Forum for Food Security in Southern Africa

The Role of Market Based Economic Development in

Strengthening Food Security

Colin Poulton and Andrew Dorward

Centre for Development and Poverty Reduction,

Department of Agricultural Sciences, Imperial College London

August 2003
This Theme Paper is one of four commissioned by the Forum for Food Security in
Southern Africa. Each Theme Paper covers a specific topic: institutions; markets; human
vulnerability; and social protection. The aim is to contribute to analytical and strategic
thinking on longer term food security options following the Southern African crisis.

The Theme Papers use economic, institutional and political analysis applied to the
evidence base in the region, and informed where relevant by evidence from other parts of
the South, to explore the policy implications of the priority food security issues raised in
Country Issues Papers produced by resident experts in Lesotho, Malawi, Mozambique,
Zambia and Zimbabwe.

The Forum for Food Security, its consortium members and funders do not necessarily
subscribe to the views expressed in this paper.

To access the other Theme Papers, or to send comments on them, and to find out more
about the work of the Forum for Food Security in Southern Africa, visit
www.odi.org.uk/food-security-forum

2
EXECUTIVE SUMMARY
This paper is intended to stimulate debate about (a) potential roles for market based
economic development in strengthening food security and (b) means for promoting such
development in southern Africa. It provides a theoretical and international review of these
topics, examining the potential of economic growth in different sectors to stimulate and
sustain wider growth, poverty reduction and increased food security, and then difficulties
in achieving pro-poor growth. It focuses on issues not addressed primarily by other theme
papers.
Processes for economic growth, food security and poverty reduction
Market based economic growth is a critical element in medium to long term poverty
reduction and food security. In a poor economy this involves two closely related
processes: the development of markets, and economic growth through those markets.
However, market based approaches to food security and poverty reduction do not work if
markets are not functioning effectively (as the current crisis has demonstrated in many
areas of Southern Africa), and short/medium term policy then has to work within the
constraints imposed by current poverty and lack of market development, and it must
provide alternative, non-market mechanisms to promote secure and low cost food
availability and access. Market based economic development then needs sectors and
activities which meet three conditions: growth must have significant poverty reduction
and/or food security benefits; there must be potential market demand (for goods and
services produced) to support such growth; and there must be potential capabilities for
growth in supply in response to market opportunities.
Different sectors’ and activities’ potential to deliver pro-poor and food security enhancing
growth is evaluated in terms of their characteristics as ‘growth drivers’ and ‘growth
supporters’ in the rural economies in which most of the Forum countries’ poor people
obtain their livelihoods. Both ‘drivers’ and ‘supporters’ are needed, as each is ineffective
without the other: balanced pro-poor growth must match the complementary benefits and
opportunities that these types of growth provide each other. Similarly, to achieve food
security, both food availability and food entitlements are needed, at household level and
at higher levels (local communities and national economies).
Diversity between and within the five Forum countries makes it difficult to identify specific
activities or sectors with the greatest potential for poverty reducing and food security
enhancing growth, but some broad principles emerge. Where supply constraints can be
overcome in medium and high potential areas (with development of technologies and of
input, output and finance markets and appropriate extension services), then smallholder
crop production (cereals, root crops and export crops) provides the greatest opportunity
for driving critical pro-poor, food security improving rural growth. Non-farm activities also
have a critical role to play as growth supporters, to sustain and spread (to poorer
households) the benefits of such growth within the local economy.
Policies for Economic Growth
For a sector to deliver pro-poor growth or food security, it must not only have both high
potential contributions to growth or food security and sufficient opportunities for market
expansion, it must also be able to expand supply in response to these market
opportunities. The structural adjustment and market liberalisation policies of the last two
decades have largely attempted to address structural problems affecting market
opportunities and demand and prices for goods and services: these policies have not,
however, paid sufficient attention to the very serious problems that often inhibit producers’
capacities to respond to and take advantage of market opportunities and increased

3
demand and prices delivered by these policies. Attention to supply constraints is an
urgent priority for economic growth, and the paper examines, for different sectors,
constraints and options for addressing these constraints - although the bulk of attention is
directed to smallholder agriculture, and in particular smallholder food crop production,
where the constraints are particularly problematic but the pay-offs to their solution
potentially very high.
Growing businesses require market opportunities, access to resources to invest, a
reasonable expected return from investments, and acceptable risks. Business
opportunities that meet these requirements are generally thinly distributed and not widely
accessible in poorer, stagnant economies. The lack of business opportunities across and
within different sectors in a poor economy can reinforce each other in the web of
relationships necessary for market based economic growth, resulting in a ‘low level
equilibrium trap’, with particular but easily forgotten problems of economic coordination
risks and risks of opportunism.
Sound and stable macro-economic conditions are a basic pre-requisite for growth in all
sectors of the economy together with good infrastructure and clear and enforceable rules
and procedures for property rights and contracts. Progress on these fronts, however, will
not by themselves be enough to stimulate the business investments needed for rapid and
widespread growth in smallholder food crop production, where the low level equilibrium
problems of poor communications, remoteness, seasonality, small production units, and
poverty are particularly severe.
The record of both market intervention and liberalisation policies in addressing these
problems shows that both approaches have often failed to deliver up to expectations, but
it also demonstrates the need for substantial central non-market coordination and
investment to kick-start growth in poor rural areas. This is not an argument for a return to
the African parastatals of the past: what is needed are new institutional arrangements for
transparent coordination and investment. These should involve a greater role for,
contribution by, and protection of the interests of rural people, farmer organisations,
private businesses, NGOs, and donors, and must avoid the old problems of patronage,
political interference and inefficiency. There are other major challenges too, notably in the
development of lower cost (and lower external input) and less risky but more productive
food crop technologies, managing access to and the distribution of land and, most
tellingly, in the debilitating effects of HIV/AIDS on households, communities and economic
activity.
Investment in smallholder agricultural development, however, needs to be judged and
justified not just in narrow economic and financial terms against alternative investments:
the ‘without investment’ scenarios used in policy analysis must allow for the likely
economic stagnation of these areas without such investment, for the consequent
economic and social costs of safety net and welfare services, and for the dependency
they encourage. Options for agricultural development need to be considered together with
those for social protection, not just to ensure protection for those that are unable to
participate in economic growth, but also to ensure that social protection mechanisms
support the processes of wider economic growth
Rural non-farm activities on the whole face a different set of problems from smallholder
agriculture. The key problem with many of these activities is limited demand for non-
tradeable goods and services unless the economy, and consumer incomes, are growing.
However, difficulties in accessing financial capital, information, skills, and business
networks also prevent poorer households from engaging in these activities. Micro-finance
and business development services can successfully address some of these problems.

4
Many larger scale, formal urban and industrial enterprises are also affected by problems
associated with the low level equilibrium trap, and again improved macro-economic
management, property rights, infrastructure and governance need to be accompanied by
some central coordination to address these problems.
Ensuring Stable Food Supplies: Markets and State Interventions
The paper examines food import and storage strategies as options for improving the
stability of national food supplies and prices, detailing major difficulties facing reliance on
either private sector or state action in each of these areas. Strategies coordinating
importation and storage are needed, without which the incentives for large scale private
sector involvement will be very weak: the state has a key role to play in providing both
strategic direction and critical finance. This presents a paradox, as state intervention itself
often also weakens incentives for private sector involvement. Nevertheless, there remains
a case for a degree of state intervention in staple food markets that goes beyond the
minimal contingency stock to protect against delays in private importation. This conclusion
is not uncontroversial and faces important practical and historical objections which need
further investigation. However, this must recognise the urgent need for rapid and
widespread stabilisation in food availability and prices, and the current lack of alternative
policy approaches to address the critical need for rapid and effective improvements in this
area.

Conclusions
The paper outlines critical elements needed in (particularly) poor rural areas. These
include
• non-market coordination and investment mechanisms addressing low level equilibrium traps
and food security problems in the absence of effective markets ;
• robust and transparent institutions protecting investors against opportunism and rent
seeking;
• attractive business opportunities, especially in smallholder agriculture, which generally
offers the best prospects for driving broad based growth (although it faces significant
institutional and technological challenges and rural non-farm activities are critical in
supporting rural growth);
• stable and transparent macro-economic and (in critical markets) intervention policies; and
• improved communications infrastructure.

A number of principles are suggested to guide the search for and design and
implementation of development strategies to take this agenda forward. We pick out here
two critical challenges:
• Central coordination and investment to address market failures needs to be given much
greater attention. There are strong a priori reasons for only slow and narrow progress in
critical areas of economic growth in very poor and stagnant economies if the market is
largely left to fend for itself. Despite examples of dire failure, the historical record of central
coordination and investment includes dramatic instances of success, whereas reliance on
uncoordinated market development in poor rural economies has few, if any, significant
success stories to its name. Given the serious governance issues facing many of the forum
countries, a major challenge is to develop new models for central coordination and risk
bearing investment to kick start markets.

5
• HIV/AIDS is a major economic problem, apart from all its other dimensions, and must be
addressed as such. We have not addressed them sufficiently in this paper, but the impacts of
HIV/AIDS must be built into all policy analysis and action.

6
1 Introduction
This paper is intended to stimulate debate about (a) potential roles for market based
economic development in strengthening food security and (b) means for promoting such
development in southern Africa. We begin with a theoretical and international review of
these topics. In this we examine the potential of economic growth in different sectors as
regards their ability to stimulate and sustain wider growth and to contribute to increased
food security and poverty reduction. We also consider difficulties in and conditions for
achieving pro-poor growth.. This provides a context against which to examine experience
across the five countries. We finish with some conclusions and questions about the
means and potential for promoting these processes as mechanisms for promoting growth,
poverty reduction and improved food security.

2 Possible processes for economic growth, food security and poverty reduction:

2.1 Introduction
In order to examine the potential for market based economic growth to contribute to food
security in Southern Africa, we need to consider the relations between food security and
other closely related objectives of poverty reduction and of economic growth itself. This
then needs to be related to the potential for growth in different sectors to contribute to
these wider policy objectives, and to the means by which growth may be encouraged in
these sectors. In this section of the paper we develop conceptual frameworks to aid
discussion of each of these issues

2.1.1 Food security, poverty reduction, economic growth and market development
A key distinction at the root of our analysis is that market based economic growth in a
poor economy involves two processes: the development of markets, and economic growth
through those markets. This involves basic presumptions about the importance of
economic growth in livelihood development, poverty reduction, and improved food
security, and of the importance of markets (and of the private sector) in pro-poor
economic growth. While the rest of the paper will explore different aspects of the relations
between these processes and aims, we make here five observations underpinning our
analysis1,2:
• first that economic growth is critical for expanding people’s opportunities to improve
their livelihoods (in terms of incomes, assets and security, for example) and for
governments’ ability to afford welfare service provision (for example health,
education and safety nets);
• second the livelihoods and livelihood opportunities of most poor people are directly
(but not solely) dependent on their involvement in a range of markets as private

1
See Dorward et al. 2003
2
We believe that these observations are consistent with an understanding of poverty that is multi-dimensional -
encompassing vulnerability, (lack of) human development and empowerment, and not just incomes. For example, as
suggested in the related paper on Vulnerability by Ellis, food insecurity may helpfully be seen as a form of vulnerability
and we argue in section 2.1.3 that, particularly where markets function imperfectly, entitlements to food depend on
much more than incomes alone. However, given our theme of market-based development, we do focus primarily on
income generation, touching on human development and empowerment only insofar as they contribute to increased
income generation opportunities.

7
 agents or as employees (and are indirectly dependent on the wider economy for
the demand and supply of goods and services);
• third that most major current and historical poverty reduction processes have
depended on equitable private sector economic growth (but we also note the
importance of actions by other stakeholders – such as CBOs and the state – in
market development);
• fourth that poor people themselves often identify problems with markets as critical
to their livelihoods (but these problems may concern both the absence of markets
and the effects of markets);
• and fifth that in supporting economic growth, markets can provide a highly efficient
mechanism for exchange, coordination and allocation of many resources, goods
and services, but they often fail.

Recognition of both the successes and failures of markets to serve the interests of the
poor is critical to the arguments of this paper. We examine some of the reasons for these
failures and argue that conventional promotion of liberalised competitive markets is
currently misplaced in some sectors in Southern Africa given their current state of market
development. We stress that although improved market access can and should be a
critical driver of sustained and broad based poverty reducing development, it is neither a
magic bullet nor a sufficient condition for such development: other social, political and
technical processes of change are also vital.
Table 2.1, from Dorward and Kydd 2003 sets out a broad framework describing the
requirements for food security, poverty reduction and economic growth to be achieved in
the short/medium term and in the medium/long term in the context of the difficulties that
many of the poor face in accessing important markets in Southern Africa. Recognising
that market based approaches to food security and poverty reduction do not work in areas
where markets are not functioning effectively (as the current crisis has demonstrated in
many areas of Southern Africa), short/medium term policy has to work within the
constraints imposed by current poverty and lack of market development in rural areas and
provide alternative, non-market mechanisms promoting secure and low cost availability of
and access to food. These policies must, however, be designed to promote rather than
undermine development of markets and wider rural growth. Two further, general points
emerge from this: first the need for consistency and coordination of policies across
different policy goals and time periods, and second the need for policies to take account
of and address the context in which they must operate – not only the lack of market
development in the rural economy, as discussed above, but also the historical context
(affecting institutions and people’s expectations and behaviour) and opportunities and
constraints arising from governance, resources, infrastructure, health and education
services and status, HIV/AIDS, gender relations, the environment and current activities in
the rural economy.

8
Table 2.1: Policies and Their Requirements
Policy Goals Requirements for Requirements for
Short/Medium Term Medium/Long Term
Achievement (Policy purpose) Achievement (Policy purpose)
Food security : Secure & Increased food self-sufficiency Increased household &
affordable access to food (household & national) with national food market access
food delivery &/or (low & stable cost, secure,
productivity enhancing safety timely) through wider
nets & humanitarian response entitlements with (mainly)
market based safety nets &
humanitarian response
Poverty reduction: Real Safety nets to increase/ secure Broad based growth with
incomes of the poor increased real incomes & develop/ protect opportunities & wages for
& more secure, through low assets (see above) unskilled rural labour, low food
food costs, higher returns to prices, and safety net &
labour, & safety nets. humanitarian response as
above
Rural economic growth: N/A Macro economic stability & low
Increased levels of local interest rates; growth in agric.
economic activity, with stable & non agric. sectors tightening
income opportunities labour markets and raising real
supporting poverty reduction & incomes with stable /
food security affordable food prices.

With these caveats regarding the roles of market based economic development in
promoting food security, we now consider in more detail types of economic activity that
can drive and support poverty reducing and food security enhancing economic growth.

2.1.2 Characteristics of poverty reducing growth opportunities

In looking for wider economic growth that (first) promotes poverty reduction (we examine
growth and food security later, in section 2.1.3), we need to identify particular types or
characteristics of economic activities that will drive or support such growth. This is the
subject of a long-standing theoretical and empirical literature on growth processes, which
we summarise using figure 2.1 (for recent more detailed reviews see Delgado et al. 1998;
Dorward et al. 2001). A key concept is that some activities have the potential to provide
stimuli to a local economy that will in turn generate strong second-round growth effects,
known as multipliers. Below we term the activities with the potential to provide initial
stimuli “growth drivers” and the activities at the centre of the second-round growth effects
“growth supporters”.
Beginning at the left hand side of Figure 2.1, exogenous change in policies, technologies,
markets, infrastructure and capital (including human and social capital) may set off
changes in prices and productivity in an economy (a region, country, or particular area or
community within a country). Tradable goods and services (those that may be imported or
exported to or from the economy, like primary commodities such as cash crops or
minerals, manufactured goods, and migrant labour) are distinguished from non-tradables
(where there is no external market, like local vegetable production, tailoring or brick
3
making where there are no external markets importing or exporting these) . Productivity

3
In practice the distinction between tradables and non-tradables is often not distinct, varying with (a) the scale or the
boundaries of an area (the larger the area the greater the proportion of non-tradables), (b) its accessibility (the less
accessible the greater the proportion of non-tradables) and (c) the comparative production costs inside and outside the
area. These factors together determine the relationship between local costs on the one hand and the spread between

9
increases in non-tradable activities normally lead to a price fall, as local demand will be
constrained by local incomes. If the good or service accounts for a large share in
households’ average expenditure (a high average budget share, as is the case for food,
and particularly staple foods, for poor households) then this price fall will lead to an
increase in consumers’ real incomes. A consumption (or expenditure) linkage or
‘multiplier’ may then kick in as increased real incomes lead to increased demand for local
(non-tradable) goods and services and this expanded demand generates local
employment opportunities. This further raises incomes, contributing to a virtuous circle
multiplying the benefits from the original gains in real consumer incomes. A tradable price
reduction has similar effects.

4
Figure 2.1. Growth Linkages and leakages in a local economy

Capital/ import intensive

production

Increased
productivity or Increased
prices for employment Supply response
tradables

Consumption,& Inelastic
production supply of non-
linkages tradables
Growth in
productivity for Demand for local non-
Rises in real tradable goods and
high average
wages. services (high marginal
budget share non-
tradables budget shares)

Inequitable assets & Demand for

incomes tradables

These gains, however, are limited by ‘leakages’, also shown in Figure 2.1. If local
consumers use their extra income to buy tradables then this reduces the local demand
stimulus. Even with increased demand, if local producers cannot respond to this (due to
resource constraints, poor markets or high transaction costs), inflationary pressure on
prices will reduce (in real terms) consumers’ increased nominal incomes. Finally, if a local
supply response is achieved from capital or import intensive processes that provide

‘import’ and ‘export’ parity prices on the other. Although these terms are often associated with international trade, they
are equally applicable to intranational trade between different districts or between rural and urban areas.
4
Grey arrows indicate leakages out of the system.

10
returns to only limited numbers of local people, then there will be reduced gains from
increased local employment and earnings.
The effects on producers of increases in non-tradable productivity also need to be
examined and are more mixed. Lower prices may largely off-set producers’ gains from
higher productivity, unless demand is relatively elastic or cost reductions or changes in
technology are sufficient to allow significant expansion of supply with expanded labour
demands and/or entry of new (perhaps poorer) producers into the market. Lower prices
for tradables harm existing net producers of these tradables, with associated losses of
producer income imposing a drag on the positive consumption linkages that may arise
from lower consumer prices and increased consumer real incomes. Higher prices for
tradables have opposite positive effects on producer incomes, similar to the positive
effects of increased productivity in tradable activities. The latter, however, does not have
the same negative effects on consumers as do price increases.
Finally, savings and investment linkages may arise where increased real incomes allow
increased savings and investment in capital, reducing vulnerability and increasing both the
productivity of local activities and the potential elasticity of supply responses crucial to
consumption linkages ‘Leakages’ arise if the returns to local savings and investment are
very low, due to lack of secure investment opportunities or of local financial markets
linking savers with investment opportunities. They may also arise if there are effective
financial markets linking the local economy with other economies, so that either local
activities are already able to access outside sources of capital or locally generated capital
is invested outside the area.
Two further types of linkage associated with economies of scope5 may arise from growth
in production of tradables (Govereh et al. 1999). First, increasing trade flows may lead to
improvements in a range of services, particularly in communications (telecommunications
and transport services for example) with both investment in improved infrastructure and
greater demand for and frequency of services, with greater volumes allowing lower unit
6
costs . These linkages may be described as economies of scope within the local
economy. There may also be economies of scope within livelihoods, such as the
purchase of farm equipment for production of tradables also being used in the production
of non-tradables.
The effects of particular changes on a rural economy and on poor people within it
therefore depend crucially upon the characteristics of goods and services subject to the
initial price or productivity change. This allows us to distinguish between ‘growth drivers’
and ‘growth supporters’ in an economy (see table 2.2), where the ‘growth drivers’ are
those processes that provide an initial stimulus increasing the flow of income in figure 2.1.
The main ‘pro-poor growth drivers’ are
(a) price or productivity increases in tradable products with a high labour input by the
poor. Examples here might include higher prices for cotton, tobacco or textiles,
through greater production, transport or market efficiency or improved quality in
production;
(b) productivity increases in non-tradable products (or falls in price for tradable
products) which have a high average budget share in the poor’s expenditure. An

5
Economies of scope occur where fixed investment costs are shared across more than one economic activity.
6
These greater trade and information flows will also increase the proportion of tradables in the economy, increasing
consumption leakages and reducing linkages. They may also cause producers of non-tradables (for example traditional
goods) to lose market share to imports. These negative effects should be offset by gains to consumers from cheaper or
better goods and perhaps by new opportunities for expanded tradable production.

11
 example here might be adoption of improved varieties of maize, sorghum or millet
varieties by producers;
(c) changes in technology or reduced barriers to entry allowing the poor to produce
non-tradables with high average budget shares or tradables, which they could not
produce on a competitive basis before. Examples here might include the technical
assistance programmes and investment in micro-irrigation that have allowed
smallholder production of horticultural products to increase dramatically in some
forum countries over the past decade, or expansion of micro-credit programmes
that allow small-scale enterprises to compete in areas previously controlled by a
few large suppliers;
or
(d) gains (through mechanisms such as in (a) or (b) above) to significant numbers of
non-poor expanding demand for goods and services produced by the poor.
Productivity and price changes may arise in a variety of ways: productivity may be
affected not only through technical change but also by, for example, improvements in
health, education, development of and access to capital, and skills and information; prices
may be affected by infrastructure and marketing systems as well as market supply and
demand changes, and non-price considerations and incentives should also be taken into
account – for example production and price risks, and social norms.
Table 2.2 Tradable and on-Tradable Characteristics for Driving or Supporting
Growth
Tradables Non-tradables
Characteristics Price falls in external markets for Increase in productivity for high
driving growth high average local budget share average budget share goods and
goods and services lead to gains in services so low prices increase
consumer real income outweighing consumer incomes.
losses to producers High average budget share goods
Prices increases in external markets and services affected by technical
or local technical or institutional or institutional change that reduces
change lead to gains in producer barriers to entry for the poor and/or
incomes increases unskilled labour demands
in production.
Characteristics Affected consumers or producers High marginal budget share goods
supporting have high marginal budget shares & services have low barriers to
growth for non-tradables. entry and high labour demands in
Affected production has high non- production.
tradable input & local labour
demands.
Increased trade flows improve
communications and other local
services.
Affected consumers or producers invest income gains into local activities.
Economies of scope in livelihoods and local services

‘Growth supporters’, on the other hand, are those processes which increase linkages and
reduce leakages, thus increasing multiplier effects to give wider and more sustained

12
growth impacts from growth drivers. ‘Growth supporters’ generally involve consumption
linkages through non-tradables on which people spend a large share of extra income (a
high marginal budget share) provided that production has a high labour content and low
barriers to entry (allowing an elastic supply response providing employment for the poor).
‘Growth drivers’ are essential to get economic growth going, but growth supporters are
also essential to get greater, more sustained, and more widely distributed (and hence pro-
poor) benefits from initial growth increases. Significant and sustained poverty reducing
growth therefore requires the presence of both growth drivers and growth supporters.

2.1.3 Characteristics of growth opportunities that enhance food security

We examine economic growth characteristics needed to improve food security by
breaking down the components of household food security and relating these to incomes
and markets as shown in figure 2.2. Drawing on Swift 1989, figure 2.2 sets out alternative
strategies by which households may achieve food security (defined as secure access to
sufficient food), and the necessary conditions for those for strategies to succeed. Different
households of course use different mixes of these strategies at any one time, depending
on their status and activities, and the effectiveness and stability of each strategy in
delivering secure food access. They often try to retain flexibility to vary these strategic
mixes to match changing conditions.
Household food
production or
stores

Household
access OR
Household
claims/transfers
Household non-food
Household OR assets
entitlements Household OR
exchange
Household non-food
AND income
Local / national
food production
Local/ national or stores Local / national
economy claims/ transfers
availability OR
Local / national OR
entitlements Loca l/ national
exchange (effective
demand for
commercial trade)
Figure 2.2 Determinants of Household Food Security

Although it over-simplifies some important interactions between different elements of food

security, figure 2.2 stresses the need for considering and addressing households’ access
to food in terms of both availability and entitlement (the ability to obtain food from others)

13
at both household and different (local and national) levels of the economy. Entitlements at
these different levels may then involve claims (social, moral or political demands on other
parties able to provide transfers of food or of other resources with which to obtain food)
and/or exchange of assets or income held by the household or (in aggregate) by the
community and sufficient to purchase (or exchange with) food. This helps us to consider
how different aspects and types of market based economic growth may impact on food
security. Thus the development of markets themselves will have far reaching impacts on a
number of elements in food security strategies through their impacts on household food
production; on options for (and prices in) household exchange; on local food production;
on the ability of food markets to meet demand; and on wealth (income and stores)
required to import food into the area. The food security impacts of growth in different
types of economic activity will tend to be limited to more specific elements in figure 2.2:
different activities may affect differently, for example, food production, non-food income,
or accumulation of non-food stores in a specific set of households or in a locality or
community. This analysis raises questions about the roles of markets in promoting food
security in economies with different degrees of market development.
In the next sections of the paper we examine different sectors as regards (a) the potential
value of their contributions to growth or food security and (b) opportunities for market
expansion. Before this Table 2.3 gives a sense of the contributions of agriculture and
various forms of industrial activity to overall GDP in the forum countries, in both 1990 and
2000. We note the particular importance of agriculture in Malawi and Mozambique,
although the share fell between 1990 and 2000 in both countries, and also the fact that
the contribution of industry (manufacturing and other sectors such as mining) fell in
proportionate terms over the period 1990-2000 in Malawi, Zambia and Zimbabwe.

Table 2.3: Relative Contribution of Different Sectors to GDP, 1990 and 2000
Share of GDP to: Agriculture Industry Manufacturing Aid
Country 1990 2000 1990 2000 1990 2000 1990 2000
Lesotho 21.4 17.4 33.7 41.3 14.3 16.6 13.8 4.0
Mozambique 39.4 29.1 20.1 24.2 9.7 12.2 42.8 24.9
Malawi 45.5 38.5 27.9 18.7 18.6 13.9 26.6 25.9
Zambia 19.8 24.2 51.6 26.2 35.7 12.6 18.6 20.4
Zimbabwe 15.6 19.9 35.2 24.3 25.2 16.3 4.0 4.0

South Africa 4.9 3.4 39.7 31.2 23.3 18.9 . 0.4

Rest of SADC 18.4 19.1 33.1 34.1 13.8 14.4 7.5 4.6

Source: World Development Indicators 2002

2.2 Small scale agriculture

The experience of the forum countries with regard to smallholder agricultural production
over past 20 years or so has been mixed. In Zambia, despite enormous potential, per
capita production has been on a gradual decline for most of this period. In Zimbabwe
optimism in the 1980s was replaced by stagnation in the 1990s, whilst in Malawi dramatic
increases in maize production in the late 80s and very early 90s were rapidly lost in the
mid-90s, and neither growth in smallholder burley tobacco nor in root crop production
have averted the slide crisis in the early 2000s. In Mozambique, post-war reconstruction

14
has generated a more positive story, but growth has been from a low base and the
country faces many of the same challenges in achieving sustained agricultural growth as
the other countries just mentioned. With a very different socio-economic and agro-
ecological context, smallholder agriculture in Lesotho has not been able to respond to the
extra demands made of it with the collapse of migrant labour opportunities in South Africa.
The challenges to achieving sustained agricultural growth are dealt with in section 3.
Here, we document the potential for smallholder agricultural growth, if it happens, to
contribute to broader rural growth and food security. Before looking at different crop
categories in turn, we note the findings of Bautista and Thomas 1999, who conducted
CGE analysis for Zimbabwe (based on a 1991 SAM) and calculated GDP multipliers7 for
four different loci of growth. They found that increased income for smallholder farms was
associated with a GDP multiplier of 1.92, whereas for export agriculture (predominantly
the preserve of commercial farms, the subsequent expansion of smallholder cotton
production notwithstanding) they estimated a multiplier of 1.5 - 1.6. Even this was higher
than the highest multiplier associated with a non-agricultural activity, which was that for
labour-intensive light manufacturing (1.49). Bautista and Thomas report that the key
component of the multipliers they calculated were consumption linkages, with
smallholders exhibiting much higher expenditure on locally produced non-tradables than
8
either commercial farm households or urban workers . However, they also note that two
poor groups within the country – workers on commercial farms and low income urban
9
households - benefit less from smallholder agricultural growth than from any of the other
growth paths that they simulated.

2.2.1 Staples

2.2.1.1 Cereals
The contribution of cereal production to both rural growth and food security has often
been considered self-evident and many would argue that this has led to an uncritical and
perhaps unhelpful preoccupation with increasing cereal (and particularly maize)
production within national agricultural policies. Although the Asian Green Revolution
experience is that cereal-based intensification can be a powerful driver for rural growth,
experience with efforts to intensify smallholder maize production in Malawi, Zambia and
Zimbabwe in the 1980s showed that it was a small proportion of (often wealthier)
households in higher potential and/or more accessible areas that produced most of the
resulting marketed surplus. They thus reaped the direct benefits from attractive producer
prices and improved access to inputs and credit, whilst the majority of rural households
(including most of the poorest) were faced with higher consumer prices for maize10. The

7
These are the total increases in GDP that are stimulated by a given increase in income accruing to households or
enterprises in a given production sector. Thus a multiplier of 1.92 indicates that a 1% (or $1) increase in income
accruing to a particular sector leads to an additional 0.92% (or $0.92) increase in GDP over and above the original
stimulus.
8
There is some debate as to whether or not their interpretation of available consumption data overstates the crucial
difference in expenditure patterns.
9
Through the 1990s there were around 300,000 farm workers employed on commercial farms in Zimbabwe. According
to the 1992 census, around 3.3 million people (31% of the total population) resided in urban areas. The proportion of
urban residents classed as poor rose rapidly in the first half of the 1990s, reaching around 40% in 1995-6. By contrast,
over 5.3 million people resided in communal areas in 1992. Around 80% of these were classed as poor in 1995-6
(Nkum 1998).
10
There were a number of reasons why the intensification push was accompanied by higher consumer prices. In both
Zimbabwe and Malawi, producer prices were deliberately set so as to give a remunerative return to producers who
invested in the full package of production inputs, including credit. Meanwhile, in Zimbabwe and Zambia the centralised
marketing system necessary to achieve credit repayment fed into a high-cost, capital-intensive, urban-based milling
industry that produced highly refined flour. Given the choice at liberalisation, poor consumers opted for coarser but

15
indirect benefits to poorer households of this expansion in maize production (increased
demand for agricultural labour, consumption multipliers, greater local maize availability –
at least in the immediate production areas) were clearly not negligible. However, in the
short term at least11, it still seems likely that as many households lost from the drive for
intensive maize production as gained from it.
The lesson from this is not that intensified maize production is a bad thing. Rather, there
needs to be a search for:
• production technologies that permit higher yields and returns at existing producer
prices (for example mixing organic and inorganic soil improvement methods, with
lower use of costly inorganic fertilisers)
• mechanisms for recovering credit that do not tie production activities into high-cost
marketing structures.
• protection for groups who may be disadvantaged by the higher prices needed to
stimulate initial adoption of more intensive, costly and risky cereal production
technologies
We examine these issues in more detail in section 3.
Meanwhile, in lower potential areas, the focus of research and extension should be on
assisting households to reduce their food deficits (by sustaining or raising yields of maize
or other food crops in the face of population pressure), so as to reduce their dependence
on underdeveloped maize markets. The focus is thus likely to be on improved soil and
water conservation, drought tolerance, crops that respond to very low levels of external
inputs etc. The “without intervention” scenario is a bleak one of declining soil fertility,
increasing food deficits and, certainly for households without access to significant non-
farm income streams, increasing food insecurity.
Turning to cereals markets, high transport costs, combined with Western protectionism,
preclude profitable international exports from Forum countries (even South Africa
struggles to be price competitive). However, high transport costs should work to the
advantage of “neighbouring” producers in the southern African regional market. The
regional slide into “net grain deficit” status presents opportunities for any country that can
turn itself around to generate regular surpluses, particularly if initiatives under SADC
and/or COMESA lead to freer regional trade (see section 4). In Zimbabwe the immediate
impact of land redistribution has been to dramatically reduce maize production capacity.
In due course we may expect many of the newly resettled households to become
significant maize surplus producers. However, based on the 1980s resettlement
experience, this may well only happen after 5-10 years (depending on the speed with
which support services can be provided).
Finally, sorghums and millets, as well as being more drought tolerant than maize, can
have high marginal budget shares amongst poor rural households, so may function as
useful supporters of growth. There are also some market opportunities related to brewing
(both traditional – see recent work by Bryceson - and industrial) and to livestock feed,
which would tend to increase the tradability of small grains. Whilst some higher yielding

cheaper hammer-milled flour, whilst rural consumers were able to buy their maize (grain) direct from neighbouring
producers, rather than from shops supplied by the urban industrial millers.
11
Note that, in Green Revolution Asia, benefits also took time to diffuse through the population, starting with better-off
producers, then spreading to smaller producers in the same areas and gradually also to labourers and consumers. As
time went on, new technologies were adapted for (higher potential) rainfed areas. Work on technologies appropriate for
semi-arid areas has proceeded with a further lag. By contrast, it should be remembered that the cereal-based
intensification experiences in southern Africa only lasted 5-10 years.

16
varieties have been developed, making these available to producers is a major challenge
(Tripp 2000).

2.2.1.2 Root crops

These, especially cassava, are increasingly important in the forum countries. However,
with the possible exception of drought years, they remain more important for own
consumption than for sale. For households with low labour availability (due to age,
absenteeism or the impacts of HIV/AIDS), root crops have the major benefit of requiring
low labour input. Their production is also less seasonal than that of grains. Whilst
supplying high quality planting material is a challenge, as are post harvest and marketing
issues associated with storage, processing and transport, they are less demanding than
cereals in terms of their requirements for other inputs, credit and extension support.

2.2.2 Cash crops

There is a longstanding debate over the relationship between food and cash crop
production (Maxwell and Fernando 1989, Govereh, et al. 1999). Whilst acknowledging in
particular that there can often be undesirable gender impacts to expanded cash cropping
activities, the current authors see significant potential for expanded cash crop production
to contribute both to rural growth (through consumption multipliers, expanded labour
demand, and economies of scope mentioned earlier in section 2.1.2) and to household
food security (through generating cash with which to buy food or inputs, improving
household access to fertiliser and animal traction etc).
As with intensifying maize production, expanding cash cropping activity will often bring
most direct benefits to middling and wealthier smallholder households. Benefits to poorer
households are more likely to be indirect (through multiplier effects). However, ongoing
household modelling work for Zimbabwe by Poulton (using 1996 prices) suggests that,
whereas the benefits of higher maize prices to producers were generally offset by the
accompanying higher prices for poor consumers, increases across all commodity prices
(including groundnuts, sunflower, market horticulture and cotton):
• Led to significantly higher demand for hired labour than raising maize prices alone
• Compensated many households for the impacts of higher maize prices
• Led to a decline in estimated income poverty figures, compared with a rise when
maize prices alone were raised12.
Dorward 2003 reports similar findings in Malawi. This suggests that smallholder cash
crops can be significant drivers of growth provided that there are sufficiently large
international markets ready to buy smallholder produce at competitive prices, and that
smallholders can gain an increasing share of those prices through better infrastructure
and access to more efficient marketing systems. These requirements are explored in
more detail in what follows.

2.2.2.1 Traditional cash crops

Cotton is the most important smallholder cash crop in the forum countries taken as a
whole. Over the past decade production has expanded significantly in Mozambique,

12
The poverty estimates primarily capture the direct impacts of changed cropping activities within different household
types. Second order impacts, through labour markets and consumption multipliers are not allowed for in these results,
but are likely to lead to greater poverty reduction impacts.

17
Zambia and Zimbabwe (including in parts of semi-arid NR4), despite the current problems
of depressed international markets as a result of chronic over-supply. (These are due
largely to protectionism in the US, China and a few other countries). By contrast, the
cotton sector in Malawi is in decline. If international prices do not remain depressed for
too long, the cotton sectors in Zimbabwe and Zambia are well run and capable of
continued expansion. Indeed, cotton is likely to be adopted by many of Zimbabwe’s
recently resettled households, particularly if prices begin to recover. The Mozambique
sector faces greater challenges, related mainly to sector organisation.
Burley tobacco is the major smallholder cash crop in Malawi. Although the worst de-
regulation fears have not been realised - of generally depressed prices due to a collapse
in quality when the market was opened up to smallholders – the burley tobacco market in
Malawi faces many challenges regarding quality, prices, market structures, and the
capacity of the market to absorb expanded production given US regulations on use of
imported tobacco.

2.2.2.2 Non-traditional cash crops

The difficulties for smallholders of accessing international horticultural supply chains are
increasingly widely recognised. However, domestic opportunities are available for
smallholders with reliable water supply and well served by roads in both Zambia and
Zimbabwe. Like root crops, fruit trees would appear to hold potential for labour-scarce
households (such as those affected by HIV/AIDS), if they can afford the initial
establishment cost and take a sufficiently long-term investment view13. In Zimbabwe, the
ousting of most commercial farmers could present opportunities for smallholder
horticultural producers in NR1-3, as well as for producers of soyabean and sunflower,
industrial demand for which has been met largely by commercial producers up to now.
A major threat to smallholder access to domestic horticultural markets may be the growing
market share and importance of large supermarket chains in the Southern Africa,
particularly as they target poor urban consumers. With demands for regular and high
specification deliveries, these generally prefer to source their supplies from highly
organised large scale producers or tightly organised smallholder producer groups with
whom they develop a close relationship (Weatherspoon and Reardon 2003).
Interest has been created by the export of legumes to India (see Jones et al. 2002 for
example on pigeon peas from Malawi and Mozambique), and these have soil fertility and
nutritional, as well as cash generating, benefits for the households concerned. It remains
to be seen, however, how far private marketing structures can be encouraged that also
provide ongoing support services to the fledgling producer base, and there are also fears
of over reliance on the Indian market and hence vulnerability to protectionist action by
Indian producers. Production and exports of paprika and birds eye chillies have grown
rapidly in Malawi and Zambia. Valuable though these developments are, there are likely to
be market limits to the extent to which they can expand, and it is important to consider
how to avoid boom and bust prices, as experienced in the past with soyabean prices in
Malawi, for example.

2.2.3 Livestock
There is considerable diversity across forum countries with regard to the importance of
smallholder livestock keeping in rural livelihoods. In Zimbabwe there is broad consensus
that investment in livestock is central both to the path out of poverty for many smallholder

13
The other side of low labour requirements, however, is that multiplier benefits from labour hire are also low.

18
households (e.g. Jackson and Collier 1991) and to food security strategies. However,
whilst most households keep some small stock and a disputed proportion (25% to 50%)
keep some cattle, the importance of livestock keeping lies not so much in its contribution
to income generation as in:
• its contribution to direct household nutrition (supply of milk and meat) and to crop
production activities (the contribution particularly of cattle through draft power and
manure)
• the reliance on livestock as a savings mechanism and as an insurance policy to be
cashed in during a drought14.
A recent study on the control of foot and mouth disease in Zimbabwe estimated that only
2% of smallholder households – mostly in Matabeleland South province and mostly
wealthy - rear cattle for sale on a regular (commercial) basis. Thus, despite expanding
export opportunities to the Democratic Republic of Congo15 (also open to livestock
keepers in western Zambia), smallholder livestock keeping does not constitute a
significant direct driver for growth. On the other hand, a much larger proportion of
livestock keeping households do make unplanned (sometimes distress) sales of single
animals, often to local buyers, and would benefit from generally expanding market
opportunities. This ties in with evidence from Eastern Zambia (Hazell and Hojjati 1995)
that livestock (and horticulture) have high marginal budget shares in local consumption
patterns. Smallholder livestock production may thus represent an effective supporter
(rather than driver) of growth.
At the other end of the scale, in Malawi the importance of livestock (especially cattle) to
rural livelihoods has declined over time, the victim of increasing population pressure and
thus diminishing grazing land, as well as increased insecurity. This presents a grim picture
of the effects of increasing population pressures in the absence of wider economic growth
and increased agricultural productivity, with increasing competition rather than
complementary between livestock and crops.

2.3 Rural non-farm economy

The growing recognition in recent years of the diversity of rural livelihood portfolios
(Reardon 1997; Ellis 2000, Barrett 2001) has highlighted the importance of the wide range
of activities that fall under the general heading of the rural non-farm economy. Given the
diversity of activities that this term encompasses, it is perhaps not surprising that there is
less agreement about their potential contribution to future rural growth than there is about
their current importance to livelihood strategies and food security. Key issues here
include:
• Whether people are currently engaging in non-farm activities because these represent
attractive income earning opportunities in their own right or because the agriculturally-
based activities on which they have traditionally depended are in decline (see, for
example, Bryceson 2000)
• The extent of barriers to entry into more remunerative non-farm activities, excluding
poor households and determining how many rural households can respond to new
non-farm opportunities (Reardon 1998)
• The volume of employment generated by rural non-farm enterprises

14
… the fact that livestock prices can fall precipitately when many households try to sell at once notwithstanding.
15
By contrast, smallholders are unlikely to be able to replace commercial livestock exports to the EU market, where
quality standards are extremely high.

19
• The extent to which demand for the goods and services produced within the rural non-
farm economy is independent of the fortunes of the agricultural sector.

2.3.1 Tradables
In the view of the current authors, a question that has received much less attention than it
should have done in recent literature is the extent to which the rural non-farm economy
produces tradable goods and services and so has the potential to act as an independent
driver of rural growth16.
Considering the forum countries, the main internationally traded good or service that may
be said to fall within the rural non-farm economy is (eco-)tourism. Here we note the
following:
• As is the case with minerals (see below), the basic resource endowment on which
rural tourism depends is largely a gift of God. (People can develop or destroy an
attraction, but not create one). This endowment is not equally distributed across areas,
but the focus countries are all generally well endowed.
• However, even eco-tourism requires a fairly well developed basic infrastructure (to get
close to the sites without major delay or discomfort), even if part of the experience of
the attraction itself is then its undeveloped or “unspoilt” nature. This basic
infrastructure is lacking in many of the rural areas of the focus countries.
• The tourism business as a whole is also critically dependent on political and economic
stability. Thus, in the last couple of years there has been a major decline in arrivals to
Zimbabwe. There is also the question of whether the global terrorist threat will spread
to Southern Africa, having already impacted countries in East Africa.
The international tourism business as a whole is also renowned for its high “leakages”,
with many tourists expecting to be served with similar goods during their travels as they
are accustomed to back home. The eco-tourist movement has attempted to promote
greater local linkages (within a more framework of generally more responsible tourism)
and, within the forum countries, the CAMPFIRE programme provides an example of how
to channel benefits directly to communities. The proportion of tourist attractions that have
the potential to distribute large benefits to local communities in this way is, however,
limited.
Beyond tourism, it is hard to think of many other internationally tradable products:
Amarula, or carbon trading permits? Perhaps more importantly, more research needs to
be done into the current status of, and potential for, goods and services to be sold in
major urban centres. Fisheries, construction materials and fuelwood / charcoal are
obvious examples, but where their harvest is not sustainable, their medium-long term
contribution to growth and poverty reduction is questionable. This highlights the
importance of effective management of common property resources.
A key question given the general lack of production of non-farm tradables concerns how
they may be promoted. In parts of Asia textile producers have taken advantage of low
rural labour costs and good rural infrastructure (roads and power) to invest in production

16
A notable exception to this is Haggblade et.al. (2002), who, in common with this paper, see identification of “key
engines of rural growth” as the starting point for a rural poverty alleviation strategy. They observe that the rural non-farm
economy (RNFE) produces largely non-tradable products and comment that, “… the local demand for the goods and
services produced by the RNFE is conditioned by the level of output and income generated by the tradables sector (or
economic base) of a region’s economy. … In rural areas of the Third World [sic], agriculture typically forms the core of
the rural tradables sector.” (p62-63)

20
units in rural areas. Rural areas in Southern Africa generally lack the necessary
infrastructure and, as discussed below, light industry faces severe challenges even in
urban areas. Nevertheless, this is an issue that needs continued attention.

2.3.2 Non-tradables
The perception of the current authors is that much rural non-farm activity produces non-
tradable goods and services, where supply (for example petty trading of agricultural
products) and demand are dependent on the fortunes of the agricultural sector (this is
discussed in a little more detail in section 2.6 below). If this perception is correct, then
these activities are largely supporters of, rather than drivers of, growth.
An area of rural activity that has been largely ignored until recently is the exploitation of
environmental goods (see for example Cavendish 2000 and Shackleton et al. 2000). In
his work in Shindi ward in southern Zimbabwe (in 1993/4 and 1996/7), Cavendish found
that: “roughly 35% of average total income came from freely-provided environmental
goods” (p16). With the exception of gold panning (see below), almost all of these goods
were apparently non-tradables and/or intermediate inputs into own agricultural or livestock
production activities (e.g. termitaria as fertilisers, livestock fodder). The share of total
income derived from environmental goods was highest for the poorest households,
showing the importance of these activities to food security (entitlements), but in absolute
terms the income derived from environmental goods was highest for better off
households. Nevertheless, Cavendish comments that: “… it is clear that the process of
enrichment in Shindi involves shifting into much more lucrative economic activities. The
counterpart to the declining environmental cash income share is the rising share of cash
derived from remittances and, less significantly, high value crops and large livestock.”
(p14)

2.4 Commercial agriculture

The multipliers calculated by Bautista and Thomas 1999 show that commercial agriculture
does have some potential as a driver for overall economic growth. Whilst generally not as
labour intensive as smallholder agricultural production, certain high value activities (e.g.
floriculture) can be very labour demanding. Moreover, commercial producers can access
lucrative international markets that smallholders find very difficult to break into. Where
effective links can be developed between core commercial estates and surrounding
smallholder outgrowers (a model that was belatedly being explored in Zimbabwe just
before the redistribution programme was accelerated in 2000), the potential for more
broad-based rural growth is high.
Commercial agricultural production may also have a role to play in the opening up and
development of previously under-exploited areas or districts, providing an initial volume of
tradable goods production to justify upgrading of local infrastructure, and a boost to local
employment, on the basis of which the provision of other services can develop. This
highlights the question of land availability, another area where there is considerable
variation within the focus countries. In Zimbabwe, commercial farms have traditionally
made unjustifiably extensive use of the bulk of the highest potential land, much of which
would have been better reallocated to smallholders (Weiner et al. 1985). By contrast, in
parts of Mozambique and Zambia, good quality land is currently underutilised and the
case for encouragement of commercial farm activity may be much stronger (see for
example Duncan et al. 2003).

21
2.5 Industrial and Urban activities
Although our prime focus in this paper is on rural growth (as the bulk of poor households
will continue to live in rural areas in most, if not all, forum countries, for the foreseeable
future), it is important to recognise the general interdependence of rural and urban
activities and, more specifically:
• the importance of urban centers as markets for rural (especially agricultural) products
• the importance of wage employment and remittance flows from urban residents to the
livelihoods of many thousands of rural households.
• the importance of the some industrial sectors in producing tradables and earning
foreign exchange that (if not out of balance to the rest of the economy and causing
‘Dutch disease’) can enhance economic growth and food security by financing food
imports and/or imports of agricultural inputs, and in .
An overwhelming finding of studies of rural livelihoods and poverty (e.g. Cavendish 2000,
Jackson and Collier 1991, Nkum 1998) is that non-poor households are first and
foremost those with wage or remittance income. Such income can be used to fund
consumption (food, clothing, medical expenses etc) and also investment. In high potential
agricultural areas, it appears that this money is often invested in the purchase of animal
traction or inputs (Jayne 1994)17, whereas this is less likely in lower potential areas. (In
Matabeleland, remittances from South Africa in the 1990s have apparently encouraged
investment in solar paneling amongst wealthier households!). By contrast, in Malawi,
where urbanization is low (and migrant labour opportunities to South Africa’s mines dried
up several years ago), few farm households have access to remittances and struggle to
find any means of acquiring cash to purchase inputs. Some (e.g. Barrett 2001, Orr and
Orr 2002) now argue that, in areas of particularly high population density, such that most
farms are too small to generate surpluses even for their own intensification, increasing
access to non-farm employment is the key to stimulating agricultural and rural growth.

2.5.1 Minerals
Mineral wealth is unequally distributed across the forum countries, but it does exist
(primarily Zimbabwe, Zambia, Mozambique). If managed badly, such wealth can be a
curse. There is a danger that profitable extraction of minerals might encourage conflict
(Angola, Democratic Republic of Congo) and/or rent-seeking and generally bad
governance (Nigeria) as rulers do not feel the need to go to the general populace to raise
much in the way of tax revenues. Large mineral exports are also often associated with
overvalued exchange rates. Within the forum countries, the copper story in Zambia is one
of negative trends for most of the past 30 years – is this the curse of dependence on a
“national asset”?
On the other hand, if well managed, minerals can provide the basis for a national growth
strategy (as in Botswana). Whilst the linkages from large-scale mineral extraction to rural
growth are almost certainly lower than those from agriculture, tax revenue from mineral
extraction can fund rural services and foreign exchange earnings can finance imports of

17
Some observers argue that much of the investment in high yielding maize technology in Zimbabwe, even in the
1980s, came from remittances and not as a result of AFC credit, as is claimed by others. Available data suggest that, at
its peak in the mid-1980s was lending to around 77,000 communal farmers (representing c9% of all communal
households) as well as to 13,000 to resettlement farmers. This is probably as many as were producing significant maize
surpluses. However, maize production then rose to a new peak in mid-1990s, despite the absence of credit (Durevall
and Mabugu 2000). It seems plausible that, even if credit played an important role in encouraging the initial technology
adoption, remittances were an important part of the story of continued production expansion in the mid-1990s.

22
food and agricultural investment, whilst the wage and remittance contribution to rural
livelihoods can also be important. On a much smaller scale, Cavendish 2000 found that
gold panning accounted for around a quarter of the “environmental income” in his surveys
in Shindi. Again, however, it is only certain areas that have access to the necessary
resources.

2.5.2 Industrial growth

According to Fafchamps M et al. 2001, rapid and sustained growth in national GDP is only
possible when the manufacturing sector is growing strongly: “This is because the rates of
growth achievable in manufacturing and modern services are much higher than the rate of
growth achievable in agriculture and mining. The historical record shows that industry can
easily grow at 10-20% a year for extended periods of time, while achieving 4% long-term
growth in agriculture is extremely difficult.” (p8). However, there are important debates as
to whether such manufacturing growth can be achieved without prior growth in agriculture.
Fafchamps M, Teal et al. 2001 clearly believe that it can, citing the example of Mauritius.
Others (e.g. Mellor 2000) argue that the more common historical record is one of prior
agricultural growth providing the preconditions for industrial take-off: food supplies to feed
urban workers at low cost, foreign exchange for importation of capital goods, demand for
local manufactured products.
Unfortunately, the recent history in forum countries is one of industrial contraction, rather
than of industry acting as a driver for growth. It has been argued (e.g. by Davies and
Rattso 1996) that trade liberalisation in the 1990s was carried out too rapidly, causing
inefficient, import substituting industries to go out of business, rather than giving them
chance to prepare themselves to face new foreign competition. The resulting decline in
formal employment in Zimbabwe in the early 1990s was an important contributor to the
rise in poverty in both urban areas (Central Statistical Office 1998) and in rural areas,
through the reduction in remittance flows (Campbell and Mukamuri 1998). Similar
problems were also experienced in Malawi, although here accentuated competition from
Zimbabwean goods is seen as part of the problem.
Looking forward, we note that, even with several years of strong agricultural growth,
domestic markets in individual forum countries are unlikely to provide sufficient demand to
stimulate rapid industrial growth. Therefore, industrial growth will have to be built either on
international exports or on regional trade. Unfortunately, there is little chance of
establishing international export industries in landlocked African countries (Fafchamps M,
Teal et al. 2001). This is primarily a question of transport cost. Within our focus countries,
therefore, the only possibilities lie in Mozambique.
However, poor infrastructure, whilst bad for international export prospects, could offer
some protection for regional marketing efforts, if SADC trade liberalization efforts do
achieve a significantly more open regional market. The key issue in regionally-focused
industrial development is: how many countries will gain? A common expectation with
regional trade liberalization is that most industrial and financial investments will gravitate
to one or two countries that offer the best operating environment (and location?)
(Fafchamps M, Teal et al. 2001). Therefore, regionally-focused industrial development is
likely to be good for poverty reduction in some countries, but not in others. This fear has in
the past caused many African politicians to back away from regional trade agreements.
Will SADC liberalisation efforts eventually overcome this or will it still cause current efforts
to stall?

23
2.5.3 Migration and Remittances from within the Region
In southern Africa there is a long history of intra-regional labour migration, which has been
an important determinant of livelihood strategies for numerous rural households.
However, many of these opportunities are now either past or under threat.
Given its size and economic strength (and the presence of its mining industry) South
Africa has historically been the focus for most labour migration. Whilst its importance for
Malawians (mining labour) was already declining by the early 1980s, migration remains
central to many livelihood strategies in Lesotho. The same is true for many households in
southern Zimbabwe. Informal (largely illegal?) migration from Matabeleland increased
rapidly after 1994, but was then curbed by the South African authorities towards the end
of decade. It has probably increased again since the political and economic troubles in
Zimbabwe intensified in 2000. However, such “distress” migration – overwhelming the
capacity of South African immigration authorities to control it - is probably not a
sustainable driver for rural growth.
At the same time, there is the threat of the mass repatriation to Malawi of up to 300,000
Malawian commercial farm workers and their families currently living in Zimbabwe, but
affected by the land redistribution programme. It is not clear how the stretched Malawian
economy could re-absorb them, forcing President Muluzi to become more active in
seeking dialogue with President Mugabe.
On a more positive note, there appears to be a steady flow of migrants (or returning
refugees) from Malawi to Mozambique, and in general this is welcomed northern
Mozambique (Whiteside 1998). A large-scale exodus (involving hundreds of thousands of
people) would be needed to take some of the pressure off the land in rural Malawi,
allowing holding sizes to expand and making intensification of production more plausible.
One of the key issues appear to be access to services in Mocambique as compared with
Malawi, but greater understanding is needed of the reasons why people decide to move
or stay, and of the impacts of migration on recipient communities, on recipient economies,
and on the sustainability of current agricultural practices and systems (Whiteside 1998).

2.6 Conclusions
In this section of the paper we have examined different sectors as regards (a) the
potential value of their growth in contributing to wider growth or food security and (b)
opportunities for market expansion. This is summarised in table 2.4. The ‘grading’ is
necessarily broad and subjective, using criteria discussed in sections 2.1.2 and 2.1.3.
Food security impacts are graded in terms of likely numbers of people affected and the
security of access provided (income from non-tradables is generally considered less
secure than income from tradables, as non-tradable prices are more likely to fall during
times of general food shortage). It must be stressed that table 2.4 shows impacts of
growth in different sectors, it does not assess the current importance of these sectors in
the economy and in supporting food security. The current importance of each sector will
be largely determined by its scale and share in household incomes and livelihood
strategies and in the economy (determining drivers’ importance in underpinning the
economy and supporters’ importance in spreading linkages).
An immediate conclusion suggested by the discussion of this section and by an initial
examination of table 2.4 is that smallholder agriculture has the best linkages for pro-poor
growth, the best food security impacts, and among the most hopeful market expansion
opportunities. Two objections, however, may be made to this conclusion: first that

24
agriculture often accounts for only around 50% of incomes in African rural economies and
livelihoods 18 (and that therefore non-farm activities are critical for growth in rural Africa,
where most of the poor live), and second that large investments in smallholder agriculture
have already been tried, in the 1980s, and they failed to deliver sustained growth. We
briefly consider the first of these issues below, and leave to section 3 a more detailed
examination of the difficulties facing smallholder agriculture (and other sectors).
As noted above, there is increasing evidence of the importance of non-farm income in
African rural economies and in the livelihoods of poor households (see also Barrett 2001).
However, while higher income households often have access to higher return non-farm
activities, diversification out of agriculture may be due as much to declining opportunities
within agriculture as to increasing opportunities in non-farm opportunities (a process of
‘push’ into low return, crowded non-farm activities rather than ‘pull’ into high return non-
farm activities, Bryceson 2000). Dorward and Kydd 2003 describe complex farm / non-
farm interactions both within different household livelihood strategies and (more
importantly) within the rural economy in Malawi, as these affect supply and demand (and
hence prices) for non-farm goods and services and for labour and grain (whose prices are
critical for the livelihoods and food security of the poor). Dorward 2003 estimates that
smallholder agriculture in Malawi may account for 60 to 70% or more of growth drivers in
the informal rural economy, even though it may account for only 30 to 40% of net income.
Ongoing work by Poulton yields similar conclusions for Zimbabwe. However, smallholder
agriculture often benefits from non-farm activities (which may be an important source of
capital for on-farm investment). More fundamentally, while farm activities are critical in
driving growth, non-tradable rural non-farm activities are largely responsible for supporting
growth (although livestock and vegetables have an important role here too), supporting
linkages and reducing leakages to increase and spread benefits from growth drivers.
Complementary growth in both farm and non-farm activities is needed therefore, and the
dichotomy between policy emphasis on farm and non-farm growth is not helpful. However,
without smallholder agriculture driving rural growth, expansion in the rural non-farm
economy has very limited scope for promoting pro-poor growth and food security on its
own.
For a sector to deliver pro-poor growth or food security, therefore, it must satisfy three
necessary conditions: it must have a high potential contribution to growth or food security,
together with sufficient opportunities for market expansion, and it must also be able to
expand supply in response to these market opportunities. It can be argued that the

18
Reardon 1998 summarising results from a large number of case studies in the 1970s to 90s finds average non farm
income shares of 42% in Africa (45% in East and Southern Africa, and 36% in West Africa), although this may mask
wide variation in the importance of non-farm income between households with different incomes and livelihood
strategies in the same area, and between households in different areas (Barrett et al. 2000). Bryceson 1999 finds even
higher non-farm income shares of 55 to 80% across a range of case studies in sub Saharan Africa, with evidence that
non-farm income shares have increased dramatically in many areas in the late 1980s and in the 1990s. Unfortunately
different authors’ definitions of non-farm and off-farm income are not always consistent. Whereas farm income
generally refers to income from a household’s own farming activities, and non-farm income refers to income that is not
gained from direct engagement in agricultural activities, the status of agricultural wage employment is often ambiguous.

25
structural adjustment and market liberalisation policies which have dominated Sub
Saharan Africa in the last two decades have largely attempted to address structural
problems affecting market opportunities and demand and prices for goods and services:
these policies have not, however, paid sufficient attention to the very serious problems
that often inhibit producers’ capacities to respond to and take advantage of market
opportunities and increased demand and prices. It is to these problems that we now turn.

26
Table 2.3 Pro-poor Growth and Food Security Impacts of Expansion in Different Sectors
Market
Linkage Effects Food Security Impacts
Opportunities
Economy/

Lower Food
Household food

Consumption

International
community food

Supporters
sources
sources

Domestic
Unskilled

Regional
Demand
linkages
Drivers/
Growth

Labour
Own

Prices
Exchang Own Exchang
SECTOR productio
e production e
n
Driver (semi-tradable **
Cereals *** ** ** *** *** *** - ** **
staples) *
Smallholde Driver (tradables) *** *** - - *** - *** * ** -
Cash
r Supporter (non-
crops *** *** - - ** - ** - - ?
Agriculture tradables)
Largely non-tradable
Livestock ** * - ** *** * ** - - ?
supporter
Tradables Driver ** ** - - ** - ** * -
Rural Non-
Supporter if high
Farm Non-
marginal budget ** ** - - * - * - - ?
Activities tradables
share
Driver (semi-tradable **
Cereals * * - ** *** *** ** **
staples) *
Commerci
al Driver (tradables) * ** ** - ** - *** ** ** -
Agriculture Other Supporter (non-
* ** - - * - ** - - ?
tradables)
Minerals Driver (tradable) * - - R R - R/*** *** -
Driver (tradable) * - - R R - R/*** * *
Industrial Supporter (non- R R R
* - - - -
tradable)
KEY: *** Very strong / very good; ** Strong / good; * Moderate; - Very small ; ? If economy growing; R remittances.
Note: The table shows linkage and food security impacts of growth in different sectors, not current impacts, which will be largely determined by
sectors’current scale. Impacts of growth in supporters assumes wider processes of growth stimulated by drivers.
3 Policies for Economic Growth

3.1 Conditions required for market based economic growth

We now examine factors that inhibit firm or household businesses’ responses to market
opportunities. We examine these first in terms of general requirements for businesses to
grow, and then consider issues specific to different sectors.
Growing businesses require market opportunities, access to resources to invest, a
reasonable expected return from such investment (taking account of expected sales
revenues, costs of inputs and other investments), and acceptable risks. Although there
are some highly profitable business opportunities in most economies, and people able to
take advantage of them, business opportunities that meet the four requirements outlined
above are generally thinly distributed and not widely accessible in poorer, stagnant
economies: market opportunities are limited by the lack of economic activities and the
limited flow of incomes and expenditures in the economy, and the lack of business activity
itself not only limits demand for final and intermediate outputs, it also constrains input
supply and raises input costs (as low volumes lead to high unit costs and missing
markets), raising costs and lowering returns. These constraints affect all sectors to a
greater or lesser extent – transport, primary industries, business and consumer services –
and reinforce each other in the web of relationships necessary for market based
economic growth. The result is a ‘low level equilibrium trap’ (Rosenstein-Rodan 1943).
The low level equilibrium problems relate particularly to problems of risk. Following
Dorward and Kydd 2002 we identify four basic categories of risk that businesses face:
production risks (from natural shocks, mechanical failures, etc); market price risks (from
general changes in supply and demand affecting input or output prices); economic
coordination risks; and risks of opportunism. The first two risk categories are widely
recognised as being particularly problematic in poor economies which are highly
dependent on natural resources with limited protective capital investment and thin
markets. Coordination risks and risks of opportunism are, however, less widely
understood but are a particular feature in situations of low economic activity and poorly
developed and/or thin markets. They describe respectively the risk of failure of a
businesses’ investment due to the absence of necessary suppliers or buyers when
needed, and the risk of a supplier or buyer who is present taking advantage of their
effectively monopolistic or monopsonistic position to drive a very hard bargain that in the
end makes the business unattractive. Dorward and Kydd describe in more detail these
types of risk and the problems that they pose, with a particular focus on rural areas and
smallholder agriculture, but these risks are found in other sectors, wherever markets are
thin – even if that thinness only affects one critical resource in the supply chain. Risks of
opportunism also arise where businesses are subject to uncertain controls and ‘taxes’
(from bureaucratic regulations, corruption, licensing requirements, and protection rackets).
What then are the critical changes and policies needed to allow businesses to respond to
market opportunities and to contribute to pro-poor growth and food security as discussed
above? We consider first issues common to all sectors, and then specific issues with
regard to agriculture, the rural non-farm sector, and urban and industrial growth.

3.2 Macroeconomic and general policies for market based economic growth
It is widely recognised that sound and stable macro-economic conditions are a basic pre-
requisite for growth in all sectors of the economy. Key elements of this are low and stable
interest rates, low inflation, and stable and realistic exchange rates. Low interest rates
allow cheap and accessible access to finance and hence promote equity investments and
access to production inputs, while low inflation is important for financial markets and
businesses’ financial management and ability and confidence to predict opportunities and
difficulties. Realistic foreign exchange rates, and access to foreign exchange, are
essential for access to input and output markets and to foreign investment, and to the
attractiveness of investments in tradables – seen earlier as critical drivers for growth.
Stability in each of these is necessary for business confidence and reduction in market,
coordination and opportunistic risk.
Such conditions are, unfortunately, notable for their absence in most of the Forum
countries, which have experienced high interest and inflation rates, with large
devaluations, despite nearly twenty years of structural adjustment rhetoric and, to a lesser
extent, policies. Current high interest rates in Malawi, for example are widely recognised
to be a major cause of economic decline. These are the direct result of the government’s
large budget deficit, which is financed by treasury bills yielding 40% or more: very few
business opportunities can provide a risk free return of 40% to match that, so firms and
smallholder farmers are crippled by lack of access to affordable finance.
Achieving macro-economic stability is not easy. Although unnecessary government
expenditures may be rife and easy to identify as a major cause of unsustainable budget
deficits, there are very severe challenges to reducing fiscal deficits, with regard to
governance and to (related) economic difficulties faced in poor economies. Governance
issues are considered in detail in Theme Paper 1, and we do not consider them further
here. Economically, governments face very large demands to provide services and to
invest in infrastructure, with a very weak tax base and limited capacity and efficiency in tax
collection, service provision and expenditure. Attempts to increase the tax base are likely
to bear heavily on businesses and undermine investment incentives (with decentralization
these may be particular problems for poorer districts). There are therefore very heavy
pressures for large budget deficits, which encourage inflation and/or high interest rates.
Reliance on donor financing has its own difficulties. There are related difficulties with
foreign exchange earnings.
In addition to sound and stable macro-economic policies, the need for a favourable
business environment is widely recognised. This requires good infrastructure (roads and
railways, tele-communications, power, water) to reduce general business costs, and clear
and enforceable rules and procedures as regards property rights and contracts, to reduce
risks of opportunism. Petty but onerous business regulations and poor services pose a
serious and widespread problem, raising both costs and risks (Fafchamps M, Teal et al.
2001). While the need for improvements in these areas is generally accepted, they are
difficult and costly to achieve, and often involve challenging entrenched and powerful
private and political interests. Infrastructure, once built at very high cost (for example
US$250,000 per km of new paved road) must be maintained, so greater attention must be
paid to types and locations of infrastructure (for example lower cost rural feeder roads)
and to the institutions and mechanisms for financing, constructing, operating and
maintaining them. Promising examples here include privatization in the
telecommunications sector, with the growth of mobile phone networks, and local
involvement in rural road construction and operation. Property rights are equally
problematic to develop, requiring effective change in formal and informal institutions, the
development of a strong financial and banking system, and clearly defined, transparent
and trusted judiciary and government roles in regulation and in economic management
These interact with each other: the strong banking sector in Zimbabwe, for example, has
been unable to perform its developmental role because of the wider economic crisis.

29
3.3 Sectoral policies for market based economic growth

3.3.1 Agriculture
Smallholder farming activities, and particularly those carried out by poor people in poor
areas, experience to a greater extent than most other sectors problems discussed earlier
that lead to a low level equilibrium trap. Poor physical transport and communications,
remoteness, dependence on an uncertain and seasonal environment and production
processes, small production units, and poverty all lead to thin markets with associated
high costs and coordination and opportunistic risks, not just for farmers but also for
providers of agricultural services to farmers. These problems are particularly important
when purchased inputs are required to raise agricultural productivity, since this increases
farmers’ investment risks in agricultural production, while at the same time demanding
greater coordination along the supply chain (involving input suppliers and input finance, as
well as output traders).19

Past Experience
Many African countries set up monopolistic marketing parastatals in the immediate pre- or
post- independence period, partly as a mechanism to address these problems: private
sector involvement in smallholder agriculture was weak (as regards access to capital and
human resources, and in organisational capacity) and the poor market and infrastructural
development in rural areas presented highly risky and unattractive investment
opportunities, as discussed above. There was also implicit recognition of the major
coordination challenges facing private investors in smallholder agriculture. State
intervention, was seen as a means of addressing all these problems, in that it could
provide a coordination mechanism across trading, infrastructural, research and extension
investments and activities; it could access official finance sources; it could coordinate with
farmers; and it could invest in the organisational and human resource development
20
necessary to develop working systems. At the same time government policies to fix
exchange rates and to control agricultural markets allowed price stabilisation and price
setting to reduce price risk to farmers and to set finance, input and output prices to give
risk adjusted returns high enough to attract investments in intensified crop production, at
least by better off smallholders. Pan territorial pricing allowed these benefits to extend
even to remote rural areas. At its height this approach led to the integrated rural
development projects of the 1970s and 80s, attempting to extend coordination into health,
education, and roads as well as agricultural research and extension, input supply, crop
marketing, and seasonal finance.
The parastatal system can therefore be seen as a specific ‘institutional fix’ (Kydd et al.
2001b) that enabled governments to address the low level equilibrium trap problems
discussed earlier, of low returns and high natural, market, coordination and opportunistic
risks. By committing themselves to coordinating and making investments themselves and
to controlling and stabilising prices, governments took on the risks involved in developing
and delivering financial and input and output marketing services, encouraged coordinated
commitment by farmers, and took over price risks from farmers. Coordination across

19
These problems are reduced when agricultural production can be expanded by increasing extensive production
(increasing area and labour inputs), and where there is strong external demand for crop outputs. This appears to have
been the case in northern Mozambique in supplying maize into Malawi. It is not, however, a situation that predominates
in other Forum countries, except in Zambia: ‘sustainable intensification’ (Reardon et al. 1999) is needed in most areas,
and will become increasingly important in areas which are currently able to engage in more extensive production.
20
In addition to these very practical problems facing private sector led agricultural development, there were other wider
political motives which were very important for the development of parastatals (Dorward and Kydd 2002).
30
credit provision and recovery, input supplies and crop marketing also allowed the
development of mechanisms to reduce incentives for farmers to default on loans, and
thus reduce risks of opportunism. This was the basis for the very high loan repayments
rates achieved in Malawi although the political economy in Zambia and in Zimbabwe did
not take advantage of this feature of the system.
As is well known, these parastatals had a mixed record. They supported, at different
times, large increases in maize production in more favoured maize growing areas, a
growth dynamic in some rural areas, and national (though not household) food security.
These gains were, however, achieved at considerable cost. The parastatals were often
inefficient, ineffective monopolies and state organs of patronage. In Malawi ADMARC
operated for a number of years with commendable efficiency, and the Smallholder
Agricultural Credit Authority maintained for many years an outstanding repayment record
on farmer lending. However, cross subsidisation from cash crops to maize depressed
smallholder cash crop production and earnings and became increasingly difficult to
finance, and this led to steady decline in its effectiveness. Progressive partial privatisation
of maize markets further undermined the system, and then the combination of the
aftermath of the 1991/92 drought with the advent of multiparty politics in the early to mid
1990s dealt the final blow and the system collapsed. In Zambia, and especially
Zimbabwe, the parastatal systems supported large increases in maize production, but the
burden of subsidies, loan defaults and price controls led to unsustainable drains on
government fiscal resources, and, with increasing cash flow problems, inability to deliver
effective services. Direct benefits tended to accrue to better off farmers in more favoured
areas (favoured as regards lower land pressure and more reliable climate) and by-passed
more challenged rural areas where large numbers of the rural poor are located – in
Zimbabwe Natural Regions IV and V were largely excluded from the benefits of the maize
revolution (Poulton et al. 2002). There was also a tendency to rely on state and party
power to command top down coordinated action rather than on positive incentives rooted
in players’ perceived self interests.
However, these problems should not mask the institutional problems that the parastatals
addressed, nor the successes that they sometimes achieved in addressing these
problems. In particular their record needs to be judged against the achievements of the
liberalised markets that succeeded them.
There is an extensive literature describing the different processes of liberalisation in
Southern Africa (see for example Jayne and Jones 1997; Kherallah and Govindan 1999;
Jayne et al. 2002; Chilowa 1998; Deininger and Olinto 2000). There continues to be
considerable debate about the effects of liberalisation, largely due to difficulties (a) in
establishing counterfactuals as regards the effects of alternative policies to liberalisation,
(b) in agreeing how far liberalisation has been achieved, and whether continuing problems
with market development are the result of too little or too much liberalisation, and (c) in
separating out the effects of different elements of liberalisation and of other simultaneous
changes, in, for example, national governance and international markets (for example
Kherallah et al. 2000; Jayne, Govereh et al. 2001; Dorward et al. 2002; Orr and Mwale
2001). It is, however, generally agreed that by the late 1980s the parastatal system was
unsustainable, as it was becoming increasingly inefficient and ineffective, and imposed
growing fiscal demands on government. By pulling back the state from commitments to
carry investment, price and exchange risk, liberalisation solved some problems, removing
the price distortions and operational inefficiency of state managed systems, reducing
fiscal strain, and reducing scope for rent seeking. Positive developments noted include
benefits for maize consumers from competition in maize processing, with expansion of
local hammer mills and reduced transport and processing costs (Jayne and Jones 1997),
31
increased smallholder opportunities in growing some previously regulated cash crops
(burley tobacco in Malawi), and the development of successful private institutional
arrangements supporting smallholder production of certain cash crops (for example
cotton, Gordon and Goodland 2000). Generally building on these cash cropping
opportunities, a new breed of farmer organisations have also achieved some successes in
developing coordination across farmers’ investments and input, output, financial and
extension services (for example NASFAM in Malawi, CLUSA in Mozambique). There have
also been a (small) number of examples of private buyers of cash crops (for example
legumes, sorghum) entering into contractual agreements with (generally larger scale)
smallholder farmers, and providing inputs, seasonal finance, technical advice and
guaranteed end of season prices in return for guaranteed supply of produce (Rusike et al.
2000; Jones, Freeman et al. 2002). Important and significant though these initiatives are
in expanding the opportunities for and volumes in smallholder cash crop production in the
last 10 years, it is often not clear how their achievements and costs compare with those of
the parastatal systems at their peak.
As regards maize crop production, liberalisation appears to have solved the problems of
parastatals’ high cost and patchy service delivery largely by removing these services.
Investment in financial and input service delivery, in produce trading, and in farm
production has withered away, as private sector investment has not replaced the
parastatal system that aspired to support rural investment in maize production. Not
unexpectedly, rural economies are now caught in a low equilibrium trap with systemic
investment risks, and farmers face an absence of financial services and large uncertainty
about maize prices and hence risks as regards profitability of investments in maize
production. Rural financiers face problems of widespread borrower opportunism and
strategic default, with limited investment opportunities for borrowers, against very high
interest rates. Input traders face low demand and output traders face low and uncertain
supply. Consumers also face very uncertain maize prices, making it dangerous to diversify
out of maize production into other more profitable farm or non-farm activities. All investors
also face high degrees of uncertainty from macro-economic instability (with rapidly
changing exchange rates and inflation, and high interest rates), and from often erratic
government and donor policies and interventions affecting food and other markets.
The critique that liberalisation policies have failed to address key low level equilibrium trap
and coordination problems in poor rural areas is supported by experience elsewhere in
the world. Dorward, Kydd et al. 2002 examine irrigated and non-irrigated agricultural
transformations in the 20th century and put forward evidence that external (government)
action played an active role in market coordination in almost every case, with the
establishment of specific institutional arrangements for finance, input and produce
markets, not a reliance on liberalised competitive markets. They postulate a set of phases
through which development proceeded. In Phase 1 (starting from a poor pre-
transformation agriculture and economy) basic interventions were made to establish
technological, infrastructural and institutional conditions for productive intensive cereal
technologies (see Figure 3.1). Once these were in place, uptake was limited to a very
small number of farmers with access to seasonal finance and markets. Agricultural
transformation was then ‘kick started’ (in Phase 2) by government interventions which
enabled more farmers (but still a privileged minority) to access seasonal finance and
seasonal input and output markets at low cost and low risk. Subsidies were required
primarily to cover transaction costs, in the sense that the technologies would have been
basically profitable if farm gate prices were not affected by very high transaction costs and

32
 21
hence marketing margins . Once farmers became used to the new technologies and
when volumes of credit and input demand and of produce supply built up, transaction
costs per unit fell, and were also reduced by growing volumes of non-farm activity arising
from growth linkages. Governments could (and should) have then withdrawn from these
market activities and let the private sector take over (in Phase 3), transferring their
attention to supporting conditions to promote development of the non-farm rural economy.
Difficulties arose in managing these interventions effectively and efficiently, and from
political pressures to include price subsidies with transaction cost subsidies and to
continue with these market interventions and subsidies when they were no longer
necessary (and were indeed harmful).

Figure 3.1: Policy Phases for Agricultural Transformation in Favoured Areas of

Extensive, low
Phase 1. Roads / Irrigation Systems / productivity agriculture.
Research / Extension / (Land Reform)
Establishing the
basics

Profitable intensive technology.

Uptake constrained by inadequate
finance, input and output markets
Phase 2. Seasonal finance
Kick starting Input supply systems
markets Reliable local output markets

Effective farmer input demand

and surplus production.
Phase 3.
Withdrawal
Effective private sector Large volumes of finance and
markets input demand and produce
supply. Non-agricultural
growth linkages.

from Dorward et al, 2002

Asia
Fan et al. 2003, testing this hypothesis, find high agricultural growth and poverty reduction
payoffs from government investments in India in the 1960s to investments in fertiliser
subsidies, in roads, in agricultural research on HYVs, in power subsidies and in credit
subsidies (in order of descending returns). Returns to all of these decline over the three
succeeding decades, to the extent that they become non-significant or negative. Roads

21
An alternative way of addressing the drag from high transaction costs is to provide agribusinesses with direct support
in the form of low cost access to financial and other services while they are building up the volume of their business.
Although this is a potentially very valuable approach, we question whether on its own it can stimulate business activity
on the scale and schedule required to address Southern Africa’s need for agricultural development.
33
are the exception to this, showing consistently high (indeed the highest returns) in the
later decades, while returns to educational investments rise from initially low levels (see
Annex 1).
Dorward, Kydd et al. 2002 also consider reasons for the widespread failure of this
development model in Africa and explain it in terms of more difficult agro-climatic
conditions, lower population densities (although there are also high population density
areas and these are often too crowded and poor to support processes of intensification),
lower human capital, and poor communications infrastructure. Varied and complex agro-
eco systems - with a high proportion of cultivated land subject to soil fertility constraints,
and lack of irrigated land and of land with ‘drought proofing’ irrigation - demand a wider
range of more challenging technological solutions. These in turn suggest higher unit costs
(per hectare and per capita) of agricultural research, information and other services and
greater risks and lower returns to investment (Kydd et al. 2001a). Further difficulties have
arisen from the small size of most African economies (affecting their capacity for
independent action), from failure to make sufficient progress before processes of
patronage and inefficiency became entrenched in and undermined the parastatal system
(Sachs concept of ‘premature greying’ or ageing of state organisations is pertinent here),
and perhaps also from lack of recognition of the importance of women in African
agriculture, and consequent difficulties in developing effective research, extension and
other institutions recognizing this (Gladwin).
Today’s poor rural areas not only carry forward these difficulties from the past, they also
face a number of new difficulties, as a result of changes in global economic and policy
conditions (Dorward, Kydd et al. 2002). Thus world prices for primary agricultural
commodities have fallen steadily since the 1960s, and the globalisation of markets within
the world economy (as semi-tradables become tradables and local prices fall towards
world market prices) has further reduced the terms of trade for poor farmers and may
weaken local demand for non-tradables and its positive effects on consumption linkages
and growth. It is not clear what the overall relative balance will be for poor rural
households between the direct benefits of low food prices and the (indirect) effects of low
product prices on employment and growth in the agricultural sector, although recent
modelling results in Malawi suggest that higher prices accompanied by compensating
welfare transfers to the poor may offer positive long term growth and poverty reduction
opportunities (Dorward 2003). Smallholder agriculture also faces threats from increasingly
high specifications in global (and even local) supply chains (Weatherspoon and Reardon
2003; Kaplinsky 2000; Kydd and Poulton 2000) and from the likelihood of exclusion from
the bio-technology revolution (Pingali 2001; Kydd et al. 2000). Finally, and critically, the
spread of HIV/AIDS threatens and undermines the fabric of national and rural society and
economic activity, drastically increasing the dependency ratio, undermining savings and
services, and attacking the social, human and financial capital of the rural poor and of the
organisations that might support their welfare and economic development (de Waal
2002).
Future Options
What then are the options for overcoming the low level equilibrium trap in smallholder
agriculture in Southern Africa today when faced with historical failure of government led
coordination in the region and current failures of liberalised market coordination? It seems
clear that some non-market coordination is needed in the development and delivery of a
range of services, to reduce risks and engender confidence and coordinated investment.
These services should include research and extension (which should be enabling rather
than prescriptive, recognising diversity and encouraging and supporting experimentation,
34
adaptation and fine tuning), input supply, credit, secure output markets, and, where
livestock are important, veterinary services. Where credit is being provided for the
purchase of inputs in rainfed agriculture, then some explicit or implicit insurance should
also cover loan repayment in the event of crop failure as a result of drought. Some degree
of price support may also be necessary – in terms of producer subsidies for seasonal
credit, inputs, or outputs – in order to raise producers’ expected returns to make risky
investments sufficiently attractive22.
Although the need for these services may be clear, their financing and coordination
presents a major challenge. For cash crops there may be commercial incentives for
buyers and processors to take on the risks of coordinating and delivering these services,
for the food/cash crop system as a whole. Dorward et al. 1998 discuss the conditions
which are likely to promote or undermine such ‘interlocking’ arrangements, and farmer
organisations can play a key role in developing these conditions, provided that broader
institutional, economic, agro-ecological and commercial/ market conditions are supportive.
The challenges are much greater in providing coordinated services to food crop producers
– and in coordinating these with welfare services and national and household food
security systems.
In medium/high potential areas the experience of the 1980s suggests that more intensive
production is possible. However, progress is needed with both technological and
institutional development. Technologies that were more productive with lower risks and
lower costs, would both increase incentives for farmers’ investments and reduce the
reliance on coordinated supply chains, and hence the need for more effective
coordinating institutions (which we discuss below). Some progress has been made with
agronomic practices that involve greater use of organic fertilisers produced on-farm, to
supplement and limit dependence on timely purchases of costly inorganic fertilisers, but
more progress is needed23. In the meantime we urgently need but do not have, to our
knowledge, sustainable institutional models that can deliver to farmers the coordinated
services required to support increased food crop productivity on a large scale with existing
technologies. Our analysis does, however, suggest some pointers:
• the old parastatal system is flawed, but central government resources are needed to take on
some of the costs and risks inherent in the system, and some central authority is necessary to
define and regulate the necessary market structures;
• the private sector cannot be expected to develop and operate such systems on its own (the
costs and risks are too high, the private returns too low, while the balance of power, risks
and benefits between farmers and service providers must be finely balanced) but private
sector capital, expertise and discipline needs to be encouraged and tapped;
• farmer organisations are likely to play a key role in local coordination, but they are delicate
institutions which tend to fail if they are made to carry too many activities and expectations
or if they are scaled up too fast;

22
We emphasise again that these proposals are consistent with experience in successful agricultural transformations in
Asia, discussed above, but recognise that there are very serious questions about (a) how these services should be
delivered, (b) how they should be financed, and (c) how poor consumer welfare and food security should be protected
where grain prices are raised to encourage producer investments. We address these issues in the remainder of this
section, and in section 4 but argue that the importance of these issues means that recognition of these practical
difficulties should not preclude a search for their resolution.
23
Our view is that, in most areas of moderate agro-ecological potential and above within the forum countries, there is a
need both for greater use of organic soil fertility enhancement technologies and greater use of inorganic fertiliser.
35
 • there is no doubt a role for NGOs as advocates, facilitators and service providers, although
again realistic roles and expectations need to be defined and their performance watched
carefully;
• and finally, donors can play an important role supporting in many different ways these
players in their search for effective, transparent and balanced institutional mechanisms for
coordinated service delivery.

The development of successful models therefore needs a willingness of different parties

to cooperate and invest in imaginative ‘institutional innovations’, recognising
complementary strengths and weaknesses. It also needs active processes which
encourage diversity, seek out examples of different models that have been tried in the
field, are open to and share lessons from success and failure, and keep adapting and
responding to emerging lessons and changing conditions. These are complex issues,
involving issues of broader economic policy, agronomy, and, crucially, the governance of
complex and controversial systems with the potential for inefficient, ineffective and
dysfunctional performance from manipulation by particular interest groups seeking large-
scale private gains.
Another factor to be considered here include the relationship between increasing
decentralisation in many Southern Africa countries and the coordination challenge
presented above. It is possible to see different effects of decentralisation on the
development of coordinated service delivery: supporting it through, for example, increased
roles for local knowledge, commitment or control24; or inhibiting it through, for example,
lack of resources, discipline, or wider coordination. We suggest another key issue is the
current weakness of Ministries of Agriculture in many Forum countries. These have been
affected by a particularly marked reduction in resources, power and roles in the last 20
years, beyond the more general difficulties facing the Civil Service as whole. A revitalised
smallholder agriculture needs a much clearer definition of the role(s) of Ministries of
Agriculture, and the development of the necessary financial and human resources to fulfil
these roles. Again, this is not to argue for a return to large scale Ministry of Agriculture
extension and investment projects, but for them to recognise and deliver core coordination
and regulation functions in smallholder agriculture, even if this mainly involves
coordination of other service providers. This may require a major review of the core roles,
structures and activities of Ministries of Agriculture, and of the human and other
resources they need to fulfil these roles.
In low potential areas we suggest that there are currently no technologies for significant
intensification of food crop production and, as suggested earlier in section 2, research and
extension should focus on assisting households to raise food crop yields and reduce their
food deficits through low cost, low external input technologies. A further difficulty for these
areas is that they if they are food deficit areas without the potential to significantly expand
food production, then they will suffer from policies that raise grain prices to promote
agricultural development in high and medium potential areas. This is a problem shared
with poor food deficit households within high and medium potential areas, and in urban
areas. In India this was addressed using ‘fair price shops’, providing low priced grain to
affected households. Recent modelling work in Malawi suggests that higher maize prices
together with compensatory support to poor food deficit households can be effective in
promoting economic growth, poverty reduction and food security (Dorward, 2003).

24
It is possible to imagine, in a situation where decentralization did lead to more accountable government, that there
may be a role for local government development planners in encouraging coordination amongst diverse service
providers (commercial, state, NGO) within their areas of jurisdiction.
36
We conclude that in both medium and low potential areas, but to a much a greater extent
in the latter, smallholder agricultural development must be seen as a costly, long term and
risky undertaking, with no guarantee of a viable and sustainable agricultural economy at
the end of it25. However, the costs of not investing will be that the natural resource base
on which current livelihoods depend will become even more depleted for future
generations, resulting in even greater poverty and food insecurity. We also note that, if
smallholder agriculture is to continue to perform a basic food security function in such
areas, there is the same (possibly greater) need for the long term subsidised supports to
coordinated services and prices that were necessary even in the more favourable
conditions that supported the Asian agricultural transformations. This is an unpalatable
conclusion in an era that emphasises market liberalisation and free trade, but, we argue,
is soundly based on historical experience and on broader theoretical understanding of
development processes (Dorward et al, 2002).
Investment in smallholder agricultural development needs to be judged and justified not
just in narrow economic and financial terms against alternative investments. Rather, the
‘without investment’ scenario used in such analysis must take account of the likely
economic stagnation of these areas without such investment, and of consequent costs of
safety net and welfare services, and of the dependency they engender. The costs of
investment in smallholder agricultural development outlined above may be high, but they
will often be lower than the costs of relief26. In addition, they provide some wider stimulus
to economic growth (through the linkages and multipliers discussed earlier), as well as
greater freedom and dignity for rural people. They can also be designed to promote,
rather than undermine, the development of markets. This suggests that options for
agricultural development need to be considered together with those for social protection,
not just to ensure protection for those that are left out of or unable to participate in
economic growth, but also to ensure that social protection mechanisms support the
27
processes of wider economic growth .
To conclude our discussion of critical production issues in smallholder agriculture we
recognise the importance of smallholder access to land, particularly in Malawi and
Zimbabwe. A critical issue in Zimbabwe is how to get services to people in resettlement
areas. An important opportunity here may be the retention and/or recovery of “core”
estates as a way of getting lower cost and lower risk services to surrounding smallholders.
This may go beyond traditional food and cash crops and allow links into high value, often
international, markets. Where core estates do not exist then if resettled farmers are given
freehold rights it might be possible to buy out a limited number of them on favourable
terms, so as to recreate “core” land holdings for commercial production. In Malawi there is
much less potential for redistribution of under-utilised to land to poor households. A key
question here concerns the minimum scale of holding necessary for viable agricultural
intensification. While this is an important long term consideration, it is essential that any
who relinquish access to land do so on favourable terms that allow them to establish a
viable and food secure livelihood without access to any land. This does not seem to be
possible in the short to medium term, given the current lack of growth in the non-farm

25
It is risky partly because of the nature of smallholder agriculture in these areas, partly because of the coordination
and opportunistic risks inherent in a low level equilibrium economy, and partly because we do not have proven
institutional models for implementing such development.
26
For example, Clay et.al. (2003) report the cost of the 1991-2 drought in southern Africa as US$500 million in logistical
costs of importing cereals into the affected countries, on top of the US$1 billion in actual cereal losses. “There were
also severe wider GDP and agricultural sector impacts over 12 months of at least double this magnitude.” (pviii).
27
For example public works schemes being developed in Malawi promote ‘inputs for work’, where households
constructing roads or schools may opt to receive payments in cash or agricultural inputs – and these systems can be
designed to support private markets in food, inputs and even seasonal borrowing.
37
economy and the variability of maize prices (which makes own production of maize a
critical part of most households’ food security strategies). For areas in Zambia and
Mozambique with large tracts of under-utilised land, we mentioned earlier the potential for
smallholders to benefit or lose from the growth of large scale commercial estates, which
on the one hand may deprive people of land and put increasing pressure on fragile
natural resource, but on the other hand offer employment, skills and access to a variety of
productive and welfare services and opportunities: these issues need proper attention and
careful handling.
Finally, it is important to recognise the potential for increased irrigation in many areas. In
Zambia there is large untapped potential for large scale irrigation, although the past
record of the difficulties and dangers of large schemes is not encouraging. There is much
more widespread potential for small scale irrigation to allow dry season cultivation in
valley bottoms, and using small dams. The latter have been very successful in Zimbabwe.

3.3.2 Rural Non-farm

Rural non-farm activities on the whole face a different set of problems from the
smallholder agricultural problems discussed above, although there are more similarities
with small scale income generation from local sales of horticultural and livestock products.
The key problem with many of these activities, recognised in section 2, is limited demand
for non-tradeable goods and services unless the economy, and consumer incomes, are
growing. However, there are also constraints preventing households from engaging in
these activities: principally access to financial capital, information, skills, and access to
business networks (a form of social capital). Micro-finance services can successfully
address some of these problems, particularly if accompanied by wider business
development services that address thee other constraints. In the absence of such
services there is a tendency for these constraints to be effectives barriers preventing entry
of poorer households to higher return activities (Reardon et al. 2000), and under such
circumstances there tends to be crowding in of these households to low barrier to entry
activities, depressing the prices and returns they achieve.

3.3.3 Urban / Industrial

Many larger scale, formal urban and industrial enterprises are affected by problems
associated with the low level equilibrium trap and with macro-economic instability, as
discussed in sections 3.1 and 3.2. Macro-economic, political and economic stability are all
important for attracting and retaining investment, as is improved infrastructure – some of
which is improving with privatised utilities, although the record may be mixed. It is helpful
to consider competing (but not incompatible) explanations for the limited industrial growth
in Africa. Wood argues that Africa’s basic factor endowment favours primary commodity
production and hence Africa has very limited comparative advantage in industrial
development. Fafchamps et al. 2001 place greater emphasis on high high costs for
utilities and basic services, together with a heavy burden of regulatory and bureaucratic
costs. Possible solutions proposed to these problems are liberalization and, implicitly
addressing low level equilibrium problems, the development of industrial enclaves so that
packages of services can be developed and delivered in one place.

4 Ensuring Stable Food Supplies: Markets and State Interventions

This section is first and foremost about options for ensuring national availability of staple
foods in the five forum countries. However, it is also recognised that:

38
• The choice of strategy for ensuring national availability of food influences the range of
instruments that are available for distributing that food to (poor) consumers within the
country;
• The questions of availability of food and the price at which that food is available cannot
be entirely separated, and price stabilisation is itself an important objective: as already
seen, the price at which staple foods are available to poor consumers is critical to their
food security, and food producers are also negatively affected by high price volatility
between seasons;
• Village level grain banks can ensure food availability and a fair degree of price stability
at local level without reliance on state agencies (which can be problematic!)28.
However, there remains the challenge of scaling-up what works in one area to ensure
more generalised coverage.
At national level, food can be obtained either from domestic production or from imports.
The challenges of raising domestic food production were addressed in section 3. Ensuring
food availability where a significant proportion of food is produced in-country requires
measures to allow for the effects of climatic variability on production (see figures A2.1 and
A2.2 in Annex 2). This can be addressed in three ways: increased diversification of
national food production (to reduce annual variability in food production), importation (to
make up varying annual deficits), and/or inter-year storage of own production. In this
section we examine large scale (national and regional) food import and storage
29
systems . Each of these can be pursued by reliance on markets and/or by state
intervention, the latter involving some combination of state regulation of markets, state
financing of market interventions, and implementation of market interventions by public
agencies.

4.1 Food import systems

We consider first the difficulties and options for importing maize. A number of factors
combine to make the business of importing food particularly challenging. We consider in
turn difficulties associated with
(a) the location of the forum countries in Africa (with consequent effects of high
transport costs on price variability);
(b) the strong regional preference for white maize;
(c) correlation in annual maize production across countries within the region; and
(d) national policies promoting food self sufficiency.
First, four of the five forum countries are landlocked (only Mozambique has a coastline).
This means that grain imported from the world market has to travel long distances to
reach consumers. These long distances are compounded by poor basic road and rail
infrastructure, meaning that there can be delays in importation and also that transport
costs tend to be extremely high. Thus Coulter and Poulton 2001 present figures showing
that the cost of importing maize from the US into Zambia can be more than the basic cost

28
This point was made in the e-discussion based around the earlier version of this paper.
29
We do not address in this paper the large debate in some countries about diversification out of heavy reliance on
maize, beyond linking it to the wider issues of agricultural growth development discussed earlier in sections 2 and 3,
and noting producers’ and consumers’ strong emphasis on maize, an emphasis which is often appropriate to the
circumstance and will not be easily or rapidly changed. This is not to deny that there is scope for greater diversity in
staple food production and consumption, and an important technical, institutional and economic agenda here, but to
recognise the likely continued dominance of maize in food security for the foreseeable future.
39
of the maize itself. Whilst imports from within the region can be considerably cheaper than
imports from the world market, the transport cost element can still be large. Even
transporting maize from Gauteng zone in South Africa to Lusaka can cost more than the
basic cost of the maize in Gauteng. We do note, however, that informal trade over shorter
distances – for example, from northern Mozambique into Malawi – may not incur such
high costs30.
High transport costs mean that there is a wide differential between import and export
parity prices within most of the forum countries (in Mozambique they mean that the north
and south of the country should really be treated as two separate entities for food policy
purposes). Thus, without some form of price intervention, domestic prices can fluctuate
widely both within seasons and between years of good and bad harvest – as
demonstrated dramatically by recent events. This has major consequences for the
incentives facing producers of staple foods. Price uncertainty tends to reduce the
incentive to produce surplus for market (and the region desperately needs more surplus
producers). At the same time, it increases the incentive for households to devote some –
often a large proportion – of their own resources to producing staple foods for own
consumption. If this effort comes at the expense of activities with higher mean returns,
then price instability carries a high inefficiency cost, reducing the chances that households
will be able to lift themselves out of poverty. At the same time, large price variability can
be bad for poor consumers, especially if their coping strategy at times of price spikes
involves selling off assets that they have struggled for years to accumulate.
There are also a number of other problems associated with relying on imports of staple
foods (especially white maize) within the forum countries. Firstly, white maize is not widely
traded on the world market. Hence, not only can availability be a problem (Maasdorp
31
1998), but it carries a price premium over the more common yellow maize .
Within the southern African region, there are questions about the availability of white
maize when it is most urgently needed. An important consideration here is the extent to
which inter-seasonal production variation is correlated across the different countries. If all
countries experience bad harvests, then this limits the extent to which regional trade can
be relied upon to guarantee national food availability in bad years. This is a debated area
and we are not able to give a definitive judgement here. However, Table A2.1 and Figures
A2.1 and A2.2 shed some light on the subject.
From Table A2.1 we see that maize production in some countries is closely correlated.
The strength of correlation across South Africa, Zimbabwe and Zambia is particularly
noteworthy32. By contrast, there are no significant negative correlations within the table.
However, we do note that production in Tanzania shows very little relationship to that in
South Africa, Zimbabwe or Zambia, indicating that trade between these countries and
Tanzania has some potential to smooth inter-seasonal availability within them. Moreover,
there may be important regional differences within countries. For example, production in
north-eastern Zambia is much less susceptible to drought than that in the south of the
country (Maasdorp 1998)33.

30
On the other hand, it may be discouraged by official government policy. We return to this below.
31
With increasing US production of white maize, these problems (the price premium and thin market for white maize)
may be declining.
32
Similar findings are reported by Maasdorp (1998) and Jayne et.al. (1994).
33
The question of transport costs does, however, become critically important when considering trade between southern
Tanzania or north-eastern Zambia on the one hand and Zimbabwe, Lesotho or southern Mozambique on the other
hand.
40
Figure A2.1 plots production in the five forum countries, plus the two central African
countries over the same 30 year period (1972-2002). The huge variability in production in
Zimbabwe throughout this period is clearly seen. Meanwhile, variability appears to have
been increasing in Malawi, Tanzania and, to a lesser extent, Zambia. In Mozambique,
recovery from war appears to have dominated the impact of climate. The critical point to
note, however, is that, in times of recent severe drought (1991-2, 1995-6), all the southern
African countries have experienced major production declines, irrespective of their
experience at other times. An additional dimension to the problem comes when we
consider per capita production trends, rather than total production. As noted in the country
reports, these have been on a downward trend for a decade or more in most of the
countries considered (the clear exception being Mozambique). Thus, within the southern
African region, South Africa is probably now the only country that is a reliable net surplus
producer (across the climatic cycle). However, as Figure A2.2 shows, South African
production is strongly correlated with the combined production totals of the other
countries. Thus, its ability to provide food to the region when it is most needed may be
limited.
Equally importantly, given what was said earlier about the additional cost of importing
grain from the world market, rather than from within the region, there is a major jump in
the cost of ensuring national food availability when the region as a whole switches from
being a net maize surplus area to a net maize deficit area. Arguably, this has now
happened (or is very close to happening). This indicates that national food availability
strategies cannot be considered entirely in isolation from each other. The region needs
some of its countries to be net food surplus producers, but how can net food deficit
countries “pay” net food surplus producers to encourage them to sustain production?
This brings us to another issue of regional policy coordination. Given the difficulties of
relying on trade (international or intra-regional) to guarantee national food availability,
most countries in the regional have fallen back on national food self-sufficiency strategies
that have severely restricted private trade within the region (through administrative
restrictions and simple bureaucratic delays in granting import/export permits) or just failed
to create conditions – e.g. availability of market information – that would encourage
private trade34. Although there is some sign of movement through SADC and COMESA
negotiations and commitments, there is still a way to go before free private trade in staple
foods is a reality and the progress to date could yet stall or be reversed. Maasdorp (1998)
notes that agricultural trade liberalisation has lagged behind liberalisation of trade in other
sectors in almost all regional trade agreements around the world. The coordination
challenge comes in because the benefits of intra-regional trade are most likely to be felt if
all countries jump at once. This is particularly true for small countries. One of the possible
dynamic benefits of regional liberalisation is that it might encourage entry and investment
by large, multinational grain trading companies, whose scale of operations could
significantly reduce the costs associated with storing and transporting grain. However,
such companies are unlikely to be interested in investing in a country such as Malawi
alone. Rather, they might set up operations there if they could be confident that, in years
of surplus, they could move grain out of Malawi and, in years of deficit, they could source
it from neighbouring countries and bring it in. The private investment response to

34
Recent experience with maize exports from northern Mozambique to Malawi illustrate some of these difficulties, with
Mozambiquan maize producers adversely affected by the wide fluctuations in Malawian maize prices, which may
themselves have declined in the late 90s and in 2000 due to the starter pack initiative in Malawi (distributing substantial
quantities of free inputs for maize production). There are important questions here about the impacts of Malawian policy
both upon the welfare of Mozambiquan maize producers and upon the incentives for them to supply food into Malawi.
41
domestic policy reform could, therefore, rest partly on what is simultaneously happening in
neighbouring countries.
There are therefore current serious difficulties with reliance on maize imports to address
inter-year variation in regional and national maize deficits. As a result of the basic
geography, these are likely to remain significant problems even if regional transport
systems are improved and white maize loses its price premium and is more widely traded
35
in world markets . Given these difficulties, we need to carefully consider possible roles
for markets and for governments in importation. Although it is rightly argued that
government and donor interventions often contribute to the problems and risks faced by
the private sector, we would argue first that it is unrealistic to expect good governments to
abrogate all responsibility for involvement in food supply systems, and second that even if
governments and donors were to completely withdraw from these markets, the costs, risks
and potential returns for private actors in supporting large scale and intermittent (but
timely) regional maize imports largely preclude rapid private sector development of the
much thicker markets needed to deliver these imports. Furthermore, due to the large
differences between import and export parity prices, this would not eliminate large price
fluctuations between years (unless the region as a whole was a consistent importer of
maize, not a scenario we should currently be working towards). Recent events in
Southern Africa would seem to support these arguments.
If we cannot (and should not) look to reliance on markets and the private sector alone for
delivering intermittent regional imports of maize, this does not mean that governments
should not look for new ways of working with the private sector in their pursuit of more
effective and lower cost means of facilitating food imports to stabilise national supplies
and prices. The rapid growth and development of SAFEX, with futures and options
markets, and the capital and expertise of large, multinational grain trading companies
mentioned earlier both present potential opportunities which need to be explored, taking
account of different options for government intervention mentioned at the start of this
section: regulation, financing of market interventions, and implementation of market
interventions. We suggest that some state financing of imports and of coordinated import
systems is unavoidable and indeed desirable, but that there are very significant
challenges in developing systems for this.

4.2 Food storage systems

We now turn to consider the possible role for and means of using inter-year storage to
address inter-year variability in national and regional food production. We do not suggest
that this is a strategy that is likely to be pursued on its own - it relates closely to national
food production and to food imports. Again, we face market and state options, with
possible private sector storage by households, small firms, and large grain trading
companies, and state options involving regulation of private sector storage activities,
financing of storage activities by other players, and direct involvement in storage activities
itself.
Although the private sector is heavily involved in storage within each year, the large intra-
year price fluctuations show that there are large costs involved in grain storage and trade,
and there is significant scope for reducing these. There is, however, very little inter-year
storage, and following the discussion of regional and national maize markets in section
4.1, it is difficult to see how large scale inter-year storage by could become attractive

35
Possible impacts of trade liberalisation on grain and white maize markets could also compound importation
difficulties.
42
enough to attract large scale private sector investment, without substantial financial
incentives from the state. We therefore consider now the options facing states when
deciding how far to support inter-year food storage to guarantee national food availability
and price stability. The choice has to take into account technical (e.g. transport and
logistics), economic and political considerations. Our discussion considers the relationship
between policy objectives, storage volumes, and their financial demands on the state. We
are not assuming any particular institutional mechanisms by which the policy may be
implemented (for example, although in the past state interventions have involved direct
action by parastatals, we envisage that storage policies would be implemented better by
private sector agents acting under transparent long term contracts immune to short term
political interference). Our discussion also considers the relationship between storage and
imports, and recognises the impacts of these upon returns and coordination risks in
domestic production and trade.
We distinguish three basic levels of state involvement and address them in ascending
order of interventionism.
The first level is for the state simply to keep a minimum quantity of basic stocks, to tide
the country over in case of delays in importation in response to national or local
shortages. Essentially, however, there is a free private trading regime for grain import and
export. The level of stocks should, therefore, be small enough to send clear signals to the
private sector that they can invest in storage and importation without fear of state
intervention in the market. Private sector confidence would also be assisted by a
transparent policy with respect to stock turnover and what would trigger the release of
stocks onto the market, the aim being to minimise opportunistic political interference in
such decisions, so as to maintain private sector confidence in the trading regime. Whilst
the assumption is generally that a public or contracted agency, typically a strategic grain
reserve (SGR), will store the stocks on behalf of the state, this may not be necessary
where a working system of transferable warehouse receipts is established. In theory,
these could allow the SGR to be reduced to little more than a single office comprising a
couple of analysts, plus the managerial and budgetary capability to buy and sell
warehouse receipts for grain stored in private warehouses! However, this assumes that
the private sector – commercial or NGOs - can then handle distribution when stocks are
released.
The problems with this approach should be readily apparent from our earlier discussion of
problems with heavy reliance on private sector imports: difficulties in achieving regional
consensus on pursuing a free market approach to national food availability, the questions
over regional correlation of harvests, the political risks for politicians if they are seen to
leave the main responsibility for providing grain in the hands of traders whose main
interest is private profit. At root, this approach leaves the state with little influence over the
extent of price fluctuation that will occur. If, in a bad year, the price approaches import
parity (and the imports have to be brought in from the international market), the most
needy consumers may not be able to afford it. (This is particularly likely if they depend for
their entitlements, either directly or indirectly, on agricultural production activity).
Once we assume the need to subsidise food prices in some way in the worst years, then it
is possible to conceive of an economically optimal holding of stocks. On the one hand, if
the level of stocks held is small, the costs of holding these stocks (direct storage costs,
plus opportunity cost of capital bound up in storage) are also small. On the other hand, in
bad years the costs of subsidising maize prices will be higher, as more will have to be

43
 36
imported at a price above the domestic consumer price . In other words, there is a cost
to not holding stocks, as well as a cost to holding them. Critical determinants of the
optimal stock size then become:
• Whether or not food can be reliably imported from within the region (might more
expensive world market supplies have to be relied upon?)
• The prevailing interest rate, which affects the opportunity cost of storage
• To what extent donors will assist in subsidising food prices in bad years.
This leads us to our second level of state intervention: price stabilisation. As well as
limiting the level to which consumer prices can soar in bad years, intervention can also
serve to reduce the fluctuations (both intra- and inter-seasonal) in prices received by
producers wanting to sell their maize. Unfortunately, the considerations surrounding
interventions for price stabilisation are more than just economic. Most obviously, the
governance requirements for the intervention agency, in order to maintain private sector
confidence for investment in the food trading system, are even more critical here than
they are in the case of more limited intervention discussed above. The agency should be
governed by a transparent policy setting out the upper and lower price boundaries that
intervention will seek to defend. In order to prevent an enormous fiscal drain, there also
needs to be a mechanism to adjust the price band in response to medium-long term
trends in world market prices. These conditions have never existed in any of the five
forum countries, where politicians apparently cannot resist manipulating SGRs for short-
term patronage ends.
The “ideal” governance structure for an SGR with official price stabilisation objectives may
be for it to operate as an independent agency, accountable to politicians and the public
only for its performance in keeping prices within an agreed band. This is analogous to the
independent Bank of England’s accountability for meeting the inflation target that the UK
government has set it. However, can civil society or donors persuade national politicians
to agree to the establishment of independent agencies along the lines of this model?
The third level of intervention goes beyond price stabilisation to influencing the mean level
of food prices within the country over time. The level could be influenced in one (or both)
of two ways:
• Subsidising consumer prices
• Raising producer prices.
In India state intervention has aimed for both of these, with targeted consumer subsidies
delivered through state shops in poor (particularly urban) areas, and recent modelling of
Malawian rural livelihoods suggests that this is one of the more effective policy packages
promoting growth and poverty reduction (Dorward 2003). However, similar efforts in forum
countries (for example, in Zambia and Zimbabwe in the 1980s) proved fiscally
unsustainable or, if raising prices without complementary support to poor consumers (as
in Malawi), led to adverse food security and welfare impacts. How far the practice of the
intervention (e.g. how prices were set, the associated stock holding policy, managerial
inefficiency within the system) made it unsustainable in these cases, how far the problem
stemmed from high interest rates due to general macroeconomic instability or whether
such intervention is inherently unaffordable is an unexplored issue. Current conventional
wisdom is that government attempts to intervene in grain markets in Africa were a very
expensive and ineffective part of the state systems that needed major reform as part of

36
This assumes limited private storage within the country even under the liberal trading environment.
44
 37
the structural adjustment and liberalisation processes initiated in the 1980s . However, as
with our (related) discussion in section 3.2 of the problems of agricultural parastatals, we
should not let the difficulties encountered in their management hide the very real
problems that they tried to address. Rather, we should look to learn from both their
achievements and their failures (and look to Asian experience here as well), and we need
to compare their record with the liberalised systems that have followed.

4.3 Ensuring Stable Food Supplies: Conclusions

In this section we have examined food import and storage strategies as options for
improving the stability of national food supplies and prices. We describe major difficulties
facing reliance on either private sector or state action in each of these areas, and suggest
that strategies coordinating importation and storage are needed, and that the state has a
key role to play in this, providing both the strategic direction and critical finance. We argue
that there is a case to be made for a degree of state intervention in staple food markets
that goes beyond the minimal contingency stock to protect against delays in private
importation.
This conclusion is not uncontroversial and we consider below five related objections to it:
The historical record: As noted above, there is a high degree of scepticism regarding the
achievements of such policies in the past.
Concerns about governance: where governments intervene in markets, it is very difficult to guard
these interventions from abuse by opportunistic politicians, so that they do indeed contribute to
policy food security objectives, rather than becoming a costly mechanism for extending patronage.
Arguably, the willingness of politicians to create institutional arrangements that protect a strategic
grain reserve from their arbitrary interference is the key determinant of whether such a reserve is
likely to deliver benefits to the majority of the poor and food insecure.
Concerns about state capacity: government interventions to coordinate and finance food storage
and imports require information and analytical resources which are scarce and costly
Financing: the costs of such interventions can be very high, particularly if the governance and
capacity problems described above are not addressed. Forum countries have very tight budgets, and
the costs of such programmes in the past often contributed to large fiscal deficits
Opportunity costs: tight government budget constraints mean that investments in strategic grain
reserves, even if they are effective, will reduce resources available for other investments, such as
roads, which may gradually reduce the underlying problems of price instability, as well as having
broader economic benefits. (However, note also our earlier point about the opportunity costs of not
holding stocks when bad years come).
These are important issues. However, while the historical record of state interventions in
Africa was often bad, it was not universally bad, and one can also question the record and
potential of systems where the state does not play a role in coordinating and financing
private sector action to stabilise food supplies and prices. Over-reliance on states and
over-reliance on markets both have a poor record, and in low level equilibrium trap
situations where rapid and large-scale change is needed, the record of attempted state

37
However, a preliminary examination of the accounts for ADMARC in Malawi in the 1970s and early 1980s suggests
that its maize price stabilisation activities were not a major drain on its budget, as it was able to offset the holding and
occasional disposal costs of inter-year storage from profits made on intra-year storage, despite reducing the intra-year
variation between immediate post harvest prices (when it made most of its purchases) and later sales. Important factor
in keeping costs down were strong management and prices that did not lead to consistent oversupply, problems faced
by Zimbabwe in the 1980s and, more recently, by Ethiopia.
45
withdrawal may be worse. This is an empirical question, which needs to be addressed.
Irrespective of its answer, however, there is an urgent need to build on past experience to
develop new systems that coordinate and finance more stable national food supplies and
prices.

5 Conclusions, issues and questions

The recent crisis in Southern Africa has drawn stark attention to the failures of
development policies over the last 40 years to create wealth and develop a robust
economy or the markets on which such an economy must depend. Market intervention
and market liberalisation policies have both failed, in different ways, to address
fundamental coordination problems in market development. These failures can be
attributed, in part at least, to a certain degree of naivety about the weaknesses of
government and of markets. Looking forward we now have a better understanding of
these weaknesses, and of ways in which they may be addressed. However, the task is in
other ways more difficult than it was twenty or thirty years ago, as there is more pressure
on limited natural resources, the global environment is perhaps more difficult now than it
was, and there are severe challenges from the impacts of HIV/AIDS. These difficulties
should not, however, be an excuse for inaction: the crisis must be a stimulus to concerted
and committed action to learn from the lessons of the past and to develop and implement
consistent policies that will support development of the fundamentals of a working
economy. This will require, inter alia, long-term investment in institution building, and a
willingness to radically rethink current market liberalisation policies and consider costly
interventions and investments.
We conclude by suggesting some broad principles regarding the critical elements needed
to promote productive investments and wealth creation in (particularly) poor rural areas.
Our analysis suggests that the following are needed for economic development and
wealth creation in poor rural areas:
1. Non-market coordination mechanisms to reduce economic coordination risks in thin
markets, to reduce investors’ vulnerability to and risks from price shocks, and to
reduce consumers’ vulnerability to and risks from food price shocks. As noted in
section 3.3.1, non-market coordination may be encouraged at different levels
(national, district, local) and may be facilitated by different actors (central or local
government, NGOs, farmer organisations, donor interventions). There is no blue-
print for how it should be achieved; appropriate solutions will be location specific
and depend on the organisations involved and their capabilities. Examples of such
coordination mechanisms include: (on the investment side) start-up programmes to
link rural retail enterprises with appropriate wholesalers (e.g. the Care Agent
programme for agricultural input stockists in Zimbabwe), or the Malawian “starter
pack” programme (as originally conceived38), which aimed, inter alia, to boost
effective demand for agricultural inputs in Malawi; and (on the consumer side)
interventions to provide a degree of maize price stabilisation at either national or
village level. As with many other development interventions, whilst it may be
possible to achieve effective coordination at local level, it may then be difficult to
devise the institutional mechanisms for scaling such coordination up. This takes us
into broader questions of governance and accountability and the reform of state

38
In the e-discussion based around the first version of this paper, it was argued that the original objectives of the
“starter pack” programme were lost when donors took control of it, moving it from a pro-market intervention to a much
larger-scale welfare programme.
46
 institutions, without which many private rural markets are likely to remain stuck in
their low level equilibrium traps.
2. Stable and robust institutions that reduce investors’ vulnerability to, and risks from,
opportunism by other actors in the supply chain and by the state and politically
powerful rent seekers. Here again, we can consider institutional arrangements at
different levels. At micro level, examples could include the adoption of recognised
grades and standards in both input and output markets, or the promulgation of
good / best practice in the awarding of licences and contracts. At national level,
state agencies need to bind themselves to codes of good practice in how they
intervene in markets (e.g. publicising intentions regarding official food imports to
minimise adverse impacts on private storage activity) or create institutions that limit
their influence (e.g. an independent strategic grain reserve with clearly laid down
responsibilities for price stabilisation). Farmers’ associations can contribute to
reducing the vulnerability of their members’ investments both at local level, by
negotiating with traders on behalf of the whole group, and at national level, through
representation in appropriate policy fora. However, as noted in section 3.3.1, this in
turn requires that farmers’ associations reach a degree of maturity and internal
accountability that few have yet reached.
3. Business opportunities that offer significant expected returns to investors. We have
argued that agriculture generally offers the best prospects for driving broad-based,
poverty reducing growth, but there are significant institutional and technological
difficulties in raising agricultural productivity (particularly in lower rainfall/ lower
potential areas) and rural non-farm activities have a critical role in supporting rural
growth.
4. Stable and transparent policies governing macro-economic management. A major
problem in several forum countries (e.g. Malawi, Zambia and now Zimbabwe) has
been the high level of interest rates, as a result, inter alia, of fiscal indiscipline. This
is crippling for the agricultural sector, where capital is often only turned over
through a season of 6-9 months (or more). High interest rates render borrowing for
seasonal inputs unprofitable and mean that otherwise desirable strategic grain
stocks can place an enormous burden on national finances.
5. Improved communications infrastructure in terms of roads and telecommunications
linking rural areas to markets
This is a long ‘shopping list’, but we suggest that these are all necessary elements for
broad based poverty reducing growth. It is striking that, with the exception of limited cash
crop business opportunities in some areas, these elements are absent from many rural
areas in the Forum countries.
How can this agenda be taken forward? There are no simple answers to this, but we put
forward the following principles (drawing partly from Dorward and Kydd 2002) to guide the
search for and design and implementation of effective development strategies:
The fiscal costs of rural development must be set against the human, economic and financial costs
of development failure, either continuing poverty and sporadic relief (with unacceptable human
costs that are particularly apparent in the current crisis) or indefinite safety nets. Increased donor
investments should promote appropriate roles for state and other stakeholder action to develop
market based economic growth and food security, while taking account of current market failings.
Institutional innovation is needed to develop more imaginative solutions that reduce risk and
promote coordination, sustainable investment, confidence and market development, addressing the
47
twin problems of state and market failure that have each bedevilled in different ways both the
market intervention and the market liberalisation approaches to development.
Policies and interventions should be designed to be flexible and to address and match the varied and
changing opportunities and constraints of different areas, with different balances of emphasis
between wealth creation and safety nets and between different opportunities and different
institutional mechanisms in different areas. This will involve recognition of diversity between and
within countries, districts, communities and households, and an opportunistic approach seizing
opportunities as they arise and prepared to move forward fast in areas where the way forward is
clearer, while acting more cautiously where problems are more intractable.
Policies and interventions should also be mutually consistent and long term, so that different
players have time to learn how to operate in a stable economic and institutional environment, so
that they have confidence that investments will yield returns in the short, medium and long term,
and so that policies and interventions in different sectors and different areas do not work against
each other. A particularly important issue here is that short and medium term interventions focusing
on relief and poverty alleviation should recognise market failures (and not rely on markets where
markets do not exit) but should then support rather than undermine longer term policies and
processes of market and wealth creation.
Central coordination and investment to address market failures needs much greater attention. This
is critical for rapid and widespread progress in smallholder food crop production, in urban and
industrial development, and in the stabilization of food availability and prices. In each of these
cases there are strong a priori reasons for only slow and narrow progress in poor and stagnant rural
economies if the market is largely left to fend for itself, even if large investments are made in
improved infrastructure and institutional development according to current conventional policy
thinking. However, the historical record of central coordination and investment, despite many
examples of dire failure, includes dramatic instances of success, albeit in more inherently
favourable conditions than those faced in many parts of the Forum countries. On the other hand
market development in poor rural economies, without some form of central coordination and risk
bearing investment, has few, if any, significant success stories to its name. Given the serious
governance issues facing many of the forum countries, one of the major challenges we now face is
to develop new models for central coordination and risk bearing investment to kick start markets.
HIV/AIDS is a major economic problem, apart from all its other dimensions, and must be
addressed as such. The pervasive impacts of HIV/AIDS presents a new and frightening set of
problems that are continuing to emerge. We have not addressed them sufficiently in this paper, but
the impacts of HIV/AIDS on peoples’ capacities and incentives to engage in economic activities
(and on the capacities and incentives of state and other stakeholders) must be built into all policy
analysis and action.

48
Annex 1: Phased Returns to Investment in Agricultural Development in India

Returns in Agricultural GDP (Rps per Rps Spending) by investment and period
1960s 1970s 1980s 1990s
Roads 3.07 3.48 2.92 4.29
Education 1.20 1.49 0.95 1.26
Irrigation Investment 0.51 1.06 1.02 0.07
Irrigation Subsidies 0.69 1.20 -1.18 0.24
Fertiliser Subsidies 4.51 1.26 0.88 -0.65
Power Subsidies 2.26 1.29 0.30 0.07
Credit Subsidies 2.05 0.62 0.08 -0.20
HYV Agric. R&D 3.11 1.89 0.39 n.s.

Number of poor reduced per Million Rps spending by investment and period
1960s 1970s 1980s 1990s
Roads 229 722 717 474
Education 13 130 168 154
Irrigation Investment 42 125 116 6
Irrigation Subsidies 57 142 n.s 24
Fertiliser Subsidies 368 150 100 n.s
Power Subsidies 184 153 34 7
Credit Subsidies 168 73 9 n.s
HYV Agric. R&D 254 224 44 n.s

Cost per poor person lifted above the poverty line (UK£) by investment and period
(Exchange rate: 75 Rp / UK£)
1960s 1970s 1980s 1990s
Roads 58 18 19 28
Education 1,026 103 79 87
Irrigation Investment 317 107 115 2,222
Irrigation Subsidies 234 94 n.s 556
Fertiliser Subsidies 36 89 133 n.s
Power Subsidies 72 87 392 1,905
Credit Subsidies 79 183 1,481 n.s
HYV Agric. R&D 52 60 303 n.s

Fan, Thorat et al. 2003

49
Annex 2: National Maize Production
Table A2.1: Correlations Between Maize Harvests Across Forum and Other Selected
Southern and Central African Countries

Correlations

Congo, Dem Tanzania,

Republic of LESOTHO MALAWI Mozambique South Africa United Rep of ZAMBIA ZIMBABWE
Congo, Dem Republic of Pearson Correlation 1 .217 .420* .517** .060 .820** -.138 -.008
Sig. (2-tailed) . .241 .019 .003 .750 .000 .460 .965
N 31 31 31 31 31 31 31 31
LESOTHO Pearson Correlation .217 1 .364* .517** .409* .172 .104 .139
Sig. (2-tailed) .241 . .044 .003 .022 .354 .578 .456
N 31 31 31 31 31 31 31 31
MALAWI Pearson Correlation .420* .364* 1 .685** .134 .381* .065 .107
Sig. (2-tailed) .019 .044 . .000 .473 .034 .729 .567
N 31 31 31 31 31 31 31 31
Mozambique Pearson Correlation .517** .517** .685** 1 .067 .363* -.199 .012
Sig. (2-tailed) .003 .003 .000 . .722 .045 .283 .949
N 31 31 31 31 31 31 31 31
South Africa Pearson Correlation .060 .409* .134 .067 1 .083 .400* .733**
Sig. (2-tailed) .750 .022 .473 .722 . .656 .026 .000
N 31 31 31 31 31 31 31 31
Tanzania, United Rep of Pearson Correlation .820** .172 .381* .363* .083 1 -.022 -.002
Sig. (2-tailed) .000 .354 .034 .045 .656 . .906 .989
N 31 31 31 31 31 31 31 31
ZAMBIA Pearson Correlation -.138 .104 .065 -.199 .400* -.022 1 .526**
Sig. (2-tailed) .460 .578 .729 .283 .026 .906 . .002
N 31 31 31 31 31 31 31 31
ZIMBABWE Pearson Correlation -.008 .139 .107 .012 .733** -.002 .526** 1
Sig. (2-tailed) .965 .456 .567 .949 .000 .989 .002 .
N 31 31 31 31 31 31 31 31
*. Correlation is significant at the 0.05 level (2-tailed).
**. Correlation is significant at the 0.01 level (2-tailed).

Notes:
1) Based on FAOSTAT annual data covering the period 1972-2002.
2) These figures show correlations between total national production levels in the respective countries, not
per capita production levels, which may give a better indication of the capacity to export. As noted in the main
text, when per capita production is considered, the main additional consideration is that this has been on a
downward trend for a decade or more in most of the countries considered. Therefore, the total surplus
available for export at any one time is reduced.

50
Figure A2.1:

Maize Production in Southern and Central Africa (excl South Africa)

3500

3000

2500

Congo, Dem Republic of

Production (000t)

Lesotho
2000
Malawi
Mozambique
Tanzania, United Rep of
1500
Zambia
Zimbabwe

1000

500

0
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
Year

Figure A2.2:

Maize Production in Southern Africa

16000

14000

12000

10000
Production (000t)

5 Forum Countries
8000
South Africa

6000

4000

2000

0
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002

Year

51
References

Barrett, C., T. Reardon, et al. (2001). "Income diversification and livelihoods in rural
Africa: cause and consequence of change." Food Policy 26(4 (special issue)).
Barrett, C. B., Besfuneh, M., Clay, D. C. and Reardon, T. (2000). Heterogeneous
Constraints, Incentives and Income Diversification Stragies in Rural Africa. mimeo:
40.
Bautista, R. and Thomas, M. (1999). "Agricultural growth linkages in Zimbabwe: income
and equity effects." Agrekon 38(May special issue): 66-75.
Bryceson, D. (1999). Sub Saharan Africa Betwixt and Between. Working Papers. Leiden,
African Studies Centre, University of Leiden, Netherlands.
Bryceson, D. (2000). "Rural Africa at the Crossroads: Livelihood Practices and Policies."
Natural Resources Perspectives(52).
Campbell, B. and Mukamuri, B. (1998). Changing Livelihoods Strategies in Zimbabwean
Communal Areas: Implications for Woodland Use. Harare, Institute of
Environmental Studies, University of Zimbabwe: 28.
Cavendish, W. (2000). "Empirical Regularities in the Poverty-Environment Relationship of
Rural Households: Evidence from Zimbabwe."
Central Statistical Office (1998). Poverty in Zimbabwe. Harare, CSO.
Chilowa, W. (1998). "The Impact of Agricultural Liberalisation on Food Security in
Malawi." Food Policy. 23(6): December 1998.
Clay, E., Bohn, L., Blanco de Armas, E., Kabambe, S. and Tchale, H. (2003). Malawi and
Southern Africa: Climatic Variability and Economic Performance. Disaster Risk
Management Working Paper Series No.7, World Bank, Washington DC.
Coulter, J. and Poulton, C. (2001). Cereal Market Liberalisation in Africa. Commodity
Market Reforms: Lessons of Two Decades. T. Akiyama, Baffes, J., Larson, D. and
Varangis, P. Washington DC, World Bank.
Davies, R. and Rattso, J. (1996). "Growth, Distribution and Environment: Macroeconomic
Issues in Zimbabwe." World Development 24(2): 395-405.
de Waal, A. (2002). 'New Variant Famine' in Africa. Paper presented at SADC VAC
Meeting, ictorua Falls, October 2002.
Deininger, K. and Olinto, P. (2000). "Why liberalisation alone has not improved
agricultural productivity in Zambia: The role of asset ownership and working capital
constraints." World Bank Policy Research Working Paper Series 2302.
Delgado, L. C., Hopkins, J. and V.A., K. (1998). Agricultural Growth Linkages in Sub-
Saharan Africa. FPRI Research Report. Washington DC., International Food Policy
Research Institute.
Dorward, A. (2003). Modelling Farm-Household Livelihoods in Malawi: Methodological
Lessons for Pro-Poor Analysis. Paper presented at the Preparatory Meeting of the
OECD Global Forum on Agriculture: Agricultural Policies in Developing Countries,
the Scope for Using Disaggregated Analysis, Paris, May 19th 2003.
Dorward, A. and Kydd, J. (2002). The Malawi 2002 Food Crisis: The Rural Development
Challenge. Paper presented at 'Malawi after Banda: perspectives in a regional
52
 African context', a conference to mark the retirement of John McCracken, 4-5th
September, Centre of Commonwealth Studies, University of Stirling.
Dorward, A. and Kydd, J. (2003). Work in Progress: Policy Analysis for Food Security,
Poverty Reduction, and Rural Growth in Malawi., Centre for Development and
Poverty Reduction, Department of Agricultural Sciences, Imperial College London.
Dorward, A., Kydd, J. and Poulton , C. (1998). Conclusions: NIE, Policy Debates and the
Research Agenda. Smallholder Cash Crop Production under Market Liberalisation:
A New Institutional Economics Perspective. A. Dorward, Kydd, J. and Poulton , C.
Wallingford, CAB International: 240-265.
Dorward, A., Moyo, S., Coetzee, G., Kydd, J. and Poulton, C. (2001). Seasonal finance for
staple crop production: problems and potential for rural livelihoods in sub Saharan
Africa. Working paper, DFID Policy Research Programme project 'Diverse income
sources and seasonal finance for smallholder agriculture: applying a livelihoods
approach in South Africa'. Wye, Imperial College.
Dorward, A., Poole, N., Morrison, J. A., Kydd, J. and Urey, I. (2003). "Markets, Institutions
and Technology: Missing Links In Livelihoods Analysis." Development Policy
Review 21(3): 319--332.
Dorward, A. R., Kydd, J. G., Morrison, J. A. and Urey, I. (2002). A Policy Agenda for Pro-
Poor Agricultural Growth. Paper presented at the Agricultural Economics Society
Conference, Aberystwyth, 8th - 10th April 2002.
Duncan, A., Macmilan, H. and Simutanyi, N. (2003). Zambia: Drivers of Pro-Poor Change,
An Overview. Oxford, Oxford Policy Management.
Durevall, D. and Mabugu, R. (2000). Maize Markets in Zimbabwe. Stockholm, Swedish
International Development Cooperation Agency (SIDA): 38.
Ellis, F. (2000). Rural Livelihoods and Diversity in Developing Countries. Oxford, Oxford
University Press.
Fafchamps M, Teal, F. and Toye, J. (2001). Towards a growth strategy for Africa. Oxford,
Centre for Study of African Economies.
Fan, S., Thorat, S. and Rao, N. (2003). Investment, Subsidies, and Pro-poor Growth in
Rural India,. Preliminary report presented at workshop on Institutions and
Economic Policies for Pro-Poor Agricultural Growth, Imperial College London, Wye
Campus.
Gordon, A. and Goodland, A. (2000). "Production credit for African small-holders-
conditions for private provision." Savings and development XXIV(1): 55-84.
Govereh, J., Jayne, T. S. and Nyoro, J. (1999). Smallholder commercialisation, interlinked
markets and food crop productivity: cross country evidence in Eastern and
Southern Africa, Michigan State University, Department of Agricultural Economics
and Department of Economics: 39.
Haggblade, S., Hazell, P. and Reardon, T. (2002). Strategies for Stimulating Poverty-
Alleviating Growth in the Rural Non-Farm Economy in Developing Countries. EPTD
Discussion Paper No.92, International Food Policy Research Institute, Washington
DC.
Hazell, P. and Hojjati, B. (1995). "Farm / Non-Farm Growth Linkages in Zambia." Journal
of African Economies 4(3): 406-435.

53
Jackson, J. and Collier, P. (1991). Incomes, Poverty and Food Security in the Communal
Lands of Zimbabwe. Rural Development and Planning in Zimbabwe. N. Mutizwa-
Mangiza and Helmsing, A. Aldershot, UK, Avebury: 21-69.
Jayne, T. S. (1994). " Do High Food Marketing Costs Constrain Cash Crop Production?
Evidence from Zimbabwe." Economic Development and Cultural Change 42(2):
387-402.
Jayne, T. S., Govereh, J., Mwanaumo, A., Nyoro, J. K. and Chapoto, A. (2002). False
Promise of False Premise? The experience of food and input market reform in
Eastern and Southern Africa. World Development 30(11): 1967-1985.
Jayne, T. S. and Jones, S. (1997). "Food marketing and pricing policy in Eastern and
Southern Africa: A survey." World Development 25(9): 1505-1527.
Jayne, T.S., Takavarasha, T. and van Zyl, J. (1994). "Interactions between Food Market
Reform and Regional Trade in Zimbabwe and South Africa: Implications for Food
Security." International Development Working Paper No.48, Department of
Agricultural Economics, Michigan State University, East Lansing.
Jones, R., Freeman, H. A. and Le Monaco, G. (2002). "Improving the access of small
farmers in Eastern and Southern Africa to global pigeon pea markets." Agren
Network paper 120, ODI.
Kaplinsky, R. (2000). "Globalisation and unequalisation: What can be learned from value
chain analysis?" Journal of Development Studies 37(2): 117-146.
Kherallah, M., Delgado , C., Gabre-Madhin, E., Minot, N. and Johnson, M. (2000). The
road half-travelled: agricultural market reform in sub-Saharan Africa. FFood Policy
Report. Washington D C, International Food Policy Research Institute.
Kherallah, M. and Govindan, K. (1999). "The sequencing of agricultural market reforms in
Malawi." Journal of African Economies 8(2): 125-151.
Kydd, J., Haddock, J., Mansfiled, C., Ainsworth, J. and Buckwell, A. (2000). Genetically
Modified Organisms: Issues for DFID's Research Programmes. Symposium on
Genetically-Modified Organisms and Smallholders in the Developing World,
Agricultural Economics Society Annual Conference,, Manchester 15 April 1999.
Kydd, J. G., Dorward, A. R., Morrison, J. A. and Cadisch, G. (2001a). The Role of
Agriculture in Pro Poor Economic Growth in Sub Saharan Africa.
Kydd, J. G., Dorward, A. R. and Poulton, C. (2001b). New institutional economics,
agricultural parastatals and marketing policy. Renewing Development in Sub-
Saharan Africa: policy, performance and prospects. D. Byerlee and Livingstone, I.,
Routledge.
Kydd, J. G. and Poulton , C. (2000). Agricultural liberalisation, commercialisation and the
market access problem. Part II of Killick, Kydd and Poulton, The Rural Poor and
the Wider Economy: the Problem of Market Access, background paper for IFAD
Rural Poverty 2000 report, Rome.
Maasdorp, G. (1998). "Regional Trade and Food Security in SADC" Food Policy 23(6):
505-518
Maxwell, S. and Fernando, A. (1989). "Cash Crops in Developing Countries: The Issues,
The Facts, The Policies." World Development 17(11): 1677-1708.

54
Mellor, J. W. (2000). Faster more equitable growth: The relation between growth in
agriculture and poverty reduction, Harvard University CAER.
Nkum, J. (1998). Comparative analysis of poverty alleviation action plans in Ghana and
Zimbabwe, GTZ.
Orr, A. and Mwale, B. (2001). "Adapting to adjustment: smallholder livelihood strategies in
Southern Malawi." World Development 29(8): 1325-1343.
Orr, A. and Orr, S. (2002). " Agriculture and micro enterprise in Malawi's rural south."
Agren Network Paper(119).
Pingali, P. (2001). The Changing Locus of Agricultural Research: How will the poor fare?
74th EAAE Seminar, Livelihoods and Rural Poverty: Technology, Policy and
Institutions,, Imperial College at Wye, UK.
Poulton, C., Davies, R., Matshe, I. and Urey, I. (2002). A review of Zimbabwe's agriculutral
economic policies: 1980 - 2000. Working paper, Institutions and Economic Polices
for Pro-Poor Agricultural Growth. London, Imperial College.
Reardon, T. (1997). "Using evidence of household income diversification to inform study
of the rural non-farm labour-market in Africa." World Development 25(5): 735-748.
Reardon, T. (1998). Rural non-farm income in developing countries. The State of Food
and Agriculture 1998. FAO. Rome, Food and Agriculture Organisation.
Reardon, T., Barrett, C. B., Kelly V.A. and Savadogo, K. (1999). "Policy reforms and
sustainable agricultural intensification in Africa." Development Policy Review 17:
375 - 395.
Reardon, T., Taylor, J. E., Stamoulis, K., Lanjouw, P. and Balisacan, A. (2000). "Effects of
non-farm employment on rural income inequality in developing countries: an
investment perspective." Journal of Agricultural Economics 51(2): 266-288.
Rosenstein-Rodan, P. (1943). "Problems of industrialisation of Eastern and Southeastern
Europe." Economic Journal 53(210-211): 202-211.
Rusike, J., Sukume, C., Dorward, A., Mpepereki, S. and Giller, K. (2000). The Economic
Potential for Smallholder Soyabean Production in Zimbabwe. Harare, Zimbabwe,
Soil Fertility Network for Maize Based Cropping Sytems in Malawi and Zimbabwe,
CIMMYT.
Shackleton, S., Shackleton, C. and Cousins, B. (2000). "Revaluing the communal lands of
Southern Africa: new understandings of rural livelihoods." Natural Resources
Perspectives 62.
Swift, J. (1989). "Why are rural people vulnerable to famine?" IDS Bulletin 20(2): 8-15.
Tripp, R. (2000). Strategies for Seed System Development in Sub-Saharan Africa: A
Study of Kenya, Malawi, Zambia and Zimbabwe. Bulawayo, ICRISAT: 56.
Weatherspoon, D. D. and Reardon, T. (2003). "The rise of supermarkets in Africa:
implications for agrifood systems and the rural poor." Development Policy Review
21(3): 333-355.
Weiner, D., Moyo, S., Munslow, B. and O'Keefe, P. (1985). "Land Use and Agricultural
Productivity in Zimbabwe." Journal of Modern African Studies 23(2): 251-285.

55
Whiteside, M. (1998). When the whole is more than the sum of the parts: the effect of
cross-border interactions on livelihood security in southern Malawi and northern
Mozambique. A Report for Oxfam GB. Stroud, UK, Martin Whiteside.

56