopian

famine could have been prevented. NGO's, the UN, and donor governments

 

45

have been actively promoting early warning and prevention of famines, with

differing results.

Emphasis on warning systems and famine prevention methods became
heightened after the devastating Sahel drought of the early 1970's. Predictive
models were proposed to warn of impending droughts and famines using a
variety of inputs, from food supply data to socio-economic information. The

systems range from highly technical to highly social (see Table 2).

 

46

Table 2. Typology of Early Warning Systems

 

Conventional Famine Alternative Food
, Early Warning System Information System
‘ (Technically Orientew (Sacially Grimm

 

 

Scope

Determinants of Food
Security

Level of Operation

Unit of Analysis

Approach

Response

 

Famine- Oriented

Food Production

Macro, centralized

Geographic, e.g. nation I
district

Top-down, data-centered

Food-aid oriented

 

Food-security oriented

Access to Food

Micro, decentralized

Socio—economic, e.g.
vulnerable groups

Bottom-Up, People
Centered

Sustainable Improvement
in Access to Food

 

Source: Davies, et al. 1991

The examples in the above figure represent the extremes of early

warning systems, as most systems incorporate some of each extreme. Until

recently, early warning systems (EWS) have been overwhelmingly supply side

oriented. Components of early warning included monitoring food production,

analysis of satellite images, rainfall indexes, and other such indices that are then

used to highlight areas vulnerable to famine.

Most Early Warning Systems share some basic features. EWS are

designed to provide information, thus the gathering and processing of data is vital

to any EWS. Information is then relayed to decision-makers and subsequent

action is taken (or not). Most EWS go beyond simply gathering data to analysis

 

47

and refinement. In highly technical EWS this could include satellite interpretation
and GIS analysis as well as food supply prediction and in highly socially based
EWS this could include synthesis and interpretation of data based on an intensive
socio-economic study of the area. Some EWS go further and make

recommendations for action.

Indicators used by EWS differ greatly. Morgan (1985) distinguishes

between input and output indicators in EWS as noted by Campbell:

Input indicators reflect food production potential and include rainfall, soil moisture
status, and crop growth. Output indicators, on the other hand, reflect the actual
food scarcity situation, the results of shortage. Among the output variables that
have been suggested are nutritional indices, behavioral indicators such as rates
of migration and economic signals such as crop prices and livestock sales
(Campbell, 1990, p.232)

This definition of indicators comes closest to the middle ground between
the technically oriented systems and socially oriented ones. Some method of
environmental monitoring is included in almost every EWS and many are

beginning to include socio-economic indicators.

Campbell further suggests a scaled approach to EWS. He maintains that
the regional and national scale are most appropriate for technological (input)
variables and that social (output) variables are best monitored at the district and
sub-district level (with the caveat that this approach is beyond the capacity of
many African governments due to funding limitations). A scaled approach allows
for the flexibility needed in African societies where funds are severely limited.

Monitoring at the district level can often be assumed by a local structure

48

responsible to a national authority. A low cost, simple system at the district level
coupled with a broader national or regional system supported by outside funds is
perhaps more practical than a grand, one-size-fits-all system designed completely

by outsiders with no local participation.

Walker (1989) presents the components of the "perfect warning system”

as the following:

1. Detecting, evaluating, and predicting a hazard

2. Constructing a forecast or warning message

3. Spreading the warning message

4. Creating effective preparedness and mitigating responses
5. Learning from mistakes

Even with‘the "perfect warning system" Walker notes some problems. He
highlights five common problems of the warning message. First, the language is
often too technical. People do not know how to respond, or to what they are
responding. Second, different warning systems produce "conflicting messages”
and people "ignore the whole package". Potential victims are confused and adopt
a wait and see attitude. Third, messages often do not state the problem clearly
and confusion follows. Fourth, the potential victims are poorly identified. People in
danger are overlooked while those with little potential risk are alerted needlessly.
Fifth, people often fail to react (paraphrased from Walker, 1989, p.65-6). He
further argues that people can respond in five ways to the warnings. They often
ignore warnings, take a wait-and-see approach, try to "investigate... accuracy and

credibility", take action not recommended, or take recommended action (Walker,

49

1989, p.67-8). The conflicting systems and actions of those directly affected are
often blamed for the calamitous outcome experienced in the Ethiopian famine of
1984-5, for example. Suspicion about the intent of outsiders run high in normal
situations, but during a disaster when mixed messages are sent or poor
instructions cause pandemonium, those suspicions can lead to a complete lack of

faith.

Buchanan-Smith (1995) reviews the approach of four donor agencies to
early warning - the Overseas Development Administration of the British
government (ODA), the US Agency for lntemational Development (USAID), the
European Union (EU), and the UN World Food Programme (WFP). What is
important here is that most donors fund some sort of assessment programme,
and USAID is the only donor agency with a formal EWS, the Famine Early
Warning System (F EWS). Funding is limited in most aspects of early warning
projects and the major information sources for all include NGO's. The role of the
media is strong in all except the WFP. Media attention to the Ethiopian crisis was

crucial in stirring donors and Western governments to action.

Walker (1989) reviews the FAO's Global lnforrnation and Early Warning
System (GIEWS) as well as WFP, USAID,UNICEF, OXFAM, and other NGO

early warning systems. He recommends that systems must serve six purposes:

1.Warn of the impending beginning of the famine process
2. Be capable of monitoring the development of that process

3. Be able to recommend sustainable strategies for reversing the
process

4. Be able to steer ad monitor the application of these strategies

50

5. Be able to evaluate their effectiveness after they have been
applied

6. In the event of the failure of these strategies, it must be able to
switch to monitoring those parameters that will predict the likelihood
of mass starvation. (Walker, 1989, p.147)

Table 3 (page 50) summarizes the constraints faced by EWS. Perhaps
the biggest constraint to using EWS is financial. All actors are consumed by the
need to obtain funding for such a project. With hundreds of competing interests,
EWS often fall low on the political agenda. Recently, donor fatigue has
threatened the most lucrative source of funding for EWS. Many African
governments are paralyzed by strict spending constraints laid down by Structural
Adjustment Programmes and have little choice but to depend on N603 and

donors for support of even the most rudimentary EWS.

In addition, donors have recently expressed "fatigue" with the
development process, noting corruption of project funding as a major source of
this "fatigue". Buchanan-Smith classifies this as "avoidance of responsibility"
(Buchanan-Smith, p.20). "Passing the buck" also afflicts many donor and non-
governmental organizations. Unclear mandates, imprecise goals, and political
directives from home governments provide ample room for the avoidance of

responsibility.

At all levels the need for more information often overshadows objectives
at the ground level. The drive for more Information often becomes a goal in itself,

crippling any substantive action. Governments and donors often absolve

‘1

 

51

themselves of action by continually demanding more information, or more precise

information, or another report, etc.

Margaret Buchanan-Smith is highly critical of the supply side focus of
most EWS,

supply-side indicators continue to predominate over demand-side ones,
in spite of doubts about the accuracy of annual harvest
assessments...local people's coping strategies are rarely incorporated
systematically... many EWS, especially those which are donor funded,
are increasingly reliant on high-tech inputs using sophisticated
information technology and satellite imagery (Buchanan-Smith, 1995,
p.16).

While the literature includes a great deal of critical reviews of the
technical EWS, little attention is given to the limitations of the highly social form of
EWS. Highly social EWS may overemphasize certain aspects of an affected
area. As discussed above, coping strategies are not constant and building an
EWS based on specific coping mechanisms may prove to be misleading. The
social approach is also limited in the scale of coverage. While intense coverage
of one's study region would be very beneficial both to the local inhabitants and
the researcher, intense coverage is not possible for the wide range of areas
affected by drought, nor does every region affected by drought have the luxury of
an outside researcher interested in its survival. If national EWS are faced with

severe limitations caused by financial and political considerations, a micro

approach would seem to be even less sustainable than a national approach.

52

Table 3. Constraints to the Use of Early Warning Systems.

 

 

 

7 Actors Institutional Constraints Political Constraints Logistical Constraints
Government Lack of human and Conflicting political Lack of financial
financial resources. objectives. resources with which to
Inappropriate Reluctance to accept aid act.
bureaucratic procedures conditionally. Lack of Infrastructure and

Donors

NGO’s

 

and structures.
Avoidance of
Responsibility.
Lack of Institutional
Memory.

Ad hoc response
arrangements.

Distance from problem
and failure to appreciate
lag time in delivery relief.
Inadequate co-ordination
between field and HQ, or
with other donors.
Inappropriate
bureaucratic procedures
and structures.
Avoidance of
Responsibility.

Lack of Institutional
Memory.

Dependence on Donors.
Bureaucratic Overload.
Lack of Institutional
Memory.

Political incapacity.
SAP's.

Home country political
agenda.

Unwillingness to breach
sovereignty.

Lack of accountability to
local beneficiaries.

Difficulty in preserving
neutrality.

Lack of accountability to
local beneficiaries.

Own political agenda.

 

 

weak transport sector.
Lack of trained local
administrators.

Distance over which
resources must be
mobilized.

Lack of infrastructure and
weal transport sector in-
country.

Lack of financial
resources with which to
act.

Ad hoc response
arrangements.

 

Source: adapted and modified from Buchanan —Smith (1995)

This is not to argue that the micro-scale is not the most appropriate, and

that monitoring local responses is not the most accurate indicator of coming

problems. Rather, given the present financial and political situation in most

drought prone nations, a micro-scale approach is simply less feasible and impacts

potentially far fewer people than a national approach to EWS.

 

53

Entitlement Theory Rarely until the early 1980's did the early warning
literature focus on the people directly affected by famine and food insecurity. The
approach to warning systems, as noted above, had been supply oriented, i.e.
food supplies, food availability, and soil conditions. Amartya Sen, writing in 1980,
challenged the status quo of famine early warning, notably that as food availability
declines (FAD) started, a famine was sure to happen. He also challenged the
notion that famines were caused by a FAD, also prominent in the literature.

Defining the entitlement approach, Sen writes:

The entitlement approach to starvation and famines concentrates on the ability of
people to command food through the legal means available in society, including
the use of production possibilities, trade opportunities, entitlements vis-a-vis the
state, and other methods of acquiring food. A person starves either because he
does not have the ability to command enough food, or because he does not use
this ability to avoid starvation. The entitlement approach concentrates on the
former, ignoring the latter (Sen, 1980, p.45)

The entitlement approach developed as a response to FAD theory of
famine. Sen remarked about FAD at length in his manuscript Poverty and
Famines: An Essay on Entitlement and Deprivation. Sen showed, using many
famine experiences, that famine and hunger can occur without a substantial FAD.
This argument had been used to explain the 1943-4 Bengal famine and the 1973
Ethiopian famine, however, Sen highlights faults with this assumption. Political
maneuvering and lack of entitlement to food rather than a dramatic FAD caused

the Bengal famine. The same was said for Ethiopia.

 

54

Sen countered FAD theory by maintaining that access to food and the
entitlements of individuals was more at issue when examining famines than was

FAD. Two kinds of entitlement failures can cause a lack of food:

[maximum food entitlement] can fall either because one has produced less food
for own consumption, or because one can obtain less food through trade by
exchanging one's commodity for food. The former will be called a direct
entitlement failure and the latter a trade entitlement failure. The former can arise
for food producing groups, while the latter can occur for others (Le. for those who
sell their commodities to buy food) (Sen, 1 980,p.51 ).

The failure of entitlement is the true cause of famines, Sen maintains.
On the issue of food exports during famines, he states that the balance of
entitlements can cause problems. People with weak entitlements can not acquire
food even if it is available. The mechanisms they employ to secure food have
collapsed either as a result of direct entitlement failure (i.e. crops fail, herds die)
or because of a trade entitlement failure(i.e. the services provided are no longer
affordable). Those with strong entitlements, therefore, are able to access food.
They have the means to acquire food at any cost. In the Ethiopian famine of
1973, food was actually being exported from the worst hit region- Wollo. The
food entitlement in Wollo as a whole was weak, while strong outside the region,
therefore allowing export of food. The same situation prevailed in Bengal in 1943,

another region where food was being exported while thousands were starving.

The entitlement theory of famine has led many other authors to call for
more social approaches to food issues. Buchanan-Smith (1995), Davies et al

(1991), and Downing (1990), among others, have contended that a more people-

55

oriented warning system would most accurately predict the onset of famines.
Campbell (1990), Walker (1989), and Buchanan-Smith (1995) suggest
incorporating analysis of coping strategies as one possible indicator for an EWS.
Torry(1984) has urged the use of more socio-economic indicators to assist in
identifying the onset of famine. These approaches, often referred to as demand
side, are similar to, if not derived from the entitlement approach. More recently,
the literature has focused on the maintenance of livelihoods. Protecting
livelihoods during famines will enable people to better withstand future events.
Families will not find themselves in a weakened entitlement position at the onset

of the next major famine, and will therefore be more able to survive.

Gender and Food Security One of the pivotal issues in food security
research is the role of women in agriculture. Research into women’s role in the
household production cycle generally focuses on the gender division of labor.
Women’s roles vary greatly throughout Africa. Just as dryland production differs
from pastoralism women’s roles in each system differ, however, commonalties do

exist.

Women are an integral part of the household production system. Often
women’s labor provides the bulk of both farm and off-farm income, not to
mention the hours of unpaid domestic work performed almost exclusively by

women. The FAO best sums up the role of women in agriculture:

 

56

Women are generally responsible for providing certain foods for the household that
complement the foods and other goods for which men are responsible. Women
provide these foods either by producing them in separate fields, gardens, or
livestock concerns, or by selling or exchanging some of their produce to obtain
foods they need...Furthermore, men’s incomes are not often used to make up a
shortfall in women’s production. Thus, in much of Africa, thinking in terms of a total
household production or total household income is inaccurate. Two sets of
activities, men’s work and women’s work generate the household’s wealth and
meet different obligations to the household... (FAO, 1987, in Gittinger.)

Campbell and Trechter ( 1983) observe differences between men and
women’s income and coping strategies in their study in northern Cameroon. They
noted that women tend to stay near the home to cope, while men tend to migrate.
Huffman confirms much of Campbell and Trechter’s work, finding that women
tend to be more involved with subsistence crops, while men tend to be involved
with cash crop production (Huffman, 1987, p.375). She further notes that in two
villages in Cameroon “women provided 57 percent of family cash income” (ibid.).
Women’s role in the sustenance of the household should not be overlooked, but
unfortunately it has been. Staudt writes “macrostatistics detailing women’s
involvement in agriculture are far from adequate” (Staudt, 1987, p.37). She
further writes of the role of women’s labor in the household and the relationship
between men and women in regard to the exchange of labor. Women influence
greatly the subsistence of the household. Chidzonga confirms much of Staudts
observations regarding women’s work remarking that "women...have a central
function and vital stake in the agricultural sector and its development”
(Chidzonga, 1993,p.181). Attention has been drawn to the role of women, but
empirical studies of women’s work are few. The role of women in the provision of

food is commanding and must be investigated further, especially in South Africa

 

 

57

where the government has placed an interest in the empowerment of women.
This coupled with their role as food providers and preparers places women in the

center of food security issues.

 

58

Review of the Literature: Food Security in South Africa

Food security in the rural areas of South Africa has been a systematically
understudied topic. The apartheid legacy in the rural areas manifests itself as a
lack of information on the daily rigors, trails, and tribulations of the local
populations. A comprehensive body to monitor rural health and local environment
has materialized only as recently as 1993 (Vogel, 1993). The degree of ignorance
about rural food security and health issues in the rural areas is highlighted by the
fact that out of over 1,000 entries in a recent publication devoted to food security
in Southern Africa, less than 50 made specific reference to South Africa’s rural
food security situation (Geldenhuys, 1994). Furthermore, voices in the rural

development field in South Africa decry the sad state of knowledge. Vogel writes:

while much is known of such [coping] strategies in the rest of Africa, including the
areas to the north of South Africa, little local detailed knowledge is available.
Communities in rural areas, even where well intentioned drought relief has been
devised appear to lack any ability to actively mobilize themselves around droughts
(Vogel, 1994).

Vogel further argues that responses to drought in rural South Africa are
different from the general pattern described above for the rest of Africa. She

writes of a case study:

The notion of a linear, organized series of coping options in the face of harsh
droughts as mentioned in some of the literature for the rest of Africa, does not
occur in this case. Rather strategies are locally/regionally specific. The choice of
coping mechanism is influenced by variables peculiar to these areas while the
role of violence as a deterrent in migration options cannot be underestimated

(Vogel, 1994, p.2)

59

Adams (1994) also calls for a wider understanding of the issues facing
rural dwellers. She recounts the effects of drought on the rural areas in KwaZulu-
Natal and proposes several “strategies for drought relief’ (Adams, 1994,p.1). Her
first concern regards the distribution of drought relief funds. She maintains that
these funds benefited primarily white farmers with large land holdings and that
rural areas received little in terms of support during the 1992 drought, one of the
worst in decades. The distribution of drought relief was “dispersed through a
plethora of bureaucracies created as a result of the homelands system” (Adams,
1994,p.3). She maintains that this system was an inefficient agent of relief. Lastly
she addresses the issues of the allotment of water supply. She argues that there
exists no “strategy for ensuring rural water supply” (Adams, 1994,p.3). To counter
these problems she suggests the development of a “National Eany Warning
System for Food Security” be established. This system would be responsible for
alerting authorities of impending famine before it become too late. She further
advocates similar systems of monitoring for water supply and nutrition. Adams
echoes Vogel's calls for more information gathering: “the rural voice needs to be
strengthened in order to continue to lobby for more appropriate and effective

relief and development programmes” (Adams, 1994,p.6)

Some common themes throughout the Review of Literature will be
addressed in this study. First, the history of shortages will be identified. The cycle
of drought is as important as the effects of drought. Second, the coping

mechanisms and sources of food during shortage will be identified. Thirdly,

60

prevention/mitigation and expectations of future famines will be investigated as a
component of possible early warming systems for the area/region. An analysis of
the use of GIS at the local scale will also be presented in the context of EWS.
The need for such a study is clearly presented by Vogel. These most basic

questions will be addressed in this study.

Chapter Four

Results

To accommodate the wide range of data collected and the various
methods of interpretation used, this chapter is divided into four sections. The first
section presents summary statistics from the questionnaire schedule. The
second section details the results of statistical tests to determine relationships in
the data. Analysis of the qualitative data is presented in section three, and
section four explores the efficacy of using GIS in the analysis of the data from the

interview schedule. A brief review and synthesis of the hypotheses follows.

Both the summary statistics and tests were designed to investigate
themes discussed in the Review of Literature including the history of shortages,
coping strategies and preparations for future shortages. Several hypotheses
were drawn from the literature and tested through the use of both the summary
and test statistics. The results of the statistical operations are presented in this
chapter. The relevance of the tests to the thesis will be presented in the

Discussion section.

The qualitative section allows for more probing of issues that only came
to light during the course of quantitative interviewing, such as the apparent
inconsistency between the oven/vhelming response that the government has not

assisted the area, while power lines hung overhead and windmills pumped water.

61

62

Finally, concepts, ideas, and nuances lost in the translation from English to Zulu

are discussed to assist in future study designs.

Summary Statistics Summary statistics are essential to this thesis as it
was designed to be a tool for primary data collection. The scope of the study
limits the amount of statistical procedures that can be used to analyze the data.
The data is nominal in nature, which violates the assumptions of both correlation
and regression tests, and therefore, bars their use. Summary statistics provide
groundwork for understanding phenomena at the most basic level. In areas where
research has not previously been conducted, such as the study site described
herein, often the most useful information is that which describes the setting, or
context or the area. The summary statistics have been grouped and are

presented by theme.

In order to characterize respondents to the survey, several demographic
variables were collected on each respondent, including the gender, age, and
number of people in the family. OvenNhelmingly, respondents were female.
Females represented 90.3 percent of the survey. Men were often not at home or
inaccessible. The issue of female bias was addressed to some degree in the
qualitative data, where men were more willing to participate (see footnote, p.21).
The average age of respondents was 41.4 years. The youngest respondent was
16 (1 respondent) and the oldest was 87 (also 1 respondent). The age data was

grouped into cohorts for ease of analysis (see Table 4).

63

Table 4. Age Cohorts for the Sample at Zidonini

 

 

 

 

 

 

 

 

 

 

, Frequency Percent
0-19 years old 1 0.7
20-29lears old 20 13.9
30—39 years old 44 31.2
40-49 years old 25 17.4
50—59 years old 34 23.6
60 + years old 17 11.8

 

Southern African societies are comprised of younger populations, and
this study concurs finding the highest number of respondents to be in the 30-39
year old age cohort. While population per cohort in Southern Africa is highest in
the under 20 category, this study has largely excluded this cohort. The design of
the study did not actively seek to exclude the youngest cohort (or the eldest,
either). Rather, when asked for interview, persons often deferred to the “head
wife”. This is also reflected in the gender composition of the survey. After asked

for an interview, most men deferred questions to their wife, or wives.

The mean number of people in a household was 10.7. This does not
reflect the number of pe0ple per dwelling, rather the number of relatives or
dependents living in the same homestead, therefore, the number of people who
could call on the interviewee at any time for assistance. Because of inter-familial
ties, and lack of precision over the definition of a “household”, “homestead”,
“dependent”, or “relative”, this statistic may reflect a disproportionately high

number of “people living in your household”.

 

64

The final socio-economic factor relevant to the study was the number of
years an interviewee had lived in the given study area. This indicator was
extremely important as new arrivals could not comment on even the most recent
shortages and lifelong residents had extensive personal recollections of
shortages. The mean residency in the study area was 26 years. The mode was
20 years, confirming both the in- migration in the 1950’s and 60’s and the large
component of younger respondents. Several key informants corroborate the

finding that large population increases occurred in the 1960’s.

The bulk of the interview schedule was designed to illuminate the history
and causes of shortages, how people cope (or coped) with shortages, whether or
not they were preparing for future famines, and what expectations people had
regarding future assistance, particularly from the new South African

administration.

Several factors of production had to be established before reasonable
analysis could begin, such as the number of fields and gardens, and the type of
tools used. Wide variation could be anticipated if large numbers of people had no
access to the key components of production while a few had access to the entire

component of production.

The average number of fields was just over one, while the average
number of gardens was between zero and one. The majority of respondents
(n=102) had under two fields. Hoes and oxen were the primary tools used by

respondents in the study, as mechanized farm implements were available to very

- “v,—

65

few. Hoes were used by 131 people, oxen by 96, tractors by 23, and other

implements such as rakes and scythes, by 15.

Central to this thesis is the recent history of shortage, coping
mechanisms, and expectations of assistance. Respondents were asked to list
the years in which they experienced a severe drought. Only three respondents
reported never having a food shortage. Twenty-four percent of the respondents
reported experiencing a food shortage in 1983, and eleven percent noted that
they experience a shortage in 1982. In fact, the years 1980-83 were mentioned

by thirty-eight percent of the respondents as shortage years.

The early 1990’s were more often observed as shortage years. Twenty-
five percent of respondents indicated 1992 as a shortage year and eleven
percent indicated 1993. In total the years 1990 through 1993 were reported as
shortage years by forty-nine percent of respondents. However, when controlled
for length of time in residency, more respondents who had lived in the area since

1980 reported experiencing the 1982-4 drought than the 1992-3 drought.

The shortage data was compressed by case to evaluate the number of
times a respondent reported experiencing a shortage. This allows for some
rudimentary analysis to be performed on highly vulnerable groups (if any exist).
Twenty-eight percent reported experiencing only one shortage, while forty-four
percent noted two. Twelve and a half percent experienced three periods of
shortage and just fewer than ten percent of the sample experienced four or more

periods of shortage.

66

When asked about severe shortages, the situation changes somewhat.
The year reported most often for severe shortage was 1983 (33.1%), followed by
1992 (17.2%), 1982 (12.6%), and 1993 (11.3%). Far fewer people had
experienced more than one severe shortage (16.7%), unlike the same statistic for

(general) shortage.

Table 5. Occurrence of Food Shortages and Severe Food Shortages5

 

 

 

 

 

 

 

 

 

Shortage SEVERE Shortage
Frequency - (Percentage) Frequency- (Percentage)
No shortages 4 (2.8%) 3 (2.1%)
One period of shortage 30 (20.8%) 97 (67.4%)
Two periods 64 (44.4%) 14 (9.7%)
Three periods 18 (12.5%) 6 (4.2%)
Four periods 12 (8.3%) 0
5 or more periods 2 (1.4%) 0
Missing/ NR 14 (9.7%) 24 (16.7%)

 

 

 

The second major subject of inquiry was coping strategies/mechanisms.
Coping mechanisms, as noted in the Review of Literature, usually vary with
increasing intensity of shortage. The easiest coping options are exhausted first,
and thereafter, more severe measures are generally implemented. When asked
their first response to drought, forty-five percent of the respondents remarked that
they made and sold baskets. Asking neighbors for assistance and buying food in
shops were both reported as the first response by eleven percent of the sample.

Attempting to cultivate was the first response for six percent of the sample6.

 

5 This is NOT controlled for length of residency.

6 Men do not make baskets as their first coping mechanism. Most men said that if the crops failed they tried
to plant again. The first coping mechanism of men was to try to plant again. They then noted that they would
go to shops to buy food, similar to women.

 

67

When asked about the second coping mechanism thirty-eight percent
responded “none”. Twelve and a half percent made and/or sold baskets and
eleven percent relied on store bought foods. Nine percent of the sample used
stored foods. Overvvhelmingly (94.4%), coping ends after the second cption is
exhausted. The lack of a third coping mechanism will be highlighted in the

discussion section.

The most significant source of food during shortages is the purchase of
food in shops (35% of responses), followed by other sources such as selling
wood, fruit, goats, etc.(15%) and only then does basket making/selling appear as
a source of food (12.5%), followed closely by the use of stored foods (11.7%) and

dependence on neighbors/relatives (10.8%).

The difference between the coping mechanism and sources of food
question is interesting when examined in light of the basket trade. Forty-five
percent of respondents said that they made and sold baskets as their first coping
mechanism, however, only 12.5% report it as a source of food during food
shortages. This may indicate how many baskets were actually sold. This research
is insufficient to draw such conclusions firmly, but indicates such a phenomenon

may exist.

Because of the proximity of the Hluhluwe-Umfolozi Game Reserve a
question was incorporated which would attempt to access the state of hunting in
the study area. Wild game is often used in other parts of Africa as a source of

food, but has a long history of “protection” in the apartheid period in South Africa.

68

The majority of respondents (67.8%) reported that they do not hunt while 28.5%
reported that they do hunt. The policies on hunting in the homelands were similar
to those in apartheid South Africa, however enforcement was drastically different.
Both in homeland and “white” South Africa, special costly permits were needed to
hunt game. This actively discouraged illegal hunting in “white” South Africa, but
enforcement in the homelands was scant at best. The situation changed when
the Hluhluwe and Umfolozi Parks were consolidated. Stricter enforcement in the

KwaZulu homeland areas around the park curtailed hunting once more.

In order to follow up on the progression of coping mechanisms, a
question was included which asked about the level of assistance from the
community in general, as community assistance often represents an indirect form
of a coping mechanism. However, in this case, seventy-five percent of
respondents reported that no one in the community assisted them, while twenty-
one percent noted the opposite. A fuller explanation of problems related to the
outcome of this question can be found in the discussion section. A second
question was asked to find the influence of remittances and labor migration. The
majority of respondents did not work elsewhere or have family members working
elsewhere (88.2%). Of those that did work outside the study area, the most
notable locations were Johannesburg, Hlabisa, Empangeni, Durban, Richard’s

Bay, and Nongoma.

The final question that dealt with coping related topics queried
respondents on assistance from the Zulu homeland government. Governmental

assistance was recognized by just fewer than thirty-two percent of respondents.

69

Infrastructure in the area was noticeable and the apparent conflict will be

addressed in the Discussion section.

When asked why the last shortage occurred, two environmental factors
overwhelmingly dominated — heat/sun and drought/rain (47.2% and 36.1%
respectively). Only eleven percent reported that they “did not know” what caused

the last drought.

The final emphasis of the study revolved around expectations. Questions
included topics that probed for any preparations for future famines and perceived
needs of the community. Basket trade dominated the question on preparations
with fifty-nine percent of the respondents reporting that they were actively making
and/or selling baskets to safeguard against the next shortage. Storing food was
the second most important preparation being undertaken to prevent future

shortages (25.4%).

Community input was gathered on the perceived needs of the community
- what they thought they needed. Just over nineteen percent of the respondents
maintained that water was needed, just fewer than seventeen percent said
schools/pre-schools were needed, fourteen percent said clinics, and eleven
percent said jobs were needed. (Note, the question did NOT ask for a ranking of
what was needed, rather just for a list of things that people thought were

important).

In addition to the questions presented above, a composite variable was

created which helped to emphasize the most vulnerable component of the

70

population (see Methods for a more detailed description). The Vulnerability Index
proved to be a useful indicator of the number of highly vulnerable people in the
study area and will also be used in the next section dealing with statistical tests.
In this variable, the numbers six through four (six being the most vulnerable)
represented the highest level of vulnerability. These individuals met over half of
the requirements to be classified as “vulnerable”. The numbers three through
one represented those people with moderate to no vulnerability. No individual
met all six requirements to be classified as the “most vulnerable group”, however,
40.3% of the population fell into the “highly vulnerable” grouping. Just fewer than
60% of the population was moderately to slightly vulnerable. The percentage of
the population in the higher categories is significant and will be examined in the

Discussion section.

The summary statistics have provided the framework from which the
statistical tests in the next section are drawn. Many of the hypotheses formulated
from the Review of Literature section have been tested in the above section and

will be further reviewed in the Discussion section.

 

71

Table 6. Vulnerability among Sample Population

 

Frequency Percent Classification

 

1.00 9 6.3% Slightly vulnerable, met only one
characteristic of a “vulnerable
individual”

 

2.00 27 18.8% Somewhat Vulnerable, met two
characteristics of a vulnerable
individual

 

3.00 50 34.7% Moderately Vulnerable, met three
characteristics of a vulnerable
individual

 

4.00 43 29.9% Highly Vulnerable, met four
characteristics of a vulnerable
individual

 

 

5.00 15 10.4% Highly Vulnerable, met five of the six
characteristics of a vulnerable
individual

 

 

 

 

Analysis of Statistical Tests In this section the results of statistical
tests to determine the existence of relationships in the data are presented. These
tests originated from themes in the literature and were incorporated into the
interview schedule. The variables were then summarized and analyzed in the
last section to determine the efficacy of their use in this section. Many of the
variables presented in the last section were used to uncover relationships in the
data. This section explores those relationships using the chi-square test. This test
was chosen because the nominal nature of the data prevented the use of
correlation or regression tests. The confidence limits were set at 90 percent for all
tests. The Statistical Program for the Social Sciences (SPSS) was used for the

chi-square testing.

Results of Testing In order to highlight the major themes in this thesis,

several tests were conducted to support or question hypotheses regarding the

 

72

state and nature of food security as it pertained to rural Zululand in early 1997.
These themes, introduced at the end of the Review of Literature section, include
understanding coping mechanisms, past history, and the expectations of the
residents. Statistical tests to uncover relationships were utilized to further build
upon these themes and begin to better understand the meaning of the data

collected.

Determining the causal factors of shortages included the testing of
several variables against the number of shortages. The shortage question is
open-ended and only requires the dates of shortages be listed. While helpful in
understanding the cyclical nature of drought in southern Africa, this did not prove
useable until a variable was created that compressed the open-ended dates into

a simple numerical counting of the number of shortages per respondent.

The first test was a cross tabulation- chi square test to determine whether
or not a relationship existed between the number of shortages and the distance
from the Nduna. As the center of power for the study area, the Nduna controlled
the distribution of food in times of stress, when it was available from the KwaZulu
government. Respondents were coded according to their distance from the
Nduna, i.e. 1-kilometer. From this test a chi-square value of 9.209 was attained
with four degrees of freedom. The critical chi value for the ninety- percent
probability range is 77.79, thus the null hypothesis is rejected. There is a

relationship between distance from the Nduna and the number of shortages.

73

Table 7. Chi-Square Results for Proximity to Nduna Test

 

 

 

 

 

One Shortage Two Shortages , Three or Mbre
f , * ~ ~ . ~ * » Shortages
Nearest to the 9 33 18
Nduna (1 Km)
Second Nearest to 15 14 9
the Nduna (2 Km)
Furthest from the 10 17 5
Nduna

 

 

 

 

 

 

 

 

 

Test Chi-Square df

Pearson 9.209 4

Likelihood 9.571 4
Ratio

 

 

 

 

The second relationship explored with the chi-square test was between
the number of fields and the number of shortages. Often those farmers with more
fields, or more access to fields, fare better during times of shortage. In fact as
evidenced from the Review of Literature, a political ecology approach would
support the proposition that groups of individuals or households with more access
to fields are better off than those with less access. A political ecology perspective
would maintain that it is the differentiation of farmers that promotes food
insecurity. However, this research shows that in this particular situation the
number of fields has no relationship with the number of shortages. A chi-square
statistic of 5.420 was attained with six degrees of freedom. The critical chi value
at the ninety- percent level is 10.645 — again, the null hypothesis must be

accepted. Statistically, no relationship exists.

74

Political factors were next tested to determine any influence outside
assistance had on the intensity and prevalence of shortages. One of the more
important questions considered the degree of help received from the KwaZulu
government, both during and after apartheid and the homeland system of
governance. With a chi-square of 0.638 and two degrees of freedom, this test

also failed to reject the null hypothesis.

One of the final factors of food insecurity reported by key informants in
the qualitative section and queried in the qualitative section was the impact of the
loss of hunting rights due to the encroachment of the Hluhluwe-Umfolozi park
officials into the lives of residents. Quantitative testing shows that indeed a
relationship exists between hunting and shortages of food. The chi-square for this
test (10.386, df=4) exceeded the critical chi-square leading to the rejection of the

null hypothesis.

Finally, the vulnerability index created to investigate the idea that chronic
shortages are caused by a “package” of vulnerabilities rather than disconnected
chance relationships, was tested for the ability to provide useful insight into
relationships in the data. The main function of the index was to further investigate
any relationship between vulnerability and distance from the Nduna. However, the
chi-square test showed that the two variables are independent, suggesting more

investigation of the Vulnerability Index.

This section presented the results of the statistical tests used to highlight

relationships in the data. The tests helped to prove or disprove assumptions

 

75

made at the end of the Review of Literature. The findings and shortcomings of

the statistical tests will be reviewed in the Discussion section.

Qualitative Analysis The qualitative analysis of key respondent data
provides much needed insight to issues of particular interest that arose in the
course of interviewing. This section focuses on the provision of basic
infrastructure, the role of the Nduna, history of shortage in the earlier part of this

century, and the role of the basket trade.

When a respondent to the quantitative section felt the need to elaborate
before, during, or after the interview, the statements were recorded. Often,
people who had not been randomly selected felt the need to express their views
and in these instances their information was recorded in a qualitative format’.
More men agreed to participate in the informal interviews. Thirty percent of the
qualitative interviews were with men, providing a mechanism for comparing men

and women’s responses.

One of the most often discussed topics throughout the fieldwork was the
basket trade. Often while conducting interviews or waiting to be interviewed,
women would be sitting and weaving baskets or engaged in exchange of baskets
for other goods. The older key informants maintained that only recently had
baskets become a medium of cash exchange. In the past baskets were primarily
for use in the household itself for carrying water, storing beans, and other

purposes. Exchange with whites in earlier times was very limited. The only means

 

71D fact, many of the refusals to the quantitative portion would answer informal questions.

 

76

of sale was for migrant workers to transport them to interested employers. With
the “opening up” of the rural areas during the establishment of the homelands,
and especially in the late 1980’s, women began to be able to sell baskets to
tourists. Major relaxation of apartheid laws in the early 1990’s further facilitated
the basket for cash trade. As the study area is rather far inland, one option for
sale of baskets was for the women to set up stands along the N-2 road, which
runs from Durban to the northwest into current day Mpumalanga. However, the
over 75 km trek to the freeway, coupled with the high cost of transportation
resulted in few trips to sell baskets. Women were needed more in the home. The
only other method for the sale of baskets was for the women to rely on visitors
traveling through Hlabisa to Nongoma. The returns from the market at Hlabisa
were not as profitable as those from the N-2 road. At the time of my arrival in late
February, 1997, the women at the Zidonini site reported having not sold any
baskets since the previous July. The women at Sangoyana, Ximbakazi,
Madlinjini, and Zagomfe reported similarly dismal sales. Waiting by the main
north-south road through the area proved to be the only means of selling baskets.
Children were often sent to the road to waive down passers-by in an attempt to
sell baskets. Women outside the study area closer to Hlabisa City reported
generally better sales along the road to Nongoma and during the three month
observation period were often engaged in the sale of baskets, although of lower
quality than those produced in the Zidonini-Ximbakazi area. This supports the

suggestion made in the summary statistics section that the difference between

77

the use of baskets as a coping mechanism and their use as a source of money

for food represents the number that were actually sold.

The role of baskets was only one of many topics that needed further
investigation. In the quantitative section of the field research sixty percent of
respondents noted that the government had never helped them, yet power lines
hung overhead and water pumps were nearby. After speaking with several key
informants, including a lengthy conversation with the Nduna, the situation became
clearer. The KwaZulu homeland government had erected power lines through
about half of the study area and had plans to erect more in the near future,
according to the local administrator”. The homeland government administered a
project to drill boreholes throughout the more arid and drought-prone regions of

the homeland. Several pump systems could be found throughout the study area.

The apparent discrepancy between the actual government assistance
and response to the question was uncovered in follow-up interviews after
quantitative questioning was completed. When asked, “Did the Zulu or national
government ever help?” most respondents replied that it hadn’t helped them
personally. Although they knew that the government provided the infrastructure,

they considered this a form of “community improvement” and not a direct benefit

 

8 The KwaZulu homeland government ceased to exist in 1994 with the election of the ANC and end of
apartheid laws. The homeland government had jurisdiction over a disjointed territory that is now entirely
part of the KwaZulu-Natal province. Administrative control is still exercised from Ulundi, about 100 km to
the southeast of the study area.

78

to them, particularly in times of drought.9 This incongruity in the interview was

never resolved. Semantics were adjusted after the pilot study to no avail.

The history of drought was a fertile topic for discussion and older
informants were rather helpful in this respect. Droughts occurred in the late
1930’s, 1945-48, 1969, 1974, 1983, and 1992. The late 1930’s drought was
remembered as being particularly severe as was the 1969 drought. Several of the
oldest respondents referred to the use of famine foods such as figs and sour milk,
which incidentally are no longer identified as coping mechanisms. Settlement of
the Zidonini area had occurred at the turn of the century as a result of a drought

and famine in the Nongoma area.

Numerous informants reported filling out forms from the government for
food assistance in both the 1992-3 and 1982-4 droughts but to no avail. This was
an exceptionally contentious issue with the local population. Many people were
still upset that the government had promised them food and had not delivered,
which may help explain the fact that many respondents reported that the
government had not helped them. The severity of the 1982 hunger was worsened
by the fact that people were waiting for food from the government rather than

making other preparations.

The impact of the Hluhluwe-Umfolozi Game Reserve was a lively topic of
conversation among key informants. Several key informants were relocated

because of the merging of the two parks in 1989 and had faced food insecurity as

 

9 It must be noted here that some of these same respondents had electrical appliances running in their homes

 

79

a result. The two game parks, Hluhluwe and Umfolozi, had been separated by a
strip of land known as the Corridor (see Figure 5, p.17). In 1989, the Natal Parks
Board joined the two parks and evicted the residents from the land. The only road
from the east to the Hlabisa area ran through that strip of land. Eventually, the
Parks Board consented to allow public access on the road, thus keeping Hlabisa
from being isolated from the east. The consolidation of the Parks is a very sore
issue in the minds of the people near the Park. Hunting restrictions outside the
Park and the loss of access to game within it were sited as major causes of food
insecurity. lnformants noted that they were allowed, under certain conditions, to
hunt game that escaped from the Park, but were often harassed by Park officials.
Cattle kraals10 within the Park were one of the concessions made for the loss of

access to game.

GIS Analysis This section explores the use of GIS as a tool for
understanding the spatial relationships of socio-economic data from a rural
southern African setting. Included in this section are the particular challenges,
obstacles, and design issues required for the effective use of a GIS in such a
setting. The data from the previous sections in this chapter will be used to

illustrate the major topics mentioned above.

To help avoid some of the problems noted by critics of GIS (particularly
Pickles,1995), the relationship between variables used in overlay procedures

were proven through the use of the Chi-square test. This helped to prevent the

 

as they were saying this.
1° Fenced areas in which cattle are kept for the night.

80

impression of relationships in the data that did not exist. Any two or more

variables can be used in the overlay process, related or not.

Before using variables in the overlay process, some basic spatial patterns
were identified. The percentage of respondents with one or fewer fields was
mapped (see Figure 6). Madlwaleni and Sovane had the highest percentages of
peOple (upper end at 90%) in this category while The Ridge, Madlinjini, and
Ximbakazi had the fewest (lower end at 33%). At the opposite end of the
variable, the Ridge, Madlinjini and Ximbakazi had the highest percentage of
respondents with three or more fields (see Figure 7). Both of these sites are

furthest away from both the Game Park and the Nduna in hilly terrain.

The areas with the greatest percentage of respondents reporting no
assistance from the KwaZulu government were rather randomly distributed (see
Figure 8). The highest proportion was at the Ridge and Sovane, followed by the
Tree, Ximbakazi, and Sangoyana. Surprisingly, one of the most inaccessible and
remote locations, Madlinjini, had the lowest proportion of respondents who had

not received assistance.

The spatial distribution of sites with high proportions of respondents
reporting no hunting was not surprising given the history of game wardens
patrolling the areas near the game reserve. The four sites closest to Hluhluwe-
Umfolozi Park, Khalweni, The Ridge, The Tree, and Zagomfe, had the fewest

hunters, while the most distant sites, Sangoyana and Ximbakazi had the most.

81

Next, the number of shortages was plotted (Figure 9). While the
overwhelming response in the survey was two shortages, Madlwaleni and
Madlinjini sites had the highest proportion of respondents reporting one or fewer
shortages. Twenty to forty percent of respondents at Khalweni, the Nduna, the
Ridge and the Tree reported having experienced three or more shortages. Of

these sites, all except the Nduna border the Game Park.

Some very basic overlay operations were conducted in the GIS to help
show spatially the relationships uncovered in the previous sections of this
chapter. These overlay operations included the presentation of only statistically

significant relationships between variables.

One of the more important findings of this chapter is that a relationship
exists between hunting and the number of shortages. While the Chi—square test
cannot indicate causality, a hypothesis was made that as hunting declines, the
number of shortages rises. When queried for the sites with greater than twenty-
five percent of respondents reporting three or more shortages and greater than
seventy-five percent of respondents reporting no hunting, Khalweni, The Ridge,
and The Tree were shown to have met the criteria (Figure 10, represented as

triangles). These three sites are the closest to the game reserve.

Another theme in this thesis is that distance from the Nduna is related to
the number of shortages. Proven significant at the ninety- percent level, this
makes for an interesting map (Figure 11). The hypothesis made for this variable

was that those furthest from the Nduna would have more shortages due to lack of

82

political contact with him. However, the spatial picture does not support this. Sites
with more than twenty-five percent of respondents having experienced three or
more shortages were actually closest to the Nduna. Respondents at the more
inaccessible and distant sites such as Ximbakazi, Sangoyana, and Madlinjini

actually fared better.

Finally, the junction of respondents reporting both no hunting activity and
no assistance from the KwaZulu government in times of shortage was mapped
(Figure 12). A significant level of vulnerability in the population is expressed in
these two variables. Upon examination, the four sites closest to Hluhluwe-
Umfolozi have greater than half the sample population that neither hunts nor has
received assistance. Closely related is the vulnerability index map (Figure 13).
The four sites near the park have some of the highest levels of vulnerability.

Sovane and Madlwaleni also rank high.

GIS and Food Security At the local scale of analysis, GIS can both
enhance and detract from a food security project. This section is a summary of
the challenges of incorporating GlS into the field research component and the

potential benefits of its use in early warning systems.

The most acute problem in this study was the lack of positional accuracy
in the location of households. Lacking a GPS", accuracy on the precise location

of households was low and, therefore, households were rejected as the unit of

 

1' Global Positioning System — A system utilizing a series of satellites to locate precise points on the earth’s
surface. Location of a particular phenomenon is accomplished through the use of a hand held positioning
unit.

83

analysis for the mapping component of the study. To compensate, a centroid was
located for each study area using an orthophoto and landmarks. Data from each
site was then aggregated to the centroid. This process admittedly lowered the
resolution of the data, but the accuracy was not as severely compromised as it

could have been if an attempt had been made to plot each individual household.

Secondly, a common problem with research in lesser-developed areas
became apparent rather early in the GIS portion of the study. The most recent
orthophotos of the study area dated from 1982. A number of important landmarks
were missing and new landmarks had not been captured. The number of
households had increased since 1982, roads had changed or been constructed,
power lines erected, and water reservoirs built. Corrections to the orthophotos
took a considerable amount of time because they were used to identify pockets of

people for inclusion in the random sampling procedure and for basic navigation.

The primary benefit of using the GIS technology was the efficiency with
which the data could be queried and spatial patterns identified. Immediately upon
examination of the map showing the relationship between hunting and the
number of shortages, one gets a clear impression that the game reserve has a

large impact on life in the communities near it.

GIS use for early warning systems at the local scale of analysis, thus,
must be investigated further. The findings help to stress some of the benefits of
the technology, such as the display of spatial patterns, and some of the

challenges, such as positional accuracy. Through effective use at the local scale,

GIS can help present a spatial picture at regional 1 national scales of analysis.

Geographic data from local studies can be merged using GIS technologies and

analyzed at the regional scale. Cross comparison becomes easier and local

variation stands out. However, more research is needed on the use of this

technology at this scale of analysis for it to be widely beneficial to food security

studies.

The Hypotheses and Findings A brief review of the hypotheses is shown

in Table Eight.

Table 8 : The Hypotheses and Findings

 

Hypotheses

Findings

 

What is the history of drought and famine in
the Zidonini study area?

Droughts and Food Shortages occurred in
1982-1984 and 1992-1993 as many other parts
of South Africa.

 

ls drought or entitlement failure more
responsible for food shortages in Zidonini?

In this case, drought is more to blame for food
shortages.

 

How do the local and provincial political
institutions influence the level of food security?

Provincial political institutions have little effect,
while the local traditional authority has a
negative effect on food security.

 

What are the coping mechanisms to
entitlement failure?

Basket Making (45%), Purchase in Shops

(1 1%), Moral Economy (1 1%). However,
asking about the sources of food during
shortage provided better data on the
origination of food supplies during shortages.

 

What is the role of labor migration in the food
security equation in Zidonini?, Does it hold to
the convention wisdom on homeland areas
(the homelands were primarily labor reserves)

Very few respondents rely on the migration

economy during food shortages (12%). This
challenges the conventional wisdom on the

role of the (former) homelands.

 

What are the people expecting from the
government regarding food?

Food distribution in times of stress (98%)

 

 

What do the people believe will help them
secure food?

 

Clinics, shops, places to sell baskets, schools.

 

 

 

85

The results of data analysis have been presented in this chapter from
both quantitative and qualitative sources. The data was presented in a concise
format with little discussion included in the reporting of the statistics or key
informant responses. The next chapter will present a discussion of the findings in
the context of the local area, the Review of the Literature, and themes of the

thesis.

86

 

 

 

mogw 8226.295 .20 22:00 ‘ om. 02 5. .
mbwom I
5. 9mm. .
co_2m>o_m
mounumwh "mend hon—um mo. 23. .
No. 28. O

     

35 scene com—3:328
.22.... 2.82... 9F: we .2 E 2.

ottommN 2&2... __ 2.1.2.... __..._.

_
22— .2
.. . x. ..
. .5 ..
\

 

 

1.1.: ...
.2 .3 . ..

.fl._...__c_E__vm_2 _

 

_memo. E.x-.....
f.....tll..ll..2.vx m2”. 2” .._2. .H 2..- .2.. :21. 2.... 22.22.
mcm>omcmm
m . .2. ..\\\..
., _f .2.. ..2.\

2.. 2

 

mEoE Lose“. 2o 2.0 5225 mucoucoamom ho mama—23.2mm 2 o 9.39....

 

87

 

 

 

   
   

  

.. - .2 ......
2225 22:23.29... .5 2250 4 2.922881 9.2220 16.8%: w¢=2m22_2.._.... .
_ .,..........\o/ .2... 2.“ -......2.._- 2.\ .2... r ..

2... .._..HH.....\., -/. \2 .22 22.2.2.2...

       

 

     

    

 

 

“ONCE I ./... ....2.\._...... 2f ../ ..... 2 ..\ .- .... ff... _
2.. .._.._._...2_..2._2..-..._-_ 2 _ . . €922ch2 _.
..:|2..l.1:_ 212.222.22w221? ,1... .22 .222 222.....9..r2<_ ..... 2 l.“—
co22m>om2 l 2.2.0.. .2 .....
22.2.2”. .722/ _r .
consummh “pond 26.22pm .2.... .2.}. (xx--

 

k. ._ .2.-2.22: were.

22.22..

 

..__2.2.2 ...2.... .....~ .12.
_

 

l .. 22.22. . .22.“... 2;
.. 2f: immhlil .f! 2......—

2..

i..- \ 32.2223. .22
22

00. 02mm.
mm. oumw.
me. ouvo.

V0. 0200.

3:22:83”. .0 35:020.".

 

 

 

mEoE $.25. 52: 222 £225 mace—9:22am”. ho managemen— 2 2. 2:9“.

88

 

mm22w 92262802222 20 2250 4

mmunumwh ”awn/w hvnpm

 

       

.. .. .2
32.2222

 
   

_ . 82 9 3.

wumom I ..
co_2m>m_m mm. 9 on.
on. 9 Nm.
mm. 02 m 2.

mucuucoamwm ho 33:020.".

 

. 26.2 mi...

  
  

    

2....-

 

2 _memmr22x

..
__/

 
 

..22

 

EuEEgow :_:N 9: :8: 3532362 oz mEtonmm mucwuconmom 2 a 9:52”.

 

89

No. 9 :u.

_v. 09 ON.

macmucoammm ..o 38:00th

ON. 8 oo. .

 

 

1.1m: ,
>.

./..1...:.....

mmmmtofi 22>. 5 $5...

 

 

~5\ n/a m

_

mmmwtocm 02F

 

 

 

mmmwtocm 53m... 8 mac

 

83> cm... “man. 9: E mwmmtocm u_o 59:52 N a 2:9".

 

9O

 

 

              
        

J... \:V 1.... Dian... /, \_.\.1\.,_ . ;
3:0 8.322256 5.50 ‘ 020.00% 02.00 _No_o.—ED..0>>:_::_IK

21.4.. £41.... .L \VKM. M- —. ......

\1......\_ \\J ...JH: . ...\._ .11 vli 1

_ ... .n .\..,.,.... ... ”ff/I. 115\.. .1...\ .1.\1...H.1.H_W\.. ._....1\ :15|_/./rnl_
. ~ I ..

V...
.. . 1......0OU—m

®£n_-:WT”

wbmom l

        

       

.._.........~_. r....J .fiflw. ...._

cozm>2m . .. _ 00 ._ ._.

mwnnumwm :32 33.50

              

 

.. _ .
... d ....
_ _ _.

... .....x._.. J. _. J
{1.1.
—. .

Lei/xix»... MB_...1M.. %—

   

1.-.... 5:.
WEBER. _
A.rf..:11.m_...._.__.r.w.._.1..\

 

.71.. 1.1.1.1.

  

....................

«fit 03.3 u 20

EC 38: ‘

wcmEU ..me .02 800 .

 

:5... .02 on «a? as. 38:05 +m

 

mwmmtocw noon. H6 .3632 m5 :o 9.32:: 00 83:: ms..."

2 23mm

 

91

 

 

85 33223:. ,6 .250 4

was 8% u 20
mumom II

 

cozw>o_m ..__.w

8063. h
- .. \m‘ d: .._m“

/.~
r
x/
x /
~ ._. J."
mmuaummm “wand 5.35 ..x. .,,
x. .. . ..., . m,
._. fl.
_0 fl.
. _
k .
W w
_ _
. m
.._ a.
, m‘
r _—
_. u
. _
_y n
u owmm cmgtmfimgo ‘
1.: .m .. _w
_Hm. _. . mmmw cmztmgwi .
”H m .< mucmucoamom “.0 33:3th

 

mczuz 2: E0: 3565 new mmmwtocw 205. .0 3.....— 525 952033. "E. 2:9".

 

92

 

 

magw mEBoEoE. .o .250 <

mcmEU .3me ‘

mcmEU 6m: 62 $00 .
20—.— OZ u0>2omm can «:3: uoz OD o\oomA

mumom l

 

cozm>2m

mmunfimmm 3.82 $3.5m

_memngx

..Iq .....

.. .
22.1.... ... .x\ X.

.L.\.

 

.300 9: Eat 2w: oz tan 33:04. 95:3... oz mEtoamm mucwucoammm ”we 2:9“.

 

93

 

 

motw OC_>>0_>LO«C_ #0 .5“ch ‘
mumom .III
Coam>0_m

mwunummm um¢H4 bud—um

 

. in; ....2\/. ..... \\§....,_
...1. ., ......r .0 .w ......i. .....a...

        

   

.....11... 111...“.

.. .... if.
:.5}. 3.11.... .. 11.. 1. 111.111.?

.«t.....\.\1..\I.V.

..mmEm 0.._._.....__.

 
 

2‘
a“

    
       

i.) _

      

......1).._..1.. ._ 4.1.1..
.... ...). _
r. _ ..
.... . _ .._ .._. __..
... ... ... . . .fi
.r ......_ J _~ a.
. . ._. _ I .
. _ _l...,.... ... ....
_ ._ ~ ~ . _ .—
.. ... .. ~ .
._. _. w .....\1.\\\ .
_. \ ._.
_r. ._ _i... ... 1...:
.1... _ .
_ ._ . _ a .
_ _ _ _. ... . ~
. . . .
. ... ... .1 .L
. .... J _ fl .
F. "J _ .._ x
..A» / ... a...
. ... _ x... ...
_ _. . x... 1“
..1 . ..x
_ ...}. \WGI.
_ ..J. g
._
...“: m

...). _

_cmfiéumz _ .... .. ......

.tuk.
..

...:rk .._... ... .
_. ._.... . , . mcgomcmw
- , m... H
.H. w . .1 WK 1
_. . m ,_ _
,_ . . _ .
. 4 . ._.».
.
_ .

S. o. ow. .

on 9. om. .

0m. 9. mm. .
0

mm. 9 .m.

mucoucoamam ._o 3352?.

 

 

b___ne¢c_:> .0 £23 629: 2: £5, 8.5 ”2 2:9".

 

94

 

 

  
 

  

:1”... ... .1... . . a. .
.. ..
I ._ .

        

kw \1.? 1.-.-. x 211...... J J... x. .....J. xv...) ...-....IJP,

mama-mm mEmO. _N_o_£E3om>>:_I:_I .. 5m...
I I I .
_ .1 II. _I

1r.-
\..H.....m.: ..-I f... ...
./

...... I

 

    
      

   

f..m.m..._u_~...._ mIb.

   
  

I. __ ...I. . :1... \1

. 21......
... 1 1 1... . r
I I .._0.0._.._I 05F: .1. _\ J...
F. J... n. ..... 1......

   

E29232

I... ..\..h..

 

1.....
...-.3 .

      

 

I
_I. . ...... ...... .........,._ _wamnchx .I
5...... -. I...
. .._”: .-

mm. 9 mm.
mo. 8 m _.
m _. 9. 00.

mo. 9- 00.

8:03:33”. .6 3352.45

 

Es: oc>> mucmucoammm ho wmmucmosn. ” 3 2:9".

 

Chapter Five

Discussion

This thesis is based on the integration of the elements of food security /
shortage issues in the particular case study area. Political ecology urges an
integrative approach, rejecting the notion of singular causal factors. Rather, the
culmination of a series of variables leads to a more accurate understanding of
phenomena. The intent of this thesis is not to point out a particular cause to the
shortage history and response in the Zidonini area, but to understand the

complexities that underlie the “package” of variables relating to food security.

This chapter will emphasize the processes underlying food security.
Explanation of the findings presented in the Data Analysis chapter will be
included in two sections. The first section will detail issues related to the design
of the study- what went wrong (and right) and the meaning of the statistical tests
and interpret the results in the context of the study area. The second section will
address larger issues of relevancy to theory and the interaction between the

study findings and theory.

Findings and the Study Area Generally the history of drought in Zidonini is
similar to that found elsewhere in South Africa (Tyson,1986; Vogel, 1994). More
respondents identified 1992/3 as drought years, although 1982/3 was a more

devastating drought. Influx of newer residents coupled with the fact that the

95

96

1992/3 drought is in recent memory, explains the apparent contradiction of the
finding that the 1982/3 drought was far worse than the 1992/3 drought. The two
most recent drought periods are about equal in the terms of loss of food
acquisition mechanisms. However, upon closer examination, the 1992/3 period of
shortage may be envisaged as more severe because of the loss of previous
rights of hunting. Indeed, one of the major coping mechanisms to shortage
suffered a severe constriction in 1989 when the Hluhluwe and Umfolozi Game

Parks were joined and hunting was curtailed.

The Sangoyana and Madlwaleni sites provide an excellent example of
some of the processes undenlvay in the past decade. The number of
respondents in both Sangoyana and Madlwaleni with fewer than ten years of
residency is in the upper quartile at twenty-one and thirty-eight percent,
respectively. Qualitative research confirms the presence of relatively high
numbers of newer residents. However, Sangoyana does not rank in the upper
quartile of vulnerability while Madlwaleni does rank in the second to highest
quartile, thus ruling out length of residency as a cause. The question poses an
interesting challenge. Both groups of settlers originated from the “Corridor”, but
one is more vulnerable. Even more puzzling is the fact that the Sangoyana site
has steep slopes, while Madlwaleni is in a relatively flat river valley. Two factors
can help to clear up the puzzle in this instance. First, Sangoyana is less densely
populated than Madlwaleni. Because Madlwaleni is a fertile area, more people

want to farm and live there, thus reducing the available cropland. Examination of

97

Figures 6 and 7 helps to clarify the situation. The percentage of respondents with
one or fewer fields in Madlwaleni is greater than that of Sangoyana. Figure 7
shows the reverse of this— Sangoyana has a higher percentage of respondents

with three or more fields.

Second, when asked about hunting more respondents reported hunting
activity in Sangoyana (Figure 14). In fact, Sangoyana, being one of the furthest
sites from the Park, has the best hunting opportunities. The combination of more
land, even though at lower quality, and more hunting indicates the better position
of Sangoyana on the vulnerability index. In this context, residency has less to do
with vulnerability than previously assumed in this study. Although Madlwaleni is
slightly less well off in these terms than Sangoyana, it ranks with the four sites
closest to the Park in most of the indicators of vulnerability. The sites closest to
the Park, however, are not handicapped in terms of land, as all except Khalweni
have high percentages of respondents with access to three or more fields. This
leaves the role of hunting as a strong suspect in the understanding of the factors
of vulnerability to shortage. The research here, however, is simply inadequate to
draw a firm conclusion on the role of hunting”. While strongly suggestive, this
research does not satisfactorily delineate the loss of hunting rights and game as a

sole causal factor of shortages in this area.

 

'2 The “Corridor” is the strip of land between the two Parks from which people were forced in the years
leading up to the merging of the two units in 1989.

13 For more information, see 8. Brooks forthcoming dissertation on the role of hunting in areas surrounding
Hluhluwe-Umfolozi Game Reserve.

 

98

The coping mechanisms identified were not extraordinary or unusual.
However, the heavy dependence on basket making is intriguing. Two notions are
particularly interesting here. First, the sale of baskets as a coping mechanism is
overwhelming, yet access to markets is very limited. Many respondents at the
Nduna’s site, "The Tree", Zagomfe, Madlinjini, and Ximbakazi reported that a long
period of time had elapsed since they had last sold handicrafts. The most severe
case was at “The Tree”, where some residents noted not having sold baskets in
over nine months. At the same time, residents of Sovane reported that they were
able to sell far more baskets along the main north-south road. In fact, basket
trade near Hlabisa was far more brisk than anywhere in the study area. Lack of
access to markets is the key factor preventing widespread trade. The women are
not able to leave to sell baskets because of their responsibilities on the farm.
Often, women complained that the market was too far and to unpredictable for

them to leave. However, basket making continued.

The heavy reliance on the basket trade as a coping mechanism seems
contradictory. However, women control the basket trade, a task exclusively for
women. Many women maintained that even though they are not always selling
baskets they are at least ready to seize the opportunity to enhance their incomes.
This makes them less dependent on their husband in times of stress. Men may
always be at the Nkosi’s or have to “go to town” to look for employment or may
spend all of the extra cash in good times on beer. The basket trade is a very
personal form of food security. Women can ensure their family’s well being and

have control over the income in many cases. Selling baskets is the first response

99

to dwindling food stocks. Women are less dependent on others at first, which

extends the range and options of coping strategies.

Not only is the basket trade a form of coping with food shortage, but also
a form of family security. The sale of baskets went not only to the purchase of
food, but helped pay schools fees, purchase raw materials for dressmakers, and
a variety of other uses. In the long run, this is a diversification of assets designed
to widen the number of ways in which insecurity (not just in food stocks) can be
minimized. For example, some women who the profits of the basket trade to
make dresses, giving them an alternative source of income during periods of
stress. This supports the notion that coping mechanisms are intensifications of
practices that are used in non-stress periods as income diversification techniques
(Campbell and Trechter, 1982). Coping mechanisms are not practices only used

during food shortages.

The secOnd finding relating to the basket trade as a coping mechanism is
a little unsettling. When asked for second or third coping mechanisms many
respondents said “nothing” or “don’t know". The first coping mechanism to
shortage is largely the basket trade, however, after that, the large proportion of
respondents reporting “nothing” was disturbing. Pursued further in the qualitative
follow up, many respondents reported that they often only tried to sell baskets.
They reported no other source of income and no other options. Thirty-eight
percent of respondents reported “don’t know” as a second coping mechanism and
ninety-four percent reported it as a third coping mechanism. The other common

response in the qualitative follow up to this phenomenon was that in the last

100

shortage of 1992/93 people actively sold baskets in Hlabisa and other areas. The
stored foods helped to mitigate shortage and the basket trade was successful

enough to prevent resorting to other strategies to obtain food.

Examining the question regarding the sources of food during shortages
provides some clarification. This question was designed precisely to clear up any
inconsistencies in the coping question. During the pilot study, a strong suspicion
that the role of the basket trade might fog the memory of what was done to cope
during shortages was confirmed. The high degree of importance people place on
the basket trade could have obscured the role of other coping strategies. In the
first response to the query “ What were your sources of food during shortages?”,
basket making/selling was noted by thirty-six percent of respondents, a higher
rate than for any other response. However, this question was more successful at
identifying second, third and fourth responses than the coping question. When
the sources of food were tallied for first through fifth responses, buying food in
shops became the most widespread source of food (35%) followed distantly by
“other” (15%) and then basket making/selling (12.5%). While basket making is
important, other factors, contribute substantially to obtaining food in times of

shortage.

The divergence between women and men is illustrated best through the
analysis of the basket trade. Women are more obligated to the farm and '
therefore, when off farm strategies of income diversification are sought men are
the seekers. The traditional role of men in this respect is no different in the

Zidonini area. The men worked on farms in the area or went to the Nkosi or

101

Nduna to seek assistance. The women, however, utilized the local resource base
to make baskets. Many female respondents remarked with a great deal of
amusement that when first diversifying they tried to hunt. After complete failure,
they turned to basket making for sale outside the area. This intensification is a
response particular to women. Obligated to the maintenance of the family unit
requires women to invent new ways of using the available resources. Nowhere
was this more evident than in the basket trade. Men simply do not make or trade
baskets. This supports Campbell and Trechter’s finding that coping mechanisms
are every day income diversification techniques that are intensified during times

of stress.

The findings that distance from the Nduna and the number of shortages
are related, from a political ecology point of view, is not extraordinary. However,
when mapped the relationship is the opposite of hypotheses made in this
document. Those sites nearer the Nduna are, in fact, the ones with more
shortages (Figure 11). While the Chi-square test does not indicate directionality,
mapping the data clearly shows the relationship. This suggests that the Nduna
has less control over food issues than previously assumed. This source of the
erosion of the traditional authority of the Nduna is unclear. However, a cursory
explanation draws upon South Africa’s past. The apartheid government of South
Africa tried to discredit traditional leaders and cause division in the homelands.
People filled out forms for food, but the Nduna never received food promised to
him. In addition, during the last food shortage South Africa was a net food

exporter. Serious questions are raised about the role of the apartheid state during

102

food shortages in the homelands. However, more research is needed on the
structure of local governance and relations with apartheid South Africa before any

clear explanation can be provided.

The distance decay of vulnerability from the Game Park is remarkably
steep, as the effects of the park are lessened drastically within five to six
kilometers. To understand this phenomenon, a closer examination of the Park is
needed“. Game wardens often patrol outside the border of the park through
Khalweni, The Ridge, The Tree and Zagomfe sites. Tales of harassment were not
uncommon. However, at Sangoyana, Ximbakazi, and Madlinjini, the park was
less impressive on daily life. Respondents replied that rarely had game wardens
bothered them at their homes, although isolated incidents had occurred. Fear of
the wardens is extensive and pervasive in the four sites nearest the park‘s.
Hunting is widespread only in the sites furthest from the park. Respondents often
took to complaining about the poor quality of hunting in the area compared with
the past. The influence of the park is great, but mitigated by distance, poor roads,

and a hilly terrain.

Remittances were significant neither as a c0ping mechanism nor as a
source of off farm income. In general, South African homelands have served as
sources of significant labor and subsequent out-migration. This study suggests

that not all areas of the old homelands have played that role in the national

 

14 This data was provided exclusively through qualitative methods. Nearly sixty people in close and distant
residence to the Park were asked to describe its effects on their life.

15 In fact, until the conclusion of the pilot study the author was often suspected of being a spy for the Parks
Board.

103

economy. The area studied is relatively near Durban, yet migration and
remittances are minor aspects of the economy. While not entirely surprising, the
low level of migrations/remittances was unexpected. More research on this topic

is needed to clarify and validate this finding.

Finally, the number of respondents in the “highly vulnerable” category of
the vulnerability index is disturbingly high (40%). When coupled with evidence
that preparations for future famines are lacking, the potential of a future food
shortage becomes a pressing issue. The vulnerability index indicates from the
data the respondents who have experienced the most shortages, have the fewest
fields, have been residents the least amount of time, do not hunt, and have
received no assistance from the KwaZulu government. It is interesting that the
four sites near the Park are among the most vulnerable according to the above

categories (Figure 13).

This study found that pe0ple in this area are somewhat able to mobilize
themselves around drought. Vogel’s (1994) statement that communities appear
to lack any ability to actively mobilize themselves around drought is not entirely
incorrect, however, since the context of this study is so vastly different (See
Review of Literature). More research is needed for a strong conclusion on this

hypothesis.

Broader Questions and Issues The results of this study, presented in
the context of the local area only, would indeed provide some insight into food

security processes. However, when related to larger theoretical issues, the study

104

gains a broader relevance, helping to strengthen or weaken a variety of
theoretical expectations. Arguing that one case study will drastically alter theory
would be shortsighted, however, the study can provide a platform for testing
theory. This case study has been set in just such a context. Several hypotheses
from the literature have been used in the Zidonini case study, with mixed results.
Detailed in this section is the relationship between the findings in this case study
and the larger body of theory on food issues. Questions posited include: Is
political ecology an appropriate approach to use in this situation?, and How does

this case study fit into theory?

An Appropriate Approach? The ecology in political ecology is often
marginalized in social science work. No claim was made here that biophysical
issues such as soil structure, precipitation levels, or wildlife management issues
would be covered. However, the junction of humans and environment and the
dependence of the former on the latter are paramount. This study has the same
flaw as many other political ecology studies. The use of local natural resources is
a key coping mechanism manifested through basket making (the use of local
grasses) and hunting, not to mention the most basic dependence of people on the
soil and weather. Basket making and hunting are local responses to
environmental conditions, mitigated by political, economic, and geographic forces
(i.e. game wardens’ restriction of hunting even outside the park and lack of
markets due to isolation and political indifference). In a broader context, debate
over environmental policies has also affected local conditions. The

conservationist approach to the environment adopted by apartheid South Africa

 

105

was compounded by a paternalistic attitude toward local people. The resentment
many residents feel toward the game park stems from the paternalistic policies
enforced by the apartheid regime. In discussing the Park, local residents
acknowledge the need to protect game, but question why they are the ones
absorbing the brunt of the negative effects. Their land was seized, their cattle lost
foraging land”, and they are constantly pressured by game wardens. What have
the people who enjoy the Park for aesthetic reasons suffered? Why aren’t the
Northern Suburbs being preserved? The Park is part of the environmental
“package” with which the residents have to grapple. Basket making is a response
to pressures put on the population to cope with the loss of resources dedicated to
the park. Hunting is no longer a viable coping mechanism to drought, thus in
some part, inducing basket making. Basket making has become the most
common coping mechanism to drought. Further research could provide
confirmation or challenges to this statement. Research in other areas near the
park that confirm this finding could indicate a Boserupian situation where
pressure on agriculture induces innovation. A lack of evidence on this matter has

precluded such analysis on this front, but evokes provocative questions.

Food security is influenced to a small degree by the moral economy.
Almost eleven percent of the 120 valid responses to the question about sources
of food during shortage were “relatives” or “neighbors”. Ranked fourth in terms of

importance, the moral economy follows buying food in shops, basket making, and

 

16 The Natal Parks Board has recently conceded this point. Cattle are now allowed in the Park in confined
areas to graze.

106

the use of stored foods. When phrased in terms of coping mechanisms, the
degree of reliance on the moral economy is more pronounced. When asked
about the first coping mechanism, relying on relatives and neighbors is second
only to basket making. The divergence between the two statistics is rather
meaningful. As a coping mechanism, the moral economy is indeed important, but
as a source of food less so. People may attempt to get food from neighbors and
relatives, but with a lesser degree of success than through food purchase, basket
making, or using stored foods. This divergence between expected and actual
assistance may indicate a weakening of the moral economy, an explicit goal of

the apartheid regime.

Environmental Security Food security is one part of a larger
environmental security that includes soil, wildlife, plant, and political security.
Food issues are a very large part of the equation and as such need to be
considered not only in their own right but also in combination with the other
components of security. Food security alone cannot provide total security. The
multiplicity of factors involved in the food security process attest to the inter-
relatedness of food security with other forms environmental security. Typical
queries in a food security study include topics relating to the weather and climate,
quality and/or quantity of land, the role of local authorities, etc. The role of food
security in environmental security is large, but for an accurate assessment of the
level of food security other forms of environmental security must be carefully
analyzed, especially in light of the fact that Africans were moved to the most

marginal land during the rule of Apartheid.

107

Scale Issues How does this study relate to the region and nation? The
question of integration of scale was brought up early in the Review of Literature.
This study addresses issues of migration, drought, food security, coping
mechanisms, and preparations at the local scale, but with broad implications for
the region and the nation. If one study area was found that calls into question the
role of migration, are there others? What does this say about models that base

assumptions on a migration economy?

The generally held belief that smallholder agriculture in South Africa is
largely unviable and that rural people are dependent on urban processes is
seriously questioned. Rural people in this setting are able to mobilize themselves
around food shortages. The coping mechanisms, however, do rely heavily on
industrial agriculture to produce surplus as evidence by the dependence on
“buying food in shops”. These dynamics between the local (the study), the
regional (migration), and the national (reliance on agricultural surpluses) provide

fertile ground for research.

The El Nino Southern Oscillation (ENSO) raises many questions for the
southern African region, as ENSO leads to warmer, drier conditions with
substantial implications for agriculture. The degree of success in predicting ENSO
events has increased in the past decade. At the local scale, coping mechanisms
can become proactive rather than reactive. Nationally, governments can prepare

for shortages (as the South African government began to do in the summer of

 

108

1996) while regionally organizations such as SADC can prepare plans for
cooperation and assistance across national boundaries. The questions in this
study that addressed preparation for the next shortage was an exploratory
exercise in anticipation of the development of a mechanism which allows
researchers to ask questions such as “If you knew that there was going to be a
drought in X months from now, would you begin coping now and what would you

do”. This, perhaps, is one of the most fertile areas for future research.

Related to this thesis, research should focus on what farmers in areas
such as Zidonini should or could do to prepare for a forecast food shortage. How
would people adapt and cope? What about false predictions of an ENSO effect?
Will this cause more disruption than preparedness? At the national scale, how will
governments react to such information? Will governments prepare far enough in
advance as South Africa has begun to do? How will nations cooperate regionally
to lessen the effects of a predicted ENSO event? These questions are beyond

the scope of this research, but answers are needed urgently.

Problems and Opportunities The were a number of problems in the
development and implementation of this study. Translation problems became
apparent in the pilot study. Despite being translated and back-translated several
times, question design problems became apparent early in the project after two
difficulties arose. The first regarded assistance from the outside. The idea behind
the question regarding community assistance failed dismally. The prime culprit
was translation. During the pilot study the wording of the question was changed

several times but to no avail. The goal was to gain an understanding of

 

 

109

assistance from the community as a whole. The question was meant to ask “Did

people receive help from members of the community outside of their

homestead/family?” but the term “community” and “family" were imprecise at best.

Many respondents assumed that this meant people outside the immediate family
while others included relatives separated by generations. Others thought
“community” referred to the Hlabisa-Zidonini area, while others did not include

people next door. The difficulty with the question excluded it from analysis.

The second problem in study design pertained to the question on
government assistance. The question was intended to probe for forms of
govemment assistance in the past. However, respondents did not include
obvious government assistance programs in the consideration of their responses.
Individuals would respond that government had not helped them while at the
same time sending children to draw water from a well sunk by the government or
using an electrical appliance. Subsequent interviews uncovered the source of the
inconsistency. Respondents assumed the question was asking about assistance
in the time of shortage, as that was the theme of the questionnaire. Respondents
also noted that the electrification and water projects were not intended for one
individual only, and therefore, the project was for everyone — one person could
not claim that the government had directly helped them. A question regarding
government assistance to the community at any time could have helped to clear
up the inconsistency, if a definition Iterm for “community” could be identified. The

question, however, was exceedingly useful as the focus of government

l

110

assistance was narrowed to times of shortage, and therefore, able to be used to

highlight differences between respondents and the relationship to vulnerability.

Overall, the design of the study satisfactorily accomplished the goals of
this thesis. Other than the two items noted above, the questionnaire schedule
was rather successful in isolating issues and conveying ideas. Improvement of
the interview schedule would entail probing to uncover the more complex issues,
such as the exact role of hunting, and expansion of the core of the schedule to

include more biophysical variables and relationships.

The use of a pilot study, follow up questions, and the qualitative method
is highly recommended. The qualitative portion provided valuable information
both about the quantitative schedule and information not included in the schedule.
Discrepancies noted above were identified and corrected through the use of
qualitative methods after the pilot study. The short quantitative interview did not
always allow the necessary flexibility to pursue topics of interest. The pilot study
elicited participant reaction to the actual structure of the questions. Unclear or
unanswerable questions were re-translated after meanings intelligible to the
participants were identified. The process of clarification in the pilot study, though
initially time consuming, allowed for a much smoother interview process

afterward .

Cha ter Six

Conclusion

Food security results from a dynamic process and understanding it
requires careful examination of the history of drought and shortages, coping
mechanisms, expectations of future shortages, and prevention, both by local
peoples and through the use of Early Warning Systems. Food issues affect a
variety of people, from the rural smallholder struggling to provide enough food for
her or his family to the government decision-maker trying to implement

development programs.

The method for examination of food security issues is drawn from the
vast literature, particularly Sen (1980), Watts (1983), Corbett (1988), Lambert
(1994), Hesselberg (1994), Campbell (1990), Buchanan-Smith (1995), and Vogel
(1995). Central to the analysis was the political ecology framework from which
most of the authors examined food security issues. The contestation between
coping mechanisms as linear and progressive and coping mechanisms as

context specific provided further insight.

This study focused on the level of food security in one area of rural South Africa.
The food shortages and coping mechanisms were identified. The study
confirmed the hypothesis that drought and shortages occurred in 1981-84 and

1992-94. These droughts and shortages set off a series of coping mechanisms

111

112

including basket making, buying food in shops, and the use of stored food. The
causes of food shortages in the study area were presented as a package. No
single cause was cited as the most important, or having the most explanatory

power.

,f The study confirmed other research that has shown coping mechanisms
in rural South Africa to be context specific, while following general patterns of
coping common to the rest of Africa. The study further confirmed that there is
some degree Of mobilizationmargggd droughts and shortages inuz‘idonini. Coping

mechanisms were poorly developed and obscured by the role of the handicraft

trade, but clarified by determining the sources of food.

The study evoked questions about the role of the Hluhluwe-Umfolozi
Game Reserve in the lives of people near its borders. The study found that the
role of hunting is significant when addressing the level of food security. The
inability to hunt affected the population nearest the park border dramatically. The
tests indicate that the loss of hunting led to a higher percentage of shortages and
a higher level of vulnerability in the study sites near the reserve. However, the
lack of hunting was not conclusively determined to be the primary cause of food

insecurity.

The study confirmed the hypothesis that shortages are related to distance
from political authority. Proximity to local authority was found to have a positive
effect on shortages, that is, as proximity to the local authority increases, so does

the number of shortages, which indicates political incapability.

113

The use of Geographic Information Systems was found extremely useful.
Positional accuracy was the largest challenge, while the display and manipulation
of spatial data was the primary benefit. This research may provide some insight

into the development of Early Warning Systems for South Africa.

Further research on the political ecology of the area is recommended.
Several issues are intriguing and need exploration. First, the role of the game
reserve needs further investigation. The role of game wardens, loss of land and
hunting rights, and the spatial relationships between the number of shortages and
proximity to Hluhluwe-Umfolozi are just a few of the topics that need further
investigation. Secondly, the basket trade needs to be examined to determine the
precise impact on coping mechanisms and the local economy and ecology, and
the methods of incorporating the trade into the larger economy. Third, further
research into the role of the political authorities is needed. The study addressed
none of these issues in great detail. Fourth, the geographic area was limited by
poor infrastructure among other factors and could be widened to include not only
more sites but also physically larger sites as well. Fifth, the bio-physical elements
relative to food security need to be investigated. Particularly, exact
measurements of soil quality and field size are needed in place of the surrogate
that was used (number of fields). Drought needs to be more clearly understood.
How long after the onset of drought does coping begin? What are the effects of
recurrent drought on the basket trade? ls drought more significant than
entitlement failure? This study focuses heavily on drought as a causal factor of

food insecurity. Are there other areas of KwaZulu-Natal and South Africa where

114

this is not the case? Finally, more research is needed in the analysis of how
these local processes affect the regional and national scales and how regional
and national processes affect the local, especially in the case of the ENSO

phenomenon.

This study focused on the exploration of themes often ignored in South
Africa. While the local context is meaningful, the findings of this study are broadly
indicative of the food security situation in rural South Africa. The political
economy of the area is not unique to South Africa and this study has provided
illustrations that can be used as a context for research in other areas of rural

South Africa.

 

Appendix A

A

. How has the weather been?

. How many people live in your house] household?

. How long have you lived here?

. How much land do you typically farm for yourself?

. How much land do you farm for someone outside your family?

. What tools do you use to farm?

#0301wa

. Did any h/h members receive wages from off farm labor?

If yes, who & how much

CD

. Did the h/h receive remittances?

If yes, how much?

(0

. Have you experienced a food shortage in the last 10 years?
In the last 5 years?
Please give dates if possible
10. Have you experienced a SEVERE food shortage?
If yes when?
11. What was the first thing you did when food was scarce? Next? and Next?

12. What were the household’s major sources of food during the shortage?

13. What do you think caused the last shortage?

14. Did you receive assistance from others in the community?
15. Did you receive assistance from a political party?

16. Did you go to town to work during the shortage?

17. Are you doing anything to prepare for the next shortage?

1'15

116

18. Do you expect the new government will help you in the next shortage?
18a. Do you want them to?

19. Did the Old government help?

19a. Did the KwaZulu government help?

20. What do you need / how can we (the outside) help?

Bibliography

ACARTSOD.1990. Understanding Africa '5 Food Problems : Social Policy
Perspectives. London: Hans Zell

Adams, L. 1993. " A Rural Voice : Strategies for the Poor" In South African
Outlook 122

Ahmed, l. 1985. Technology and Rural Women: Conceptual and Empirical
Issues. London: George Allen & UNWlN

Alamgir,M. and Arora, P. 1991. Providing Food Security for All. New York: New
York University Press

Amin,S. 1974. Accumulation on a World Scale. New York: Monthly Review Press

Anderson, M. and Woodrow, P. 1991. " Reducing Vulnerability to Drought and
Famine: Developmental Approaches to Relief " ln Disasters 15: 43-54

Babu, S. and Mthindi, G. 1995."Developing Decentralized Capacity for Disaster
Prevention : Lessons from Food Security and Nutrition Monitoring in
Malawi" ln Disasters 19: 127—139

Bawtree,V. 1978. Ideas and Action. Rome: UN FAO

Bay, E. 1982. Women and Work in Africa. Boulder,Co.:Westview Press

Beneria, L. 1982. Women and Development: The Sexual Division of Labor in
Rural Societies. New York : Praeger

Berck,P. and Bigman, D. 1993. Food Security and Food Inventories in

Developing Countries. Oxon: Cab International

117

118

Bernstein, H. 1996. The Agrarian Question In South Africa. London: Frank Cass

Bernstein,H. 1993. "The Maize Industry : Business as Usual" In South African
Outlook 110

Bernstein,H. 1990. "Agricultural Modernization and the Era of Structural
Adjustment" In The Journal of Peasant Studies. 8: 3—35

Blaike, P. 1985. The Political Economy of Soil Erosion in Developing Countries.
London: Longman.

Blaikie, P. and Brookfield, H. (1987) Land Degradation and Society. London:
Metheun

Bryant,C. 1988. Poverty, Policy, and Food Security in Southern Africa Boulder:
Lynne Rienner

Bryceson, D. 1990. Food Insecurity and the Social Division of Labour in

Tanzania, 1919-85. Oxford: Macmillan

Buchanan-Smith, M. and Davies, S. 1995. Famine Eariy Warning and Response:

The Missing Link. London: Intermediate Technology Publications

Buchanan-Smith, M. et al 1991. A Guide to Famine Early Warning and Food
Information Systems in the Sahel and Horn of Africa. Brighton: University of
Sussex

Bundy, C. 1988. The Rise and Fall of South African Peasantry. Cape Town:
David Philip

Bush, J. 1995. "The Role of Food Aid in Drought and Recovery : Oxfam's North
Turkana (Kenya) Drought Relief Programme, 1992—4" In Disasters 19: 247-

259

119

Campbell, D. 1990. "Strategies for Coping with Severe Food Deficits in Rural
Africa: A Review of the Literature" In Food and Foodways 4(2): 143-162

Campbell, D. and Olson, J. 1986. "The Prospect for Desertification in Kajiado
District, Kenya" In Geographical Journal 152: 44—55

Campbell, D. 1990. "Community based Strategies for Coping with Food Security:
A Role in African Famine Early Warning Systems" In GeoJoumaI

Campbell, D. and Trechter, D. 1982 “Strategies for Coping with Food
Consumption Shortage in the Mandara Mountains Region of North
Cameroon"” in Social Science and Medicine 16: 2117-2127

Campbell, D. 1977. "Trends in Man—Land Interaction in the West African Sahel"

In Drought in Africa, Eds. Dalby, D., Harrison—Church,P.J., and Bezzaz, F.

London: International African Institute

Cekan, J. 1992. "Seasonal Coping Strategies in Central Mali: Five Villages During
the 'Soudure' In Disasters 16:66-73

Chambers, R and Moris, J. 1973 MWEA: an Irrigated Rice Settlement in Kenya.
Munich: Weltforum Verlag

Chazan,N. and Shaw, T. 1988. Coping with Africa's Food Crisis. Boulder: Lynne
Rienner

Chigodo, T. 1993. “Drought Takes its Toll” in New Ground. 12, Winter 1993.

Chole, E. 1990 Food Crisis in Africa : Policy and Management Issues. New
Delhi: Vikas Publishing House

Cloud, K. 1977. Sex Roles in Food Production and Food Distribution Systems in

the Sahel. Tucson: University of Arizona

120

Commander,S. 1989. Structural Adjustment and Agriculture. London: Overseas
Development Institute

Corbett,J. 1988. "Famine and Household Coping Strategies" In World
Development 16: 1099-1 1 12

Cunningham, T. 1984. Indigenous Plant Resources :A Buffer against Rural
Poverty. From the Second Carnegie Inquiry in to Poverty and Development
in Southern Africa. CapeTown.

Davies,S., Buchanan-Smith,M.,and Lambert,R. 1991. Eariy Warning in the Sahel
and Horn of Africa: The State of the Art Brighton: University of Sussex

Derman, P. and Poultney, C. 1984. The Politics of Production and Community
Development in Rural South Africa. From the Second Inquiry into Poverty
and Development in South Africa. CapeTown

DeJanvry, A. 1995. ”Farm-Nonfarm Synergies in Africa: Discussion” in American
Journal of Agricultural Economics 76: 1183-1185

DeKlerk,M. 1991. A Harvest of Discontent : The Land Question in South Africa.
CapeTown: IDASA

Delgado, C., et al. 1994. “Promoting lntersectoral Growth Linkages in Rural Africa
through Agricultural Technology and Policy Reform” in American Joumal of
Agricultural Economics 76: 1 166—1 171

Dias, J. 1981. "Famine and Disease in the History of Angola" In Joumal of
African History. 22:349-378

Dirks, R. 1980. "Social Responses during Severe Food Shortages and Famine"

In Current Anthropology 22:21 -44

121

Downing,T.,Gitu,K. and Kaumau,C.1989. Coping with Drought in Kenya.
Boulder: Lynne Rienner

Downing,T. 1986. "Monitoring and Responding to Famine: Lessons from the
1984—85 Food Crisis in Kenya" In Disasters 14:204-225

Dreze, J. and Sen, A. 1991. The Political Economy of Hunger. New York: Oxford
University Press

Fuggle,R., and Rabie,M. 1992. Environmental Management in South Africa.
CapeTown: Juta and Co.

Geldenhuys,G. 1994. Food Security in South Africa. Johannesburg: South African
Institute of International Affairs

Gibbon,P. 1992. "African Agriculture under Structural Adjustment with Special
Reference to Kenya and Ghana." in The Journal of Peasant Studies. 20: 50-
95

Giorgis,D. 1989. Red Tears: War Famine and Revolution in Ethiopia. Trenton:
The Red Sea Press, Inc.

Gittinger. J. et al. 1987. Food Policy: Integrating Supply, Distribution, and
Consumption. Baltimore: John Hopkins University Press.

Glantz, M. 1987. Drought and Hunger in Africa: Denying Famine a Future.
Cambridge: Cambridge University Press

Hay, R. 1988. "Famine Incomes and Employment: Has Botswana Anything to
Teach Africa? In World Development 16: 1113-1125

Hay, R. and Rikuni, M1988. "SADC Food Security Strategies: Evolution and

Role" in World Development. 16: 1013-1024

122

Hinchey, M. 1978. Procedings of the Symposium on Drought in Botswana.
Worchesten Clark University Press

Hjort Af Omas, A. "Pastoral and Environmental Security in East Africa." In
Disasters 14: 115-122

Hollist,W., and Tullis,F. 1987. Pursuing Food Security: Strategies and Obstacles
in Africa, Asia, Latin America, and the Middle East. Boulder: Lynne Rienner

Huss-Ashmore, R. and Katz, S. (1989) African Food Systems in Crisis. New
York: Gordon and Breach Science Publishers.

Hyden, G. 1980. Beyond Ujamaa in Tanzania. Los Angeles: University of
California Press

Hyden, G. 1983. No Shortcuts to Progress. Los Angeles: University of California
Press

lsaacman, A. 1990. "Peasants and Rural Social Protest in Africa." In African
Studies Review 33:1-120

Jowkar, F. et al. 1991. Gender Relations of Pastoral and Agropastoral
Production: A Bibliography with Annotations. Binghamton, NY: Institute for
Development Anthropology

Jones, S. 1994. "Food Security Reserve Policy in Ethiopia: A Case Study of
Experience and Implications." In Disasters 18: 140-151

Kelly, M. 1992 " Entitlements, Coping Mechanisms and Indicators of Access to
Food: Wollo Region, Ethiopia,1987-88." In Disasters 16: 322-338

Kelly,M. and Buchanan-Smith, M. 1994. "Northern Sudan in 1991: Food Crisis

and the International Relief Response." In Disasters 18: 16-34

123

Kennedy,E. and CogilI,B. 1988. "The Commercialization of Agriculture and
Household Level Food Security: The Case of Southwestern Kenya." In
World Development 16: 1075-1081

Kepe, T. And Filson, G. 1993. “ Perceived Levels of Farming Skills and Problems
of Small Scale Commercial Farmers in Ciskei” in Development Southern
Africa. 10(3), p.427-436

Khan,M., et al. 1992. "Composite Indicators for Famine Early Warning Systems."
In Disasters 16: 195-206

Kowet, D. (1978) Land, Labour Migration, and Politics in Southem Africa:
Botswana, Lesotho and Swaziland. Uppsala: The Scandinavian Institute for
African Studies

Krueger, A. et al. 1991. The Political Economy of Agricultural Pricing Policy.
Baltimore: John Hopkins Press

Lambert, R. 1994. "Monitoring Local Food Security and Coping Strategies:
Lessons from Information Collection and Analysis in Mopti, Mali." In
Disasters 18:333-343

Larson, T. 1966. "The Significance of Rainmaking for the Mbukushu." In African
Studies. 25: 23-36

Lenta, G. 1981. Land Shortage and Land Unused: The Paradoxical Pattems of
KwaZulu. Durban: University of Natal Press

Leys,C. 1975. Underdevelopment in Kenya. London: Heineman

Liedholm, C. 1994. “Small Enterprise Employment Growth in Rural Africa” in

American Journal of Agricultural Economics 76: 1177-1182

 

124

Mamdani, M. 1976. Politics and Class Formation in Uganda. New York: Monthly
Review Press

Manetsch, T. 1979. "On Strategies and Programs for Coping with Large Scale
Food Shortages" In Ecology of Food and Nutrition. 8: 209-218

Manona, C. 1984. Migration from Farms to Towns and its Implications for Urban
Adaptation. From Second Carnegie Inquiry into Poverty and Development in
Southern Africa. CapeTown

Marsh, 8., et al. 1994. "Development of a Computer Workstation for Famine
Early Warning and Food Security." In Disasters 18: 117-12

Matlala, P. 1992. “Turning Flowers into Food” in New Ground, 8, Winter 1992

May J. and Nattrass, J. 1986. Migration and Dependency: Sources and Levels of
Income in KwaZulu. Durban: University of Natal Press

May, J. and Peters, A. 1984 a. A Study of Income and Expenditure in Rural
KwaZulu -— Mbongalwane. Ulundi: KwaZulu Finance and Investment
Corporation Limited

May, J. and Peters, A. 1984 b. A Study of Income and Expenditure in Rural
KwaZulu — Madadeni. Ulundi: KwaZulu Finance and Investment
Corporation Limited

May, J. and Peters, A. 1984 c. A Study of Income and Expenditure in Rural
KwaZulu —Nqutu. Ulundi: KwaZulu Finance and Investment Corporation

Limited

125

May, J and Stavrou, S. 1989. The Informal Sector: Socio-Economic Dynamics
and Growth in the Greater Durban Metropolitan Region. Durban : Rural
Urban Studies, University of Natal

McCorkIe, C. 1987. "Foodgrain Disposals as Early Warning Famine Signals: A
Case from Burkina Faso." In Disasters

Melmed, J. and Santiago, C. 1996. “ The household and Employment in Small
Scale Nonfarm Enterprises” in World Development 24 (4) : 749-764

Minnaar, A. 1994. Access to and Affordability of Land in South Africa: The
Challenge of Land Reform in the 1990’s. HSRC: Pretoria

Mkandawire, R. And Matlosa, K. 1993. Food Policy and Agriculture in Southern
Africa. Harare: Sapes Books

McIIer, V.,et al. 1984. Poverty and Quality of Life among Blacks in South Africa.
From Second Carnegie Inquiry into Poverty and Development in Southern
Africa. CapeTown

MoIIer, V. 1996. Perceptions of Development in KwaZulu-Natal : A Subjective
Indicator Study. Durban : Indicator Press

Moock, J. 1986. Understanding Africa’s Rural Households and Farming Systems.
Boulder,CO: Westview Press

Moore, D. 1993 “Contesting Terrain in Zimbabwe’s Eastern Highlands” in
Economic Geography, 69 (4), p.380-402

Mpanza, 2. And Nattrass, J. 1987 Poverty, Migration, and Unemployment in
Dumisa: A Rural Area of KwaZulu. Durban : University of Natal

Muller, J. et al. 1995. GIS and Generalization. London: Taylor and Francis

iii...

 

 

126

Murray, C. And O’Regan, C. 1990. No Place to Rest: Famed Removals and the
Law in South Africa. Cape Town: Oxford University Press

Nesman,E. 1981. Peasant Mobilization and Rural Development. Cambridge,MA:
Schenkman

Neocosmos,M. 1993. The Agrarian Question in Southern Africa and
Accumulation from Below. Uppsala: Nordiska Afrikainstitutet

Olomola,A. and Nwoso,A. 1993. Perspectives on Food Security in Nigeria.
Lagos: Nigerian Rural Sociological Association

Olson, J. 1990. The Impact of Changing SocioEconomic Factors on Migration
Pattems in Rwanda. East Lansing : Michigan State University

Onimode, B. et al. 1990. Alternative Strategies for Africa. London: IFAA

‘Patnaik, U. 1991. "Food Availability and Famine: A Longer View." In The Journal
of Peasant Studies. 19: 1-25

Peters, A. And May, J. 1984. A Study of Income and Expenditure in Rural
KwaZulu — Mapumulo. Ulundi: KwaZulu Finance and Investment
Corporation Limited

Pickles, J. 1995. Ground Truth: The Social Implications of Geographic
Information Systems. New York: The Guilford Press

Pinckney,T. and Valdes,A. 1988. "Short Run Supply Management and Food
Security: Results From Pakistan and Kenya." In Worid Development 16:
1 025-1034

Pottier, J. 1985. Food Systems in Central and Southem Africa. London: School of

Oriental and African Studies

 

127

Prah,K. 1987. Food Security Issues in Southern Africa. Maseru: Institute of
Southern African Studies

Rahmato, D. 1988 "Peasant Survival Strategies in Ethiopia." In Disasters 12:
326-344

Ranger, T. 1983. "The Invention of Tradition in Colonial Africa." In The Invention
of Tradition. Eds. Hobsbawn and Ranger. London: Cambridge University
Press

Reardon, T. 1997. "Using Evidence of Household Income Diversification to Inform
Study of the Rural Nonfarm Labor Market in Africa," World Development, 25
(5), 735-748.

Reardon, T. and J. E. Taylor. 1996. "Agroclimatic Shock, Income Inequality, and
Poverty: Evidence from Burkina Faso”, World Development, May, Vol. 24,
no. 4: 901-914.

Reardon, T., et al. 1994. “ Links between Nonfarm Income and Farm Investment
in African Households: Adding the Capital Market Perspective” in
American Journal of Agricultural Economics 76: 1172-1176

Reardon, T., et al. 1992. "Determinants and Effects of Income Diversification
Amongst Farm Households in Burkina Faso", in Joumal of Development
Studies, 28(January): 264-296.

Reardon, T., et al. 1988. "Coping with Household-level Food Insecurity in Drought

Affected Areas of Burkina Faso." in World Development, Vol. 16, (9): 1065-

1074.

 

128

Rikuni, M. and Wyckoff, J. 1991. Market Reforms, Research Policies and SADCC
Food Security. Harare: University of Zimbabwe

Rikuni, M. et al.1989. Food Security Policies in the SADCC Region. Harare:
University of Zimbabwe.

Rikuni,M. and Bemsten, R. 1987. Southern Africa: Food Security Policy Options.
Harare: University of Zimbabwe

Rikuni,M. and Eicher, C. 1987. Food Security for Southern Africa. Harare:
University of Zimbabwe

Salih, M. 1994. Inducing Food Insecurity: Perspectives on Food Policies in
Eastern and Southern Africa. Stockholm: Gothab

Sen, A. 1981. Poverty and Famines. Oxford: Clarendon Press

South African Outlook, 1994. "How to Beat the Drought." In South African
Outlook 100-102

Spencer, D. 1976. African women in Agricultural Development: A Case Study in
Sierra Leone. East Lansing: Michigan State University Press

Sperling, L. 1987. "Food Acquisition during the African Drought of 1983-1984: A
Study of Kenyan Herders." In Disasters 16:263-272

Staudt, K. 1987. "Uncaptured or Unmotivated? : Women and the Food Crisis in
Africa” in Rural Sociology. 52(1),p.37-55

Stavrou, S. 1987. Black Farm Labour Attitudes and Perceptions on Employment
and Migration patterns in the Natal Midlands. Durban: University of Natal

Taylor, J. 1992. Food Science in the New South Africa .' Fueling the Economy

and Feeding the Nation. Pretoria: The University of Pretoria.

129

Teklu, T. 1994. "The Prevention and Mitigation of Famine: Policy Lessons from
Botswana and Sudan." In Disasters 18:35-47

Thimm,H.et al. 1993 Regional Food Security and Rural Infrastructure. Hamburg:
Lit Verlag

Tyson,P. 1987. Climatic Change and Variability in Southern Africa. Cape Town:
Oxford University Press

van Zyl, J. 1994. "The Last Straw: Drought and the Economy." In South African
Outlook

Vogel, C. 1993 “ Dimensions of Drought : South African Case Studies” in
GeoJoumaI. 39: 51-58

Vogel,C. 1994. "A Prayer for Rain in Boputhatswana." In South African Outlook

Vogel, C. 1994. “South Africa : Case Study”. In M.Glantz (ed.) Drought Follows
the Plow: Cultivating Marginal Areas. Cambridge : Cambridge University
Press

Walker, P. 1989. Famine Early Warning Systems: Victims and Destitution.
London: Earthscan Publications

Watts,M. 1983. Silent Violence: Food, Famine and Peasantry in Northern
Nigeria. Berkeley: University of California Press

Webb,P. ,et al. 1992. Famine in Ethiopia: Policy Implications of Ceping Failure at
National and Household Levels. Washington: International Food Policy
Research Institute

Webb,P. 1993. "Coping with Drought and Food Insecurity in Ethiopia." In

Disasters 17: 34-47

 

 

 

130

Webb, P. and T. Reardon. 1992. "Drought Impact and Household Response in
East and West Africa", Quarterly Journal of lntemational Agriculture, (3)
July-September, pp. 230-246.

Weiner, D. 1997 No More Tears: Struggles for Land in Mpumalanga, South
Africa. Trenton, NJ: Africa World Press

Weiner, D. and Levin, R. 1994. “Land Reform in the New South Africa”. in Earth
Island Journal 23 (64): 391

Weiner, D. et al. 1992 “ Energy for Sustainable Agricultural Development in
Zimbabwe”. in Growth and Change 23 (3): 333

Weiner, D. and Levin, R. 1994 Community Perspectives on Land Reform.
Johannesburg: University of the Witswatersrand

Whiteside, A. 1988. Labour Migration in Southern Africa. Braamfontein: South
African Institute of International Affairs

Wilson, F. 1972. Migrant Labour in South Africa. Johannesburg : Christian
Institute of Southern Africa

Wuyts, M. 1994. "Accumulation, Industrialization and the Peasantry: A
Reinterpretation of the Tanzanian Experience." In The Journal of Peasant
Studies. 21: 159-193

Young, H. and Jaspars,S. 1995. "Nutritional Assessments, Food Security, and

Famine." in Disasters 19: 26-36

 

 

 

 

 

R

llllli

36

HHHHHHHHHHH UN

lilllllIllllllllllliljlj

 

312930 706