ArticlePDF Available

Abstract and Figures

Despite much policy attention to agricultural development in South Africa, efforts since democratisation have failed to raise smallholder engagement in agriculture and to break the trend of persistent rural poverty. This paper presents results from a study of the Massive Food Production Programme (MFPP) in three villages in Eastern Cape Province, South Africa. The MFPP aimed to reduce poverty by raising maize yields. Following a trend of introducing maize varieties developed for large scale farming, the MFPP introduced hybrid and genetically modified maize varieties suited to high-input farming environments. These varieties did not perform well under smallholder conditions. In particular, they were highly sensitive to local storage conditions. Furthermore the restrictions on saving and sharing seed associated with new genetically modified varieties were resented locally. The results show how farming was most important for the poorest households who depended on it for their food security. While these households were in most need of agricultural support, they were also the least supported by the programme. Support with fencing, cattle traction, and locally attuned agricultural advice, which was not prioritised in the MFPP, would have been beneficial across wealth groups. Such support could, in contrast to the MFPP, lead to sustained and positive impact on smallholder livelihoods. In contrast, the strong emphasis on raising yields in the programme did not prove to have the desired effects on poverty.
Content may be subject to copyright.
Land
Use
Policy
46
(2015)
304–313
Contents
lists
available
at
ScienceDirect
Land
Use
Policy
jo
ur
nal
ho
me
pag
e:
www.elsevier.com/locate/landusepol
Does
raising
maize
yields
lead
to
poverty
reduction?
A
case
study
of
the
Massive
Food
Production
Programme
in
South
Africa
Klara
Fischer,
Flora
Hajdu1
Department
of
Urban
and
Rural
Development,
Swedish
University
of
Agricultural
Sciences,
P.O.
Box
7012,
SE-750
07
Uppsala,
Sweden
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
22
May
2014
Received
in
revised
form
16
February
2015
Accepted
23
March
2015
Keywords:
Agriculture
Maize
Development
Poverty
Smallholder
South
Africa
a
b
s
t
r
a
c
t
Despite
much
policy
attention
to
agricultural
development
in
South
Africa,
efforts
since
democratisation
have
failed
to
raise
smallholder
engagement
in
agriculture
and
to
break
the
trend
of
persistent
rural
poverty.
This
paper
presents
results
from
a
study
of
the
Massive
Food
Production
Programme
(MFPP)
in
three
villages
in
Eastern
Cape
Province,
South
Africa.
The
MFPP
aimed
to
reduce
poverty
by
raising
maize
yields.
Following
a
trend
of
introducing
maize
varieties
developed
for
large
scale
farming,
the
MFPP
introduced
hybrid
and
genetically
modified
maize
varieties
suited
to
high-input
farming
environ-
ments.
These
varieties
did
not
perform
well
under
smallholder
conditions.
In
particular,
they
were
highly
sensitive
to
local
storage
conditions.
Furthermore
the
restrictions
on
saving
and
sharing
seed
associated
with
new
genetically
modified
varieties
were
resented
locally.
The
results
show
how
farming
was
most
important
for
the
poorest
households
who
depended
on
it
for
their
food
security.
While
these
house-
holds
were
in
most
need
of
agricultural
support,
they
were
also
the
least
supported
by
the
programme.
Support
with
fencing,
cattle
traction,
and
locally
attuned
agricultural
advice,
which
was
not
prioritised
in
the
MFPP,
would
have
been
beneficial
across
wealth
groups.
Such
support
could,
in
contrast
to
the
MFPP,
lead
to
sustained
and
positive
impact
on
smallholder
livelihoods.
In
contrast,
the
strong
emphasis
on
raising
yields
in
the
programme
did
not
prove
to
have
the
desired
effects
on
poverty.
©
2015
Elsevier
Ltd.
All
rights
reserved.
Introduction
Improvement
of
smallholder
farming
is
a
high
priority
in
South
Africa’s
fight
against
rural
poverty
(Aliber
and
Hall,
2012;
Kepe
and
Tessaro,
2014).
Substantial
efforts
and
large
sums
of
money
have
been
devoted
to
poverty
reduction
through
agriculture,
but
research
shows
that
these
efforts
have
failed
to
raise
smallholder
engagement
in
agriculture
and
to
break
the
trend
of
persistent
rural
poverty
(Aliber
and
Hall,
2012;
O’Laughlin
et
al.,
2013).
This
paper
seeks
some
explanations
for
this
by
analysing
the
reasons
behind
the
failure
of
the
Massive
Food
Production
Programme
(MFPP)
to
reduce
rural
poverty
through
raising
maize
yields
in
three
villages
in
Eastern
Cape
Province,
South
Africa.
The
MFPP,
which
was
run
by
the
Eastern
Cape
Department
of
Agriculture
(ECDA2)
in
close
cooperation
with
the
agro-industry,
was
organised
into
424
different
projects
and
planted
over
Corresponding
author.
Tel.:
+46
671771.
E-mail
addresses:
klara.fischer@slu.se
(K.
Fischer),
flora.hajdu@slu.se
(F.
Hajdu).
1Tel.:
+46
672162.
2Since
6
May
2009,
the
former
Eastern
Cape
Department
of
Agriculture
(ECDA)
has
changed
its
name
to
Department
of
Agriculture
and
Rural
Development.
15,000
ha
in
the
province
between
2003
and
2009
(Mtero,
2012).
It
aimed
to
raise
maize
yields
through
subsidisation
of
hybrid
and
genetically
modified
(GM)
seed,
fertiliser
and
mechanisation.
The
idea
behind
MFPP
was
that
stimulating
agricultural
growth
through
raising
smallholders’
maize
yields
would
lead
to
poverty
reduction.
This
idea
is
strongly
rooted
in
agricultural
development
policy
in
Africa
today
(Collier
and
Dercon,
2013).
In
this
paper,
we
analyse
how
the
focus
of
the
MFPP
on
max-
imising
maize
yields
affected
the
possibility
of
the
programme
to
make
agriculture
more
important
for
rural
livelihoods
and
to
meet
the
needs
of
the
poorest
in
particular.
Maize
is
the
staple
crop
for
many
smallholders
in
Africa
and
pro-
motion
of
new
maize
hybrids
and
GM
varieties
is
viewed
by
many
as
the
solution
to
low
yields
in
smallholder
farming
(Brooks
et
al.,
2009).
South
Africa
in
particular
has
promoted
the
introduction
of
GM
maize
(James,
2013).
An
analysis
of
smallholder
experiences
in
South
Africa
might
therefore
provide
insights
that
can
be
used
in
other
smallholder
contexts
across
Africa
where
GM
crops
are
considered.
The
present
study,
conducted
in
three
villages
in
Eastern
Cape
Province,
was
based
on
two
important
facts
about
South
African
smallholder
farming.
Firstly,
farming
is
commonly
just
one
of
many
activities
that
smallholders
employ
to
make
a
living.
There
is
a
http://dx.doi.org/10.1016/j.landusepol.2015.03.015
0264-8377/©
2015
Elsevier
Ltd.
All
rights
reserved.
K.
Fischer,
F.
Hajdu
/
Land
Use
Policy
46
(2015)
304–313
305
strong
historical
interdependency
between
smallholder
farming
and
urban
wage
work
in
South
Africa,
making
it
particularly
rel-
evant
to
draw
on
the
wider
livelihoods
situation
to
understand
farming
in
a
South
African
context
(Aliber
and
Hart,
2009;
Carr
and
McCusker,
2009;
Hebinck
and
Lent,
2007;
Slater,
2002).
Secondly,
research
in
South
Africa
shows
high
levels
of
social
dif-
ferentiation
even
within
areas
of
widespread
poverty
(Carter
and
May,
1999;
Neves
and
du
Toit,
2013).
The
heterogeneity
of
rural
poverty
results
in
farming
playing
different
roles
within
rural
com-
munities,
which
means
that
the
MFPP
is
unlikely
to
have
been
of
equal
benefit
to
all.
Therefore
the
analysis
specifically
examines
how
the
capacity
to
engage
in
agriculture
and
to
benefit
from
the
MFPP
differs
between
households
of
different
wealth
status,
defined
here
through
a
local
wealth
ranking
exercise.
It
is
acknowledged
here
that
an
important
reason
for
the
failure
of
the
MFPP
to
stimulate
agricultural
growth
through
raising
yields
was
that
the
programme
did
not
support
smallholders
in
reaching
markets.
Access
to
markets
is
a
key
factor
determining
whether
smallholders
in
can
in
fact
benefit
from
raised
yields
(Andersson
Djurfeldt,
2013;
Poulton
et
al.,
2010).
Due
to
the
particularly
comprehensive
suppression
of
commercial
smallholder
farming
in
South
Africa
historically
(Bundy,
1988),
supporting
smallholders
to
reach
and
compete
on
agricultural
markets
might
be
even
more
important
here
than
elsewhere
in
the
region.
The
importance
of
facilitating
market
access
is
also
acknowledged
in
the
new
agri-
cultural
policy
for
the
province
(Jacobson,
2013).
We
argue
in
this
paper,
however,
that
even
if
smallholders
were
supported
in
reach-
ing
markets,
the
narrow
focus
on
raising
yields
is
problematic
in
itself
as
it
is
not
attuned
to
the
local
role
of
farming
in
the
wider
livelihoods
situation
of
Eastern
Cape
smallholders.
As
the
focus
on
maximising
yields
through
hybrid
and
GM
maize
remains
also
in
more
recent
interventions
in
the
region
(Iversen
et
al.,
2014),
a
crit-
ical
analysis
of
the
role
of
maize
yields
in
poverty
reduction
is
still
needed,
and
is
the
focus
of
the
present
paper.
The
Massive
Food
Production
Programme
The
MFPP
was
introduced
in
the
study
villages
in
2003.
Like
in
the
pre-democratic
mechanisation
schemes,
land
was
ploughed
collectively
to
benefit
from
economies
of
scale.
Tractors
ploughed
across
all
fields
in
the
field
areas,
including
the
fields
belonging
to
smallholders
who
did
not
participate
in
the
programme.
At
the
same
time,
fields
located
outside
main
field
areas,
in
mechani-
cally
inaccessible
terrain,
were
excluded.
Furthermore,
to
avoid
the
heavy
weed
problems
resulting
from
fallowed
fields,
only
small-
holders
currently
actively
engaged
in
farming
could
participate.
Subsidies
for
input
purchases
were
conditional
upon
compli-
ance
with
programme
terms.
In
the
first
year
all
inputs
were
free
to
participants,
while
the
subsidies
were
reduced
stepwise
in
follow-
ing
years.
It
was
anticipated
that
when
the
programme
finished
after
5
years,
participants
would
be
able
to
pay
the
full
cost
for
inputs,
as
their
income
would
rise
with
increased
yields.
Programme
terms
included
applying
specified
minimum
levels
of
fertiliser
and
selecting
high-yielding
(in
essence
hybrid
or
GM)
maize
varieties
through
dialogue
with
the
seed
industry.
Other-
wise,
no
government
advice
on
seed
was
provided,
as
MFPP
policy
was
explicitly
to
leave
choice
of
varieties
to
the
market.
Seed
com-
panies
were
encouraged
to
promote
their
products
at
fairs
to
which
participants
were
invited.
Other
extension
services
and
agricultural
advice
were
not
prioritised
in
the
programme.
During
the
first
3
years,
the
study
villages
planted
genetically
modified
Bt
maize
(CRN
4549B
and
DKC
7815B)
from
the
agro-
chemical
company
Monsanto.
Due
to
misunderstandings
during
the
seed
order,
a
conventional
hybrid
(SNK2551)
was
planted
dur-
ing
the
fourth
season.
The
village
chief
explained
the
choice
of
Bt
maize
by
the
fact
that
Monsanto
was
the
only
company
that
had
demonstrated
its
seed
in
a
local
trial.
Villagers
objected
to
the
MFPP
excluding
households
with
unplanted
or
mechanically
inaccessible
fields,
so
the
chief
dis-
tributed
seed
and
fertiliser
from
the
MFPP
to
households
formally
excluded
from
the
programme.
People
also
continued
to
recycle
seed
from
the
programme,
although
advised
against
this.
These
adaptations
of
the
programme
to
local
circumstances
were
inter-
preted
by
programme
leadership
as
disobedience
and
are
also
one
reason
why
yields
and
incomes
did
not
rise
as
much
as
expected.
Moreover,
there
was
clear
disagreement
between
MFPP
managers
and
smallholder
participants
in
many
villages
about
the
contribu-
tion
to
input
costs
that
smallholders
were
willing
and
able
to
make
and
the
two
parties
appeared
to
have
fundamentally
different
views
about
the
role
of
agriculture
for
livelihoods.
The
study
villages
and
many
other
villages
in
the
region
dropped
out
of
the
MFPP
after
four
of
the
five
intended
years
due
to
disagree-
ment
with
the
management
about
the
terms
and
conditions
of
the
programme.
Maize
and
South
African
smallholders
The
comprehensive
political
suppression
of
rural
livelihoods
in
pre-democratic
South
Africa
led
to
a
reorientation
of
agriculture
towards
subsistence,
with
significant
reliance
on
migrant
labour
(Bundy,
1988;
Hendricks,
1990).
Commercial
smallholder
agricul-
ture
was
made
virtually
extinct
in
the
region
and
the
shortage
of
labour
in
agriculture
that
resulted
from
the
enforced
labour
migra-
tion
led
many
households
to
focus
on
gardening
over
farming
in
the
more
distant
fields
(Aliber
and
Hart,
2009;
Andrew
and
Fox,
2004;
Fraser
et
al.,
2003).
The
switch
to
maize
from
sorghum
as
a
staple
crop
in
the
region
has
also
been
described
as
a
reaction
to
labour
constraints,
as
maize
is
less
labour
intensive
(Beinart,
1982).
To
address
the
severe
poverty
resulting
from
the
political
sup-
pression
of
the
majority
population,
smallholders
in
the
region
have
been
subjected
to
repeated
agricultural
development
interventions
(De
Wet,
1990).
It
has
been
described
how
this
pre-democracy
focus
on
raising
yields
in
effect
was
an
efficient
strategy
for
avoiding
dealing
with
the
root
of
the
problem,
that
smallholders
were,
due
to
political
suppression,
short
of
land,
labour
and
market
access
(Jacobson,
2013).
Despite
repeated
introduction,
hybrid
maize
has
so
far
been
adopted
to
a
very
limited
extent
by
South
African
smallholders.
A
key
reason
for
this
is
that
the
majority
of
hybrids,
including
all
of
the
more
recently
introduced
GM
maize
varieties,3sold
in
South
Africa
are
adapted
to
the
agricultural
practices
and
environments
of
large-scale,
capital-intensive
and
commercially-orientated
farm-
ers.
Hybrids
are
generally
high-yielding
under
optimal
agricultural
conditions
and
high
fertilisation.
Furthermore
hybrid
seed
need
to
be
purchased
new
every
year
to
retain
yields.
Estimates
by
the
South
African
seed
industry
in
2003
suggested
that
90%
of
smallholders
planted
open-pollinated
varieties
(OPVs)
of
maize
or
recycled
seed
from
OPVs
or
hybrids,
and
that
only
10%
purchased
hybrid
seed
in
any
given
year
(Gouse
et
al.,
2005).
While
(non-
hybrid)
OPV
maize,
the
type
of
maize
mainly
used
by
South
African
smallholders,
commonly
produces
lower
yields
than
hybrids,
they
are
generally
better
adapted
to
less
optimal
agricultural
condi-
tions.
OPV
seed
can
also
be
recycled
without
major
effects
on
yields
(Chimonyo
et
al.,
2014).
The
genetically
modified
Bt
maize
was
introduced
in
South
Africa
in
1998
and
was
first
introduced
to
South
African
smallhold-
ers
through
field
trials
in
smallholder
communities
conducted
by
3All
genetic
modifications
in
maize
have
so
far
been
incorporated
into
hybrid
varieties.
306
K.
Fischer,
F.
Hajdu
/
Land
Use
Policy
46
(2015)
304–313
the
multinational
seed
company
Monsanto
in
2001
(Gouse
et
al.,
2005).
Such
trials
in
the
study
villages
prompted
the
chief
to
order
Bt
maize
to
be
planted
during
the
MFPP.
Bt
maize
is
toxic
to
the
African
maize
stemborer
(Busseola
fusca)
and
the
Chilo
stemborer
(Chilo
partellus),
insects
which
if
uncontrolled
may
cause
signifi-
cant
damage
to
maize
crops
in
South
Africa
(Van
Wyk
et
al.,
2008).
There
are
few
previous
studies
on
South
African
smallholders’
experiences
of
Bt
maize
(Assefa
and
Van
Den
Berg,
2009;
Ezezika
et
al.,
2013;
Jacobson
and
Myhr,
2013).
Some
studies
show
that
on
average,
smallholders
have
obtained
increased
yields
and
eco-
nomic
gains
by
planting
Bt
maize
compared
with
conventional
hybrids
(Gouse,
2012).
As
pointed
out
by
others
(Jacobson,
2013),
reports
of
average
yield
gains
however
say
little
about
the
ben-
efit
to
smallholders
when
some
years
result
in
economic
loss,
as
smallholders
cannot
spread
gains
and
losses
across
seasons
like
capital
intensive
farmers.
The
fact
that
currently
available
Bt
maize
varieties
all
are
bred
into
hybrids
adapted
to
large
scale
capital
intensive
farming
has
also
been
argued
to
limit
their
usefulness
to
smallholders
(Fischer
et
al.,
2015).
Furthermore,
research
from
the
region
has
shown
that
Bt
maize
needs
to
be
accompanied
with
comprehensive
advisory
support
to
be
made
useful
to
smallhold-
ers
(Assefa
and
Van
Den
Berg,
2009;
Jacobson
and
Myhr,
2013).
In
the
light
of
these
aspects,
and
the
fact
that
farming
is
only
one
part
of
smallholders’
livelihoods,
studies
of
effects
beyond
changes
in
yield
have
been
called
for
(Dowd-Uribe,
2013;
Stone,
2011).
Description
of
the
case
study
region
and
methods
for
data
collection
and
analysis
The
present
study
was
based
on
data
obtained
during
repeated
field
work
in
three
villages
between
2006
and
2012,
which
is
described
in
detail
in
Jacobson
(2013).
The
study
villages
are
located
in
Ngquza
Hill
Local
Municipality,
one
of
the
two
poorest
munic-
ipalities
in
OR
Tambo
district.
The
district
comprises
80%
of
the
former
Transkei
homeland
and
is
itself
the
poorest
district
in
the
Province.
Like
the
majority
of
rural
producers
in
South
Africa’s
former
homelands,
people
farm
small
plots
of
land,
commonly
around
1-ha
fields
and
small
home
gardens.
Maize
farming
is
gen-
erally
combined
with
other
livelihood
activities
and
is
often
not
the
main
source
of
income.
Many
households
in
the
district,
including
those
in
this
study,
still
lack
access
to
basic
services
such
as
electricity,
water
and
sanitation.
As
a
result
of
long-term
marginalisation,
OR
Tambo
district
still
only
has
a
peripheral
role
in
the
national
economy
today.
Limited
access
to
urban
centres
because
of
poor
infrastruc-
ture
characterises
the
province
at
large.
In
the
study
villages,
the
nearest
local
town
is
a
40-min
drive
away
on
a
dirt
road
of
variable
quality.
Transport
(mainly
local
minibus-taxi)
to
town
is
unreli-
able
and
it
is
considered
impossible
to
commute
daily
from
the
villages
for
work.
Apart
from
a
handful
of
job
opportunities
at
the
local
healthcare
centre,
schools
(largely
staffed
by
people
from
out-
side
the
villages),
local
informal
taxi
firms
or
spaza
shops
(kiosks),
especially
young
people
migrate
to
more
distant
urban
areas
to
seek
employment.
Agriculture
and
welfare
payments
are
domi-
nant
livelihood
strategies
in
the
study
villages.
Welfare
payments
(including
pensions,
childcare
grants,
foster
care
grants
and
disabil-
ity
grants)
are
accessed
by
83%
of
households,
compared
with
75%
of
households
in
the
municipality
as
a
whole
in
2011
(OR
Tambo
District
Municipality,
2011).
Despite
farming
being
a
key
livelihood
activity
in
many
rural
households,
many
fields
in
the
region
lie
unplanted.
The
particularly
comprehensive
and
long-term
under-
mining
of
smallholder
farming
that
occurred
in
pre-democratic
South
Africa
has
had
major
effects
on
the
role
of
agriculture
today
(Bryceson,
2004;
O’Laughlin
et
al.,
2013).
To
grasp
the
role
of
farming
and
maize
in
the
wider
livelihoods
context
this
paper
draws
on
data
collected
through
participant
observation,
individual
and
group
interviews,
and
household
ques-
tionnaires.
The
village
where
the
chief
resided
was
subjected
to
more
in-depth
field
work
because
all
information
meetings
on
the
MFPP
took
place
in
this
village
and
it
had
the
highest
par-
ticipation
rate
in
the
programme
(46%
of
households
took
part
in
the
MFPP,
compared
with
22%
and
36%
in
the
other
two
villages).
The
household
served
as
the
smallest
unit
of
data
collection
and
analysis.
Diverse
activities
and
assets
are
commonly
brought
together
at
household
level
to
create
livelihood
outcomes
for
its
members,
and
farming
is
often
organised
at
household
level.
Gen-
eral
descriptive
data
were
gathered
in
surveys
of
all
265
households
in
the
three
villages
between
March
and
May
2008.
All
(105)
house-
holds
in
the
village
where
the
chief
resided
completed
an
extended
survey
in
which
respondents
were
asked
to
rank
the
most
impor-
tant
livelihood
activities
and
household
costs.
Local
understanding
and
use
of
maize
were
recorded
through
open-ended
questions.
Respondents
were
also
encouraged
to
compare
their
experiences
of
the
performance
of
maize
from
the
MFPP
with
locally
used
varieties.
The
responses
were
subjected
to
multiple
response
analysis
in
SPSS
(Statistical
Package
for
the
Social
Sciences,
a
software
system
that
allows
statistical
analysis
of
numerical
data
sets).
To
cross-check
information
about
local
use
and
understanding
of
maize
obtained
in
surveys
and
interviews,
a
focus
group
discussion
was
performed
with
eight
smallholders
selected
for
their
long
experience
of
maize
farming.
To
grasp
wealth-related
differences
between
households
and
capture
local
perceptions
of
relative
poverty
and
wealth,
a
wealth
ranking
activity
(modified
from
Pretty
et
al.,
1995)
was
performed
in
each
of
the
three
villages.
Locally
relevant
classifications
of
poverty
were
discussed
and
participants
decided
on
four
wealth
groups:
rich,
middle,
poor
or
very
poor.
Local
classifications
showed
that
a
household’s
wealth
status
was
judged
by
a
qualitative
assessment
of
its
combined
assets,
household
history
and
social
relations.
The
associated
attributes
and
meaning
of
each
of
the
wealth
groups
were
described
in
similar
ways
in
the
three
wealth
ranking
activities.
Participant
observation
also
showed
that
house-
holds
belonging
to
a
particular
wealth
group
in
any
of
the
three
villages
experienced
similar
livelihood
situations.
Therefore
it
was
deemed
appropriate
to
pool
the
wealth
ranking
data
for
all
three
villages.
In
order
to
identify
effects
of
poverty
that
were
not
highlighted
in
group
discussions
and
to
deepen
understanding
of
the
percep-
tions
of
the
MFPP
in
households
with
different
experiences
of
poverty,
the
heads
of
11
households
that
had
participated
in
the
MFPP
were
approached
for
repeated
interviews
between
2008
and
2012.
The
households
were
selected
using
multivariate
analysis
of
variables
on
the
age
and
sex
of
the
de
facto
head
of
household,
the
varieties
of
maize
planted
in
garden
and
field,
and
the
wealth
group
of
the
household.
The
aim
was
to
capture
the
variety
of
smallhold-
ers
from
different
poverty
groups
with
first-hand
experience
of
the
project.
Of
the
11
households,
four
were
classed
as
very
poor,
two
poor,
three
middle
and
two
rich.
Seven
where
female
headed
and
four
were
male
headed.
The
Chambers
and
Conway
(1992:
6)
definition
of
livelihoods
as
the
combination
of
capabilities,
assets
(stores,
resources,
claims
and
access)
and
activities
required
for
a
means
of
living
was
used
to
provide
some
structure
to
the
data,
with
farming
analysed
as
one
of
many
livelihood
activities.
The
analysis
was
guided
by
research
questions
(RQ,
1–4):
1.
What
is
the
role
of
farming
in
the
wider
livelihoods
situation?
K.
Fischer,
F.
Hajdu
/
Land
Use
Policy
46
(2015)
304–313
307
2.
What
are
the
effects
of
poverty
on
how
smallholders
organise
their
livelihoods
in
general
and
how
they
practise
farming
in
particular?
3.
How
do
people
use
and
understand
maize?
4.
What
were
the
relationship
between
smallholders’
practices,
the
MFPP
and
Bt
maize?
Each
household
was
assigned
a
numerical
code
during
field
work,
which
was
used
throughout
the
data
set.
This
permitted
coupled
analysis
of
data
from
household
surveys
and
interview
data
for
specific
households.
Entering
wealth
ranking
categories
for
each
household
into
SPSS
allowed
quantitative
household
data
(e.g.
number
of
cattle,
household
members,
monetary
income)
to
be
analysed
in
relation
to
wealth
category.
Results
Capabilities
and
access
to
assets
in
the
four
poverty
groups
In
the
wealth
ranking
exercise,
40%
of
households
were
classed
as
‘very
poor’,
while
the
remaining
households
were
distributed
evenly,
with
20%
in
each
of
the
groups
‘poor’,
‘middle’
and
‘rich’.
It
should
be
noted
that
these
represent
locally
defined
categories.
The
villages,
located
as
they
are
in
the
poorest
district
of
the
Eastern
Cape,
suffer
from
low
integration
into
the
national
economy
and
low
educational
levels.
While
poverty
plays
out
differently
in
every
household,
some
common
trends
emerged
from
survey
data
and
interviews.
Key
assets
for
all
households
in
the
villages
were
money
(from
welfare
payments
or
employment),
cattle
and
family
labour.
As
expected,
overall
monetary
income
declined
with
rising
poverty
(Table
1).
Conversely,
welfare
payments
increased
from
rich
to
poor,
although
the
very
poor
received
the
least
wel-
fare
payments,
highlighting
their
particularly
vulnerable
situation.
Interviews
revealed
a
widespread
perception
of
welfare
payments
as
the
most
important
monetary
income,
in
particular
because
of
their
reliability
compared
with
other
types
of
incomes.
Con-
firming
what
has
been
found
by
others
(Bryceson,
2004;
Hull,
2014;
Ngonini,
2007),
young
people
in
the
villages
continue
to
strive
for
urban
employment.
However
migrant-working
family
members
are
becoming
less
important
as
a
source
of
regular
income,
both
due
to
high
urban
unemployment
and
to
weakening
ties
between
urban
and
rural
since
the
end
of
apartheid.
Having
cattle
was
strongly
correlated
with
wealth
grouping,
even
more
so
than
monetary
incomes.
This
highlights
both
his-
torical
connections
between
migrant
labour
and
cattle,
the
cultural
significance
of
cattle
as
well
as
the
important
agricultural
invest-
ment
that
cattle
represent,
as
detailed
in
much
previous
research
(Beinart,
1992;
McAllister,
2001;
Shackleton
et
al.,
2001).
As
will
be
discussed
further
below,
having
cattle
was
central
to
agricultural
engagement.
There
was
no
statistically
significant
difference
in
number
of
people
or
number
of
adults
between
households
from
different
wealth
groups
(Table
1).
It
is
likely
however
that
a
survey
recording
more
detailed
information
on
household
members’
age
and
health
status
would
have
been
more
informative
regarding
actual
house-
hold
capabilities.
Interviews
and
participant
observations
revealed
that
poorer
households
suffered
more
from
ill-health,
and
thus
had
less
capability
to
farm
their
fields
than
richer
households.
There
was
also
a
clear
difference
between
wealthier
and
poorer
households
in
how
labour
time
was
spent.
Households
that
could
afford
it
often
paid
others
to
perform
tedious
and
time-consuming
work,
such
as
fetching
water
and
firewood,
which
released
their
own
labour
time
for
other
activities.
More
resource-constrained
households
did
all
this
work
themselves
and,
in
addition,
often
performed
such
work
in
other
households
to
meet
immediate
needs
for
money
or
food.
This
significantly
reduced
time
left
to
engage
in
agriculture
in
these
households,
an
important
explanation
why
they
more
seldom
planted
their
fields
(Table
2).
Effects
of
poverty
on
agricultural
engagement
In
2008,
77%
of
households
in
the
study
villages
had
a
field,
com-
monly
1
ha
in
size,
and
in
total
69%
of
fields
were
planted
(thus
54%
of
all
households
planted
a
field).
All
households
had
a
home
gar-
den
and
83%
of
gardens
were
planted.
Access
to
fields
was
evenly
distributed
between
the
wealth
groups
(Table
2),
but
poverty
had
significant
negative
effects
on
ability
to
plant
the
field.
The
widely
held
perception
of
farming
in
the
villages
was
that
it
is
mainly
a
subsistence
activity
and
not
something
that
will
get
you
out
of
poverty.
This
reflects
the
long
term
historical
undermining
of
farming
as
a
commercial
enterprise.
This
perception
of
farming
was
also
reflected
in
that
many
of
those
who
had
the
opportunity
to
do
so
(young
family
members
with
limited
perceived
obliga-
tions
to
engage
in
farming,
and
wealthier
households
with
other
options)
directed
their
labour
towards
earning
an
income
through
other
means.
Indeed,
interviews
confirm
that
respondents
classed
as
rich
who
did
not
plant
their
field
exclusively
did
so
by
choice,
because
they
had
other
income
alternatives
that
they
considered
more
profitable
and
preferred
in
favour
of
farming,
such
as
running
a
local
taxi
business
or
owning
a
local
shop.
In
sharp
contrast
to
this,
respondents
from
poor
and
(in
particular)
very
poor
households
never
had
a
fallow
field
by
choice,
but
repeatedly
emphasised
the
importance
of
farming
for
their
food
security.
However,
our
results
showed
how
it
was
these
same
households
that
faced
most
difficul-
ties
in
actually
farming.
In
particular,
constraints
related
to,
labour,
cattle,
fencing
and
maize
seed
were
important
reasons
cited
for
unplanted
fields
in
these
households.
While
it
is
difficult
to
rank
the
relative
importance
of
access
to,
labour,
cattle,
fencing
and
maize
seed
for
ability
to
engage
in
farming,
as
they
are
qualitatively
different
and
also
have
effects
on
each
other,
it
might
be
said
that
labour
was
the
most
limiting
factor
to
planting
the
field
in
poor
and
very
poor
households.
As
discussed
above,
particularly
poverty-stricken
households
often
had
to
sell
their
own
labour
to
others
to
meet
immediate
needs.
This
reduced
the
time
possible
to
invest
in
agriculture.
There
was
no
significant
difference
between
wealth
groups
regarding
planting
the
garden,
which
was
located
within
the
homestead
and
thus
required
less
time
and
labour
input
than
farming
the
field.
An
obvious
solution
for
a
poor
labour
constrained
household
in
need
of
money
might
seem
to
be
to
lend
the
field
to
someone
able
to
plant
in
exchange
for
money
or
harvest.
The
MFPP
management
also
said
in
interviews
that
the
long
term
plan
of
the
programme
was
that
this
indeed
would
happen,
so
that
more
successful
farmers
could
benefit
from
economies
of
scale.
However,
as
described
ear-
lier
in
this
section,
an
unplanted
field
is
seldom
the
same
as
a
lack
of
will
to
plant
the
field.
Because
of
the
important
role
of
agricul-
ture
for
food
security
in
the
most
resource
constrained
households,
these
households
were
clearly
reluctant
to
let
go
of
their
field.
Fur-
thermore,
there
was
a
widespread
perception
that
if
you
lend
your
field
to
someone
else,
you
might
not
get
it
back.
Fields
are
inher-
ited
within
households.
However,
agricultural
land
is
held
jointly
by
the
village
and
a
household
can
only
claim
a
field
when
it
is
in
use.
Thus
while
common
practice
in
the
villages
was
to
let
households
keep
currently
unused
fields,
the
perceived
risk
of
losing
the
field
if
someone
else
used
it
also
prevented
households
from
lending
their
field
to
others.
Cattle
were
shown
to
be
vitally
important
for
farming.
House-
holds
with
a
planted
field
had
significantly
more
cattle
(on
average
3.4
head)
than
those
with
unplanted
fields
(0.7
head;
t-test
for
independent
samples,
equal
variances
not
assumed,
p
=
0.000).
In
fact
cattle
had
an
even
stronger
effect
on
agriculture
than
the
308
K.
Fischer,
F.
Hajdu
/
Land
Use
Policy
46
(2015)
304–313
Table
1
Data
on
incomes
and
assets
in
different
wealth
groups
(265
households).
Wealth
group
Type
of
income/asset
Welfare
payments
Rand/month
Mean
(range)
Regular
income
excl.
welfare
paymentsa
Rand/month
Mean
(range)
Total
regular
income
Rand/month
Mean
(range)
Household
members
Mean
(range)
Adult
house-hold
members
(over
18
years)
Mean
(range)
Cattle
Mean
(range)
Rich
850
(0–3450)
760
(0–12,950)
1651
(0–
129,500)
5
(1–15)
2
(1–6)
4
(0–
24)
Middle
948
(0–3465)
496
(0–2879)
1455
(0–4465)
6
(1–16)
3
(1–6)
3
(0–
24)
Poor
988
(0–3030)
196
(0–1800)
1183
(0–3030)
6
(1–12)
3
(0–8)
2
(0–
12)
Very
poor 655
(0–3900)
100
(0–1200)
1136
(0–3900)
5
(1–12)
2
(1–7)
1
(0–10)
One-way
ANOVA
testing
of
differences
in
means
between
wealth
groups
(significant
differences
(p
<
0.05)
are
marked
in
bold)
0.013
0.000
0.000
0.363
0.290
0.000
aThese
include
regular
income
from
remittances,
paid
work
and
business.
Note
however
that
most
respondents
who
reported
receiving
remittances
could
not
count
on
these
on
a
regular
or
predictable
basis.
Such
irregular
incomes,
which
overall
represented
a
comparatively
small
part
of
yearly
incomes,
are
not
reported
here.
quantitative
data
indicate,
as
households
owning
cattle
would
plough
their
own
fields
first,
resulting
in
those
relying
on
borrowing
cattle
from
others
having
to
plant
later
than
they
wished.
Ploughing
late
reduces
yield
and
also
runs
the
risk
of
the
maize
not
ripening
before
being
damaged
by
frost
or
eaten
by
cattle,
which
are
eager
to
get
into
the
fields
to
eat
the
maize
stalks
at
the
time
of
year
when
grass
is
becoming
dry.
While
households
without
cattle
and
with
little
money
to
pay
for
ploughing
most
often
failed
to
plant
their
field,
even
households
with
the
capacity
to
purchase
ploughing
ser-
vices
sometimes
failed
to
plough.
Many
smallholders
pointed
out
that
the
fact
that
there
were
fewer
cattle
in
total
in
the
villages
today
made
it
difficult
to
borrow
cattle
for
ploughing.
While
access
to
tractors
was
equally
constrained,
tractors,
which
have
been
the
repeated
focus
of
development
interventions,
cannot
completely
replace
the
role
of
cattle
in
the
region.
Tractors
were
unable
to
plough
steep
and
rocky
fields
which
could
be
accessed
by
cattle.
Furthermore
new
rain
patterns
might
make
tractors
even
less
suit-
able
than
in
the
past.
Many
farmers
reported
that
too
hard
rains,
excessively
wet
years,
and
years
when
rains
came
too
late
were
becoming
more
common.
In
excessively
wet
years
the
tractors
that
are
small
enough
to
enter
smallholders’
fields
get
stuck
in
wet
soils
(large
farmers
get
around
this
problem
by
dual
wheels
on
tractors).
Those
who
used
cattle
could
plough
in
wetter
years,
reinforcing
the
importance
of
having
access
to
cattle.
Having
access
to
cattle
also
meant
increased
flexibility
in
the
timing
of
ploughing
to
suit
the
weather.
The
lack
of
adequate
herding
of
animals
was
frequently
cited
as
a
problem
by
farmers
as
animals
frequently
entered
fields
and
ate
the
crops.
When
the
fields
were
reorganised
into
the
current
location
during
the
Betterment
in
the
1970s,
fences
were
provided
by
gov-
ernment.
These
were
now
commonly
damaged.
At
the
same
time
cattle
was
often
not
herded
properly,
one
central
reason
being
that
children
who
used
to
do
this
before
the
fields
were
fenced
today
are
in
school.
Poor
and
very
poor
households
were
particularly
hard
hit
by
damaged
crops,
both
because
of
their
more
constrained
situation,
and
because
they
commonly
were
less
successful
than
wealthier
households
in
demanding
reimbursement
from
animal
owners
for
crop
damage.
Initially,
the
MFPP
fenced
part
of
one
field
area.
However
support
with
fencing
was
subsequently
abandoned
in
the
programme.
The
Eastern
Cape
Department
of
Agriculture
has
generally
been
reluctant
to
provide
fencing
support,
as
fencing
often
has
been
stolen
or
the
subject
of
local
conflict
and
corruption
(Hajdu
et
al.,
2012).
Smallholders
nevertheless
repeatedly
men-
tioned
mending
of
fences
as
a
key
priority
when
they
were
asked
to
suggest
government
support
to
farming
that
would
be
most
helpful
to
them.
Fencing
a
single
field
is
costly,
and
also
prevents
the
field
owner
benefitting
from
government-supported
plough-
ing
schemes,
where
whole
field
areas
are
commonly
ploughed.
Two
(one
middle
and
one
rich)
of
the
265
households
surveyed
here
had
fenced
their
own
fields
in
2012.
Effects
of
poverty
on
investment
in
agricultural
inputs
The
most
common
annual
investments
in
agriculture
were
fer-
tiliser,
maize
seeds,
pesticide
treatments
for
grain
weevils
and
ploughing
services.
Investment
in
these
inputs
differed
between
the
wealth
groups
(Table
3).
While
very
poor
households
purchased
less
fertiliser
than
the
other
groups,
it
is
notable
that
a
substantial
number
of
households
across
wealth
groups
did
invest
in
fertiliser.
Even
in
small
quan-
tities,
it
was
seen
as
a
good
investment,
with
significant
effects
on
yields.
It
was
also
much
less
labour-demanding
to
transport
to
the
field
than
cattle
manure.
Cattle
manure
was
frequently
used
in
home
gardens.
Investment
in
maize
seed
was
less
frequent
across
wealth
groups,
reflecting
the
practice
of
recycling
and
sharing
seeds
and
also
the
lack
of
knowledge
about
the
qualities
of
the
differ-
ent
seed
material
available
in
local
agricultural
supply
stores
(see
“Smallholders’
maize
varieties
and
the
new
‘project’
maize”
sec-
tion).
Table
2
Proportion
of
households
active
in
farming
within
wealth
groups
(265
households).
Wealth
group
Involvement
in
farming
Owning
a
field
Planting
field
(of
those
owning
fields,
n
=
199)
Planting
garden
Rich
76%
81%
83%
Middle
72%
80%
82%
Poor
86%
77%
89%
Very
poor
76%
56%
80%
Pearson
Chi-square
testing
to
determine
whether
wealth
group
affects
agricultural
investment
(significant
differences
(p
<
0.05)
are
marked
in
bold)
0.397
0.006
0.549
K.
Fischer,
F.
Hajdu
/
Land
Use
Policy
46
(2015)
304–313
309
Table
3
Proportion
of
households
making
agricultural
investments
within
wealth
groups
(265
households).
Wealth
group
Type
of
agricultural
investment
Buying
fertiliser
Buying
maize
seeds
Buying
pesticide
treatment
for
grain
weevil
Paying
for
tractor/cattle
for
ploughing
Rich
69%
36%
83%
26%
Middle
70%
37%
82%
24%
Poor
67%
13%
82%
20%
Very
poor 47% 20% 64%
22%
Pearson
Chi-square
testing
to
determine
whether
wealth
group
affects
agricultural
investment
(significant
differences
(p
<
0.05)
are
marked
in
bold)
0.006
0.005
0.014
0.887
Table
3
also
shows
that
the
most
common
investment
across
wealth
groups
was
pesticide
treatment
for
grain
weevils,
which
attacks
maize
in
storage.
Despite
this,
participant
observation
con-
firmed
that
under
treatment
of
storage
pests
was
widespread,
resulting
in
damage
to
stored
maize.
Relatively
few
households
paid
for
ploughing,
but
there
was
no
significant
correlation
between
poverty
and
paying
for
ploughing,
reflecting
the
lower
availabil-
ity
of
cattle
for
ploughing
with
increasing
poverty.
The
mean
cost
reported
for
ploughing
the
field
was
520
Rand
(range
120–2600
Rand),
which
represents
a
substantial
investment
for
resource-
constrained
households.
The
fact
that
many
resource-constrained
households
still
chose
to
pay
for
ploughing
gives
an
indication
of
the
value
these
households
placed
on
ploughing
their
field.
Working
the
field
by
hand
is
labour-intensive,
which
makes
this
a
particu-
larly
challenging
option
in
poor
and
very
poor
households.
Smallholders’
maize
varieties
and
the
new
‘project’
maize
As
many
households,
in
particular
the
poorest,
frequently
failed
to
access
sufficient
amount
of
seed
to
plant
their
field,
the
focus
in
the
MFPP
on
distributing
maize
seed
might
seem
appropriate.
However,
as
will
be
described
in
this
section,
the
maize
varieties
distributed
during
the
MFPP
did
not
provide
the
increased
food
security
needed
as
they
were
not
adapted
to
local
farming
condi-
tions.
The
majority
of
smallholders
relied
completely
or
partially
on
locally
recycled
seed
material.4Recycling
of
seed
kept
input
costs
down
and
was
also
a
way
of
keeping
preferred
varieties.
However
damage
in
storage
or,
particularly
for
resource-constrained
house-
holds,
having
to
eat
the
intended
seed
due
to
food
scarcity,
were
common
reasons
for
not
having
sufficient
amounts
of
seed
left
at
planting
time.
Those
who
could
afford
it
topped
up
their
local
seed
supply
with
purchased
seed.
Friends
and
neighbours
were
other
central
sources
of
seed.
Smallholders
referred
to
their
own
recycled
maize
as
‘Xhosa
maize’,
Xhosa
being
the
dominant
ethnic
group
in
the
region,
indi-
cating
a
perceived
closeness
to
the
local
maize.
Some
smallholders
could
name
a
range
of
different
Xhosa
maize
varieties
with
different
qualities.
However
‘Xhosa
maize’
is
by
no
means
an
isolated
or
sta-
ble
variety.
Locally
recycled
and
purchased
varieties
are
not
planted
separately
and
thus
cross-pollinate
in
farmers’
fields.
Practices
of
recycling
and
sharing
both
local
and
purchased
seed,
often
without
clear
information
on
the
origin
of
the
seed,
contribute
further
to
the
mix
of
local
seed
material
(this
is
described
in
more
detail
in
Iversen
et
al.,
2014;
Jacobson
and
Myhr,
2013).
Some
farmers
also
used
the
word
‘Xhosa
maize’
for
purchased
maize
varieties
express-
ing
certain
features,
such
as
suitability
for
local
storage
conditions,
4While
76%
of
the
households
that
had
planted
maize
reported
planting
recycled
seed
material
in
2008,
in-depth
interviews
and
participant