Smallholders and the "Walmart Effect" in South Africa

Full text

i
Smallholder and agro-food value chains in South Africa
Smallholder and agro-food value chains
in South Africa: Emerging practices,
emerging challenges
Edited by Stephen Greenberg

ii Smallholder and agro-food value chains in South Africa
Contributors
Michael Aliber – Project Co-ordinator, Department of Agriculture, Forestry and Fisheries (DAFF)
Jack Armour - Operations Manager (Transformation, Local Government, Environment and Agricultural
Economics), Free State Agriculture
Davison Chikazunga – Researcher, Institute of Poverty, Land and Agrarian Studies (PLAAS), University of
the Western Cape
Ben Cousins - DST/NRF Chair in Poverty, Land and Agrarian Studies, University of the Western Cape
Nerhene Davis – Department of Geography and Geoinformatics, University of Pretoria
Stephen Greenberg – Freelance researcher
Lusito D. Khumalo - Enterprise Development Manager, Westfalia Fruit Estates, Modjadjiskloof, Tzaneen,
Limpopo Province
Marc Lewis – MA student, Institute of Poverty, Land and Agrarian Studies (PLAAS), University of the
Western Cape
Andre Louw - ABSA Chair of Agribusiness Management, University of Pretoria
Mandla Nkomo - Programme Manager, Agribusiness and Food Security, TechnoServe South Africa
Gaynor Paradza – Senior researcher, Institute of Poverty, Land and Agrarian Studies (PLAAS), University of
the Western Cape

iii
Smallholder and agro-food value chains in South Africa
Contents
Page 1 Introduction: smallholders and value chain integration in South Africa
Stephen Greenberg
2 Conceptualising approaches to smallholders and markets
Michael Aliber
3 Determinants of smallholder farmers’ participation in modern food markets: the
case of tomato supply chains in Limpopo
Davison Chikazunga
4 Sustainable policy support for smallholder agriculture in South Africa: key
issues and options for consideration
Andre Louw
5 Big Business for Small Farmers: the Case of Venda Avocado Growers
Lusito D Khumalo
6 Experiences and Insights on smallholder farmer value chain integration
Mandla Nkomo
7 Aspects of the citrus value chain explored in the context of Moletele
restitution initiatives, Hoedspruit
Nerhene Davis
8 Smallholders and the ‘Walmart eect’ in South Africa
Stephen Greenberg and Gaynor Paradza
9 Private sector and commercial farmer involvement in smallholder integration
Jack Armour
10 Urban agriculture, livelihood strategies and commodity networks in inner-city
Johannesburg: a case study of a vegetable co-operative in Bertrams
Marc Lewis
11 ‘Accumulation from below’ and the Tugela Ferry irrigation farmers
Ben Cousins

1
Smallholder and agro-food value chains in South Africa
Introduction:
Smallholders and value
chain integration in
South Africa
Stephen Greenberg
In many cases, integration of smallholders into corporate
value chains can make protable economic sense. The
sugar, poultry, cotton, tobacco and forestry sectors have
been doing this for a long time already, without any gov-
ernment compulsion. The arena opening up now is the
small-scale production and delivery of fresh fruit and veg-
etables to supermarkets. These new opportunities have
emerged as a result of the expansion of supermarkets
into more distant rural areas, previously only served by
informal markets, and government black economic em-
powerment (BEE) procurement policies have added to
the logic.
A key emerging strand in the development of smallholder
agriculture in South Africa is the effort to integrate small-
holders into corporate food retail value chains. In this, the
private sector and government have a common agenda,
which is to build a commercial smallholder class that
does not require ongoing nancial support for survival,
but which is able to stand on its own feet and compete
in the market. Both government and the private sector
recognise the need for some kind of ‘start-up’ support,
and Walmart-Massmart’s recently announced supplier
fund will put pressure on other food retailers to deepen
their own activities in this regard.

2Smallholder and agro-food value chains in South Africa
Many ‘development’ voices argue that incomes for
producers will increase if they can meet the conditions
for entry into corporate value chains, which will have a
positive impact on livelihoods (e.g. Brown and Sander
2007; Seville et al 2011). The argument therefore is
that resources should be used to facilitate this entry.
This simple narrative has been challenged on the basis
that not all smallholders can enter these chains because
of a lack of resources, high transaction costs or biases
against smallholder production in policy and in corporate
procurement practices. Others refer to ‘adverse incorpo-
ration’ (Hickey and du Toit 2007) to indicate that even
where smallholders might get a foothold into these value
chains, they do not always benet because power rela-
tions are skewed against producers in the buyer-driven
chains that characterise most food products. Others sug-
gest that local food economies and so-called ‘informal’
systems of production and distribution establish a basis
for an alternative that is less resource intensive, more
benecial to producers and which has a higher likelihood
of making food more accessible to the poor (e.g. the food
sovereignty and agro-ecological movements).
The papers in this publication come out of a workshop
hosted by the Institute for Poverty, Land and Agrarian
Studies (PLAAS) in Johannesburg in November 2011.
The workshop brought together academics, government
ofcials, a few representatives from black smallholder
farmers and the private sector, and generated a number
of case studies on efforts to integrate smallholders into
formal or corporate value chains. A selection of these is
included in this publication. They focus on private sector
initiatives and raise key issues around who smallholders
are and what strategies can best be employed to build a
layer of productive smallholders in South Africa.
Defining smallholders
In 2009 the African National Congress (ANC) identied
rural development as one of its ve priorities for South
Africa. To some extent breaking with the past, agricultural
production was placed at the centre of such a strategy,
displacing the welfarist development strategies that had
characterised rural development since 1994, and which
had failed to make any signicant inroads into rural pov-
erty apart from the extension of social grants.
As a result, the idea of smallholder production in South
Africa was given a new lease of life after 2009. The
ANC’s 1994 election platform, the Reconstruction and
Development Programme (RDP), referred to a shift from
“the inefcient, debt-ridden, ecologically-damaging and
white-dominated large farm sector to… a more sustain-
able agricultural system” without dening what the latter
meant. Further down, the document went on to propose
the government should support “part time activities,
including small-scale farming, which can increase pro-
ductivity, incomes and household food security” (ANC
1994: 84). In practice this commitment dissolved as
agricultural budgets were slashed in efforts to stabilise
the overall economy and as agriculture was modernised,
leading not to more small-scale opportunities but to the
consolidation of land and agricultural resources amongst
fewer large-scale producers.
But over the years a consensus emerged amongst many
in the state and in civil society organisations working on
issues of land and agriculture that smallholder agricul-
ture was required to rebalance the rural economy and
to open opportunities for those disadvantaged under
apartheid and the economic system that emerged from
it. Disagreements remain, in particular about the most
sustainable path to creating and sustaining smallholder
production. We are currently at a point where the corpo-
rate agro-food system is widely considered indispensable
for food security in South Africa. Consequently, any efforts
to expand black smallholder production must be done
in tandem with the increasingly concentrated agribusi-
ness sector. Supermarkets are seen as the way of the
future, and conversely ‘informal’ markets or systems of
distribution are seen as the product of historical neglect
and marginalisation that should eventually be eliminated
and replaced with a modernised food distribution system
based on the logic of capital. This logic goes hand in hand
with the commodication of food and the nal detach-
ment of food producers from food consumers.
So far we have used the terms small-scale and small-
holder without dening them. These two terms will be
conated in this paper under the name of smallholder.
There are many different denitions currently doing the
rounds but there is general agreement that a useful de-
nition will encompass differentiation within the category
of smallholder.
Large-scale agriculture and agribusinesses currently
favour a denition based on turnover rather than land
size. Johan Kirsten (2011) argues that if measured by
gross farm income, small-scale farmers should be con-
sidered as any producer with a gross farm income below
R500 000 a year. The 2007 Census data shows that this

3
Smallholder and agro-food value chains in South Africa
included 56.5% of what are called ‘commercial farmers’
in the census.
Two points of differentiation emerge from this denition.
The rst is that smallholders are economically diverse,
and by this denition can incorporate subsistence pro-
ducers using land in their backyards to purely commer-
cial producers on large amounts of land. The second dif-
ferentiation is by race, indicating that smallholders can
be black or white. This is a necessary corrective to the
dualism that has characterised the discussion to date
which equates smallholders with black producers and
large-scale farmers with white producers.
But this is not sufcient as a denition of a smallholder.
We also need to take into account the amount of land
producers have at their disposal. A producer may be in
possession of a very large amount of land but is only
using part of it or is using it unproductively. Can they
then be called smallholders? It seems that we must take
some account of land holdings in the denition. There is
no practical value in dening an unproductive farmer with
a large amount of land in the same category as very pro-
ductive farmers with very limited land at their disposal.
What is a reasonable cut-off with regard to land size?
Vink and van Rooyen (2009: 32) talk of farmers with less
than 20ha as being small scale, but ultimately measure
scale on the basis of income. The size might have to
differ according to type of production (e.g. extensive live-
stock vs. intensive horticulture), or perhaps we can nd
a happy medium that is able to incorporate livestock as
well as horticulture (including orchards) and eld crops.
If we aim for 60-80ha as the upper limit of small-scale
production, this can incorporate all types of production
with the possibility of at least deriving a substantial
portion of household income from agriculture. This can
then be used in conjunction with Kirsten’s gross income
denition to exclude producers with relatively large in-
comes from the denition of smallholder.The essential
argument from PLAAS researchers (Hall 2009; Aliber et
al 2009; Cousins 2010; Cousins in this volume) is that
smallholders incorporate a range of different classes
and social groups and production systems, and that they
have differential relationships with markets. Cousins
(2010) argues that the term ‘smallholder’ is problematic
because it disguises these differences, and should be
used only in conjunction with a qualifying adjective (e.g.
‘subsistence’ or ‘commercially-oriented’). Without this ex-
plicit internal differentiation, policy and practice will end
up treating smallholders as a homogenous group.
However, there is something inherent about small-scale
production that may have some value in its own right,
especially in relation to the ecological crisis. In agricul-
ture this crisis is very much linked to a production model
built on fossil fuels, technology removed from the direct
control of producers, and integration into global markets
where food travels long distances to reach the end-user.
Farms are consolidating (growing in size and decreasing
in number) because of the logic of economies of scale
which are enabled by these factors. If we are to move
away from this production model, the scale of produc-
tion becomes an important part of the solution. There is
general agreement on the need for vibrant smallholder
production from the World Bank and the UN to Via Camp-
esina (e.g. IAASTD 2009; ETC Group 2009; de Schutter
2010). This is not to say all small-scale production is in-
herently ecologically sound, but it is a necessary compo-
nent of ecologically sustainable production. Gender, race
and class differentiation amongst smallholder producers
remains critical, especially if resources for support are to
be targeted. But small-scale production has its own value
beyond these differences.
Processes of smallholder
formation
Ben Cousins (2007; 2010; this volume) focuses on the
class differentiation of farmers. For all intents and pur-
poses, this can be applied to differentiation within the
smallholder category. He introduces the terms ‘accumu-
lation from above’ and ‘accumulation from below’ to indi-
cate different ways in which new farmer can be formed.
‘Accumulation from above’ refers to sponsored accumu-
lation in the interests of established capitalist entities,
including the state. Primary production is outsourced
to smallholders, often with a high degree of institutional
involvement, e.g. input provision, insurance, credit, sec-
ondary transport, sales and distribution directly man-
aged by agribusiness or the state where it is involved.
‘Accumulation from below’ refers to farmers using their
own resources to expand into capitalist producers with
eventual possible absorption into agribusiness value
chains. The notable aspect of this is that it remains within
the framework of accumulation. This is consistent with
Cousins’ suggestion that there is need for a bias towards
the estimated 200-300 000 existing commercial black
farmers, but without neglecting the importance, both
economically and socially, of providing support to others

4Smallholder and agro-food value chains in South Africa
to produce food and to connect into the formal economy
where they can.
In the case studies presented at the workshop and pro-
duced in this publication, we have no cases of accumula-
tion from below. In all instances, there is signicant infra-
structural and technical support either from agribusiness
or from the state to connect smallholders to corporate
markets. In two cases, irrigation farmers in Msinga in
KwaZulu-Natal (Buthelezi and Cousins) and urban food
producers in Johannesburg (Lewis), the producers are
generally marginalised or excluded from resources, but
where resources do come in, they still come from above
mainly via state welfarism. In these cases there is limited
or no capital accumulation, placing the majority of these
farmers outside the dichotomy of accumulation from
above or below.
The agribusiness or commercial farming perspective of-
fers a linear relationship between ‘backward’ subsistence
agriculture and ‘successful’ commercial production. Are
producers sustainably integrated into circuits of capital
accumulation or not? That is the measure of advance or
development of food producers. The articles in this pub-
lication raise the question of whether there is room for a
diversity of production types that all receive appropriate
support, whether private or public. That is, it raises the
question about whether subsistence agricultural produc-
tion is a valued part of the landscape that warrants a
strategy and support in its own right, not necessarily and
only as a precursor to commercial production, but as a
part of diverse food production and distribution systems.
Hall, Cousins and others at PLAAS have highlighted the
‘missing middle’, a category of medium-scale black
farmers that can ll in the gap between strategies of
accumulation from above and accumulation from below
and that are ‘commercially oriented’. The importance of
lling this gap is to create a diversied production base
which can spread risk in the sector, and which brings
racial and scalar balance to production. However, the
focus on methods of capital accumulation (from above or
below) limits our analysis to those producing directly into
concentrated agribusiness markets, or at least into mar-
kets overshadowed by this concentration. Input supply,
storage, processing and food retailing are all highly
concentrated. It makes little difference from the overall
point of view of the reproduction of capitalist social and
economic relations whether accumulation occurs from
above or below. It seems appropriate, in the current
capitalist crisis, to begin to raise practical questions of
how to transcend accumulation as the driving logic of
agricultural production.
Contemporary
agrarian capitalism and
smallholder integration
South Africa’s agrarian structure is characterised by the
concentration of resources and a dualistic structure of
production. The National Development Plan (NDP) (Na-
tional Planning Commission, 2011), which seems to be
the product of a very distant government, has proposed
integration of smallholders into corporate value chains
as a key objective in the rural areas. One of the main
contradictions in the NDP is that South African agrarian
capitalism is in crisis, yet the NDP insists that the only
way forward for smallholders is to be integrated into it (to
paraphrase Peter Jacobs1).
Looking at the big prots agribusiness corporations are
reaping, there does not appear to be a crisis. But under-
neath those prots is an increasing cost-price squeeze,
precisely for small- and medium-scale commercial pro-
ducers, that threatens their long-term survival. Although
food prices are rising, producers often get a small share
of the nal price, while input costs have risen dramatically
as natural resources become scarcer and the logic of per-
manent growth constantly increases demand for these
resources. The protability of corporate agribusinesses is
also built on the back of sharply rising consumer prices
for food and the consequent rise in hunger. The structural
underpinnings of the system which produced this deterio-
rating situation are ecological, social and nancial.
Ecologically, the reliance on fossil fuels and the impact
of climate change is causing deterioration in the mate-
rial base on which production is built – the land, water,
livestock and vegetation. Yet integrating smallholders
into corporate value chains, directed towards supermar-
kets especially, requires a duplication of these same
production methods. This reinforces a path dependency
at a time when this path is receding into unknown terri-
tory. The capitalist response to this is as it always was:
a faith in technology to solve tomorrow’s problems, for
example, through the use of biotechnology, irrigation and
mechanisation. However, these technologies themselves
are heavily rooted in a fossil fuel economy and cannot
be separated from growing social inequalities in the

5
Smallholder and agro-food value chains in South Africa
form of greater concentration of land ownership in fewer
private hands; growing hunger amongst those unable to
purchase the bounty being produced by capitalist agricul-
ture; and the radical separation of food producers and
consumers.
This technological response also deepens agriculture’s
reliance on the credit economy. The provision of credit
sits at the centre of any effort to integrate smallholders
into corporate value chains, since agricultural production
has been credit-driven for decades now, in South Africa as
well as in any place where consistent surpluses are pro-
duced for sale. Money in the form of credit is a necessary
condition for entry into capitalist commodity relations,
and is also necessary for capitalism because indebted-
ness is the driver of ‘at’ money (Rowbotham 1998; The
Agonist 2012) – the production of virtual money by the
nancial institutions to sustain capital growth. The vast
majority of money in circulation is in the form of credit.
Producers require credit not only for immediate produc-
tion, but also for a reserve to act as a buffer in conditions
where prices drop unexpectedly. Historically, the state
in South Africa provided buffer services for key crops
through the operation of a oor price (a minimum price
the state would guarantee in the event the market did
not absorb all the production). The state therefore car-
ried grain reserves for example, which stabilised prices,
but also encouraged overproduction because the state
was a guaranteed market of last resort. Deregulation has
transferred the price risk to individual farmers, whether
black or white, funded through credit. The larger enti-
ties can take losses in their stride, either because they
have diversied economic activities which spread risk or
because they have cash reserves or established lines of
credit or both.
For new entrants into the capitalist markets, credit is
hard to come by, especially in the context of the past
three years, where liquidity in the global economy as a
whole has dropped drastically and the provision of credit
has dried up. A few new entrants will be hand-picked
for integration into the credit economy but for those out-
side circuits of accumulation and even for the majority
of those accumulating ‘from below’ using their own re-
sources, the lack of credit will prevent their growth and
expansion. Those with control over the provision of credit
(the banks and other nancial institutions) determine
the direction of investment and consequently the shape
of commercial agriculture in South Africa. The state his-
torically made some attempts to extend credit to new
entrants, especially via the Land Bank, whose mandate
was nevertheless to prot from loans. However, some
of the loans they made were questionable (‘sub-prime’)
which is why the Land Bank got into such trouble – by
loaning to people who could not or did not pay back.
The discipline of the credit market is required to ensure
a constant ow of revenue to the lending institutions,
over and above the amount that was lent (bank charges,
interest and penalties). This additional revenue is con-
sidered to be a return on the risk they took in lending
(i.e. producing money). For smallholder farmers (consid-
ering income and land size as the primary criteria) these
charges are higher because they are not always in a posi-
tion to repay the loans and they have few assets that can
be seized. This means they are considered a higher risk.
A greater share of the repayments will therefore accrue to
the lender as returns to take account of that risk. Banks
will not currently lend to risky borrowers unless under
compulsion or state guarantee. Is the solution to expand
credit, nding innovative ways of integrating smallholders
into the credit economy, or to nd other ways of over-
coming this conundrum such as a way that does not rely
on the same ow of resources?
Everyone is caught in the web of capitalist relations,
which means production (even the production of ideas)
is driven by growth and the accumulation of capital.
Even if producers are not selling into markets, they are
producing in the context of the commodication of agri-
cultural products which establishes an alternative source
for what they are producing. So even if food producers
are locked out of the credit economy, the exchange value
of their products is still benchmarked by those inside the
credit economy. That means they can sell only in rela-
tion to market prices. There are other values attached
to their products though such as ‘use values’ which are
directly realised through own consumption or distribution
i.e. not for sale to neighbours and social networks. This
remains an important component of local and household
level food security even while it functions in the shadow
of capitalist markets.
Experiences to date
The articles in this publication consider some of the prac-
tical experiences of integrating smallholder farmers into
agro-food value chains in South Africa. The emphasis is
on private sector initiatives, with welfarist government
support to smallholders as a counterpoint. The workshop

6Smallholder and agro-food value chains in South Africa
had presentations on government efforts to integrate
smallholders into value chains via (corporatised) fresh
produce markets, but we were unable to secure papers
from the participants for inclusion in this publication. This
is unfortunate because the fresh produce markets offer a
potential alternative to corporate value chains based on
food retailers. Hopefully, that story can still be told in the
near future.
Speaking in his personal capacity, Michael Aliber from the
Department of Agriculture, Forestry and Fisheries (DAFF),
opens with a paper on the current wave of ‘inclusionism’
and its relation to multiplier effects in agricultural produc-
tion. Aliber offers some valuable insights and intriguing
statistics which suggest that subsistence production and
local distribution of surpluses may add more real value to
the food economy than the high multipliers of the formal
economy, which may “signify the non-trivial inefciency of
over-developed value chains”. Aliber points to two current
government initiatives that may carry these ideas for-
ward: the Zero Hunger Programme with preferential pro-
curement from smallholders for public food purchases
(e.g. schools, hospitals) and the decentralisation of agro-
processing. These are both worth watching carefully.
The next two pieces, by Davison Chikazunga and Andre
Louw, who both worked on the Regoverning Markets
Programme a few years ago, discuss the conditions
under which smallholders may be integrated into cor-
porate value chains. Chikazunga highlights production
infrastructure, in particular irrigation and greenhouses,
and collective action in the form of commodity associa-
tions as key considerations in the case of tomatoes in
Limpopo. He shows how the local wet market performs a
vital role in stabilising farmer incomes in the off season.
He conducts an income analysis which shows that those
earning the most per hectare supplied the local wet mar-
kets, whereas those earning the least per hectare sup-
plied to supermarkets over the entire year. This reveals
a trade-off between stability of demand and income.
Louw points out that agricultural restructuring has led to
increased risk for farmers, and he highlights the role of
credit, especially for on-farm investments. He indicates
that cognisance needs to be taken of the heterogeneity
of smallholders, and emphasises the importance of inter-
mediaries who can provide efcient services to support
smallholders and connect them to markets.
Lusito Khumalo and Mandla Nkomo offer lessons from
their experiences in providing support in agribusiness
smallholder procurement programmes in Limpopo.
Khumalo writes about his experience at Westfalia Fruit
Estates in Limpopo, which has developed a model of
smallholder integration for avocados, where the retailer
provides credit, with management by the agribusiness
Westfalia. It is essentially a contract farming scheme.
The project shows that although a number of producers
are making ends meet, only one of ten producers in
the agribusiness mentorship programme managed to
sustain the business of supplying Westfalia. Khumalo
indicates that economies of scale, business efciency
and product quality are key issues. Mandla Nkomo offers
seven lessons for smallholder integration into corporate
value chains. His paper highlights the role of credit in
the agricultural economy, and the importance of on-farm
extension, which cannot be parochial or “invoice driven”.
Nkomo works at Technoserve, the company that was
recently granted a R15m smallholder supplier develop-
ment contract for Walmart-Massmart.
Nehrene Davis then offers a slightly different perspective
on ‘accumulation from above’ in her analysis of the role
of agribusinesses in supporting the Moletele land restitu-
tion claimants to maintain the commercial citrus produc-
tion they inherited. Four separate strategic partnerships
were formed on different farms, each with different
results. Davis’s preliminary research ndings indicate
that knowledge and control over key processes allow
strategic partners to dictate the terms of engagement
with resource holders in the form of the claimant’s Com-
munal Property Association (CPA). The commercial logic
pursued on the restitution land led the partnerships to
adopt the same approach to agriculture as that of other
commercial producers, such as outsourcing of labour
and production and a shift to exible work. As a result,
despite maintaining commercial production on the land,
claimants have not beneted much materially.
Stephen Greenberg and Gaynor Paradza contribute a
piece on the possible implications for smallholders of
Walmart’s entry into South Africa. A supplier fund pro-
posed by Walmart-Massmart to assist producers (not
only in agriculture) to meet Walmart’s requirements was
a condition of approval of the merger by the competition
authorities. Greenberg and Paradza indicate the changes
in supplier relations that Walmart might bring. They con-
clude that Walmart’s entry will benet a relatively small
elite among smallholders. However, even this may come
at the long-term cost of increasing dependency on a
single large buyer, and depreciating terms of trade and
quality, if global experience is anything to go by.

7
Smallholder and agro-food value chains in South Africa
Jack Armour from Agri-Free State offers a piece from a
commercial farmer’s viewpoint that highlights key fac-
tors required for successful integration of smallholders
into corporate value chains. These include training, ‘in-
centivising’ mentorship models, risk reduction through
safety nets and marketing strategies that identify niche
markets. He presents a public-private partnership model
based on cellphone technology and decentralised agro-
processing hubs that is currently under development
in the Free State. Armour highlights the importance of
involving experienced commercial producers, the use of
new technology and local agricultural associations as
critical factors for a successful intervention.
Marc Lewis and Ben Cousins conclude the case studies
with preliminary analyses of eld research on a collective
urban garden project in Johannesburg and among small-
scale irrigation farmers in the Tugela Ferry-Msinga mu-
nicipality in KwaZulu-Natal respectively. In Johannesburg,
the urban food producers receive sporadic support from
the provincial government but production is very low and
even though they are oriented to selling, the producers
struggle to move beyond survivalist production. These
cases show the experience of the majority of subsistence
producers, who do not have much market information,
lack the resources for consistent production of surpluses
and are not selected for participation in agribusiness
smallholder programmes. Cousins concludes this volume
with initial research ndings on production in Msinga and
offers further insights into the underlying agrarian struc-
ture in relation to ‘accumulation from below’.
References
Aliber, M., Baiphethi, M. and Jacobs, P. (2009) “Agricul-
tural employment scenarios”, in R. Hall (ed) Another
Countryside: Policy Options for Land and Agrarian
Reform in South Africa. Institute for Poverty, Land and
Agrarian Studies: Cape Town.
African National Congress (ANC) (1994) The Reconstruc-
tion and Development Programme. ANC, Johannesburg
Brown O and Sander C (2007) Global Supply Chains and
Smallholder Farmers, IISD: Winnipeg.
Cousins B (2007) ‘Agrarian reform and the two econo-
mies: transforming South Africa’s countryside’, in L
Ntsebeza and R Hall (eds) The Land Question in South
Africa: the Challenge of Transformation and Redistribu-
tion. HSRC Press: Pretoria.
Cousins B (2010). ‘Agrarian structure and accumulation
from below: Re-thinking the role of “smallholder farmers”
in addressing structural poverty in rural South Africa’,
presentation to the Overcoming inequality and structural
poverty in South Africa: towards inclusive growth and
growth and development conference, Johannesburg,
20-22 September.
De Schutter O (2010) ‘The right to food’, report submitted
by the Special Rapporteur to the UN General Assembly
Human Rights Council, Sixteenth session, Agenda item
3. Report No. A/HRC/16/49. UN General Assembly:
Geneva.
ETC Group (2009) Who Will Feed Us? Questions for the
Food and Climate Crises. ETC Group: Ottawa.
Hall R (2009) ‘Land reform for what? Land use, produc-
tion and livelihoods’, in R Hall (ed) Another Countryside:
Policy Options for Land and Agrarian Reform in South
Africa. Institute for Poverty, Land and Agrarian Studies:
Cape Town.
Hickey S and du Toit A (2007) ‘Adverse incorporation, so-
cial exclusion and chronic poverty’, CPRC Working Paper
no. 81, Institute for Development Policy and Manage-
ment: University of Manchester, UK.
International Assessment of Agricultural Knowledge,
Science and Technology (IAASTD) (2009) ‘Agriculture at
a Crossroads: International Assessment of Agricultural
Knowledge, Science and Technology’ for Development
Global Report, Centre for Resource Economics: Wash-
ington DC.
Kirsten J (2011) ’Most SA farmers are small-scale’,
Farmers Weekly, 23 September 2011: 38.
National Planning Commission (2011) National Devel-
opment Plan: Vision for 2030. Ofce of the President:
Pretoria.
Rowbotham M (1998) The Grip of Death: A study of
modern money, debt slavery and destructive economics.
Jon Carpenter: Oxfordshire.
Seville D, Buxton A and Vorley B (2011) ’Under what
conditions are value chains effective tools for pro-poor
development?’ IISD/Sustainable Food Lab.

8Smallholder and agro-food value chains in South Africa
The Agonist (2012) ’Peak Money Arrives’, 2 January
2012. The Agonist, accessed 5 March 2012 at http://
agonist.org/numerian/20120102/peak_money_arrives
Vink N and Van Rooyen J (2009) ‘Development Planning
Division’, Working Paper Series no.17, DBSA: Midrand.

9
Smallholder and agro-food value chains in South Africa
Prologue
This contribution is not an academic exposition, still less
an ofcial statement of government policy or thinking. In
the rst place, the author’s command of the issues is too
weak to justify the pretence of an academic treatment;
and in the second, he is not authorised to speak for gov-
ernment in any ofcial capacity.
Conceptualising
approaches to
smallholders and
markets
Michael Aliber 3
Rather, this is a ‘personal reection’ on the theme of
smallholders and markets, in which I hope to do little
more than raise useful questions and maybe share a
half-insight or two, relying on a combination of my (gen-
erally differently oriented) empirical work and exposure
to recent policy discussions. My tentative suggestion is
that part of our challenge is merely in articulating the
conceptual challenges, which clutter our way towards a
pragmatic vision for the smallholder sector and the rural
economy of which it is a part.

10 Smallholder and agro-food value chains in South Africa
Value chains, multipliers
and the prevailing
‘pro-inclusionism’
The explosion of interest worldwide in ‘value chains’ is
astonishing. Within South Africa, much of the recent
interest in value chains is focused on nding ways of
ensuring that marginalised small-scale farmers, and the
poor in general, are able to ‘link’ into existing value chains.
This for example was the focus of the USAID-funded Agri-
link programmes of the late 1990s and early 2000s and
to some extent of the UK Department for International
Development (DFID)-funded “Making Markets Work for
the Poor” programme (Ferrand, et al 2004).Further, it
is sometimes claimed that both the job and/or wealth
creation potential of the agricultural sector lies only
partially in primary agriculture, but more importantly
elsewhere in the value chain. Thus the Land Bank’s more
recently proposed “value chain nancing model”, and
the Agricultural Broad Based Economic Empowerment
(AgriBEE) Fund, which was created to provide grant
funding through which historically disadvantaged
individuals (HDIs) could purchase equity in downstream
beneciation activities. The New Growth Path (NGP) sets
targets for job creation in agro-processing which are more
than half as great as its target for increasing the size
of the smallholder sector. It also calls for “[s]upport for
market and nancial institutions, especially co-ops, that
enable small producers to enter formal value chains…”
(Economic Development Department (EDD) 2011: 18).
In other words, the prevailing approach among policy-
makers in the agricultural sphere is “pro-inclusion”. There
are, or may be qualications to this characterisation, but
we will come back to those below.
‘Value chains’ are closely related to another concept
often used in discussions of economic development
in the agricultural sector and in other sectors, namely
‘multipliers’. Simply put, a multiplier is a mathematical
relationship between the direct impact of, say, an
investment in the agricultural sector – i.e. the number
of primary jobs created together with the secondary job
creation caused by that investment further increase
demand for these agricultural products. The higher the
‘multiplier effect’, the greater the ‘bang for the buck’;
investment decisions should take the secondary impacts
into consideration as well, not least because some
sectors have higher multipliers than others. Agriculture,
in particular, is often touted as having a high employment
multiplier. The National Planning Commission’s National
Development Plan: Vision for 2030 (NPC 2011) posited
that one million jobs could be created by means of
targeted investments in agriculture; of these, one third
would be secondary. Although the researchers (Bureau for
Food and Agriculture Policy (BFAP) 2011), who produced
this particular analysis, regarded this multiplier of 1.5
as ‘conservative’ (i.e. for every 100 direct jobs created,
an additional 50 jobs would be created by means of
multiplier effects), it is still high.
It is worth pausing to ask where estimates of multipliers
come from? Most come from constructing and then
manipulating input-output tables or elaborations thereof.
An input-output table is a matrix in which the rows and
columns refer to sectors. Typically, the columns of the
matrix show how much is spent by each sector (e.g. in an
average year) on intermediate goods produced by other
sectors. The rows correspondingly show for each sector
how much it receives from the other sectors. As such, the
table seeks to capture the inter-relationships between
sectors. Mathematical manipulation of the input-output
matrix, in conjunction with an array of sector-specic
labour coefcients (e.g. how many jobs are implied by a
certain amount of income in a sector), yields the labour
multipliers. Various other types of multipliers can also be
derived.
What is the purpose of this brief exposition? Firstly, to
clarify the relationship between value chains and mul-
tiplier analyses: the way in which the products of one
sector become intermediate goods/inputs for another
sector is in essence a different way of conceptualising
the relationship of different agents along a value chain.
Of course, that does not mean that value chain and mul-
tiplier analysis are the same: rather, they are analyses of
the same relationships, although conducted in different
ways and for different reasons and typically, also at dif-
ferent levels of detail/aggregation.
The second purpose is to show that enthusiasm both
for ‘linking into value chains’ and for ‘multiplier effects’
appears to be different perspectives on the same phe-
nomenon. The former perspective emphasises the vir-
tues of smallholders with better access to existing value
chains, as well as the advantages of smallholders and
other HDIs having an ownership stake in value chains
beyond the level of primary production. The latter per-
spective maintains that an investment resulting in ad-
ditional agricultural production means more agricultural
product being fed into the chain, which will have positive

11
Smallholder and agro-food value chains in South Africa
economic spin-offs beyond the immediate impact of the
investment. It also means, among other things, that the
smallholders are integrated and perhaps better sup-
ported. In sum, the rewards of linking to markets and the
benets of the multiplier effects are different aspects of
the ‘advantages of inclusion’.
Finally, the third purpose is to indicate that the under-
lying logic of multiplier analysis is both highly mechanical
yet abstract. While a particular input-output table may
be rooted in careful data gathering and estimation, the
application of a multiplier derived from an input-output
table to a specic situation in a particular place (e.g. a
proposed investment) is necessarily an abstraction and
the true implications can be grossly misread. This is
masterfully illustrated by Hart’s (1998) critique of sta-
tistically-based regional impact analysis. Hart compares
the ndings of such analyses to careful case-studies in
Malaysia, Taiwan, and China and shows that the true
implications of, say, an investment into agriculture,
depends crucially on local context. These contexts are
ignored at the cost of drawing false conclusions based
on spurious statistical relationships4. To understand the
applicability to South Africa, it is important to distinguish
between different scenarios that could be contemplated.
The table below speculates in broad, qualitative terms as
to the likely ‘linkage effects’ and ‘multiplier effects’ ac-
cording to six different scenarios and shows that even at
a high level of abstraction there is reason to expect that
these effects vary a lot from one case to another. Not only
that, but a positive ‘linkage effect’ does not necessarily
imply a positive ‘multiplier effect’ as it depends on the
scenario. In other words, as closely related as they may
be, they still operate separately, to the extent they are
differentially sensitive to context.
Doubts about the
inclusion agenda
For the most part, the table above assumes that linkage
effects and multiplier effects are ‘good’. But particu-
larly on the linkage side, this assumption is frequently
contested, especially by those who are sceptical of the
advantages of inclusion, that is, those who are sensitive
to the dangers of ‘adverse incorporation’. In du Toit’s typi-
cally eloquent turn of phrase: “Couched at this general
level, the concept of adverse incorporation thus func-
tions as a fairly broad critique of neo-liberal accounts of
poverty and development, accounts that underplay the
risks and disadvantages of inclusion and participation
in unregulated capitalist markets” (du Toit 2009: 2). The
implication is that adverse incorporation can also mani-
fest itself in more specic ways as well. Arguably the pro-
totypical instance of adverse incorporation is that which
takes place via the process which Ponte refers to as “re-
structuring of value chains with continued participation”:
This refers to changes in an existing value chain that
alter the terms and conditions of participation for
chain actors already in the chain. For example, when
supermarkets impose stricter quality standards,
require conformity to Faair Trade standards, or simply
squeeze prices, this can signicantly change invest-
ment demands, rewards or risk exposure for small
(and/or marginal) producers, and salary levels and
employment opportunities for permanent and casual
workers (Ponte 2007: 13).
While such changes are not necessarily to the disadvan-
tage of the farmer, they often are, particularly as the pur-
chaser shifts more and more costs on to the farmer. This
scenario helps explain at least one important feature
that sets apart pro-inclusionists from those concerned
with adverse incorporation. The pro-inclusion perspec-
tive considers the advantages accruing to farmers who
freely avail themselves of a new opportunity: it is in effect
a ‘revealed preference’ argument – that the opportunity
is advantageous is suggested by the fact that farmers
seek it and, if possible, take it up. The particular situa-
tion of adverse incorporation indicated above, however,
transpires over time – and may well start in a way that
justies the pro-inclusion camp’s perspective.
While the purpose here is not to take a stand on the
choice of the more pertinent perspective for Africa, it is
worth venturing one observation: the adversely changing
value chain scenario has already played out quite con-
siderably in South Africa’s agricultural sector, particularly
in horticulture, via the imposition of private standards by
supermarkets. The irony is that it is into this situation that
inclusionists are still hoping to retroactively assist small-
holders to link up to formal value chains; especially the
supermarkets, for the simple reason that supermarkets
have captured so much of the consumer spend. This is
not to say that there is no possibility that the terrain will
shift further to the detriment of producers, but in reality
much of the shift has already taken place and yet these
markets still appear to offer attractive opportunities to
new entrants by assuming some of the hurdles can be
cleared.

12 Smallholder and agro-food value chains in South Africa
Scenario Scenario
description
‘Linkage effect’ ‘Multiplier effect’ Net implications?
Commercial
expansion
E.g. irrigation
expansion, better
transport infra-
structure, lower
production costs,
etc.
Not relevant Possibly positive,
but with con-
sequences
mainly for urban
employment?
Possibly positive
implications, but
does not confront
prevailing down-
ward trajectory
in number of
farm units and
employment
Former homeland
production-led
stimulus, with
emphasis on local
markets
E.g. improved ex-
tension, access to
inputs, infrastruc-
ture; de-centralised
agro-processing
and de-centralised
procurement
Linkages to local
markets (by
construction)
Possibly signicant
at local level
Livelihood creation,
rural development,
improved local food
security?
Former homeland
production-led
stimulus, with
emphasis on non-
local markets
E.g. improved
extension, access
to inputs, infra-
structure; link-up
programmes and
incentives
Linkages to non-
local markets (by
construction)
Possibly sig-
nicant, but with
consequences
mainly for urban
employment?
Livelihood creation,
rural development,
improved local food
security?
Land Redistribu-
tion for Agricultural
Development
(LRAD)
Decline of formally
marketed surplus,
some increase in
locally marketed
surplus?
Some linkages
to local informal
markets
Somewhat
negative
Some livelihood
creation
Proactive Land Ac-
quisition Strategy
(PLAS) , ‘best case’
Maintenance of
large scale com-
mercial, production
continuity, conti-
nuity of contracts
None, no new
linkages
None Minimal economic/
livelihoods
signicance.
Political gains or
accusations over
elitism?
PLAS, short-term
case
Partial main-
tenance of LS
commercial,
production conti-
nuity, continuity of
contracts
No new linkages,
some linkages
lapse
Somewhat
negative
Some possible
damage to the
formal economy,
probably modest
Table 1: Likely ‘linkage effects’ and ‘multiplier effects’ according to
different scenarios

13
Smallholder and agro-food value chains in South Africa
If adverse incorporation is the counterpoint to the pro-
market linkage view, what is the counterpoint (element)
of enthusiasm for the multiplier effect? As far as we are
aware, the literature does not posit one, apart from the
fact that many observers have noted that the validity
of multiplier analysis depends on whether or not there
is spare productive capacity in the affected sectors. If
not, then an increase in demand may simply result in in-
creased (factor and product) prices rather than increased
activity (and thus employment) in other sectors (Stevens
and Lahr 1988). This is intuitive enough, but one might
logically ask how typical is it to have spare capacity? It
suggests either the rather odd possibility that excess
capacity was invested in and never used, or more likely,
that the capacity was created and used but then became
under-utilised due to poor performance in the sector.
In other words, the multiplier analysis is generally only
applicable to situations where the agriculture sector is
lagging behind its own previous performance.
However, I would like to suggest an altogether different
view to the pro-multiplier view by considering the table
below, which shows the estimate average (net) remu-
neration per hectare per year for different South African
‘land use regimes’. Although these gures are relatively
‘shaky’ estimates that should not be taken literally, they
nonetheless provide a suggestive, albeit unreliable order-
of-magnitude ranking. 5
For the purposes of this discussion, the important com-
parison is that between large-scale commercial and
ex-Bantustan farming. The relatively high value for ex-
Bantustan farming could reect the relatively good land
quality of some of the ex-Bantustans compared to, say,
the Karoo but the comparison is still telling in that it does
not adjust for the fact that much of the arable land within
the ex-Bantustans is under-utilised (e.g. for purposes of
Farming regime Average Rand/hectare
Large-scale commercial
- excluding Northern
Cape
368
549
Land reform 171
Ex-Bantustan 683
Source: Aliber et al (2011)
the calculations, the denominator is not the hectarage
actually used but that which is available). Moreover, even
when excluding the Northern Cape from the calculation
of the returns per hectare for the large-scale commercial
farming sector, the estimated value for the ex-Bantustans
is at least as high - if not higher.
What is the explanation for this counter-intuitive nding?
We can be fairly certain that the explanation is not due to
higher land-use intensity in the ex-Bantustans ( i.e. driven
by the application of abundant labour as opposed to
mechanisation). More likely, the reason for the relatively
high value for ex-Bantustan areas is that subsistence-
oriented producers and those who sell to local, informal
markets internalise the margins that would otherwise
accrue to the formal marketing and distribution system.
In Lipton’s terms, these producers enjoy relatively low
“unit transactions costs” (Lipton 2010) and intuitively
that seems right. But another way of saying this is that
they have weak linkages into value chains, whether by
agro-processing or by distribution and retailing, to the
extent that for subsistence and locally marketed produc-
tion, transport costs feature very little in more elaborate
chains, and it begins to seem that high multipliers signify
the inefciency of over-developed value chains. Arguably,
this is adverse incorporation from a different perspective:
the more developed the distribution system, the more the
farmer is sharing the nal value of the product with other
actors along the chain, making the terms of farming more
precarious. Some of that gap may be in the form of addi-
tional value by virtue of beneciation, but much is simply
logistics and transport. 6
Emerging themes in
government policy
This concluding section outlines some of the emerging
directions in government’s policy regarding smallholders
and marketing. It is fair to say that until quite recently,
the prevailing sentiment was very much the pro-inclusion
position noted above. This is still arguably the dominant
view or prescribed government direction.
Recently, however, there are two noteworthy develop-
ments. The rst is the proposed introduction of the Zero
Hunger Programme. Essentially, Zero Hunger is the
increasing use of preferential procurement practices
for food on behalf of government institutions directed
towards smallholders. This will particularly target the
Table 2: Average remuneration per
hectare by livelihood type

14 Smallholder and agro-food value chains in South Africa
National School Nutrition Programme, government’s
feeding scheme for government schools and public hos-
pitals. Initially, the focus will be largely on rural schools
and hospitals. The idea is that through the programme,
smallholders will furnish a progressively larger share of
the food needs of these government institutions, and in
doing so, cater to a broader range of foods in the pre-
scribed menus which means more logistical complexity
and more involved agro-processing.
The second, complementary development is the elabo-
ration of a new agro-processing strategy. As with Zero
Hunger, this policy is not yet nalised, but indications
are that one salient theme in the emerging strategy is
the state-led investment in agro-processing capacity as
a means of both de-concentrating and decentralising the
agro-processing sector. As such, an apparent goal of the
strategy is to counteract the trend towards concentration
of capacity and thus of market power in the post-primary
stages of agro-food value chains.
The unifying theme between the Zero Hunger Programme
and the emerging agro-processing strategy – even though
it does not appear explicitly anywhere in policy – is what
we might call ‘localism’. This is the attempt to promote
local self-sufciency in food production, together with
localised, albeit limited, agro-processing capacity. What
is interesting about this strategy, if it can be called that,
is that it represents a sort of intermediate form of inclu-
sion, i.e. one that gives precedence to local linkages and
limited multipliers. In this interpretation, the particular
function of Zero Hunger is to give smallholder producers
a boost for them to ultimately cater more for local food
demand. It is only partially an end in itself, given that gov-
ernment food procurement is limited relative to private
food demand, even within the ex-Bantustans.
The idea is compelling. It remains to be seen how well
government will be able to implement it. The challenges
are numerous and large, which is not to say insurmount-
able. For one, different government departments in dif-
ferent provinces have established their own procurement
systems. Thus, phasing in preferential procurement in
favour of smallholders is a very involved process. Second,
while in principle there is money available to invest in new
agro-processing capacity by merely redirecting existing
conditional grants (e.g. the Comprehensive Agricultural
Support Programme grant), in practice this is difcult to
do – established expenditure practices have their own
inertia which is not easily overcome. And third, it is one
thing to invest in new agro-processing capacity, but who
is going to manage it?
References
Aliber M, Maluleke T, Manenzhe T, Paradza G and Cousins
B (2011) ‘Livelihoods after Land Reform: Trajectories of
Change in Northern Limpopo Province, South Africa’, un-
published research report, Cape Town.
Bureau for Food and Agriculture Policy (BFAP) (2011) ‘The
Contribution of the Agro-Industrial Complex to Employ-
ment in South Africa’, unpublished manuscript, Pretoria.
Du Toit A (2009) ‘Adverse Incorporation and Agrarian
Policy in South Africa’, accessed 23 July 2010 at: http://re-
pository.uwc.ac.za/xmlui/bitstream/handle/10566/65/
duToit_Adverse2009.pdf?sequence=3
Economic Development Department (EDD) (2011) New
Growth Path Framework EDD: Pretoria, South Africa.
Ferrand D, Gibson A and Scott H (2004) ‘Making Markets
Work for the Poor: An objective and an approach for gov-
ernments and development agencies’, ComMark Trust:
Woodmead.
Hart G (1998) ‘Regional Linkages in the Era of Liberaliza-
tion: A Critique of the New Agrarian Optimism’, Develop-
ment and Change 29(1): 27-54.
Lipton M (2010) Land Reform in Developing Countries:
Property Rights and Property Wrongs. Routledge: London.
McCarthy J (2008) ‘Integrative Report: Generic Economic
and Social Impact of the Sugar Industry in the Context of
Milling Areas’ commissioned by the South African Sugar
Association: Mount Edgecombe.
National Planning Commission (2011) National Devel-
opment Plan: Vision for 2030. Ofce of the President:
Pretoria.
Ponte S (2008) ‘Developing a “Vertical” Dimension to
Chronic Poverty Research: Some Lessons from Global
Value Chain Analysis’, CPRC Working Paper 111, Chronic
Poverty Research Centre: Manchester.
Stevens B and Lahr M (1988) ‘Regional Economic Mul-
tipliers: Denition, Measurement, and Application’, Eco-
nomic Development Quarterly 2(1): 88-96.

15
Smallholder and agro-food value chains in South Africa
Determinants of
smallholder farmers’
participation in modern
food markets: the case of
tomato supply chains
in Limpopo
Davison Chikazunga

16 Smallholder and agro-food value chains in South Africa
Abstract
In South Africa, like other developing countries, there
is a debate about the implications of restructuring the
food markets of smallholder farmers. There has been in-
creased interest among policy makers, researchers and
practitioners on how smallholder farmers can be made
to participate actively in modern market channels. The
general view is that if market access among smallholder
farmers can be improved, the incidences of rural poverty
in developing countries can be reduced signicantly.
This study evaluates the factors which determine the
inclusion of smallholder farmers in modern market chan-
nels using the case of tomato growers in Vhembe and
Mopane districts in the Limpopo Province, South Africa.
The study shows that there is a low participation of small-
holder farmers. However, their increased participation
in modern markets is possible if certain conditions are
put in place at farm and institutional levels. Production
infrastructure (irrigation and greenhouse) and collective
action (commodity) are vital for smallholder farmers’
participation in formal agribusiness supply chains. In ad-
dition to farm level dynamics, both the private and public
sector can play important roles in facilitating the inclu-
sion of smallholder farmers either through application of
business modes, such as local procurement, or through
enacting pro-poor policies such as AgriBEE, which can
foster the participation of smallholder farmers in modern
market channels.
The dynamics of food
markets
Smallholder agriculture in the 21st century is at the cross-
roads especially due to the advent of the agro-industrial-
isation phenomena. According to Reardon et al (2001),
“agro-industrialization” comprises three related sets of
changes: rstly, the growth of agro-processing and distri-
bution; secondly, farm-input provision activities off-farm;
and thirdly, institutional and organisational change in
the relationship between agro-industrial rms and farms.
This phenomenon is characterised by the transformation
of food markets in which market power has shifted from
producers to buyers.
The restructuring of the food markets has led to the re-
arrangement of the food supply chains characterised by
the rise in market dominance by supermarkets and agro-
processors (Humphrey, 2007). These market channels
have been referred to in different ways: as modern,
dynamic, restructured or formal markets among many
others7 (Pote Peter et al 2007). The study uses the term
modern markets in line with the Regoverning Markets
terminologies representing supermarkets, wholesalers
and processors. Reardon (Reardon & Berdegue, 2002;
Weatherspoon & Reardon, 2003; Reardon & Timmer,
2006), the modern day father of supermarkets, hypoth-
esises that these modern markets have grown in their
dominance over the food retail markets. Today, they de-
termine how much should be produced, of which quality,
at what price, and by whom and this has been reinforced
by the introduction of private procurement standards for
coordinating food quality and safety requirements be-
tween consumers and producers.
It was postulated that the restructuring of the food
markets process was kick-started by dynamics in popu-
lation, food consumption patterns and a general rise in
household incomes (Reardon and Timmer 2006). Some
consumers are becoming more demanding in terms of
food quality and safety even in the developing world. It
is further hypothesised that the demand for food safety
and quality standards has been driven by changes in de-
mographics and disposable income (Huang and Reardon
2008). This has led to the creation of convenience-food
niche markets for more afuent consumers, such as
frozen, pre-cut, pre-cooked and ready-to-eat food items
(Reardon and Berdegué 2002). Longer working hours, di-
minishing leisure time, the greater role played by women
in the workplace and greater availability of information
also had a signicant inuence on food markets (Botha
and van Schalkwyk 2006).
According to Louw et al (2007), the restructuring of South
African food markets is at an advanced stage. It is ob-
servable through consolidation, trans-nationalisation
and the emergence and disappearance of supply chain
actors. The country has advanced stages of consolida-
tion comparable to emerging economies in Latin America
and Central Europe. Consolidation of the food industry in
South Africa is seen in the high levels of concentration
in food production, processing, wholesale and retailing
(Weatherspoon and Reardon 2003). Unlike the other
countries in sub-Saharan Africa, the expansion of su-
permarkets is driven by local investment as opposed to
foreign direct investment (FDI) (Louw et al 2004).
The transformation of agro-food markets, however, risks
the exclusion of smallholder farmers from food mar-
kets. The risk is more pronounced among small-scale

17
Smallholder and agro-food value chains in South Africa
Market conditions
-governance
-contractual arrangements
-contract forms and types
-collective action
Household factors
-land assets
-non-land assets
Policy environment
-policies
-programme
-projects
Households
Channel choice 1
Channel choice 2
Figure 1: Market channel choice framework
emerging black farmers, who have been subjected to
double exclusion. Firstly, by the country’s colonial legacy
and secondly, they were excluded on the basis of their
production capacity. Research has shown that there
are few black farmers who participate in modern market
channels (Kirsten and Sartorius 2002).
Conceptual framework
The conceptual framework presented in Fig.1 shows the
key factors which inform householders’ decisions on
whether to participate in a market channel or not. The
study hypothesises that there are three decision-making
blocks which are vital to market channel decisions: the
policy environment, the market environment and the
household environment. In each block, there are several
key variables which affect (positively/negatively) the
household on whether to supply a specic channel or
not. It should be noted that these are people who already
have access to land and irrigation.
Survey results
• Household characteristics. This section presents
an analysis of the socio-economic characteristics
based on a household survey of 220 tomato growers
in Mopane and Vhembe districts in Limpopo Prov-
ince. These two districts are responsible for up 70%
of tomatoes produced in the province. The survey
results showed that the majority of the households
are male headed (over 60%) with the average age
of the household head being 54 years. The majority
of the respondents are full-time farmers although
there are a considerable number of respondents
who are either engaged in formal employment or
private businesses. On average the household
heads have low levels of education with an average
of seven years of formal education. About 30% of
the respondents have formal training in agriculture
- mostly in crop production.
• Household endowments. Agriculture is the main
source of income (84%) among the survey re-
spondents, however there is a signicant number

18 Smallholder and agro-food value chains in South Africa
of households that consider government grants as
their main source of income. The average farm size
is three hectares, ranging from two to 50 hectares.
Few households own or have access to production
implements, less than 30% of the farmers own a
tractor and less than 20% own a vehicle. Few re-
spondents (less than 5%) have access to relevant
production infrastructure for tomatoes such as
greenhouse and packhouse. The majority of the re-
spondents have access to irrigation although some
households use rudimentary irrigation systems
(buckets).
• Tomato production and marketing. In the survey,
the average tomato production area is 3.55ha and
the average tomato yield is 19 tons/ha. Farmers
supplying to the agro-processors have the highest
yield of 26 tons/ha, whereas farmers supplying to
the local market have the least yields (14 tons/ha).
Agro-processors have the highest total tomato pro-
duction (170 tons), followed by those supplying to
hawkers (87 tons) and fresh produce markets (72
tons) whereas farmers supplying local markets have
the least production (32 tons). There are no farmers
producing tomatoes in the rst two months of the
year. Tomato production has two distinct peaks
around May and in September. On average, each
farmer produces two cycles of tomatoes.
The household survey shows that there are six main
tomato marketing channels in the study namely. retail
chains, agro-processors, wholesale markets (national
fresh produce markets), traders, hawkers and local mu-
nicipal markets. Over 70% of the farmers interviewed did
not have a xed marketing channel, most of them sup-
plied to more than one market and, in some cases, up to
four channels.
According to Table 3, the majority the farmers supply their
tomatoes to hawkers (30%) followed by those supplying
to agro-processors (23%). Wholesale Markets (Johan-
nesburg Fresh Produce Market) and the local markets
have the least number of respondents supplying them.
Over 70% of the respondents supply to more than one
marketing channel: in some cases individual farmers
supply up to four market channels. Respondents market
their tomatoes throughout the year, although there is a
signicant variation across households. The majority
of the farmers (more than 90%) supply their tomatoes
to the markets as individuals and less than 5% supply
collectively.
Determinants of market
channels
The results of an econometric estimation show that loca-
tion, education, farm size, greenhouse, market channels,
collective action, supermarket proximity, mobile phone
and irrigation type are signicant determinants of market
channel choice among smallholder tomato growers in the
study area.
• Location. The result on location implies that house-
holds in a specic geographic region have a higher
likelihood to participate modern market channels
compared to households elsewhere. In this case,
a household located in Vhembe district is more
likely to participate in modern market channels
compared with a household in Mopane District. A
possible explanation is that Vhembe district hosts
more agribusiness rms which are active in the to-
mato value chains compared to those in Mopane
district. This is evidenced by the fact that Vhembe
district is home to three quarters of the tomato
factories in South Africa, including a Tiger Brand
factory, which is the biggest tomato processor in
the country. One possible explanation may be the
location of distribution or marketing infrastructure
such as a packhouse, distribution centre, collection
depot or a factory. This has direct implications on
transaction costs incurred by farmers. If an area
has such infrastructure this will reduce transaction
costs incurred by farmers especially in relation to
transport and search costs.
• Education level. The results of this model indicate
that household heads with higher education levels
will better understand procurement demands set
by modern market channels, especially supermar-
kets and agro-processors. Food safety and quality
standards as well as contractual arrangements with
agribusiness rms require certain levels of literacy.
Given the low level of literacy among the respond-
ents (averaging seven years) education may be-
come a critical factor in market channel choice, with
the less educated individual likely to prefer informal
market channels, where transactions are a simple
handshake characterised by little or no paper work.
Education among farmers can also propagate
information asymmetry among the farmers with
those with better education being better placed in

19
Smallholder and agro-food value chains in South Africa
Table 1: Farm assets and equipment
Item N Percent
Transport:
Car 76 34.55
Lorry 18 8.18
Tractor 69 31.36
Infrastructure:
Greenhouse 3 1.36
Packhouse 1 0.45
Irrigation access: 215 97. 9
Drip 100 45.45
Sprinkler 8 3.64
Furrow 111 50.45
Bucket 1 0.45
Table 2: Tomato production
Area (in Ha) Yield (in kg per ha) Production (in kg)
Average 3.55 19 504.05 78 497.07
Supermarkets 3.83 15 873.00 54 738.00
Agro-processors 5.76 26 481.88 170 786.88
Wholesale market 3.72 16 780.43 72 658.70
Hawkers 3.98 25 619.17 87 551.67
Traders 3.06 17 333.50 64 501.50
Local wet market 1.86 14 105.36 32 718.64
Table 3: Market participation (percentage of households)
% (N=220)
Supermarkets 17
Agro-processors 23
Fresh Produce markets 8
Hawkers 30
Traders 12
Local market 10

20 Smallholder and agro-food value chains in South Africa
accessing and comprehending market information.
This is very signicant in the fresh produce sector
where there are strong price variations both spa-
tially and temporally.
• Farm size. The results on farm size show that
farmers with relatively bigger farms are more likely
to participate in modern market channels, because
they can produce large volumes. In this instance,
they are likely to have the capacity to produce suf-
cient quantities that can meet the quantity/and
supply consistency demanded by modern markets
(agro-processors and supermarkets) as well as
fresh wholesale markets. From a transaction cost
perspective agribusiness rms (supermarkets and
agro-processors) are likely to incur lower transaction
costs when procuring from farmers with relatively
large farms than those with smaller farms. Procure-
ment from smaller farmers will increase coordina-
tion costs incurred in administering supply from a
large pool of small farms compared to procuring
from a few large farms. Agribusiness rms usually
have a minimum threshold in terms of quantities
which they can procure from a farmer to address
the coordination cost problem. For example, Tiger
Brands will not buy anything less than one ton of
tomatoes and the same applies to supermarkets.
• Greenhouse. Results indicate that farmers with ac-
cess to a greenhouse (owning or renting a plastic
tunnel) are more likely to participate in modern
market channels than those without access to a
greenhouse. A possible explanation is that farmers
with greenhouses/tunnels can produce throughout
the year as well as improve the quality of their fresh
produce. Formal markets channels (supermarkets
and agro-processors) require stock on their shelves
or processing machines any time of the year hence
they require farmers to supply them consistently
throughout the year. Although supermarkets do
draw on local production more during peak season,
there remain issues with produce quality and con-
sistency of supply. Access to a greenhouse/tunnel,
among other kinds of infrastructure (e.g. pack-
houses), allows a farmer to manage production risk
as they are able to produce quality products all year
round with controlled exposure to natural elements
like hail, storm, evaporation, pests and rodents
which are detrimental to production. In this regard,
farmers with greenhouses can consistently meet
agribusiness requirements for supply consistency.
• Market portfolio. There are a number of market
channels to which farmers supply their tomatoes.
The results indicate that farmers supplying to formal
market channels are likely to adopt a diversication
strategy in order to deal with price risk. The market
portfolio of the farmer relates to prices, suggesting
that supplying more market channels is likely to
produce higher prices on average. Despite the
absence of time series data on prices, the survey
results show that although tomato prices are higher
in modern markets they also uctuate signicantly
more than in traditional markets. Hence, risk-
averse farmers are likely to adopt a diversication
strategy which includes selling to both traditional
(e.g. roadside markets) and modern markets (e.g.
agro-processors).
• Farmer collective organisations. These are a pooled
variable representing different forms of cooperation
amongst farmers. Four types of farmer organisa-
tions are presented in this study area, namely, agri-
cultural cooperatives, irrigation schemes, collective
marketing and commodity associations. The results
suggest that collective action allows farmers to gain
economies of scale and bargaining powers which
minimise transaction costs incurred in supplying
modern market channels. Collective action ad-
dresses two main market access limitations. Firstly,
it reduces transaction costs such as transport,
search, negotiation and administration costs faced
by farmers in supplying agribusiness supply chains,
especially to supermarkets and agro-processors.
Secondly, collective action through farm organisa-
tion plays a coordination role between rms and
farms. The coordination role is very important espe-
cially when it comes to the transmission of private
procurement standards between rms and farms.
The coordination role is also important in matching
the procurement requirements by agribusiness and
production capabilities by farmers.
• Irrigation technology. This variable is a dummy vari-
able with ‘one’ representing sophisticated irrigation
technologies, such as drip and sprinkler systems
and ‘zero’ representing traditional irrigation tech-
nologies like furrow and bucket systems. The results
show that farmers with high technology systems can
produce tomatoes throughout the year to allow them
to meet the supply consistency requirements set by
modern markets like retailers and agro-processors.
Reliable irrigation is a pre-requisite for securing a

21
Smallholder and agro-food value chains in South Africa
Table 4: Tomato prices (in Rands per kg)
Average year round prices In-season price Off-season price
Supermarkets 0.83** 0.95 0.57
Agro-processors 0.83* 0.62 0.47
JFPM 0.90* 0.91 1.04
Hawkers 0.89 0.95 0.89
Traders 0.90 0.81 0.71
Local wet market 0.69** 0.95 1.13
Average 0.85 0.81 0.86**
* (P<0.10) =10% signicance level
** (P<0.05) =5% signicance level
*** (P<0.01) =1% signicance level
fresh produce supply contract from agribusinesses.
Ownership of drip irrigation and sprinkler irrigation
allows farmers to water their crops efciently and
consistently. Advanced irrigation technology, there-
fore, allows farmers to grow tomatoes over larger
tracts of land enabling them to produce more quan-
tities consistently.
Net income analysis
Table 4 below, shows the price offered to farmers by dif-
ferent market channels in both peak and off-peak tomato
seasons. On average, local wet markets and the fresh
produce markets offer the highest prices. During the
season, supermarkets, local wet markets and hawkers
offer the highest prices. During the off-pick season,
local markets offer the highest prices. Price uctuations
are greatest with the local wet markets, supermarkets,
agro-processors and the national fresh produce markets
(FPMs), hawkers and traders have relatively stagnant
prices. There is no signicant difference across the cat-
egories on the average prices, however, during the off
season as prices uctuate signicantly across different
market channels.
Table 5 below, shows a comparison of the net incomes
realised by tomato growers supplying to different mar-
keting channels in the areas studied here. Farmers sup-
plying to hawkers have the highest income per hectare,
whereas farmers supplying to supermarkets have the
least net income per hectare. Overall, the net income
analysis suggests that farmers supplying to traditional
market channels receive higher incomes than those
supplying to modern market channels. The local wet
market also performs a vital role in stabilising farmer
incomes in the off season.
Summary and discussion
of results
The results presented in this paper showed that commer-
cial smallholder farmers have multiple market options
available and accessible to them. These can be catego-
rised into modern and traditional markets. The survey
results show that modern food market channels are not
popular among farmers, and that they prefer supplying
to traditional markets such as hawkers and traders. The
survey results also show that there are price differentials
across the different market channels, and that price uc-
tuations are greatest with the local wet markets. There
is no clear pattern on the price differential between
the modern and traditional market channels. Local wet
markets have relatively higher prices at both peak and
off seasons. The net income analysis suggests that
farmers supplying to traditional market channels receive
higher incomes than those supplying to modern market
channels.
The analysis presented in this paper shows that tra-
ditional markets are conducive for the majority of the
respondents. Participation in modern market channels
demands a threshold investment in relevant production
infrastructure such as greenhouses and irrigation tech-
nology as well as adequate production land. Given poor
yields, inferior quality and production risks, traditional

22 Smallholder and agro-food value chains in South Africa
Table 5: Tomato incomes (in Rands)
Net income
(Rand per ha8)
Supermarkets R26 623.20
Agro-processors R44 061.56
JFPM R32 447.28
Hawkers R54 026.67
Traders R36 873.50
Local wet market R33 935.84
Modern market channels* R34 377.34
Traditional market
channels*
R41 612.00
* average
market channels are more relevant to the majority of
the smallholder farmers in the study area. To shift the
existing marketing patterns towards modern markets
requires adequate investments for organising farmers as
well as production infrastructure.
References
Botha L and van Schalkwyk HD (2006) ’Concentration
in the South African food retailing industry’, paper pre-
sented at the 44th Annual AEASA conference in Graham-
stown, South Africa, 20 – 22 September 2006.
Huang J and Reardon T (2008) ‘Keys to inclusion of
small-scale producers in dynamic markets: Patterns
in and determinants and effects of farmers’ marketing
strategies in developing countries.’ Regoverning Markets
Project, IIED: London.
Humphrey J (2007) ‘The Supermarket Revolution in
Developing Countries: Tidal Wave or Tough Competitive
Struggle?’ Journal of Economic Geography 7: 3.
Kirsten J and Sartorius K (2002) ‘Linking Agribusiness
and Small-scale Farmers in Developing Countries: Is
There a New Role for Contract Farming?’ Development
Southern Africa 17: 4.
Louw A, Chikazunga D, Jordaan D and Bienabe E (2007)
‘Restructuring food markets in South Africa: Dynamics
within the context of the tomato sub sector’. Regoverning
Markets Project, IIED: University of Pretoria.
Madevu H (2006) Mapping the competitive food chain for
fresh produce: The case of retailers in Tshwane Metro,
South Africa. (MSc Thesis), Department of Agricultural
Economics: University of Pretoria.
Pote Peter PT, Ajuruchukwu O and Fraser G (2007) ‘Tech-
nical Constraints to Smallholder Development in South
Africa: Implication for Market Access and Food Chain
Management’, presentation to the International Food
and Agribusiness Management Association, Parma, Italy,
June 23-26.
Reardon T and Berdegue J (2002) ‘The Rapid Rise of Su-
permarkets in Latin America: Challenges and Opportuni-
ties for Development’, Development Policy Review 20(4):
317-34.
Reardon T and Timmer C (2006) ‘The Supermarket Revo-
lution with Asian Characteristics’, in A Balisacan and N
Fuwa (eds) Agricultural and Rural Development in Asia:
Ideas, Paradigms, and Policies Three Decades Hence.
ISEAS/SEARCA: Singapore/Los Baños.
Weatherspoon D and Reardon T (2003) ‘The rise of super-
markets in Africa: Implications for agro-food systems and
the rural poor’, Development Policy Review 21(3): 1-17.

23
Smallholder and agro-food value chains in South Africa
Sustainable policy
support for smallholder
agriculture in South
Africa: key issues and
options for consideration
Andre Louw
Introduction
Like many other sub-Saharan African countries, South
African agriculture is dualistic in nature with two sectors
existing parallel to each other - the small-scale farming
sector on the one hand and the commercial farming
sector on the other. The small-scale farming sector is
generally characterised by small farms that use labour-
intensive traditional production techniques and lack insti-
tutional capacity and support, whereas commercial agri-
culture is typied by farms with relatively high turnovers
that use capital-intensive modern production techniques
and have links with key input and output markets (Green-
berg 2010; Mudhara 2010). With dualism inherited from
apartheid, the small-scale sector has predominantly
black farmers while the commercial sector comprises
chiey white farmers, hence it is no surprise that the
post-apartheid government has, in the past eighteen
years, been focused on trying to establish and develop
the small-scale sector to a more commercialised small-
scale farmer level.
Policy directives, such as land reform and preferential
procurement schemes, have been made with the aim of
redressing the divide between the small-scale sector and

24 Smallholder and agro-food value chains in South Africa
the commercial sector. However, despite their efforts,
many small-scale farmers are still lagging behind the
commercial farmers. Vink and van Rooyen (2009) report
that this divide seems to be growing. This raises the
questions: What are the issues that policy fails to take
cognisance of to direct the sustainable development
of the small-scale sector? What are the key aspects
to be considered for sustained policy support for the
development of the small-scale sector? Are certain
social and cultural aspects playing a major role in this
transformation process?
To answer these questions, this article discusses
observations on the developments and operations in the
agriculture and agribusiness sector in South Africa and
also speculates on the implications to the small-scale
farming sector. In discussing the latter implications,
the importance of understanding the heterogeneity and
complexity of small-scale production is highlighted.
The heterogeneity that exists in the small-scale farming
sector is a reection of differing farming objectives. The
reasons for production of smallholder farmers vary from
making a partial contribution to the food requirements
of their households and providing a wider range of
livelihoods, to generating a primary source of income.
From this understanding, van Averbeke and Mohamed
(2006) identify three different types of smallholder
farmers:
• Subsistence farmers. These are farmers whose main
farming objective is for household consumption and
there is very limited sale of produce as they rarely
produce surpluses and sometimes not even enough
for their own consumption. These farmers make up
the majority of the small-scale farmers.
• Emerging or smallholder farmers. These are the
farmers who have the desire and are increasingly
working towards commercialising their production.
• Commercial smallholder farmers. These are the
farmers whose main objective is to earn income
from the sale of their produce. They constitute the
minority of the small-scale farmers.
These different groups of small-scale farmers commonly
have been viewed as depicting different stages of a
trajectory from subsistence-to-small-scale, emerging-
to-small-scale commercial, and ultimately commercial
farming. Entrenched in this view is the idea of what
makes a smallholder successful. Successful small-scale
farmers have been portrayed as productive farmers who
are actively engaged in marketing their produce and
earning sufcient cash income, primarily from agriculture,
to live a poverty-free lifestyle. Therefore, on this basis,
the attainment of small-scale commercial farmer status
represents success, in most cases assisted through well-
structured support and extension programmes - which
is the goal of many policy directives. Both subsistence
and emerging farmers seem to have a ‘permanent’ role
to play in South Africa and Africa in a rather diverse
production structure where they are comfortable in their
current situations.
Figure 1 below, shows the different groups and the
possibility of progressing or graduating from one
level to the next, i.e. from subsistence to commercial
farming. But in reality, progression from subsistence to
commercial is often not guaranteed due to preferences
or circumstances. Although the the gure shows that
subsistence farmers generally lack capacity, certain
circumstances would appear to be the most effective in
preventing progression to commercial farming.
Also shown on the gure is the operational diversity of
commercial small-farmers with varying levels of net
worth, creditworthiness, management experience, sup-
port and mentoring.
One also has to separate the term ‘small-scale’ commer-
cial farmers from true commercial farmers, which were
referred to earlier. The issue is not only about size or the
relative intensity of management, it is about turnover. The
small-scale commercial farmer might have a turnover of
R100 000 per annum but he produces for a market and
has a decent living. In fact, according to Hall (2008) the
majority of commercial farmers are in the lower level
of the commercial group with more than 50% having a
turnover of less than R300 000 per year.
Key observations from
the agriculture and
agribusiness sector
The agriculture and agribusiness sector, on the whole,
has been affected by factors that are within and outside
the sector. The main factors within are related to the
changes in supply chain management and in overall

25
Smallholder and agro-food value chains in South Africa
input and output markets. The most noteworthy exog-
enous factors however, are natural events (e.g. climatic
conditions) and macro-economic performance factors
(e.g. population increases, exchange rate uctuations,
climatic factors, etc).
• Structural reforms. Following the end of apartheid
and its related sanctions in 1994, South Africa be-
came integrated into the global economy and this,
among other reasons, contributed to the structural
changes that the country has undergone since then.
After the removal of marketing boards and market
controls, the fundamental structural change was
the shift from stagnant state-controlled agricultural
markets to a vibrant and open-market orientation.
Consequently, these developments have led to a
shift from low-value basic food crops to high-value
products intended for the export market. While the
commercial sector was geared for this develop-
ment, emerging smallholders became exposed to a
new environment with no prior experience and no
government support.
The World Wildlife Fund (WWF 2010) reported that with
a focus on the export market, South Africa became a
net importer of certain important food types e.g. wheat.
In fact, South African food markets became exposed
to cheaper imports, supplied at below local production
cost, from developed countries with large farm support
programmes. However, the local smaller farmers did
not have the capacity to compete with these subsidised
products.
• Increased risk in agriculture production. The re-
moval of the marketing controls and participation in
export markets left farmers exposed to uctuations
in the free market systems of the domestic and
international markets. These uctuations increased
the risk for agricultural production, especially when
changes in the market were not particularly in fa-
vour of farmers. In most cases, when uctuations
have a negative effect, small-scale farmers are the
hardest hit, for reasons that include lack of risk
assessment and mitigation capacities. However,
even with positive effects, small farmers also tend
Figure 1: The different groups of small-scale farmers
Management
experience
Net worth
Level of
creditworthiness
Access to credit
Access to markets
Access to info
Access to equity
Ability to cope with
risks
Government
involvement
Pre-screening/
beneciary selection
Support/aftercare
required
Nebtoring required
Subsidies/lower
interest requirement
Level of risk support
Project management
skills
I II IVIII
Subsistence
farmer
Emerging
farmer
Commercial
farmer
Categories on farming (subsistence to commercial)
Source: Louw (2008).

26 Smallholder and agro-food value chains in South Africa
to miss out on the benets because of lack of ac-
cess to information and capital to benet from the
opportunities.
Additionally, agriculture has also experienced increased
risk from extreme climatic conditions such as droughts
and oods. These events intensify the risk for crop dam-
ages and livestock losses and while commercial farmers
have formal and/or informal risk mitigation strategies in
place to cope, it is not in the culture of the small-scale
farming sector to provide for and manage these weather
anomalies. They do not always understand the concept
of risk mitigation and require some support and capacity
building to save for a bad season, to diversify, to cope with
sound cash ow management and also with information.
• Growing population and modernisation. The South
African population is growing at approximately 2%
per annum (WWF 2010), thus implying that food
supply must increase using the same or fewer re-
sources. Post-apartheid reforms have given rise to
increased wealth due to the expansion of the coun-
try’s middle class (measured by the LSM – living
standard method) which reportedly rose by about
30% between 2001 and 2009 (BFAP 2011). Most of
this middle class reside in the urban areas and lead
modern lifestyles resulting in changing consump-
tion patterns. However, while the agrifood market
has been restructured and modernised to meet
changing demands, farm-level restructuring has
been left behind. Farm-level modernisation requires
some infrastructural improvements, inter alia roads,
adequate water supplies, irrigation, greenhouses,
cooling tanks, etc. Some of these improvements
require on-farm investments, which emerging
small-scale farmers are unable to make, particu-
larly due to limited income or access to credit. Other
improvements, such as road and communal irriga-
tion infrastructure, are generally the responsibility
of government. However, being located in the rural
areas, where infrastructure development is limited,
commercial small-scale farmers often experience
difculty in participating in the modern markets as
a result.
Intensification of farming
According to the WWF (2010), South Africa has less
than two-thirds of the commercial farms it had in the
early 1990s. Many of the farms have merged into larger
farming units to achieve economies of scale. However,
production increased after deregulation (Sandrey et
al 2011) indicating a shift towards large-scale, high
external-input intensive farming. As such, there was in-
creased use of irrigation, fuel, fertiliser, mechanisation
and genetically modied (GM) seed. Fertiliser and seed
companies have moved in to ll the void left by poorly-
resourced government extension services to provide
private advisory services. These companies tend to pro-
vide their own extension personnel to build relations and
maintain contact with farmers thus creating a service to
sell and promote more of their products.
Closely linked to these developments is the growth of the
food retail industry. The main model of choice for retail
companies is contract farming or off-take agreements.
Contract farming involves the provision of inputs and
technical advisory services in exchange for a specied
quality and quantity of output. This has led to increased
concentration of input and output companies relative to
farmers in value chains. However, as the supply chain is
skewed towards these agribusiness and food retail com-
panies, this has brought about some power imbalances
due to the companies getting a larger share of gains.
Key policy messages
While some of the following policy messages are not en-
tirely new (e.g. notions such as infrastructure investment,
facilitation of improved access to credit markets and ac-
cess to market information), it is essential to emphasise
the importance of these factors to ensure sustainable
policy development. A particular focus on the smallholder
or emerging farmer means there is need for continued
specic policy directed to this group of farmers, such
as credit at concessionary rates and a more concerted
effort in training and capacity building of the farmers.
Despite already knowing what needs to be done ,there
is a growing realisation that new ways of thinking have
to be applied to successfully come up with sustainable
approaches.
• Understanding the heterogeneity in the small-
holder sector. Understanding the variation in the
smallholder sector is essential. It calls for a differ-
entiation in the policy process for the distinct types
of small-scale farmers. Clearly, with special needs
for the different groups of smallholder farmers
a uniform solution would not work in addressing
such heterogeneous needs. Understanding these

27
Smallholder and agro-food value chains in South Africa
differences prompts other questions for the policy
makers: Should all subsistence farmers be made
to migrate to become commercial farmers? Is this
what all smallholder producers want? Can a farmer
remain on a subsistence level and yet still be suc-
cessful? What are or should be the policy objectives
at each level of the small-scale sector? What are the
types of support systems required at the different
levels? It should be understood that subsistence
farmers do not necessarily want to become small-
scale commercial farmers and are happy with their
current circumstances. They may be successful,
given their own level of needs and requirements,
and. can therefore not be forced to migrate from
subsistence farming - where they are in a comfort
zone and happy – yet they might still require some
basic support and infrastructure. The needs must
determined for all the different levels at policy
level, The support at basic level could be for vital
extension services, infrastructure, the provision of
transport, schools, water, sanitation etc. For the
small-scale emerging commercial farmer, however,
support systems could range from the provision
of an improved extension service, supply of water
for irrigation, access to cheaper capital, access to
markets, more market information, assistance re-
garding packaging houses and collection points, as-
sistance with forming cooperatives, information on
risk management, and other management, product
and marketing support.
• Development of inclusive business models. The in-
volvment of smallholder producers is on the national
development agenda and for business to contribute
towards the national agenda they will be required to
work with the smallholder farmers. However, there
is a need to consider different and sound busi-
ness models based on a true understanding of the
smallholder sector, the markets and the roles of all
the players in the value chain (including that of the
public sector). Procuring from smallholder farmers
has always been a challenge for large businesses,
in terms of organising supply, quantity, consistency,
quality, safety and traceability because to ensure all
these involves high transaction costs. Successful
procurement from smallholder farmers has often
involved working with specialised intermediaries
who understand both smallholder farmers and the
agribusiness industry. For this reason, reinforcing
rather than ‘cutting out the middleman’ may be
the most sustainable strategy. Intermediaries have
generally been found to be key in bridging the re-
alities of small-scale production and modern organ-
ised markets.
For their part, the smallholder farmers believe that col-
lective action remains important for their increased par-
ticipation in dynamic markets, although existing producer
organisations have a mixed record for providing members
with access to the markets. Cooperatives are theoretically
an option, but much more capacity building and support
is required to make them successful and sustainable.
More government assistance is required especially in the
case of new cooperatives where the relevant experience
and management capacity is lacking and control systems
are not in place. Producer and commodity organisations
can, however, make use of policy inuence on business
strategy as well public policy to drive their agendas.
• Collaborative arrangements. Although business
models can range from relatively simple changes
to procurement policies through joint ventures, the
use of multiple actors from both public and private
sectors has proved to be a successful foundation.
Working with many stakeholders can open up space
for dialogue, build an understanding of agro-food
market trends and drivers, develop future scenarios
and dene entry points for action. Multiple stake-
holders interact directly and indirectly to shape the
structure of modern agro-food market chains.
The role of the public sector and the private sector is
changing towards facilitating business partnerships be-
tween smallholder farmers and other actors in the supply
chain. What is required is a set of new policy skills and
instruments for government to carry out economically
rational and socially valuable market interventions in the
supply chain.
New arrangements to enable dialogue among public
policy, business, support services and the farming com-
munity are also required to secure fairness in trade and
sustainable, inclusive agribusiness. These arrangements
would ideally involve collaborations between:
i. trained and organised farmers – who understand
processes and contribute meaningfully towards
dialogue;
ii. a receptive business sector – that anticipates the
benets and exhibits willingness to work with small-
holder producers;

28 Smallholder and agro-food value chains in South Africa
iii. conducive public policies and programmes – that
create an enabling environment for business and
smallholder farmers to work together.
iv. good research, extension and support services – to
provide capacity on above initiatives for farmers.
At the core of these collaborations is the need for lead-
ership and specialised partnership facilitation that sup-
ports arrangements between the groups and ensures
their objectives are met.
Conclusion
Changes, particularly in the agricultural markets in South
Africa, have tended to have a more negative than posi-
tive impact on small-scale producers. This reinforces the
need for continued policy efforts to improve the small-
scale sector. However, to successfully achieve this, a new
way of thinking is required. This includes an in-depth
understanding of the realities and complexities within
the small-scale sector before embarking on any specic
policy options. The participation of both private and
public sector is advised, thus signalling a shift from the
traditional allocation of the role of development solely
to government. Finally, the role of middlemen in linking
small-scale farmers to the markets remains important.
References
Bureau for Food and Agriculture Policy (BFAP) (2011)
‘Baseline Report’, BFAP: University of Pretoria.
Greenberg S (2010) ‘Contesting the food system in
South Africa: Issues and opportunities’, Research
Report no. 42, Institute for Poverty, Land and Agrarian
Studies: Cape Town.
Hall R (2008) ‘Dynamics in the commercial farming
sector’, technical report, PLAAS: Cape Town.
Louw A (2008) ‘Policy options for more sustainable
smallholder agriculture support’, Paper presented at
conference on Inclusive Business in Agro-food Markets:
Evidence and Action Conference, (March 5-6), Beijing.
Mudhara M (2010) ‘Agrarian transformation in
smallholder agriculture in South Africa: A diagnosis of
bottlenecks and public policy options’, paper presented
at conference on Overcoming inequality and structural
poverty in South Africa: Towards inclusive growth and
development Conference, Johannesburg, September
20-22.
Sandrey R, Punt C, Jensen HG and Vink N (2011)
‘Agricultural trade and employment in South Africa’,
International Collaborative Initiative on Trade and
Employment, OECD Trade Policy Working Paper, no. 130,
OECD Publishing .
Van Averbeke W and Mohamed SS (2006) ‘Smallholder
farming styles and development policy in South Africa:
The case of Dzindi irrigation scheme’, Agrekon 45: 2.
Vink N and van Rooyen J (2009) ‘The economic
performance of agriculture in South Africa since 1994:
Implications for food security’, Working Paper no. 17,
Development Planning Division, DBSA: Midrand.
World Wildlife Fund (WWF) (2010) Agriculture: Facts
and Trends South Africa. WWF: awsassets.wwf.org.za/
downloads/facts_brochure_mockup_04_b.pdf

29
Smallholder and agro-food value chains in South Africa
Big Business for Small
Farmers: the Case of
Venda Avocado Growers
Lusito D Khumalo
Abstract
Inclusive business models in South Africa are possible
through the identication of a strategic ‘t’ between
large agribusinesses such as Westfalia and smallholder
farmers. A complementary relationship was found in an
early market window when small-scale avocado farmers
in Venda aligned their production activities with West-
falia, one of the largest avocado producing and exporting
companies in South Africa. The farmers had a competi-
tive advantage in climate by supplying the rst fruits of
the South African season, thus playing an important role
in the reduction of imports from Spain during the off-
season. A combination of factors was instrumental in the
success of one of the ten farmers that participated in the
project. Economies of scale, efciencies in logistics and
high fruit quality were found to be key success factors.
Additional criteria included the young age of the farmer,
his low risk aversion, early adoption of new technology
and a passion for farming. A business model was devel-
oped from the successful farmer indicating the potential
for more smallholder farmers to prot from participating
in big business in the South African avocado industry.

30 Smallholder and agro-food value chains in South Africa
Introduction
Poverty is a dominant feature of small-scale agriculture
in Africa caused to a large extent by resource constraints
and technology stagnation (Ghatak and Ingersent,
1984). Since the dawn of democracy in 1994 many
attempts made by the South African government to im-
prove the level of participation of smallholder farmers in
mainstream agriculture have met with limited success.
One example is a commercialisation model, initiated by
government in 1995, which consolidated smallholdings
into larger economic units with a government-appointed
contractor to prepare and plant the land on behalf of the
farmers. The objective was to realise economies of scale
with the condition that the service provider transfers his
farming skills over time to the farmers. This model failed,
rstly because it reduced farmers to mere observers of
operations on their farms; and secondly, the huge cost
of these external operators on government made them
neither justiable nor sustainable in the long term.
Changes in policy that led to market liberalisation and
scal and governance changes eliminated support to
small farmers including the removal of subsidies (Hazell
& Diao, 2005:29; Vermeulen et al 2008).
The functioning of the agro-food chains was affected
by the dismantling of state-led enterprises (such as
marketing boards) due to privatisation and deregulation
policies. In addition, the chains were affected by the intro-
duction of new organising principles that install alterna-
tives for monopolies or oligopolies (Vellema 2011). Prior
to deregulation, farmers had only to concern themselves
with on-farm production and were not involved after the
farm gate closed. Beyond that, the co-operatives were
responsible for collecting, transporting and marketing
members’ produce. Similarly, the procurement of farm
inputs such as fertiliser, seed and chemicals was the
co-op’s responsibility, which bought inputs in bulk for all
members at lower prices by exercising bargaining power.
Although subsistence agriculture plays an important
role in food security, the need for a quick transition by
smallholders to a commercial level has become a priority
if they are to become sustainable in the long term. On-
farm income derived from small-scale avocado produc-
tion plays a signicant role in ensuring the sustainable
livelihoods of farmers in Venda. A recent study comparing
small-scale irrigated and dry-land farming in Venda found
that 83% of food-secure households had irrigated farms
compared to 53% under dry-land production (Oni et al
2011). This implies that small-scale farming contributes
positively to household food security and irrigation has a
positive effect on farm output. For instance, family deci-
sions such as adding another family member or sending
a child to school/ University depends on the nancial
performance of the avocado orchard. A farmer who de-
pends on the performance of an orchard for their entire
livelihood must have sufcient land and good yields to
secure a gross farm income to cover the costs of produc-
tion, farm debt and the living requirements of the family
(Whalberg, 1940).
The South African avocado industry has a market value
of about R400 million with the South African Avocado
Growers Association (SAAGA) as the lead association
representing 85% of all producers, small and large, in
the country (Perishable Products Export Control Board
(PPECB) 2009). Westfalia Fruit Estates has a 50% market
share of the avocado industry in South Africa. It is argu-
ably the largest producer and exporter in the country
with a vertically integrated business holding structure of
farming operations, processing and marketing compa-
nies. Small-scale farmers, some of whom are members
of SAAGA, account for less than 1% market share and
declining (Radzilani 2011).
This paper provides evidence of a private-sector led initia-
tive of inclusive business between a large agribusiness
rm and smallholder farmers. The following sections
discuss the research methodology conducted in the
study, and then describe the chosen study area and the
capabilities of the area in supporting the development
of pro-poor value chains. This is followed by a discussion
of the results focusing on the nancial feasibility of the
enterprises, and nally a summary and conclusion is
presented.
Research methodology
In 2009, Westfalia initiated a smallholder farmer devel-
opment programme aimed at the strategic inclusion of
small farmers along the avocado value chain. This was
given impetus by an increasing demand for seedling trees
by smallholders from communal farms in Venda at the
Westfalia nursery. Twenty years ago, Westfalia started
holding farmers’ days to share technical information on
crop husbandry and marketing - a relationship which
provided the farmers with market intelligence currently
lacking in the public extension service. The aim was to
address the industry problem of immature fruit landing

31
Smallholder and agro-food value chains in South Africa
on the market oor at the Johannesburg Fresh Market
Market (FPM) in the early part of the season. Most of this
fruit comes from resource-poor growers in Venda whose
poor quality fruit cannot compete with commercially
farmed fruit, so to fetch higher prices, they are tempted
to harvest immature fruit to be rst in the market at the
start of the new season.
To make these growers more competitive by improving
their fruit quality, and to demonstrate the nancial benet
of harvesting fruit later in the season rather than earlier,
the consumer dimension was introduced to their farming:
to always focus on meeting the needs and expectations
of the end consumer.
Appendix A below shows two supply chains. The pre-
intervention model indicates the situation prior to the
project: the traditional supply chain of the farmers,
which is characterised by many middlemen and a high
cost structure. However, the shortened post-intervention
value chain indicates the benets to smallholders when
Westfalia eliminated the ve middlemen and substituted
them with three of its internal divisions to perform func-
tions for maximised prot. Small farmers gained from the
improved efciencies and reduced costs after integrating
the manufacturing functions during the transformation of
raw material into high value products.
The biggest gain for the growers was the certainty of
producing for a market – an especially important factor
for pre-intervention farmers - after following a ’supply-
push’ model of growing for an unknown market. In the
post-intervention model, farmers were introduced to
the ‘demand-pull’ model where product was procured in
a contract arrangement. It is estimated that 78.5% of
the annual production of fruit and vegetables sourced by
agribusiness companies for processing is procured under
some form of contract arrangement, with the remainder
supplied through other channels such as the open
market, own estates, agents or imports (Vermeluen et al
2008). This model eliminates the marketing problem that
constrains many smallholder farmers.
A survey of 60 farmers was carried out through focus-
group interviews. A value chain analysis was then con-
ducted that revealed a number of deciencies pointing
to an immediate need to intervene at various stages
along the value chain. Technical and business training
was conducted concurrently with operations to improve
fruit quality to control blackspot, (Cercospora pupurea,)
a fungal disease of economic importance, which causes
black spots on the surface of the fruit, reducing its aes-
thetic value. The challenge of the spray project was to
improve fruit quality through applying four copper sprays
over 21-day intervals from fruit set until harvest. The im-
provement of fruit quality through chemical means was
intended to enable market access of smallholder farmers
to high-end markets. A feasibility study was undertaken
and a nancial model developed.
A business plan was then compiled with funding secured
from the Woolworths Enterprise Development Fund. The
terms were at a reduced rate of prime less 2%. Surety
and administrative control was guaranteed by Westfalia.
A contract was signed between the three parties, i.e.
farmers, Westfalia and Woolworths, committing each
partner to specic obligations. Farmers were to become
willing borrowers of the loan; Woolworths was to lend the
money and provide off-take for the fruit; and Westfalia
was to mentor and manage the repayment of the funds.
A pilot project comprising ten farmers located across
three geographic areas was then initiated. The farmers
are on communal land under the jurisdiction of a local
chief and traditional council. Each farmer has indenite
user rights in the form of a Permission to Occupy (PTO)
that requires an annual fee of R20 per hectare. This
permit gives the farmer the right to farm the land and
build a homestead. The PTO is tradable between users
and is usually transferred from parent to offspring when
the parent dies.
The rationale behind selecting different sites was to de-
termine the time of fruit maturity to assess when the fruit
arrives at the packhouse. This timing factor would also
point to a potentially protable area in which to invest
that would complement supplies from Westfalia farms.
Given the age of most of the farmers, many of whom
started farming after retirement, this change presented
a serious challenge. For an inclusive relationship to be
sustained between the parties, Westfalia needed to nd
a strategic ‘t’ along the value chain that would comple-
ment and add volume to its supply chain for both local
and export markets. A business case was made justifying
the tangible benets of prot potential and the attendant
benets of transformational agriculture in the sector.
Study area
Vhembe District (incorporating the former Venda,
parts of the former Gazankulu and parts of the former

32 Smallholder and agro-food value chains in South Africa
Soutpansberg district, bordering Zimbabwe in northern
Limpopo) has a large, concentrated population of over
1.24 million. The district has a high agriculture poten-
tial with a warm sub-tropical climate and deep well-
drained red soils that are ideal for avocado production.
Smallholdings range between one to ten hectares per
household, although it is not uncommon to nd some
families with consolidated units of up to 50 hectares.
The farming method is predominantly extensive (low or
no inputs) under dry land conditions. Production is by
default organic, since there are no synthetic inputs such
as fertilisers and chemicals used in production as the
farmers are resource-poor. As a result, output is gener-
ally low and of poor quality. Since the 1980s, the farmers
have been supplying the low end of the market (hawker
and municipal markets) due to poor quality and a lack of
market knowledge.
The introduction of an export focus created more employ-
ment opportunities in the area as inputs (capital, labour
and raw materials) were sourced locally and small farms
became more productive due to the efciency of alloca-
tions or better resource utilisation (Samen 2007). Private
sector-driven initiatives of inclusive business with small-
holder farmers provide both social and economic ben-
ets for the sector and the country as a whole. Recently,
interest has emerged among a group of investors -private
equity funders, venture capitalists, nancial services
rms - to consider “impact investment” in developing
countries (Gillam 2010).
Results
Graph 1 in Appendix B indicates that farmer no. 8 was
the only producer with a reasonable margin. Most of
the other farmers had turnovers below production cost,
indicating that their farming units were too small (i.e.
below the viable economic unit size), which resulted in
losses for those enterprises. Graph 2 in Appendix B again
shows that farmer no. 8 (Nyambeni) had the lowest costs
of production compared to the group. This was because
he produced the highest volume of 35 000 cartons and
was able, therefore, to spread his total costs over a larger
volume. This resulted in a low net cost per carton at
R8.18.
The nancial results are depicted in the Income State-
ment in Appendix C. Turnover was 26.2% less than
budget due to lower prices than anticipated. Actual
price per carton was R24.98 compared with the budget
price of R41.87. Low prices can be attributed to the high
supply of fruit in the market. Farmers need to be aware
that farming is inherently risky and one way of mitigating
price risk is through negotiations for a minimum price
guarantee with customers. In this way, a minimum gross
farm income can be estimated. A second way to mitigate
price risk can be to diversify markets to avoid depend-
ence on traditional markets. Costs of sales were 19.8%
higher than budgeted due to higher costs than antici-
pated. Gross prot was lower by 67% due to lower sales
as a result of lower prices and higher cost of sales.
Prot after tax (includes interest) the actual was lower
than budget due to higher interest incurred than what
was budgeted for. The project made a small prot as
seven of the ten farmers had a positive return i.e. their
production left little or no return as prot. Only one farmer
who had sufcient volumes was able to realise a good
prot. Two of the farmers made a loss and one was only
able to break-even. This meant a 70% success rate in the
rst year which, by all accounts, was commendable since
a prot was made in the rst year of operation. However,
since the initial aim to better their returns of their tradi-
tional supply chain was not as much as expected, there
was a contrary reaction by some of the farmers to this
outcome. The following reasons may account for the
disappointing results.
• Unusually high volumes in the market. As it was an
‘on-year’ (high yield) there was an over-supply of
fruit in the local market resulting in lower than antic-
ipated prices. However, the net income per carton
was still favourable at R17.50 (excluding waste)
yielding a good return. Net income refers to income
after all costs have been deducted, including pro-
duction, packing, transport and marketing.
• High production cost The nancial results indicate
that the spraying costs were high due to the high
cost of copper required for the effective control of
blackspot. During the initial stages when workers
were not familiar with handling the product there
was a lot of wastage.
• Low labour productivity. The project created 60
jobs for men and women that were previously un-
employed. The productivity was very low both at
initial stages during picking and packing despite
the training that was given and payment of a min-
imum wage. The minimum wage was R1 300 for an
eight-hour shift. The South African avocado harvest

33
Smallholder and agro-food value chains in South Africa
season starts from the last week of March until
the end of September and workers are seasonally
employed.
• Few trees and long distances to farm. For each of
the farms that made a loss the reasons relate to
either of the following two factors. Firstly, it was
found that some farmers produced small volumes
to offset their production costs and, as a result,
the cost of copper spray could not be recovered.
Secondly, two of the farmers were located too far
from the rest of the group and the transport costs
for the tractor and labour was prohibitive over the
long distance.
Conclusion
The avocado business has inherent commercial risk that
cannot be accurately predicted. Although the managing
partner of the project made the best possible effort to
implement the spraying exercise cost-effectively, and
sought the best possible prices for the farmers from
various markets, the result did not meet some of the
farmers’ expectations. This was despite having made a
R300 000 prot in the rst year which is unusual for new
businesses to achieve. One successful farmer had 59%
of the total volume of 300 tons supplied by the group
to Woolworths and other markets. Lessons learnt from
this research demonstrates a sound and viable business
model of inclusive business in the South African avocado
industry that can be replicated in other fruit industries.
Development practitioners, drawing up future support
programmes, need to plan sufciently before commis-
sioning projects. That entails considering the social,
physical and nancial aspects of the project. Arising from
the lessons in implementing this project was the need to
balance the various aspects with enough funding to im-
plement the whole scope of the project. Since all the ac-
tivities could not be funded, the project was scaled down
to the bare minimum which put pressure on resources
allocations to other activities.
Second, there needs to be a contingency plan. The prot
potential was decreased due to the escalation of pro-
duction costs, which included minimum wage and fuel
prices. Price uncertainty was also a factor. Transport and
logistics was found to be one of the major cost drivers of
the project. This means the selection of the target farmer
population, the proximity of the farms to the main road,
water sources, and to each other determined the cost of
farming operations and by extension the total cost of the
project.
Thirdly, on the social dimension of the project, managing
expectations of smallholders is vital. Drawing from this
experience, when farmers got involved in the project they
had high expectations. This was found to have come from
the involvement of Westfalia which created an impres-
sion that since a large player in the avocado industry was
assisting them, there would be little risk. As things turned
out, Westfalia could not absorb all the risks that the
market presented, and most of the growers felt they were
better off going back to their traditional way of farming.
The level of literacy of farmers was too low for them to
understand the complexity that commercial avocado
farming presented to them. Their comprehension of the
value chain cost factors, and how they interrelated in
the process of value creation presented a serious chal-
lenge when they reviewed their nancial statements.
Practitioners, therefore, should keep in mind the busi-
ness training needs of beneciaries along with their
understanding of the management’s responsibility for
implementing these kinds of projects.
We can conclude, therefore, that smallholder farmers
who intend to enter the avocado sub-sector should take
cognisance of the high entry barriers of xed costs pe-
culiar to this industry. The project has shown that any
farm that is less than ten hectares in size with a plant
population of less than 200 trees per hectare cannot be
viable. Furthermore, the age of the farmer, his attitude in
adopting new technology and willingness to experiment
determine the level of success. The effects of farm size
on protability were evident, indicating the importance of
economies of scale and scope as critical success factors.
These economic factors are more important than the
social factors because they determine the viability of the
enterprise. Social factors, although equally essential, can
be overcome through proper selection of farmers.
References
Gillam C (2010) Agriculture Farmland Attracting Impact
Investors, Reuters, 8 November, accessed on 13 May
2011 at: http://www.reuters.com/article/2010/11/08/
agriculture-investors-idUSN0812929920101108

34 Smallholder and agro-food value chains in South Africa
Ghatak S and Ingersent K (1984) Agriculture and
Economic Development. Harvester Press: Brighton.
Hazell P and Diao X (2005) ‘The role of agriculture and
small farms in economic development’, in IFPRI (ed)
The future of small farms: Proceedings of a research
workshop, Wye,: UK June 26-29. IFPRI: Washington DC.
Martinez S and Stewart H (2003) From Supply Push
to Demand Pull: Agribusiness strategies for today’s
consumers. United States Department of Agriculture
Economic Research Services: Washington DC.
National Agriculture Directory (2004) National
Agriculture Directory, South Africa 2004/05. Rainbow
SA: Johannesburg.
Nicol RM, Ortmann G and Ferrer SRD (2007) ’Perceptions
of key business and nancial risks by large-scale
sugarcane farmers in KwaZulu-Natal in a dynamic socio-
political environment’, Agrekon 46: 3.
Oni SA, Maliwichi LL and Obadire OS (2011) ‘Assessing
the contribution of smallholder Irrigation to house-
hold food security in comparising to dryland farming in
Vhembe district of Limpopo Province, South Africa’, Af-
rican Journal of Agriculture Research 6(10): 2287-2297.
Perishable Products Export Control Board (PPECB)
(2009) Perishable Products Export Control Board Export
Directory. PPECB: Pretoria.
Radzilani T (2011) ‘Export Market Share of Emerging
Farmers in Avocado Industry’, Subtropical Fruit Growers’
Association, Technical Advisor: Emerging Farmers. Un-
published paper.
Samen S (2007) ’Export Development and Diversica-
tion: The Role of Special Regimes, trade nance and
International Supply Chains.’ Food and Agriculture Trade
in Africa on-line course (January). World Bank Institute:
Washington DC.
Stats SA (Statistics South Africa) (2008) ‘Community
Survey 2007 Basic Results: Municipalities’ Report no. PO
301.1, StatsSA: Pretoria.
Vellema S (2011) Value chains for pro-poor development.
Wageningen University: Wageningen.
Wahlberg HE (1940) ‘Size of Avocado Orchard - Income
Factor’, California Avocado Association Yearbook 25:
91-93.
Vermeulen H, Kirsten J and Sartorius K (2008) ‘Con-
tracting Arrangements in Agribusiness Procurement
Practices,’ Agrekon 47(2): 199.

Appendix A: Pre- and post-intervention value chains
35
Smallholder and agro-food value chains in South Africa
Production Transport to
depot
Bulk trucking to
Gauteng
Deliver to
Municipal
market
Deduction of
market & agents
commissions
Proceeds Paid
to farmer
Middlemen 1 Middlemen 2 Middlemen 3 Middlemen 4 Middlemen 5
Low value
product Low or no
returns
Pre-intervention value chain
Post-intervention value chain
Production Packing &
distribution
Retail Distribution
centre/Municioal
Market
Deliver to
Municipal
market
Middlemen 1 Middlemen 2 Middlemen 3
High-value
product
Maximized
returns or
cost savings
2%
12%

36 Smallholder and agro-food value chains in South Africa
Appendix B: Production costs and turnover
Graph 1: Venda production cost versus turnover
Graph 2: Venda costs per carton versus volume produced
500 000
450 000
400 000
350 000
300 000
250 000
200 000
150 000
100 000
50 000
–
Rand
1 2 3 4 5 6 7 8 109
Actual Net Turnover Production Costs
40 000
35 000
30 000
25 000
20 000
15 000
10 000
5 000
–
35.00
30.00
25.00
20.00
15.00
10.00
5.00
–
1 2 3 4 5 6 7 8 9 10
Cartons
Rand per carton
Farmer
Costs/Carton Volume (cartons)

37
Smallholder and agro-food value chains in South Africa
TABLE 1. WESTFALIA BLACKSPOT SPRAY PROJECT
EARNINGS STATEMENT FOR THE MONTHS ENDED 30 JUNE 2010
ACTUAL 2010 % VARIANCE BUDGET 2010
GROSS SALES 1 507 -26.2% 2 044
DISCOUNT
NET SALES 1 507 -26.2% 2 044
COST OF SALES (1 045) -19.8% (838)
DISTRIBUTION EXPENSES - EXPORT
DISTRIBUTION EXPENSES - LOCAL (99) 4.9% (104)
GROSS PROFIT 364 -67.0% 1 102
OVERHEADS
OPERATING PROFIT 364 67.0% 1 102
MEDICAL AID PROVISION
SUNDRY INCOME/(EXPENSE)
PROFIT BEFORE FINANCE CHARGES
INTEREST RECEIVED/(PAID) (47) -24.8% (36)
PROFIT BEFORE GRANTS 317 1 066
GRANTS AND DONATIONS
PROFIT BEFORE TAXATION 317 -70.3% 1 066
NORMAL TAXATION
PROFIT AFTER TAXATION 317 -70.3% 1 066
Cartons (4 Kg Equivalent)
EXPORT
LOCAL 47 870 -4.3% 50 000
FACTORY 12 487.41
TOTAL 60 357 20.7% 50 000
PRICES
Price per carton 24.98 40.87
Nett return per carton 5.25 52%
RATIOS
GROSS MARGIN 24% 54%
PAT MARGIN 21% 52%
Appendix C: Results

38 Smallholder and agro-food value chains in South Africa
Experiences and Insights
on smallholder farmer
value chain integration
Mandla Nkomo
Seven lessons to share
TechnoServe South Africa is a part of global organisation
with over 40 years experience in working with entrepre-
neurial men and women in the underdeveloped parts of
the world, giving them the support they need to create
scalable enterprises. The small enterprises are assisted
to participate fully in their sectoral value chains, thereby
providing employment, earning income and impacting
local economies. In 2010, such work targeted in excess
of 2 700 businesses which, in turn, had business deal-
ings with over 280 000 smallholder producers. This
resulted in 1.5 million people benetting economically.
This work continues to grow, globally and more so within
the South African context of a society in the throes of
substantial socio-economic transformation. TechnoServe
South Africa has been engaged within South African since
2003. Their chosen practice areas have been SMME de-
velopment, agricultural value chain development, and
local economic development.
These practice areas have led TechnoServe to develop
and run programmes that assist rural-based businesses
to develop business plans, provide specialised training
and mentorship, facilitate linkage to markets and nance,
as well as virtual incubation and enterprise acceleration

39
Smallholder and agro-food value chains in South Africa
activities. Since 2007, more than 500 small businesses
have been supported within these programmes. These
activities have given insights into the inner workings and
challenges of rural-based small enterprises and have pro-
vided lessons on how to package support programmes
for such businesses.
Within Agriculture, TechnoServe has focused on working
with all actors in the agro-food value chains in order to
sustainably link farmers to secure markets and so fa-
cilitate smallholder ownership further up the value chain.
These programmes have worked with emerging farmers
in Mpumalanga, Limpopo and KwaZulu Natal, which
largely constitute black smallholders in South Africa. Les-
sons have been made in the context of working on issues
in the emotive land reform and redistribution space, as
well as former homeland farming areas.
Local economic development initiatives have indicated
that possibilities for building integrated economic ac-
tivities within specic geographies exist. This can result
in opportunities for smallholders and other small busi-
nesses to be integrated into the supply chains of larger
businesses. It is these lessons that TechnoServe South
Africa seeks to share with a larger audience of develop-
ment practitioners and academics involved in pro-poor
development and leadership.
We have learnt the importance of fully understanding
the context in which we work: clearly identifying whom
we work with; making sound choices on which value
chains make sense to work with: recognising the inherent
knowledge and skills gap and how to deal with it; being
creative in addressing the perennial access to nance
challenge for small holders; dealing with funding issues
for the interventions themselves; and choosing our
battles.
Lesson 1: Understanding
the context
The last census of South African agriculture (Statistics
South Africa, 2007) and various subsequent surveys indi-
cate a rather interesting state of affairs that must affect
the way TechnoServe understands smallholder farming in
South Africa. Although there are an estimated 2.5 million
smallholder farming households, there are only about
35 000 commercial farming units in the country. These
commercial farming units collectively operate on about
14 million hectares of farmlands, while their smallholder
counterparts are on approximately 3 million hectares of
farmland. Production statistics further indicate that 95%
of South Africa’s farming output originates from the com-
mercial farming sector. This sector in most respects is
globally competitive both in terms of uptake and use of
the latest agricultural technology (e.g. precision farming,
improved varieties and mechanisation), as well as pro-
ductivity (yields per hectare, quality specications, global
certication compliance and price competitiveness).
This context is not accidental but is a result of more than
a century of legislation and governmental policies, such
as the 1913 Land Act, apartheid laws and, of course,
the modern context of land reform such as restitution,
tenure reform, redistribution and AgriBEE..
The impact is that agri-food value chains in South Africa
favour sophisticated commercial farmers, and generally
crowd out smallholder farmers. Our interventions, there-
fore, must seek to give small- holders a ghting chance
to build long term sustainability within these highly com-
petitive agro-food value chains. Interventions that are
oblivious of this context and current dynamics will not
add value to smallholder farmers’ aspirations.
Lesson 2: Defining with
whom we work
Most practitioners in the development space constantly
have to confront issues of dening their target group as
precisely as possible. Our experience has been that this is
even more important in the agricultural context. In other
areas such as SMMEs, there is legislation that denes
small and medium enterprises on the basis of turnover
and employees, which somewhat provides a reference
point. However, the question of who is a smallholder
farmer in South Africa is a matter less agreed upon.
A useful rst lter would be a generic understanding of
what smallholders are:, “The term ‘smallholder’ refers
to their limited resource endowments relative to other
farmers in the sector. Thus, the denition of smallholders
differs between countries and between agro-ecological
zones.”(Dixon, et al., 2004) The issue of resource en-
dowments relative to other farmers is a relevant one in
South Africa, as the previous discussion on context has
shown.

40 Smallholder and agro-food value chains in South Africa
On the basis of this understanding of resource endow-
ment, the latest draft discussion document from the
Department of Agriculture Fisheries and Forestry (DAFF,
2011) comes up with a useful starting point. Three
groups: SP1, 2, and 3 are dened, based on resources:
i. the SP1 group is at the top of the pyramid, having
access to own land, some farming equipment, own
permanent labour of at least two and demonstrable
access to markets;
ii. SP2 has access to land, maybe leasing equipment
and only employs seasonal labour; and
iii. the bottom rung is occupied by the SP3 group on
leased land, share cropping or contract farming
with no hired labour.
This is a useful lter, although it must be said, it is driven
by DAFF wanting to offer specialised support to the dif-
ferent groups, and might not t entities with a slightly
different agenda for smallholder farmers.
The views of the UK Department for International Devel-
opment (DFID) on livelihoods, and livelihood assets are
also a useful set of tools for the exercise of determining
with whom to work. Analysing the South African small-
holder in the vulnerability context and through the lenses
of the ve capital assets (natural, social, human, physical
and nancial) these producers operate under help to
sharpen their disaggregation exercise.
TechnoServe’s strong private sector inuences have
led them to a point where they feel they are probably
better equipped to assist the SP1 and SP2 smallholder
farmers. This is driven by the inherent skill sets in this
organisation and the matrices used to measure success
and what one could term their ‘theory of change’. No one
organisation can be all things to all people: hence there
is a need to dene who we work with.
Lesson 3: Making choices
around value chains or
products
Another important choice to make is about which prod-
ucts or commodities to promote for the specic participa-
tion of smallholder farmers. The above-mentioned issues
of understanding the context and making a judgment call
on who to work with are both prerequisites for making
an informed choice about value chains and commodities.
It is well known that certain value chains are so domi-
nated by commercial farmers and agribusiness that
smallholders, acting individually, cannot break into
the sector. Also the critical mass of a combination of
assets and technologies to manufacture products for
demanding, and often low margin markets, is in itself
exclusionary to small holders.
TechnoServe has learnt that per commodity or sector, a
proper multi-level gap analysis is required. It is important
to determine the level of gaps with respect to quality
requirements and access to inputs. Within this cluster,
questions need to be asked about farmers’ abilities to:
i. access correct seed varieties;
ii. familiarise with the specic crop husbandry
requirements;
iii. access and use fertilisers and other crop nutrient
products; and
iv. meet minimum quality requirements and un-
derstanding these requirements from a market
perspective.
The next relevant tier of gap analysis relates to access
to farm infrastructure required for successful farming
of the specic commodity. Such on-farm infrastructure
includes:
• irrigation and tillage equipment (tractors)
• storage and produce handling facilities (storage
sheds and packhouses)
• logistics in the form of cold-chain friendly trucks to
deliver produce where it is required.
Any gaps, even at the barest minimum, should impact on
commodity choices for producers otherwise typical inter-
ventions will be required to bridge the gaps.
In our chosen target group of farmers, business skills
are an essential requisite. Ability to access key business
information such as market prices, and historical trends
and data is important. Furthermore, the ability to operate

41
Smallholder and agro-food value chains in South Africa
under medium- to long-term business plans is essential
as is basic farm and business record-keeping. These is-
sues are important as they have an impact on access to
nancial services such as loans and crop insurance.
Lesson 4: Bridging the
knowledge and skills gap
The basis of the TechnoServe smallholder farmer devel-
opment model is a combination of aggregation strategies,
skills development and improvement. The debate is al-
ways about which method is most effective for improving
specic skills levels amongst smallholder farmers, class-
room-based approaches and on-farm coaching and men-
torship. TechnoServe’s experience has taught that it is
not an either/or scenario. It is important to tailor learning
in such a way that it ts into what smallholder farmers
do - which is farming. Disruptive approaches that focus
on parachuting in ‘experts’ to lecture farmers and taking
them away from their elds, quickly leads to resentment
and results in limited assimilation of information. In the
same vein, on-farm mentorship approaches that are
parochial and invoice-driven also end up only beneting
mentors rather than smallholders.
We have attempted to adopt a learning-by-doing ap-
proach to bridge the knowledge gap. Classroom type ac-
tivities are always linked directly to activities on the farm,
and will often be delivered linked to a eld day or similar
event. An example would be teaching record-keeping
around a specic crop that the farmers are about to
plant, or in which they are already involved. Reinforcing
and offering practical advice is achieved by making use
of incentivised mentors, commonly referred to as ‘Prin-
cipal Farmers’. These are mentors who have proven
product and area knowledge, and who are contracted to
provide support to smallholders throughout the growing
season. Their remuneration is directly linked to specic
production targets being met. In this way, the mentors
are unlikely to be supercially involved as they have some
incentive.
Lesson 5: Solving the
access to finance challenge
South Africa is in a fortunate position in that smallholder
producers have a wide choice of funding sources for
various aspects of their operations, at least supercially.
The scus, through DAFF and the Department of Rural
Development and Land Reform (DRDLR), makes annual
provision for supporting smallholder farmers. Develop-
ment Finance Institutions (DFIs), such as the Land Bank,
the Industrial Development Corporation (IDC) and the
National Empowerment Fund (NEF) all have instruments
that smallholders can access. Agencies of government
such as Micro-Agriculture Finance Institute of South
Africa (MAFISA), Small Enterprise Development Agency
(SEDA) and National Development Agency (NDA) also
support initiatives in the sector. The point is that this
means there are more than adequate funds for certain
kinds of activities in the sector.
However, closer inspection reveals that as at 2009, com-
mercial banks, but not all the above-mentioned entities,
dominate the funding of agriculture. As much as 70% of
agricultural debt resides with the commercial banks and
is largely loaned out to commercial farmers. DAFF and
the Land Bank only contribute about 6% each to agricul-
tural nance. This indicates a huge challenge to ensuring
much needed nancial resources end up where they are
most required.
TechnoServe’s experience has indicated that they have
to be innovative in unlocking all these funds for small-
holder farmers. An aggregation mechanism of bringing
a group of smallholders together, to produce similar
products and receive tailored support, allows for a much
better conversation with nanciers. Bankers or funders,
as a rule, do not have the DNAor the capacity to deal with
2.5 million smallholders, but groups of smallholders in
the form of cooperatives or Section 21 companies are
more attractive, and more so if there is a support struc-
ture around them and a clear line of sight to market.
TechnoServe has piloted “management companies”, for
this aggregation function by offering loan management
services, market facilitation, technical and business
support that ensures smallholders build business track
records and local capacity to be sustainable and relate to
the business world in the long term.
Lesson 6: Funding the
intervention
TechnoServe, being an NGO, needs to raise funds to
do the work dealt with in this article and, unlike our

42 Smallholder and agro-food value chains in South Africa
colleagues in other ofces across the world, we are not
targets for big international donors such as USAID, the
Bill and Melinda Gates Foundation and DFID. This is a
consequence of South Africa’s somewhat deceptive
‘middle-income’ status which disqualies us for funding
for some development aid.
Our experience is that everyone - from politicians, busi-
ness, academics and other philanthropists - wants
smallholder farmers to be supported, and they often elo-
quently state this. However the sad reality is that very few
are prepared to pay for obviously needed interventions.
This has forced our country ofce to be creative in raising
funds from the private sector, private international foun-
dations as well as tapping into government funding for
our work.
As advocates for transformation in smallholder farming,
we have had to understand the issues that concern both
potential funders and mainstream smallholder farmers.
In some instances, it is about understanding the BEE
codes and making a case for budget allocations, either
for ‘enterprise development’, ‘preferential procurement’
and sometimes for pure corporate social investment to
fund smallholder value chain work.
Navigating tenders,and requests for proposals with
government must be undertaken in conjunction with
more creative ways of tapping into the scus. Some
departments are now comfortable in appointing NGOs
as service providers to implement specic government
programmes. A classic example is the Department of
Health using NGOs to implement HIV/AIDS mitigation
programmes when NGOs apply for state funding on the
basis of business plans. This could be a modus operandi
that the DAFF and DRDLR could adopt.
Lesson 7: Chasing your
opportunities (marathon
vs 50m dash)
A nal lesson to share is that supporting smallholders
is a full-time activity which sometimes has rather long
incubation periods. TechnoServe’s observation is that in-
terventions designed to last less than three to ve years
are simply too short to have lasting impact. A marathon
approach as opposed to a 50m dash would be more
appropriate.
Achieving the right funding outcomes is also a critical ac-
tivity. There is no formula for doing it. Some programmes
can be planned and funded in weeks, while others can
take up to a year of back-and-forth discussions before
agreement to commence.
TechnoServe’s approach is always to be at the cutting
edge of knowing what is happening in the sector, to be
adoptive to changing realities and to be open to collabo-
ration with like-minded actors who put the interests of
the producer rst.
References
Dixon J, Tanyeri-Arbur A and Wattenbach H (2004)
‘Framework for analyzing impacts of globalization on
smallholders’, in J Dixon, K Taniguchi, H Wattenbach and
A Tanyeri-Arbur (eds) Smallholders, Globalization and
Policy Analysis, AGSF Occasional Paper no. 5. FAO: Rome.
Department of Agriculture Fisheries and Forestry (DAFF)
(2011) ‘Draft Strategic Plan for SmallHolder Farmers’,
DAFF: Pretoria.
Statistics South Africa (StatsSA) (2007) Census of Com-
mercial Agriculture, 2007, Report no. 11-02-01. StatsSA:
Pretoria.

43
Smallholder and agro-food value chains in South Africa
Aspects of the citrus
value chain explored in
the context of Moletele
restitution initiatives,
Hoedspruit
Nerhene Davis

44 Smallholder and agro-food value chains in South Africa
Introduction
According to the United Nations Development Pro-
gramme, “inclusive business models aim to include poor
people into value chains as producers, employees or
consumers, in ways that are both equitable and sustain-
able” (UNDP 2010: 3). Inclusive business models in the
agricultural sector are also widely seen as a means of
providing access to capital, information and markets for
smallholders and communities who may otherwise be
marginalised from the economic mainstream. Inclusive
business arrangements, therefore, are seen by many as
an effective means of rural development (Vermeulen and
Cotula 2010) and also as a much more viable alternative
to large-scale land acquisitions (’land grabs’) as it is cur-
rently driven by institutional interests across the globe.
Vermeulen and Cotula’s (2010) denition of a business
model is used in this paper. It refers to the way in which
a company structures its resources, partnerships and
customer relationships in order to create and capture
value, in other words, all that enables a company to make
money. In cases where joint venture arrangements are
implemented in the form of business models, they are
also interpreted as a form of ‘inclusive’ or ‘collaborative’
arrangements. Such models have been variously de-
scribed as “inclusive business” (www.inclusivebusiness.
org), “mutually benecial partnerships” (FAO and CIFOR
2002) and “inclusive capitalism” (Hart 2007). Vermeulen
and Cotula assert that “the specic terms and conditions
of inclusive business model arrangements could result in
better local control of businesses activities by community
members on the one hand, but inappropriately designed,
they could deliver only nominal inuence for the com-
munity over key decisions and little or no dividends as
prots” (Vermeulen and Cotula 2010: 6). Therefore, there
is a pressing need for a better understanding of the spe-
cic agreements and design of inclusive business model
arrangements.
Inclusive business models have also been introduced in
the context of the South African land reform programme.
The stated aim of this programme is to redistribute agri-
cultural land among the wider population, restore ances-
tral lands to individuals and communities and strengthen
land rights more generally. The highly-developed nature
of the commercial agricultural sector in South Africa pro-
vides opportunities for previously marginalised groups
to engage in production of high-value commodities for
domestic and international markets. It also presents
major challenges in terms of capital, skills and competi-
tiveness (Lahiff et al 2012). The direct transfer of viable
agricultural land back to restitution communities over the
last decade has resulted in, what some commentators
have branded as “spectacular failures”, as a result of
an alarming decrease in productivity on newly restituted
farms. It is in this context that a variety of “strategic
partnerships” - a hybrid form of an inclusive business
model - have emerged in South Africa. Partnerships are
established between (largely poor) black landowning
communities and (mostly white) partners from the com-
mercial sector.
These initiatives take the form of joint ventures where
ownership of the land is transferred to the claimant com-
munity, but they are not allowed to move back onto the
land. Instead, claimant communities enter into agree-
ments with agribusiness partners who commit them-
selves to manage the land on behalf of the claimant com-
munity with the contractual understanding that benets
are shared between the partners (Department of Land
Affairs (DLA) 2008). This approach has been particularly
prominent in Limpopo where large areas of high-value ag-
ricultural land and infrastructure are being transferred to
community groups. In theory, the model should respond
to a demand from claimant communities for technical
and nancial assistance in managing large agricultural
enterprises. For private sector partners, some of them
former owners of the land in question, it might present
an opportunity to preserve or even expand commercial
activities within the agri-food sector, albeit under new
conditions (Lahiff 2007; Lahiff et al 2012).
Others contend that strategic partnerships negotiated in
terms of the restitution programme can create opportuni-
ties for existing actors in the commercial agro-food sector
to gain access to valuable land and water resources,
better control of upstream and downstream processes
and to lucrative government grants (Derman et al 2007).
In instances where these types of partnership arrange-
ments have been negotiated, as in the cases of Zebedele,
Levubu and Moletele strategic partnerships, production
on the land has continued. What is less apparent is the
extent to which real benets are being transferred to the
nominal owners of the land - the restitution communities.
Thus, in addition to the need for assessing aspects per-
taining to the ‘inclusiveness’ of these business models
in terms of ownership, risk sharing, voice and reward as
suggested by Vermeulen and Cotula (2010), we also need
to understand the role and involvement of newly resti-
tuted communities in the commodity chains from which

45
Smallholder and agro-food value chains in South Africa
they are now assumed to be beneting. In this regard, the
specic agricultural commodity being transacted and the
need to understand key aspects in the eld of high value
agricultural production are important.
Observations that some strategic partners are only be-
coming involved in these partnerships to improve their
control of upstream and downstream activities in the
value chain invokes the need to interrogate the govern-
ance structures (nature of power relations) and the range
of benets and risks being introduced to restitution com-
munities. Recent studies (Kaplinsky 2000; Vagneron
et al 2009; Bolwig et al 2010) increasingly caution
against the uncritical insertion of rural producers into
existing global value chains in terms of ‘poverty reduc-
tion’ goals. In this regard, Bolwig et al (2010) contend
that even if rural producers, workers and migrant workers
are included in global value chains, this may not be on
advantageous terms, and a careful analysis should be
made of the costs and benets of participation in a par-
ticular chain. In support of this notion, du Toit asserts
that “poverty can ow not only from exclusion but also
from processes of integration into broader economic and
social networks”. In this instance, he argues that these
tendencies are better captured as “adverse incorpora-
tion” into these value chains (du Toit, 2004). The aim of
this paper is not to make a judgement of the benets or
otherwise of the inclusion of restitution communities into
global value chains in terms of these partnership agree-
ments. Instead, it sets out to understand the structure,
agreements and outcomes in the broader context of the
‘costs and benets’ for these communities in relation to
their incorporation into the value chain.
This paper, therefore, attempts a better understanding
of strategic partnership initiatives in the context of ag-
ricultural commodity chains, A case study approach is
adopted to explore aspects of the citrus value chain from
the perspective of inclusive business model arrange-
ments introduced as part of the Moletele restitution
initiatives in the Hoedspruit area, south-eastern Limpopo
(see Figure 1).
Figure 1: Orientation map of the study area

46 Smallholder and agro-food value chains in South Africa
This paper is based on ongoing research in the Limpopo
Province and includes a detailed case study of the Mo-
letele strategic partnership initiative in South Africa.
Research methods include eld observations since
2009, analysis of the Moletele Communal Property As-
sociation’s (CPA) nancial statements and contractual
agreements, and interviews with a wide range of key
informants, community members, commercial partners
and state ofcials.
Although the central theme of the inquiry is about inclu-
sive business models involving partnerships between
highly unequal partners, it also concerns the power rela-
tions between those in the value chain. The links between
the different agents and their functions in commodity
production in its different stages are as important as the
social relations between the agents and the institutions
which characterise unequal power.
The paper begins by briey sketching the Moletele res-
titution case and the joint venture initiatives that were
introduced to enable the transfer of land to the Moletele
CPA. This section is followed by a short overview of how
these initiatives have evolved over the last few years. The
paper concludes with some of the key ndings of a de-
scriptive value chain analysis of citrus production activi-
ties on Moletele land.
The Moletele restitution
case
In 1991, the de Klerk government repealed the Land Acts
of 1913 and 1936 and appointed an Advisory Committee
on Land Allocation (ACLA) to make recommendations on
the disposal of state land, including restoration to dis-
possessed landowners. The Moletele community, which
had been trying to get back their land since 1985, lodged
a claim with ACLA in 1992. The total land under claim
covers approximately 78 000ha of land in the Hoedspruit
area. The Commission on Restitution of Land Rights ac-
cepted the validity of part of claim in 2004, and in 2007
a total of 7 652ha of prime agricultural land was restored
to 1 615 households under the Moletele CPA. The total
cost of land acquisition thus far is estimated at R194
million .
The initial expectations on the part of the state were ar-
ticulated by the then Minister for Agriculture and Land
Affairs, Ms Lulama Xingwana, at the land handover
ceremony to the Moletele community . The Minister,
framed a vision for the Moletele people asserting:
This land that we are restoring today has some of
the best oranges and mangos this country has ever
produced. As from today the people of Moletele are
now exporters. You are going to be operating from
the well-equipped pack-house that we have included
in the purchase of this land”. … This deal will also
accelerate value-adding in the produce coming from
this land of milk and honey. This will ensure partici-
pation of the Moletele Community in the entire value-
chain. These partnerships give credence to economic
empowerment because the community will not only
receive hand-outs in the form of lease rentals but will
be participating in the day-to-day management of the
farms.
Subsequent to land restoration, the Limpopo Regional
Land Claims Commission together with the Limpopo De-
partment of Agriculture and Maruleng Local Municipality
assisted the community to enter into joint ventures with
three different strategic partners forming the following
companies: New Dawn Farming Enterprise (Pty) Ltd, Di-
naledi Farming Entreprise (Pty) Ltd and Batau Farming
Enterprise (Pty) Ltd. The Moletele also leased out some
of their farms to commercial farming companies, e.g.
Richmond Estate.
Overview of partnership
initiatives on Moletele
land
When the Moletele restitution case was settled, negotia-
tions were driven by the notion that the productivity on
the farms should continue and that strategic partnership
arrangements should be the type of joint venture ar-
rangement to achieve the desired outcomes and inject
the community into the arena of global agricultural pro-
duction. Since inception, however, the benets initially
envisaged have not materialised. Some of the strategic
partners withdrew and some even led for liquidation. In
this regard, both the strategic partners and representa-
tives from the Moletele CPA blame an over-reliance on
the payment of the promised state grants for some of the
challenges they are currently experiencing. A recurring
discontent, therefore, is being expressed by key actors
involved in the Moletele claim regarding the design of the

47
Smallholder and agro-food value chains in South Africa
Joint venture
company
Total ha managed Current ha under
production
Production Employment
New Dawn Farming
Enterprise
1019 ha 405 ha Citrus, mango,
guava, and paw-paw
123 permanent and
390 seasonal
Dinaledi Farming
Enterprise
686 ha 355 ha Lemons, grapefruit,
and Valencia
oranges
650 permanent and
seasonal
Batau Farming
Enterprise
855 ha 157 ha Mango, citrus, litchi,
and vegetables
72 permanent
Richmond Estate 2434 ha 590 ha Grapefruit, Valencia
and mango
135 permanent and
440 seasonal
Table 1: Summary of Moletele partnerships
strategic partnership (SP) model. These actors observed
that the model was too complicated and that it opened
up the community to unnecessary risk. This risk became
apparent after government failed to transfer promised
grants. These could have been used to pay off the loan
account of the partnership as the ‘community’s contribu-
tion’. The community in some instances now seemingly
‘owes the strategic partner’ in terms of partnership con-
tributions. The liquidation of bankrupt strategic partners
also resulted in risk for the community with some of the
Moletele’s movable property possibly being attached to
liquidation notices.
These challenges prompted the Moletele CPA committee
to shift towards the establishment of ‘community-private
partnerships’ (CPP) – effectively rental agreements, with
some added benets for communities. In terms of the
CPP agreement, setting up a joint venture company with
investment from the community is not required. All com-
mercial operations remain in the control of the partners,
including access to nance and production risk. Large
commercial partners appear to be the desired business
partners as they would be able and more likely to under-
take the required investment and risk envisaged in the
CPP arrangements. The chairperson of the CPA observed
that the CPP arrangements might be less ambitious but
with more secure benets for the Moletele community in
the form of timeous lease payments, employment and
investment in farming production with lower risk for the
community.
In the case of the Dinaledi partnership, ideas currently
are being explored to move away from the SP model to-
wards a CPP between the CPA and the same strategic
partner. The CPP agreement compels the business part-
ners to invest in training and preferential employment
of Moletele community members, and to invest in main-
taining the quality of existing orchards, the planting of
new orchards and the upkeep of the packhouses on the
land on behalf of the community. As the CPP model re-
quires signicant investment on the part of the new busi-
ness investor, a shift in this direction may indicate a trend
towards favouring ‘bigger commercial partners’ able to
shoulder this level of investment. Key informants stated
that the Boyes Group would favour the shift towards the
CPP model because the complex arrangements regarding
decision making and daily management activities on the
farms would become somewhat less cumbersome, and
they can make the kind of capital investment required in
terms of the CPP model.
At Batau, the CPA decided to introduce a CPP agreement
with Bonosafe (the management and empowerment
company owned by South African Fruit Export (SAFE),
which has a number of joint ventures with communities
in South Africa and Zimbabwe). The Bonosafe agree-
ment would involve a CPP agreement between the Mo-
letele CPA and Bono Holdings, which forms part of SAFE,
an established company and trademark. According to its
website11, SAFE is one of the fastest growing independent
exporting companies in the world, and it has been able to
position itself as a key role player in the Southern African
fruit exporting industry. All logistical operations and com-
mercial and production activities are coordinated from
SAFE Farm Ventures in Cape Town, established in 2009.
In 2008, Bono Holdings was established as a joint ven-
ture company between SAFE Mauritius (its head ofce)
and Evans Malokisa Nevondo. Bono Holdings is a man-
agement company responsible for the operational perfor-
mance and nancial health of the empowered entities.
It provides running capital, farming implements, skills
transfer (training workers) and health care. Farms are
operated in conjunction with the trusts and beneciaries.

48 Smallholder and agro-food value chains in South Africa
Table 2: Overview of Moletele partnership initiatives, 2008-2012
New Dawn
• Partnership between Strategic Farm Management (SAFM) and Moletele CPA,
formed in 2008
• CPA owns 52% of the shares and SAFM owns 48%
• Set up for period of 15 years.
• 1,019ha of land, with 405ha under irrigation for mangoes, citrus and litchis.
• 326ha used for grazing
• 79ha leased
• Currently rent only paid very sporadically
• Development grants from the state not paid and are reected against the
loan account.
Still a strategic partnership
arrangement and strategic
partner has opted to apply for
a loan from the DBSA to stay in
business.
Dinaledi
• Formed in 2008.
This is a partnership between the Moletele CPA and the Boyes Group (which
exports oranges to Canada, EU, Russia, Middle East, Japan and Mauritius and
employs around 270 permanent and 450 seasonal workers)
• Comprises 686ha of land under citrus production:
• based on 50:50 share in the operating company
• seen to date as the “relative success story” at Moletele
• been able to invest own money in skills development towards the community
• development grants from the state not paid to date resulting in unwillingness
from the current strategic partner to continue to invest more money.
Still a strategic partnership
arrangement but the possibility
of a CPP arrangement is being
investigated.
Batau
• Partnership very similar to New Dawn contractual agreement also formed in
2008.
• 52% CPA and 48% strategic partners (comprising four of the previous com-
mercial farm owners who have structured themselves into a company called
Chestnet Holdings).
• Comprises 855ha of land under citrus, litchis, vegetables and mangoes.
• Since inception plagued by challenges.
• Could not source additional nance and development grants was never paid
– partnership collapsed end of 2009.
CPP agreement with some
challenges. Bono Holdings has
been farming and exporting
on community land for 2 years
now but the signing of the CPP
lease agreement is consist-
ently being postponed.
Richmond
• Comprises 2 434ha of land transferred at a cost of R63 million (US$9m)
• Single portion of already well-equipped land with almost 600 ha under estab-
lished citrus production.
• CPP agreement with Golden Citrus Frontier (now Bosveld Citrus - one of the
major players in the citrus industry).
CPP agreement in place and
lease payments being made
into the account of the CPA as
the representative body of the
Moletele community.

49
Smallholder and agro-food value chains in South Africa
The CPP agreement has not yet been nalised and cur-
rently a management contract with Bono Holdings is in
place. The management contract, informally referred to
as a “care-taking agreement” by CPA committee mem-
bers, takes the form of a lease agreement where the
leaseholder (Bono Holdings) pays a lease amount and
takes over the management of the land on behalf of the
owner (Moletele CPA). To provide incentives for the farm
management, the contract makes provision for prot-
sharing rather than a xed fee.
Value chain considerations
in the context of
partnership agreements
Overview of the South African citrus
industry
During the 2008/09 production season, the citrus in-
dustry contributed R6 billion to the total gross value of
South African agricultural production (Citrus Growers
Association (CGA) (2010). South Africa’s citrus industry,
unlike much of the country’s manufacturing and pro-
cessing sectors, has always been outwardly focused and
‘globally integrated’. The industry is an important foreign
exchange earner and currently is the twelfth-largest pro-
ducer and third-largest exporter of citrus globally. Some
45% of citrus is exported to Europe, with the UK as the
top destination (CGA 2010). Exports of citrus to the UK
started in the rst decades of the last century, and by
the 1960s South Africa was exporting well over half of
all southern hemisphere fresh citrus and was ranked
amongst the top ve fresh citrus exporters in the world.
By the mid-1990s, the 40 million cartons of citrus ex-
ported to over 60 countries represented one third of the
total local and export value of South African fresh fruit
production (Mather 2005; Mather 2008). Quality dif-
ferentiation sets the South African citrus industry apart
from other citrus producing regions in the world, and
buyer condence seems to be a key driver of success in
not only maintaining existing markets but also in terms of
breaking into new markets.
Since market liberalisation and a move away from single
channel marketing boards, the success in the citrus value
chain is predicated on the ability to meet fairly stringent
requirements in a vigorous self-regulated industry with
strong producer representation, research, extension and
market access service (Mather and Greenberg 2003).
While deregulation delivered a new-found freedom to
exporters and producers, this new era of independence
also brought a unique set of challenges and imperatives.
The increase in competitors led to the need to strengthen
competitive advantage for the business to succeed. The
volume of citrus supplied to the market began to in-
crease and quality and price became important tools of
diversication (Mather 2008). In an interview with one
of the members of the Citrus Growers Association12 in
Hoedspruit, the abilities needed to be a successful citrus
exporter were listed as “a sound knowledge of exchange
rates, the ability to meet buyer-driven requirements and
a general sense to ‘read’ the market (timing and plan-
ning in terms of market demands and shifts).” These
observations are informed by the evolution of the citrus
fruit sector in the highly competitive global markets with
the rise of increasingly powerful global retail chains in
fruit distribution. As a result, Fundira (2003) asserts that
citrus is evolving from a producer-driven to a buyer-driven
chain.
Input –supply considerations
The Moletele land was transferred back to the commu-
nity as the portions became available. These portions
were consolidated into four different entities/operational
farming units:
i. consolidation of land into newly-dened operational
units required new inputs;
ii. more cost-effective irrigation systems had to be
installed on some of the newly acquired land con-
solidated into continuous farming units;
iii. in some instances older, less resilient cultivars had
to be replaced with newer alternatives to make
these new units more economically viable; and
iv. a shift towards star ruby grapefruits for export, par-
ticularly to Japan, in the face of changing consumer
demand.
While the partnership deals and land transfer were being
negotiated some of the farms deteriorated and orchards
became overgrown. A great deal of money had to be
invested to clear newly acquired land and prepare it for
new tree planting. Land preparation and proper drainage
requires sound planning, considerable investment and
clearly dened implementation strategies. Cost of land

50 Smallholder and agro-food value chains in South Africa
preparation and irrigation is estimated at approximately
R50 000/ha.
In the Hoedspruit and Letsitele area, citrus producers
have two options. They can either buy nursery trees that
are already certied and registered from one of the only
two CGA-approved nurseries in the area or they can buy
seedlings/bud wood, for which they have acquired their
own certication, to be grown into seedlings. Timing and
planning for ordering the new seedlings appears to be
fairly critical as an eighteen-month waiting period is re-
quired to obtain seedlings.
The New Dawn strategic partner, SAFM, was pro-active in
this regard and started a nursery on one of the Moletele
properties where they are currently growing their own
seedlings - now targeted for new orchards on Moletele
land. At present, the nursery is growing approximately 25
000 new seedlings that will be planted on newly cleared
Moletele land once a Development Bank of South Africa
(DBSA) loan for the partnership is transferred. The ben-
et of this approach is that the eighteen-month waiting
period and the cost of sourcing the seedlings from a des-
ignated service provider would be nullied. The nursery
is turning into a very attractive endeavour, with other
farmers from the surrounding area now also approaching
the CPA and the strategic partner to explore the pos-
sibility of nursery-grown seedlings for the surrounding
farms. This pro-active approach to market demands by
the New Dawn strategic partner illustrates Greenberg’s
(2010: 17) assertion that “the role of individual actors
as active agents, who shape their own reality … and thus
alter or reinforce the function and structure of existing
value chains…” should not be downplayed as is often
the case in a value chain approach focusing only on the
structural functioning of a commodity chain. Initiating the
nursery on Moletele land and allowing the community to
manage and benet (from sales of seedlings to the sur-
rounding farmers) thus would allow the community to
become involved and benet from upstream activities
in the citrus value chain. This approach could contribute
to a previously under-explored benet stream in terms
of the partnership agreement and allow the community
more strategic positioning in terms of citrus value chains
in Hoedspruit.
The farms were transferred to the Moletele CPA as going
concerns. Existing employees were retained, leaving lim-
ited scope for new employment opportunities of commu-
nity members. Less that 5% of the workforce at the time
of transfer was from the Moletele community. Agreement
was reached between the CPA and the strategic partners
in the CPP contracts that Moletele CPA members would
be given preference for any new employment. However,
challenges surfaced. The majority of Moletele community
members live about 45km outside Hoedspruit (where the
farms are located). This means the limited transport op-
tions and escalating transport costs hinder community
members from taking up available positions on the farms.
Strategic partners and CPA members observed that com-
munity members preferred working in the packhouses
and were generally reluctant to do work on the land.
In terms of the partnership agreement, skills transfer and
employment was clearly earmarked as one of the benet
streams towards the community, but in reality this has
not materialised. The labour issue is also much more
complicated than initially envisaged. In an effort to cut
costs and comply with labour legislation, citrus growers
in Hoedspruit have responded to shifts in the agricultural
labour market by substituting permanent labour with tem-
porary and casual labour and have increased their use of
labour contracting. Therefore, apart from context-specic
hindrances to maximise employment benets to the
community, Hoedspruit’s citrus farmers’ response to the
broader trends in the agricultural industry, has resulted
in fewer employment opportunities for the community.
Production
Protable citrus production only begins in the eighth year
after planting. From the third year some yields are avail-
able but mostly for juice production. This long waiting
period must be taken into account for planning and
projections, especially where new orchards have been
planted. The waiting period must be communicated and
anticipated in terms of projected benets and prots.
Furthermore, use of additional fertilisers, pest control,
irrigation practices and even the working conditions
of farm labourers need to comply with a variety of ac-
creditation requirements, e.g. Fair Trade, Field to Fork,
and Perishable Products Export Control Board (PPCEB)
accreditation.
Despite these challenges, the scale of production at
New Dawn, Richmond and Dinaledi farms allowed for
economies of scale which enabled cartons of citrus to be
delivered to the ports at less that R50/carton (translating
into costs below the industry norm).

51
Smallholder and agro-food value chains in South Africa
Processing
Processing involves harvesting, washing, sorting and
waxing the produce in accordance with stipulated and
agreed requirements/procedures. Representatives from
various government organisations, lead rms and other
regulating authorities visit the community-owned pack-
houses to ensure compliance with a range of export
standards, regulations and accreditation specications.
In previous years, inspections were conducted at the
ports before shipment, but now representatives from all
these regulating bodies come to the packhouses.
In terms of a very crude and simplied input-output ap-
proach, the CGA representative in Hoedspruit provided
the following calculations:
• It is estimated that the value of a carton of citrus
(oranges) from the land (just after harvesting) is
around R18/carton. (This estimation is what the
carton can be sold for at this stage in the value
chain).
• Once the produce has been washed, sorted, waxed
and packed, the value of the same carton of or-
anges increases to approximately R25/carton.
• After transport and logistics, the carton is trans-
ported to the port, and the value of the carton is
now calculated at approx. R8/carton = R33/carton.
• Supply and distribution costs (cold storage and
transport) at R10/carton increase this gure to
R43.
• Price on the ship or Free on Board (FOB) i.e. cost
to port before shipping at R10 carton increases the
gure to R50.
• Delivery in port (DIP) costs include shipping to
destination. This translates into an estimated R10/
carton freight cost, bringing this to R60/carton.
Once cartons reach the export agents at the ports all ac-
tors in this value chain stop using calculations in terms of
ZAR per carton. Given that commodities at the ports are
sold in US$ and Euros and given that export agents are
allowed a 180-day waiting period to ‘read’ exchange rate
trends, it is evident that most value is not captured at the
production or even processing stages of the citrus value
chains (where there is still a semblance of community
involvement). In terms of the crude calculation above,
prots and cost incurred by logistics companies, im-
porters and marketing agents in the importing country,
further processors, retailers or other market channels
and the state (in the form of tariffs, levies and taxes)
will also need to be accounted for - or it may seem as if
exporters capture the greatest share of the value, when
they may not. But the calculations demonstrate that the
prot/benet accrued by the exporting company still out-
weighs what the community gains from the mere produc-
tion of the commodity on their land.
On Moletele land, depending on the quality of the pro-
duce, 20-30% of the citrus is channelled towards juice
production. As juicing is outsourced, there is very little
community involvement or benet from this processing.
Richmond and Bono Holdings (CPP model) used their
own subsidiary companies to do the juicing, while the
strategic partners sub-contracted juicing to independent
companies.
Export
Initially the idea was that the Moletele community would
become part of the value chain as exporters. This was
not the case in practice. In fact, some commentators cau-
tion against the idea of communities becoming exporters
in global agro-commodity chains, claiming that it might
open up impoverished rural communities to unnecessary
risk. Perhaps there is a middle ground or workable com-
promise to be reached in this instance.
Consider the following: the New Dawn strategic partner
(SAFM) was once again fairly innovative. The owner made
an offer to the community to purchase 10% of the shares
in his export company. This seems quite feasible as 10%
of the shares would provide the community with at least
some income from export activities without opening them
up to major risk. It would also open up opportunities for
mentorship. Industry specialists tend to agree that the
lion’s share of the prot is not in production or processing
of agricultural commodities, but in the export of these
commodities. Perhaps allowing communities partial
involvement in export activities might be considered a
more feasible option than simply stopping all community
involvement at production and processing stages in the
value chain.

52 Smallholder and agro-food value chains in South Africa
Conclusion
Pritchard (2000) observes that a traditional political
economy approach to the agrifood chain sees capital
accumulated through controlling the tangible means
of agricultural production: land, labour, nutrients and
chemicals, water, genetics and seeds, feed, equipment,
and capital. He continues that it is equally important to
recognise that ownership and control of intangible assets
(information, brands and patents), rather than control of
the tangible means of production, can allow the con-
centration of capital from a supply chain as well as the
conversion of that capital into mobile nancial capital. He
concludes that “the governance of supply chains hinges
on controlling the means of co-ordination rather than the
means of production.” In the case of citrus production
activities on Moletele land, it seems as if control and
ownership of the tangible “means of production” are not
delivering many benets to the restitution members. The
strategic partners and community-private partners are in
a better position to capture value in existing chains be-
cause they have a better understanding, know-how and
control, not only of upstream and downstream activities
but also of intangible assets.
New Dawn and Dinaledi did show attempts at better
horizontal integration/community involvement along the
value chain. New Dawn set up the nursery on Moletele
land, explored the feasibility of supplying seedlings for
Moletele land and to neighbouring farmers, and gave the
community the option to purchase 10% of shares in the
export company. Dinaledi invested signicantly in skills
development programmes and Fair Trade accreditation
which indicates they have tried to ensure adherence to
basic conditions and minimum wage legislation for farm
workers. Five years after the transfer of land, commercial
production on the land is continuing: a functioning man-
agement structure in the form of a business-orientated
CPA remains in place and it has an impressive bank bal-
ance. The way forward for the Moletele community seems
a bit more precarious. Production on the land might have
continued but disillusioned community members increas-
ingly ask for more benets to be channelled their way.
It is generally observed that the direct introduction of
rural producers into global value chains has delivered
mixed results and a signicant proportion of the litera-
ture focuses on the challenges and complexities of intro-
ducing rural producers into these value chains. Similarly,
it can also be concluded that the more indirect ‘inclusion’
of a restitution community into the global value chain
via inclusive business model arrangements, particularly
in the case of the Moletele community, seems to have
resulted in what can also be labelled as ‘ambiguous
outcomes’. In conclusion, the quiet discontent observed
during eldwork conducted in 2010 is also currently sur-
facing into an open challenge from community members
asking “in whose interest is production on the land?” The
CPA seems to be committed to ensure continued pro-
duction on the land, but until they come up with viable
strategies of distributing benets from the production
and other value chain related activities to the communi-
ties, their efforts appear to be promoting corporate rather
than community interests.
References
Bolwig S, Ponte S, du Toit A, Riisgaard L and Halberg N
(2010) ‘Integrating Poverty and Environmental Concerns
into Value-Chain Analysis: A Conceptual Framework’, De-
velopment Policy Review 28(2): 173-194.
Citrus Growers Association of southern Africa (CGA)
(2010) ‘Annual Report 2010’, accessed on 16 March
2011 at:, www.cga.co.za/site/les/5438/CITRUS%20
ANNUAL%20REPORT%202010%20web.pdf
Department of Land Affairs (DLA) (2008) Strategic
Sourcing and Partnership Framework. Government
Printers: Pretoria.
Derman B, Lahiff E and Sjaastad E (2006) Strategic
Questions about Strategic Partners: Challenges and
Pitfalls in South Africa’s new Model of Land Restitution.
Programme for Land and Agrarian Studies (PLAAS): Cape
Town.
du Toit, A (2004) ‘ “Social exclusion” discourse and
chronic poverty: A South African case study’, Develop-
ment and Change 35(5): 978-1010
Food and Agriculture Organisation of the UN (FAO) and
Center for International Forestry Research (CIFOR) (2002)
Towards equitable partnerships between corporate and
smallholder partners: Relating partnerships to social,
economic and environmental indicators. FAO: Rome.
Fundira T (2003) A transaction cost analysis of the fruit
supply chain in South Africa: A case study approach.
(MAgric Admin thesis), Department of Agricultural Eco-
nomics: University of Stellenbosch.

53
Smallholder and agro-food value chains in South Africa
Greenberg S (2010) Status report on land and agricul-
tural policy in South Africa, 2010. Institute for Poverty,
Land and Agrarian Studies: Cape Town.
Hart SL (2007) Capitalism at the Crossroads: The Un-
limited Business Opportunities in Solving the World’s
Most Difcult Problems. Wharton School Publishing:
Philadelphia.
Humphrey J and Schmitz H (2002) ‘How does insertion
in global value chains affect upgrading in industrial clus-
ters?’ Regional Studies 36(9): 1017–27.
Kanji N, MacGregor J and Tacoli C (2005) Understanding
market-based livelihoods in a globalising world: com-
bining approaches and methods. International Institute
for Environment and Development: London.
Kaplinsky R (2000) ‘Globalisation and Unequalisation:
What can be learned from value chain analysis?’ Journal
of Development Studies 37(2): 117-146.
Kaplinsky R (2004) ‘Spreading the gains from globalisa-
tion: What can be learned from value chain analysis’, IDS
Working Paper 110, Institute of Development Studies:
Brighton.
Lahiff E (2007) ‘Business Models in Land Reform’. PLAAS
Research Report 27, Programme for Land and Agrarian
Studies: Cape Town.
Lahiff E, Davis N and Manenzhe T (2012) Joint ventures
in agriculture: Lessons from land reform projects in
South Africa. IIED/IFAD/FAO/PLAAS: London/Rome/
Cape Town..
Mather C (2005) ‘Regulating South Africa’s citrus export
commodity chains after liberalisation’, accessed on 13
February 2009 at: http://www.tips.org.za_full.pdf
Mather C (2008) ‘The structural and spatial implications
of changes in the regulation of South Africa’s citrus ex-
port chain’, in N Fold and L Nylensted (eds) Globalization
and Restructuring of African Commodity Flows. Nordiska
Afrikainstitutet: Uppsala.
Mather C and Greenberg S (2003) ‘Market Liberaliza-
tion in Post apartheid South Africa: The Restructuring of
Citrus Exports after Deregulation’, Journal of Southern
African Studies 29(2): 391–410
Nadvi K (2004) ‘Globalisation and Poverty: How can
Global Value Chain Research Inform the Policy Debate?’
IDS Bulletin, 35: 1. Institute of Development Studies:
Brighton.
Pritchard B (2000) ‘The tangible and intangible spaces
of agro-food capital’, paper presented at the 10th Inter-
national Rural Sociology Association World Congress, Rio
de Janeiro, Brazil, (July).
Vagneron I, Faure G and Loeillet D (2009) ‘Is there a driver
in the chain? Identifying the key drivers of change in the
fresh pineapple sector’, Food Policy 34(5): 437-446.
Vermeulen, S and Cotula, L (2010) ‘Making the most of
agricultural investment: A survey of business models that
provide opportunities for smallholders’. IIED/FAO/IFAD/
SDC, London/Rome/Bern.
United Nations Development Programme (UNDP) (2010)
’The MDGs: Everyone’s Business. How inclusive business
models contribute to development and who supports
them.’ UNDP, New York, accessed on 9 July 2011 at:
http://www.growinginclusivemarkets.org/media/mdgre-
port/mdgreport_full.pdf

54 Smallholder and agro-food value chains in South Africa
Smallholders and the
‘Walmart effect’ in
South Africa
Stephen Greenberg13 and Gaynor Paradza
Introduction
In 2010, Walmart put in a bid for the acquisition of 51%
of Massmart, a South African food wholesaling and re-
tailing company. Massmart closely resembles Walmart’s
model of ‘big box’ stores, where goods are sold in bulk
at discounted prices., In March 2012, following lengthy
procedures, the Competition Appeals Court approved the
merger with some conditions, including the establishment
of a R100 million supplier fund, proposed by the merging
companies, to assist local suppliers and distributors to
meet the conditions for entry into Walmart-Massmart.
The entry of Walmart, the world’s largest company and
retailer, into South Africa potentially opens a new chapter
in food retailing in southern Africa and Africa. This paper
takes its title from a book by Charles Fishman (2007) that
essentially points to a trade-off between price and quality,
on the one hand, and volume and apparent stability
on the other. In other countries, including the US, sup-
pliers have decided to enter into supplier relationships
with Walmart because of the massive boost to volume
it provides, and hence to quick growth in the size of sup-
plier companies. However, Fishman shows that over time
these suppliers become dependent on Walmart as the
main buyer and are compelled to compress their costs

55
Smallholder and agro-food value chains in South Africa
as the retailer demands annual price cuts. The suppliers’
businesses are geared towards supplying Walmart and it
becomes difcult for them to extricate themselves from
this relationship. Over time, this results in suppliers cut-
ting corners on the quality of products, resulting in the
general trend we see today of poorer quality products
on the shelves, using poorer quality methods, without a
better alternative on offer even at a higher price.
Walmart aims to sell products more cheaply than its com-
petitors. This means some of the savings in the supply
chain are passed onto consumers. In the context of a
large base of consumers, who live in poverty, this poses a
dilemma for policy makers and regulators. Do you choose
to reduce consumer prices at all costs, or do you consider
the longer term implications of a decline in quality (for
food this is especially important) and a gradual erosion
of the production base as jobs are exported to cheaper
places in order to meet Walmart’s conditions? South
Africa’s competition authorities have explicitly endorsed
the former. The Competition Tribunal argued that “since
the evidence is that the likely consumers, who will benet
most from the lower prices associated with the merger,
are low income consumers and those consumers without
any means of support of their own, thus the poorest of
South Africans, the public interest in lower prices is no
less compelling [than the effect of the merger on local
producers and jobs]” (Competition Appeal Court 2012:
16). The Appeal Court ruled in favour of the merger and
hence in favour of low consumer prices in the short-term
over long-term production capacity.
Walmart will target lower income groups and thus build
supermarkets in townships and rural areas. This is its
primary market in South Africa, as elsewhere. It is in line
with government’s Comprehensive Rural Development
Programme, which identies rural shopping malls as an
area for development, under economic infrastructure
(DRDLR 2009: 2).
It is questionable whether prices will decrease all that
much in any case, given the rapid rise in food prices in
South Africa and globally over the past few years. These
prices rises are structural and are unlikely to decline in
future, since the basic cost structure of producing food
has increased. Concentration in food retail is not discon-
nected from these price rises. This is a story that will
unfold over time, and it will need to be tracked in South
Africa.
This paper looks in more detail at the way Walmart’s
supplier relations work and it attempts to situate this in
the context of existing corporate retail practices in South
Africa. It then seeks to consider the possible implications
of these dynamic processes on the possibility for small-
holder farmers to improve their livelihoods by entering
into supplier relations with corporate food retailers.
Fresh produce value chains exhibit buyer-driven rela-
tionships in a fairly hierarchical structure of power from
retailers down. There may be variations in the relations
between one node and another in different chains (e.g.
suppliers of high quality branded products may have
more leeway for negotiation than suppliers of undiffer-
entiated commodities). Although the ultimate buyer is
the consumer, it is a myth that consumers signicantly
shape retailer’s decisions about branding and dening
markets. More realistically, consumer preferences are
shaped by retailer strategies in differentiating markets
and products, and dening need through advertising and
sales strategies.
Although the consumer converts the product into a use-
value or throws it away once it is bought, what happens
to the product does not matter from the perspective of ex-
change value which underpins capital accumulation and
growth. The commodity disintegrates into capital growth
once its nal exchange value has been realised by the
retailer (or on some occasions, by those who add further
value after wholesaling or retailing, such as prepared
food, which then funds a further cycle of accumulation).
Much has been written on the relationships between
smallholder farmers and food retailers. Some have cri-
tiqued contract farming for the imbalances of power it
reproduces (e.g. Little and Watts 1994). Others have
identied where the power imbalances lie (e.g. Dolan
and Humphrey 2000: Gibbon and Ponte 2005; Seville et
al 2011). The technocratic response to this has focused
efforts on rectifying these power imbalances without
destabilising the overall functioning of value chains as
circuits of accumulation (e.g. Jaffee, et al 2003; Reardon,
2005). The roots of fair trade, private codes of conduct
and similar initiatives are found here (e.g. Barrientos et
al 2003).

56 Smallholder and agro-food value chains in South Africa
The South African food
retail sector
There is a long history of corporate/co-operative domina-
tion of the agro-food system in South Africa, from seed
companies, through grain storage, processors and manu-
facturers, to food retailers. Under the apartheid control
schemes, cooperatives were appointed as monopoly
agents for the receipt of the crop, payment to farmers,
storage and onward consignment to processors. Bayley
(2000) referred to “an agricultural nomenclatura” – a
privileged caste of intellectuals in the state bureaucracy
and companies/co-operatives that dictated state policy.
Amendments to the Co-operatives Act in 1993 allowed
the co-ops to convert into companies, effectively priva-
tising decades of state investment. It also allowed them
to diversify services, including into retailing of farming
requisites and nancial services. This enabled them to
retain their economic power, even while they experienced
a period of disconnection or distance from the state
bureaucracy.
In the era of state-controlled prices, not all food value
chains were buyer driven. Deregulation eliminated
single-channel markets and price controls, opening the
door for retailers to increase their power. “The demise
of marketing boards meant that the farmers’ collective
bargaining powers were drastically reduced” (National
Agricultural Marketing Council (NAMC) 2009: 1). This
led to increased competition amongst suppliers and
trade liberalisation extended competition to international
sources too. The result was a sharp rise in the global
sourcing of food products, especially processed products.
Overall, the value of imported processed food products
rose more than 6.5 times between 1995 and 2007, from
R3 billion to R18 billion. Unprocessed food products also
rose, though not as sharply, from R5 billion in 1995 to
R12 billion in 2007 (National Department of Agriculture,
2009: 84).
The formal supermarket sector is dominated by four big
corporations: Shoprite, Pick n Pay, Spar and Woolworths.
They have more than 94% of supermarket sales, and an
estimated 55-68% combined share of the food market
(Weatherspoon and Reardon 2003: 1, for the lower es-
timate; Planting 2010 for the higher estimate). Shoprite
and Pick ‘n Pay are the biggest, with a combined 64%
between them. The South African supermarket sector
has been characterised as an “extremely tight oligopoly”
(Naidoo, 2011). Each corporate chain has different store
brands which target different living standards measure-
ment (LSM) categories.
According to documents submitted to the Competition
Commission, Massmart only had a 2% share of the
formal food retail market in 2009, although it was the
second largest company (behind Metcash) in food whole-
saling, with a 22.4% share of food wholesaling in 2009
(RBB Economics 2011). Massmart has a wholesale divi-
sion called Masscash, which includes CBW Holdings and
Shield in the food sector. Shield serves 633 independent
retailers and wholesalers in South and southern Africa14.
Agro-food system dynamics are very different in South
Africa and between different countries and regions in
the rest of the continent. In South Africa, agricultural pro-
duction is concentrated amongst a small core of highly
capitalised large producers, and this concentration is
replicated at each node in agro-food chains, including
food retailing. In contrast, agricultural production on
the rest of the continent is characterised by many small
farmers, with dispersed food distribution and retail sys-
tems. There is some concentration but not to the extent
of South Africa.
According to Coriolis (2001: 55) only 20% of food retail
in southern Africa was through supermarkets in 2000.
Most food sales went through local market and home
production, small grocers and convenience stores, and
‘cash and carry’. Kenya has a concentrating indigenous
supermarket sector, with about 20% owned by large
chains. This is still small compared to South Africa, but
big in comparison with its neighbours. There has been
little ongoing research quantifying the extent of retail
expansion into Africa since Weatherspoon and Reardon’s
(2003) landmark overview. All the big South African
retailers have expanded into Africa in the past decade
or so. In 2007, 15% of stores owned by the big four
South African supermarkets in the region were outside
South Africa (Emongor and Kirsten 2009: 2). In 2011,
Massmart was operating in thirteen countries in sub-
Saharan Africa through four divisions (Massdiscounters,
Masswarehouse, Massbuild, Masscash) comprising 235
stores in Botswana, Ghana, Lesotho, Mozambique, Na-
mibia, Nigeria, Swaziland, Tanzania, Uganda and Zambia.
Shield wholesalers’ outlets are also found outside South
Africa (Fastmoving 2011). Expansion into Africa is mainly
driven by relative saturation in the South African market
and intense competition, combined with higher potential
margins in Africa (Reardon 2005: 6).

57
Smallholder and agro-food value chains in South Africa
The procurement and
distribution of fresh fruit
and vegetables
Fresh produce tends to lag overall food retail concentra-
tion, mainly because of local consumption ‘habits’, and
that freshness, convenience (closer to residential areas)
and lower cost of produce at smaller shops and fresh
produce markets outweigh the advantages of supermar-
kets until the latter are able to realise economies of scale
(Reardon 2005: 4).
Historically, the marketing of most Fresh fruit and veg-
etables (FFV) was not regulated in South Africa. Control
boards for citrus, deciduous fruit, potatoes and bananas
were established from the 1930s to the 1950s. Citrus,
deciduous fruit and bananas were controlled through
one-channel pool schemes, while potatoes had a oor
price scheme (Bayley 2000: 19). The schemes essen-
tially limited distribution outlets in the hope of achieving
price stability and increasing efciency (NAMC 1999: 20).
These schemes were abolished before or at the time of
the 1996 Agricultural Marketing Act. For the remainder of
fresh produce fourteen national fresh produce markets
(NFPMs) were established in the late 1960s under the
control of local municipalities. Agents acting on behalf
of producers negotiated privately with buyers, including
wholesalers, retailers, hawkers, consumers, processors
and institutional buyers (NAMC 1999: 22). This system
still works more or less in the same way, with registered
agents acting on behalf of producers. The extent of con-
trol of producers over agents is directly proportional to
economic size, although agents do bring specialised
knowledge about the marketing of fresh produce.
A distinction should be made between procurement and
distribution. The former is about nding the product, and
the latter is about bringing the product to the store. Nev-
ertheless, these often go hand in hand, with wholesalers
performing both functions.
There are a number of channels through which super-
markets procure fresh produce. They can either go di-
rectly to the farmer or source through an intermediary.
Intermediaries can either be independent agents, who
source produce and then approach the supermarkets
to establish a supply relationship, or the NFPMs, where
retailers or their agents may come to the markets to pur-
chase their requirements.
The fresh produce markets still exert an inuence even
though they have lost power since deregulation in the mid
1990s. In 1998, the NFPMs accounted for around 55%
of total fresh vegetable production (NAMC 1999: 26). Ac-
cording to the NAMC (2007: 16), the share of production
traded through the NFPMs of potatoes, tomatoes, cab-
bage, onions, pumpkin and carrots15 dropped from 63%
to 52% between 1993 and 2004. This has gone hand in
hand with retailers sourcing a greater proportion of their
produce directly from farmers. In 1998, direct sales to
trade (including hawkers) constituted 8% of total produc-
tion (NAMC 1999: 26). Although the major food retailers
continue to source a portion of their fresh produce from
the FPMs, this was as little as 10% of total procurement
by 2007 (Bienabe and Vermeulen, 2007: 3). Neverthe-
less, the larger fresh produce markets in Johannesburg,
Cape Town, Tshwane and Durban remain price formers
for many horticultural products.
Retailers will closely monitor procurement, increasingly
through category managers or agents contracted by the
supermarket. Category management16 refers to com-
bining similar products (e.g. fresh produce or even a line
of fresh produce) into a distinct category (“a group of mu-
tually substitutable items”) on their own, which is then
managed as a separate business within the larger retail
business. It is associated with the development of a more
collaborative relationship between retailer and supplier,
with retailers drawing on suppliers for business ideas
to grow the category as a whole (rather than just indi-
vidual products). Data analysis and the ability to turn the
information into actionable strategies are at the core of
category management. Two developments were required
to make this a reality: the concentration of retailing and
information technology.
Shoprite has its own category manager for fresh produce,
called Freshmark, dedicated to the procurement of fruit
and vegetables for Shoprite. Freshmark receives produce
from suppliers and then repackages for distribution to
Shoprite stores (Louw et al 2008: 5).
There is a shift away from an old procurement model
based on sourcing from traditional wholesalers and
wholesale markets towards a new system with ve key
features:
i. centralisation, which strips out wholesale markets
and brokers;
ii. regional sourcing networks;

58 Smallholder and agro-food value chains in South Africa
iii. specialised wholesalers (category managers); iv)
moving away from spot markets and towards pre-
ferred supplier systems; and
iv. a shift towards higher quality and private safety
standards (Louw et al 2008: 9).
Specialised wholesalers may either continue sourcing
from spot markets (the FPMs) or may enter into preferred
supplier agreements directly with farmers.
According to a US supply chain analyst, “without doubt,
the dominant trend in retail distribution [in the US] in
recent decades has been the dramatic increase in the
retailers’ control of the supply chain” (Wulfraat 2011).
Looking at distribution, he argues that direct store de-
livery (DSD) previously was very important especially for
fast moving consumer goods, including fresh produce.
In DSD, suppliers bring the produce to the retail outlet
and may even pack the shelves and do merchandising
within their category. This can benet the supermarkets
because a bigger proportion of inventory carrying costs
are held by suppliers. This ‘just in time’ and ‘lean inven-
tory’ strategy reduces logistics costs because retailers
hold the product for a shorter time before it gets to the
shelf (Wulfraat 2011). This may go hand in hand with
contractual terms that make the product the property of
the supplier until it is sold (‘buying on spec’), with unsold
products going back to the supplier. For fresh produce,
whoever owns unsold produce at the time it rots or oth-
erwise expires (i.e. its use value expires) carries the risk.
For perishables, DSD can also reduce time in the supply
chain. But DSD has high transaction costs, especially
with many small suppliers, like smallholder farmers, and
overall is a very expensive and inefcient way to distribute
many products (Wulfraat 2011).
DSD increasingly is being replaced with centralised dis-
tribution (the ‘hub-and-spoke’ system), closely linked to
category management as discussed above. Products
are brought to a central point or central points for later
distribution to stores/retail outlets. Centralised distri-
bution benets from economies of scale and allows an
increase in the variety on the shelves (whether for good
or bad). It reduces the number of deliveries to the store.
It is not just about transferring costs to suppliers, but
of taking costs right out of the system. There is uneven
‘adoption’ of these new procurement systems and fresh
produce tends to follow processed products. Simultane-
ously, these systems can be set up more quickly when
done from scratch than when existing procurement and
distribution systems have to be adapted (Reardon 2005:
25).
Wholesalers both source products and bring them to
centralised warehouses where retailers (who have either
contracted them to do this or who are independent) then
purchase from them. This can take the form of a mem-
bership scheme, like Massmart’s buyers’ clubs, or it can
take the form of franchises, where the retailer is branded
with the corporate name, but there is self-management
within the corporate framework. Franchises, like Spar,
also have centralised distribution systems and most of
the products come from the centre, although there is
some allowance for franchise owners to procure a small
proportion from elsewhere.
In franchise operations, stores may have some space
to buy on-the-spot if they so wish. This is very pertinent
for rural areas because it suggests a slight easing of the
formal contractual model followed by the supermarkets.
This can work for or against the farmer. In Limpopo, there
are examples of small farmers who manage to produce
and sell a crop to Spar supermarkets without having a
long term supply contract – such farmers may often ll in
supply gaps left by the more formal contractors (Bienabe
and Vermeulen 2007). On-going research by PLAAS has
shown how such farmers become vulnerable to price
reduction and loss of market when, for example, there
is over production of a commodity. But in general, su-
permarkets will prefer to enter into long-term contracts,
where security of supply is guaranteed. In turn, such con-
tracts serve as incentives for suppliers to invest in assets
that are specically tailored to the retailer’s requirements
(Reardon 2005: 21).
Private labels and private standards are two other fea-
tures of retailing that have grown in importance. Private
labelling, or store branding, is where the supermarket
packages the product with the supermarket’s brand. It
may cut out intermediate packaging and complete the
nal packaging with the supermarket’s brand on the
farm. Store brands are more important for processed
goods than fresh produce because, although there is