Jump to ContentJump to Main Navigation
Agricultural Input SubsidiesThe Recent Malawi Experience$

Ephraim Chirwa and Andrew Dorward

Print publication date: 2013

Print ISBN-13: 9780199683529

Published to Oxford Scholarship Online: January 2014

DOI: 10.1093/acprof:oso/9780199683529.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2021. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use. date: 06 December 2021

Malawi: political, policy, livelihoods, and market background

Malawi: political, policy, livelihoods, and market background

(p.62) 4 Malawi: political, policy, livelihoods, and market background
Agricultural Input Subsidies

Ephraim Chirwa

Andrew Dorward

Oxford University Press

Abstract and Keywords

Malawi, a small landlocked country, is highly dependent on the agricultural sector, particularly smallholder agriculture. The structure of production and of the economy have not changed significantly since independence in 1964: smallholder agricultural productivity, which is dominated by maize, remains low while agriculture continues to contribute more than a third to GDP. Over 50% of the rural population in Malawi is classified as poor with fragile livelihoods susceptible to shocks. Attempts to address these socio-economic problems through agricultural and other policies, including fertilizer subsidies, have been central to the strategies and policies of different political regimes. The chapter reviews agricultural policies in the context of changing economic and political environments and the constraints that smallholder farmers face in the use of improved inputs. It is argued that smallholder farmers are locked into a low productivity maize production trap, leading to low incomes and inability to afford inputs at commercial prices.

Keywords:   input subsidies, agricultural development, politics, Malawi, agricultural policy, maize, tobacco

4.1. Introduction

This chapter sets out the background for examination of the Malawi subsidy experience in the rest of the book. We begin with some contextual information about Malawi, including a brief outline of its post-colonial history, structural characteristics, and economic and welfare indicators. This leads on to a more detailed discussion of the interactions between Malawian politics and policy. We then examine critical features of rural livelihoods and the potential impacts of input subsidies in this context.

4.2. The context

At its independence in 1964 Malawi was a small land-locked country with a population a little under 4 million largely engaged in and dependent on smallholder agriculture, with poorly developed infrastructure and little industry or mineral resources. The infrastructure and industry that was there tended to be concentrated, along with the bulk of the population, in the south and, to a lesser extent, in the centre of the country, while the north had a very low population density and very little infrastructure or industry. Cultural and historical differences between regions, however, meant that commitment to and standards of education in the north were higher than in the centre and the south.

Many long term generic features of the country have not changed much over the last 40 years or so since independence, despite major (if somewhat erratic) investments and policies to address them (see, for example, Cammack et al. (2010)). These features include high dependence on agriculture; low productivity in production of maize (the dominant staple crop which accounts (p.63) for around 70% of cultivated area); lack of other exploitable natural resources; isolation and high import and export costs due to its land- locked location and poor external transport systems; poor physical infrastructure; chronic poor health, with very high infant mortality from malaria, water-borne diseases, and mal- and under-nutrition; low levels of literacy and education;1 and broad regional differences in many of these variables. Other elements have emerged more recently as a result of development failures or wider economic, social, and natural processes. These include high population densities and small landholdings (particularly in the south), falling soil fertility, and high rates of HIV/AIDS infection, morbidity, and mortality. A further set of problems emerged from the mid 1990s due to policy and governance failures, and these include the collapse of the industrial economy due to exposure to outside competition; poor macro-economic management with large budget deficits, high interest rates, large devaluations of the Malawi Kwacha (MK), and high inflation rates; high crime rates in urban and rural areas; and weak governance.2

As a result, in 2004/5, Malawi was one of the poorest countries in the world, with 56% of Malawi’s rural population classified as poor and 24% as ultra-poor (National Statistical Office, 2005a), a GNI per capita of around US$160, very low achievements on a range of social and economic indicators (see Table 4.1), and many people characterized by high levels of vulnerability, due to the fragility of their livelihoods, susceptibility to shocks, and large numbers of the non-poor living just above a very low poverty line (Devereux, 2006).

Understanding this depressing backdrop to the emergence of the agricultural input subsidy programme in 2005, the starting point of the programme addressed in this book, requires an understanding of agricultural and other policies in the context of wider political change in Malawi, to which we now turn.

4.3. Politics and policies

The relationship between politics and policies is best considered in relation to the periods of tenure of the three presidents of Malawi since independence: Kamuzu Banda (from 1964 to 1994), Bakili Muluzi (from 1994 to 2004), and (p.64)

Table 4.1. Social and economic indicators for Malawi, 1975 to 2005







Population, total (millions)






Rural population (% of total population)






Rural population density (n/km2arable land)







Poverty incidence (rural)



Life expectancy at birth, total (years)






Mortality rate, under-5 (per 1,000)






Prevalence of HIV, total (% of pop’n ages 15–49)







Stunting (% children 6 to 59 months)




School enrolment, primary (% gross)







GNI per capita, Atlas method (current US$)






GDP growth (annual %)






Inflation, consumer prices (annual %)






Real interest rate (%)






Agriculture, value added (% of GDP)






Industry, value added (% of GDP)






Services, etc., value added (% of GDP)







Imports of goods and services (% of GDP)






Food imports (% of merchandise imports)







Fertilizer consumption (‘000 metric tons)






Irrigated land (% of cropland)






Maize growers (% agricultural households)


(p.65) Bingu wa Mutharika (from 2004 until his death in 2012). These three periods can be examined in terms of neo-patrimonialism where politics is centred around the president who uses the power and resources of the state to dispense patronage to sustain political power (Booth et al., 2006) and, following Cammack et al. (2010), in terms of three features of government: centralization of rent utilization and time horizons (as with stationary and roving bandits (Olson, 1993)), relations of political leaders with a more or less capable technocracy, and political inclusiveness and competition. It is also helpful to explore these using a simple distinction between three different patronage client groups: the political elite, the middle classes, and the wider masses. Regional dimensions are also important in garnering the on-going support of regional elites and, in the run up to elections, of regional masses. Poulton’s analysis of the relations between political and technical considerations and processes in determining agricultural policy in different contexts also provides valuable insights (Poulton, 2012).

The first president of Malawi, Hastings Banda, held the reins of power for 30 years from independence in 1964 until he was ousted in democratic elections in 1994. He presided over a highly personalized and repressive regime, taking a strong personal interest in policy and engaging closely with a capable civil service. Two phases of policy under Banda may be considered (Booth (p.66) et al., 2006; Cammack et al., 2010), the first delivering quite rapid economic growth but achieving this through a set of ultimately economically and politically unsustainable policies. These focused on developing highly import-dependent estate agriculture producing tobacco, while the smallholder sector grew much more slowly (the estate sector grew from a very small base at an average of 17% per annum from 1964 to 1977, while the much larger smallholder sector grew at less than 3% per annum (Harrigan, 2003), below the population growth rate). Smallholders were restricted to cultivation of food crops and low value cash crops, while providing a low cost labour reserve for estate agriculture. Banda used the promotion of tobacco (Malawi’s ‘green gold’) in the estate sector as an important means of dispensing political patronage to elites and a small but emerging middle class based primarily in the central and, with time, in the northern regions. Middle class support was also garnered by investments in secondary and higher education and by growth in civil service employment, while mass support rested upon large-scale visible investments in a variety of infrastructural and development projects in all regions of the country, including fertilizer and credit subsidies channelled to better off (or less poor) smallholders, and a commitment to deliver national food security. Estate and smallholder agriculture were highly regulated, with a high degree of state intervention through generally effective parastatals and government ministries. Booth et al. (2006) characterize Banda’s approach in this period as ‘patronage following policy’. There was little explicit attention to welfare policies in this policy phase as government and the Malawi Congress Party played down the existence of chronic poverty.

The fragility of the growth developed under these policies became apparent when the economy was hit by a number of external shocks in the early 1980s. The government was then forced to recognize the need for different policies and to seek financial assistance, with policy conditions, from the IMF and World Bank. Malawi consequently entered its second post-independence policy phase, of liberalization, coinciding with Banda’s increasing fragility and a decline in his personal policy engagement. Policies then looked to increase smallholder export crop production by increasing farmgate prices while holding down maize (food) prices (Harrigan, 2003). This encouraged the substitution of smallholder maize production by cash crops which, with removal of fertilizer subsidies and unsuccessful market reforms, resulted in a food crisis in 1987, with rapid increases in maize prices. Banda’s sense of responsibility in delivering food self-sufficiency to the country (and his vulnerability to growing calls for political change and the failure of an important part of his mass patronage) led to policy reversals and the re-introduction of fertilizer subsidies and government intervention in maize markets. Despite a positive maize production response to these policy changes, maize shortages continued with two severe droughts in the 1992–4 period. At the same time (p.67) access to patronage from tobacco was extended to a much larger part of the middle classes, primarily in the central and northern regions, through the promotion of large numbers of small-scale tobacco estates.

Following the transition to multi-party democracy and presidential elections in 1994, Malawi’s second president, Bakili Muluzi, served two terms of office, from 1994 to 2004. A major change in agricultural policy in the mid 1990s was the repeal of the Special Crops Act, which had restricted smallholder cultivation of some crops, most notably burley tobacco. The liberalization of burley tobacco production was very successful, with rapid growth in the number of smallholders growing the crop, and without (initially at least) expected declines in quality (Harrigan, 2003). However the 10 years from 1994 were characterized by severe macro-economic mismanagement, rampant inflation, dramatic falls in the value of the Malawi Kwacha, and a weakening of government capacity. Opportunistic privatization, funding diversions, and the issue of bonds to finance budget deficits became an important source of patronage for a primarily southern region elite with commercial rather than agricultural interests, so that short-term financial interests of politicians drove policy with ‘policies following patronage’ (Booth et al., 2006). As the real value of civil service salaries collapsed, middle class patronage involved what Booth et al. (2006) describe as the ‘democratization’ of corruption. With the government’s political power base in the south of the country (in contrast to Banda’s base in the less populous centre), and with stagnation of the economy, growing land pressure in the south, declining soil fertility, and experience of wider use of fertilizer in the early 1990s, the politics and mass patronage of maize self-sufficiency became associated with the politics of fertilizer subsidies. In 1998 the Muluzi government introduced the universal free provision of small packs of maize seed and fertilizer under the starter pack programme. This was subsequently down-scaled to the Targeted Input Programme (TIP), and considered by donors more as a social protection programme promoting food security for vulnerable households, and less as an agricultural subsidy programme. However we note here the populist political roots of this programme and the ambiguity as to its role in promoting agricultural development, social protection, and/or short-term political patronage objectives.

‘Fertilizer politics’ has subsequently become a major feature of Malawi. In the 2004 presidential election, in which President Bingu wa Mutharika was elected as the United Democratic Front (UDF) candidate chosen by Bakili Muluzi, both the UDF and Malawi Congress Party (MCP) (parties of the two former presidents) campaigned with promises of different forms of fertilizer subsidy. Fertilizer subsidies continued to be a major political issue in subsequent political manoeuvring associated with the President’s breaking away from former president Bakili Muluzi to form his own party, the Democratic (p.68) People’s Party (DPP). Without a base in parliament, where he was vulnerable to calls for impeachment by the UDF and MCP, Mutharika introduced the highly popular Agricultural Input Subsidy Programme (AISP) to garner support from the masses, over the heads of parliament. The government also appealed to middle class professional and business people through its major emphasis and success in improving macro-economic management. This led some observers to suggest that Mutharika’s term would in some ways ‘be closer to the Banda tradition than to Muluzi’s, with patronage being subordinated to an overall vision’ (Booth et al., 2006). As time went on, however, it became apparent that Mutharika also shared Banda’s autocratic tendencies, and there were increasing concerns about corruption and patronage within his government. These concerns did not prevent him from winning a broad-based landslide second term election in 2009—indeed the subsidy programme not only allowed him to garner widespread support from rural beneficiaries, diversion of resources also provided him and his party with potential opportunities to capture and use substantial resources for political patronage (Chinsinga, 2012b).The subsequent substantial parliamentary majority, however, released him from his former precarious reliance on support from the middle classes and gave free rein to both his autocratic tendencies and to increasing corruption and patronage (Chinsinga, 2012b). This led to serious political and economic problems from late 2010 until his sudden death in April 2012.

Understanding agricultural policy changes in Malawi also requires an understanding of changing donor interventions (Harrigan, 2003; Chinsinga, 2006, 2007). These were very supportive of agricultural policies in the first phase of Banda’s dualistic policies as described above, making very large investments in integrated rural development projects. Concerns about the problems of Malawi’s dualistic and interventionist policies, as regards both economic vulnerabilities and constraints on smallholder development, then came to the fore at the same time as a wider shift in ideology to structural adjustment and the Washington Consensus. This was a major driver of the liberalization policies in Malawi as it took on structural adjustment loans in the early 1980s. Harrigan (2003) then describes a series of ‘U turns’ by the World Bank in agreeing to the re-introduction of fertilizer subsidies and then later insisting on their removal and opposing their re-introduction under the start pack programme. From 2000 onwards donors placed increasing emphasis on the development of policies for social protection, and these have since evolved as regards their relationship with agricultural policies (Dorward et al., 2006) with consequences for attitudes to agricultural policies. Chinsinga (2006) describes more recent differences between donors and changes in individual donor positions. These positions have been driven by domestic donor politics, economic ideology, humanitarian concerns, changes in international development thinking and practice, and personal (p.69) concerns of often short-term in-country staff. Changing donor policies have been important because (a) they have suffered frequent changes and inconsistencies, and (b) they have been unduly influential as a result both of the high dependence of the Malawian economy on foreign aid and of weaknesses (particularly under Muluzi) in government capacity and commitment to articulate consistent policies (which Sahley et al. (2005) explains in terms of political processes resulting from both Malawian understandings of state responsibilities in agriculture and social protection, and clashes with donor understandings of these responsibilities).

A number of important insights emerge from this discussion. We note that the use by different presidents of different approaches to delivering patronage to client groups with different regional interests has been a core determinant of the prominence and resources given to agricultural policies and the nature of these polices. A major challenge which all three presidents faced in this was the need to deliver short-term patronage without compromising the longer term capacity of the economy to support such patronage. Thus, ‘patronage policies’ were critical in the promotion of agricultural policies and investment under Banda, while failures by the policies in dealing with core poverty/vulnerability and food security problems led to their demise. Conversely, the failure of ‘commerce-based’ patronage polices under Muluzi led to a resurgent interest in fertilizer subsidies, which Mutharika used as a major means of garnering widespread political support in his first term. The challenges faced by Mutharika in his second term are more difficult to rationalize: as will be shown later, excessive and increasing subsidy costs up to 2009 were effectively reined in after the election, and agriculture received much less political attention—although the core elements of the subsidy were maintained. This ebb and flow of political interest in agriculture has revolved around the different regional and patronage group interests in food, fertilizer, and tobacco and has at times coincided with, and at times conflicted with, a different pattern of changing interests among donors.

It is also important to note, however, that an entirely appropriate and legitimate political preoccupation with food security arises not because of populism and patronage, but because food security is an important preoccupation for poor people, whether urban or rural, who spend a large proportion of their income on staple foods and who are very vulnerable to price changes. The emphasis on fertilizer subsidies as a response to food insecurity, however, is determined by recognition that

  1. (a) high price volatility in relation to domestic supply shocks is a result of lack of integration of national and international maize markets (due to poor international transport links, and also foreign exchange constraints);

  2. (p.70)
  3. (b) the majority of poor food-insecure people and of the electorate, particularly in the south, are rural deficit producers facing particular constraints in accessing fertilizers;

  4. (c) less-poor rural people also face difficulties in accessing fertilizer but have an interest in fertilizer access for the production of food and non-food cash crops;

  5. (d) most urban people have strong links with rural people and rural interests; and

  6. (e) widespread understandings among the Malawian population that fertilizers are critical to food security, that this is dependent on food self-sufficiency, and that the government has an active responsibility in ensuring food self-sufficiency and hence in enabling widespread fertilizer access.

Key to the importance of fertilizers in the food security narrative, therefore, is an understanding of severe market failures affecting rural livelihoods, an understanding which has been shared by Malawian politicians and technocrats—but less often by donors (Sahley et al. (2005) note a related divergence between donors and Malawians regarding the role of the state in maize and fertilizer provision). This difference in understanding of market failures has been an important reason for government/donor disagreements regarding instruments for pursuing the poverty reduction and agricultural development agendas of donors and the mass patronage and agriculture development agendas of domestic politics, even where their interests in these agendas appear to converge. We therefore now turn to consider briefly key features of markets and livelihoods in rural Malawi.

4.4. The livelihoods and markets context

We focus on two important features of rural livelihoods and markets in Malawi that are relevant to our analysis of agriculture policy.

First, agriculture policy is politically and economically important because of the major importance of small-scale, low productivity, and risky agriculture in the livelihoods of poor rural people. While agriculture is by no means the only source of income of poor rural people, it is critically important to their livelihoods. There are surprisingly few empirical estimates of the proportion of rural Malawians’ income coming from own farm activities. The National Statistical Office (2005a) estimates agricultural activities as comprising 50% of rural household incomes and 55% of the lowest income quartile for rural and urban households, and these estimates are consistent with the commonly cited figure of 50% of income being farm income in different parts (p.71) of Africa (Ellis, 1998; Reardon, 1998; Jayne et al., 2001). Dorward (2006) estimated figures of 33% own farm income in 1998 (closer to figures of 20 to 45% in different southern Africa case studies cited by Bryceson (1999)), although later analysis suggests own farm incomes of 50% or more in two different rural areas, but around 40% among poorer households (Dorward, 2007). However such figures underestimate the wider importance of agriculture in rural livelihoods. Food expenditures are estimated to account for just over 61% of total expenditures in the lowest income quintile in the rural population in 2004 (National Statistical Office, 2005a). A large part of the 50% or more non-own farm income of poor people is also derived from employment on other people’s farms and from providing services to other rural people whose incomes and demand for services are also heavily dependent on agriculture (Dorward, 2006). This very large importance of agriculture coupled with the low and risky nature of smallholder agriculture in Malawi means that agriculture is a major source of vulnerability in rural livelihoods.

The second major feature of rural markets and livelihoods in Malawi relevant to our analysis is the very low level of market development and economic activity. Dorward and Kydd (2004) argue that a defining characteristic of rural areas in Malawi is that low and fragile incomes and low demand lead to limited market activity based on very small transactions. The dependence on a relatively narrow range of risky and low productivity activities, which leads to increased covariant risk and vulnerability in the economy within which rural livelihoods are located, is exacerbated by poor infrastructure, services, and communications, with poor roads and transport services and poor telecommunications, leading to high costs in physical movement of goods and services in and out of rural areas, together with high costs of communication about market opportunities and prices.

The result of the low general level of economic activity, of the risks from lack of diversification, and of poor communications, is thin markets, with very low traded volumes of key commodities, manufactures, and services (notably agricultural produce, agricultural inputs, and agricultural finance). Thin markets are both a cause of and are caused by high costs and risks in trading small volumes in small transactions, requiring high risk premiums and margins to make it profitable to engage in markets. However these high margins themselves depress demand, and the result is a low level equilibrium trap and market failure (Kydd and Dorward, 2004). These problems are particularly acute in the input, output, and financial markets needed for intensification and increased maize productivity, and can be analysed in terms of a ‘low maize productivity trap’, set out in Figure 4.1.

At the heart of Figure 4.1 is low maize agricultural productivity, the direct consequence of low yields resulting from continual cultivation of maize on land without organic or inorganic fertilizers. Low yields then require (p.72)

Malawi: political, policy, livelihoods, and market background

Figure 4.1. Vicious circle of the low productivity maize production trap

Source: Dorward and Chirwa (2011c).

increased areas under maize production to meet as much of household staple demands as possible (for reasons explained below). Low productivity leads to low incomes. For each farm-household this makes it difficult to afford input purchases for raising agricultural productivity. Most farmers cannot buy inputs on credit either because of under-developed credit markets and high costs of credit administration, high borrower and lender risks, consumption rather than sale of produce (with lack of cash for repayment), and high input prices and access costs due to low input demand, poor infrastructure, and high transport costs.

At the same time low maize production, low maize sales, and a high degree of subsistence consumption lead to thin markets in maize (as a large proportion of the maize produced in Malawi is consumed within households and villages, never reaching the market) and this leads to high variability in maize prices between years (as a result of good and bad rains or access to inputs) and within years (as a result of seasonality in maize stocks, input and labour requirements, incomes, and wage rates). This contributes to risks in input use, undermining incentives for investment in higher productivity by farmers with the potential to produce surpluses. However variability in maize prices also means that deficit producers seek to produce as much maize as they can to reduce their exposure to high purchase prices for maize in years when prices are high, and thus devote large parts of their cultivated land to maize. Poverty (p.73) and vulnerability to shocks (low yields, high food prices, sickness, loss of employment or remittance income) further constrain productivity and productive investments—with women, who play a key role in agricultural production and rural livelihoods, particularly vulnerable to these shocks.

Low demand for inputs itself raises costs and inhibits the development of input supply systems in less accessible areas. Investments in maize market development are also constrained by low traded volumes and price risks in thin markets. High price variability for this critical commodity leads to government intervention in maize markets (with, for example, setting of minimum and maximum prices, export bans, and bans on private trade) but difficulties in designing and implementing interventions mean they often increase price variability for maize sellers, buyers, and traders; inhibit investment and participation in markets; and exacerbate the problems they attempt to address (see, for example, Tschirley and Jayne (2010)). These problems have then been exacerbated by unstable policies and unstable weather conditions (with particularly low maize production from 1992 to 1994, 2001/2, and 2004/5), lack of investment in roads, and a general policy, infrastructure, and economic context that has not encouraged long-term private sector investment and development.

The various influences and feedbacks described above lead to a vicious circle of low maize productivity and unstable maize prices inhibiting (a) net producers’ investment in maize production,3 (b) net consumers’ reliance on the market for maize purchases, and (c) poor farmers’ exits from low productivity maize cultivation. The result is a lock-in to widespread cultivation of low productivity maize which, because of its scale and importance in the wider Malawian economy, depresses labour and agricultural productivity and growth in the non-farm economy and of the Malawian economy as a whole.4

This analysis, which builds on wider understandings of coordination problems and low level traps (Rodenstein-Rodan, 1943; Hoff, 2001; Dorward et al., 2009) has important implications for understanding livelihood constraints and vulnerability, and in the design and implementation of agricultural policies and (p.74) instruments. It identifies low levels of investment and rural market development on the one hand and poverty and vulnerability on the other as interacting cause constraints to development and livelihood security. This then suggests that without the existence of established and functioning thick markets, markets cannot be relied on to deliver agricultural and food delivery services—but that such markets cannot develop, or can only develop slowly and fitfully, in the context of high levels of poverty. Two major questions emerge from this:

  1. 1. How can agricultural service markets (principally for inputs and credit) and food markets be developed in the medium to long term in order to address the problems of low productivity, poverty, and vulnerability that are endemic in Malawi?

  2. 2. How can agricultural services and food access be provided in the short term in the context of endemic poverty and low productivity and in ways that crowd in rather than crowd out market development?

These questions interact strongly with the political context discussed earlier and in particular with the different interests and ideologies of technocrats, politicians, and donors: in the first phase of policy under Banda there was a consensus recognizing that these questions, and development and patronage interests in agricultural policy, complemented each other. Subsequent agricultural liberalization policies have involved a lack of agreement regarding these questions (generally between the predominant Malawian analysis on the one hand and donor analysis on the other, but also between donors), leading to policy conflict and reversals as different views have prevailed. We examine these issues in the next section of this chapter. However, we consider a third question that emerges from the analysis above of livelihood and market constraints:

  1. 3. In what ways might a large-scale agricultural input subsidy impact on markets and livelihoods and the ‘low productivity maize production trap’?

This question can be answered in two ways. First, with reference to Figure 4.1, a subsidy that substantially reduces input costs can make such inputs affordable, leading to producer investment in these inputs. If their use then leads to increased maize productivity, this can raise real incomes and then, through raising real incomes and their stimulus to demand for non-agricultural goods and services, this can break the vicious circle of the low productivity maize production trap. This will be strengthened where the subsidy is accompanied by stable policies, investment in roads, support to private sector growth, and policies that encourage more stable maize prices following weather shocks. (p.75)

Malawi: political, policy, livelihoods, and market background

Figure 4.2. Subsidy impacts on beneficiary households and the rural economy

We can also investigate the possible impacts of a large-scale agricultural input subsidy by exploring possible effects on different households’ livelihoods within the rural economy. This is set out in Figure 4.2.

Figure 4.2 identifies three possible uses of the subsidy by the recipients: reselling of coupons or of subsidized inputs, incremental use of the inputs in production, or use of the inputs with displacement of otherwise unsubsidized purchases. These different uses lead to two main types of direct benefit for recipients: immediate income transfers from reselling or displacement of some or all of the input coupons or subsidized inputs, or incremental production at harvest if the inputs are used on farm.

Immediate transfer benefits should lead to a tightening of the labour market in the season of implementation, due to contraction in labour supply by poorer households (who need to hire out less ganyu—casual labour—to earn food, as a result of income from reselling some or all of their coupons or subsidized inputs) and a much smaller expansion of hired labour demand by less-poor households (who have more resources available to hire labour as a result of cash saved by subsidy displacement of unsubsidized seed and fertilizer purchases). This tightening of the labour market should lead to an increase in real wages.

Increased wages lead to immediate real income and hence welfare and consumption gains to poorer households, both recipients and non-recipients. Increased on-farm labour use by the poor (as a result of reduced need to hire out labour) also means that gains from direct transfers to poor people and (p.76) higher wages should lead to incremental production and welfare gains at and after harvest, even without any incremental input use (though these gains will be offset to some extent in the wider economy by losses of low cost labour to the less poor). Less-poor people who hire in labour may also incur a loss in net real income if they have to pay higher wages when hiring labour in and for purchasing local goods and services whose prices are determined largely by unskilled wage costs.

Impacts of a subsidy are, however, also expected in the season following its implementation, as a result of households having increased stocks of grain produced with the subsidy. Depending on the amount of extra production that is carried forward in these stocks, there should be two effects: decreased need for pre-harvest purchases of grain by households with insufficient stored grain and, as a result of this, decreased hiring out of ganyu to earn cash and food by poorer households and (slightly) increased financing to hire in farm labour by less-poor households (Dorward (2012b) provides a graphical household model analysis of this). The effects of the loosening of the grain market should be a fall in maize prices and a rise in wages.

This analysis draws out the importance of understanding different direct effects of subsidy access on different households and the different indirect effects of these as they affect labour and maize markets. Figure 4.2 shows the rather complex set of direct and indirect subsidy impacts and their relationships over time as described above. Further subsidy impacts shown in the figure are that increased real incomes should lead to greater farm and non-farm investment (in human and social capital as well as in financial, natural, and physical capital for particular enterprises), and that growing real incomes in rural areas should, through economic linkages or multipliers (Delgado et al., 1998; Dorward et al., 2002; Haggblade et al., 1989, 2007a), lead to a virtuous circle of increased demand for locally produced goods and services, including higher value non-staple foods. Increased consumption of vegetables and livestock products should lead to nutrition and health benefits and further tighten local labour demand. There will be similar human capital and welfare benefits and tightening of labour markets from investments in, for example, improved housing and sanitation, from investment in education and/or other consumption items. Many of these benefits should accrue to both subsidy recipients and non-recipients as a result of changes in maize prices and wages.

Impacts on demand for and investment in input services will depend heavily on the way that subsidies are implemented. Understanding of these potential direct and indirect subsidy benefits should guide programme design and implementation: it should also guide investigations of programme impacts. We return to these issues later in Chapters 6 to 9. In the next section of this chapter, however, we continue our examination of evolving Malawian (p.77) agricultural policies that have led to, and provide the background for, the introduction and development of the Agricultural (or later Farm) Input Subsidy Programme (AISP or FISP).

4.5. Agricultural policies

We now explore in more detail the major agricultural policies pursued in Malawi over the last 40 years or so. These policies have shaped and continue to shape infrastructure, research, knowledge, institutions, and the expectations of different stakeholders (politicians, technocrats, rural people, and voters) and, with the political analysis and context provided earlier in this chapter, are critical to understanding both the emergence of Malawi’s Farm Input Subsidy Programme and its subsequent implementation and evolution.

4.5.1. Post-independence policies

Post-independence smallholder agricultural development policy revolved around the establishment and then scaling up and out of four large donor-financed integrated rural development projects (one in the northern region, two in the centre, and one in the south) to a national programme of projects covering the majority of the country. Although elements varied between projects, there were a number of common core activities: agricultural extension; subsidized supply of improved seeds and fertilizers for maize and cash crops; construction of feeder roads and market facilities; construction of offices and staff housing; and construction of health facilities. Within the context of supporting infrastructure, the core smallholder development activities involved the promotion of farming groups which took input loans for seed and fertilizer, loans which were repaid through interlocking sales at fixed prices to the parastatal marketing board, ADMARC. The system was very successful in expanding access to purchased inputs, particularly in maize production, and in achieving very high rates of credit repayment. Fundamental to this success were (a) the role of the parastatal marketing agency, ADMARC, as a sole seller of inputs to and sole buyer of produce from smallholders, (b) a major focus on facilitation of this system by extension staff, and (c) strict enforcement of penalties for non-repayment, (with the denial of access to input purchases on both cash and credit for all members of a defaulting group and, in some cases, heavy handed confiscation of assets of defaulters). ADMARC also maintained pan-territorial and pan-seasonal prices.

These policies had complex anti-poor and pro-poor elements (Chirwa et al., 2006). The interests of the poor were damaged by food prices frequently being held above import parity, and cheaper imported food prices might (p.78) have allowed the large number of malnourished poor better access to food in some years (although lower maize prices would have depressed incentives for investment in improved seed and fertilizer use in maize). ADMARC also tended to tax the smallholder sector, and the proceeds of this were transferred to the estate sector, which also benefited from cheap labour in an exploitative tenant system of tobacco production.

However the smallholder development projects described above invested considerable sums in rural areas, and although the direct beneficiaries of the agricultural programmes were generally (but not always) less-poor farmers, they did promote national food self-sufficiency and local food availability (both through local production and through the network of ADMARC markets which sold maize) and stimulated economic growth in rural areas. Smallholder taxation was also mainly on cash crops, and the smallholder maize system was moderately subsidized by ADMARC (Kydd and Christiansen, 1982). Smallholder taxation was also offset, and with time eclipsed, by government infrastructural investment in the integrated rural development projects described above and by the implicit subsidies in the support of groups in obtaining credit and in marketing their produce.

This set of agricultural policies can be seen as setting up a system that addressed many of the demands made of it. Support for estates provided direct patronage to elites (and resources for dispensing patronage) and to emerging middle classes as noted earlier, particularly in the central region. Donor resources supported smallholder agricultural development that provided infrastructure and agricultural services and food access to smallholders (addressing the market development trap), meeting both donor developmental objectives and government developmental and patronage objectives (the latter being through regionally distributed visible project investments, civil service and parastatal employment, improved incomes to less poor farmers, and stable food availability in rural areas). There were also important social protection outcomes from stable pan-territorial, pan-seasonal food prices, and reliable food availability in most rural areas in the country. Although the direct beneficiaries of these policies were not generally the poorer members of rural communities, and there were differences between regions in the benefits produced by these policies, both the flow of seasonal finance to less-poor households and the increased incomes arising from the use of those inputs (and their multiplier effects) should have increased seasonal liquidity in rural communities, raising demand and wages for casual labour and increasing community resources for informal local social protection measures. The discussion of pro- and anti-poor elements of these policies also, however, illustrates conflicts over maize prices (low prices are good for poor, food-insecure consumers but high prices are needed to stimulate investment), while the longer term failure of the government to sustain (p.79) these policies illustrates the difficulties governments face in allocating limited resources between the short-term demands for distribution of benefits to different interest groups on the one hand and longer term demands for investment in growth on the other.

4.5.2. Liberalization policies

As liberalization and, later, multi-party democracy and currency devaluation led to the demise of the interlocking smallholder agricultural credit system and integrated rural development approach at the core of the agricultural policies described above, subsequent agricultural policies were not part of such a comprehensive vision of rural development. Agriculture, and indeed individual crops, were seen as needing crop- and commodity-specific market solutions. The best example of this is probably the development of smallholder tobacco, which, as discussed earlier, was very successful. Harrigan (2003) reports a number of benefits from this expansion: a major cash injection with multipliers feeding through into the rest of the non-farm rural economy, the use of tobacco income to buy seed and fertilizer for maize production, and market development. However she also notes that middle income smallholders were the predominant direct beneficiaries and while there were significant numbers of poorer smallholders with very limited land growing tobacco, tobacco began to crowd out maize on these farms. This led to severe declines in maize production when devaluation of the Malawi Kwacha and the removal of input subsidies made the use of fertilizer on maize uneconomic. At the same time, growth in smallholder tobacco production was mainly in the central and northern regions, not in the southern region where the holdings are smallest and the extent, incidence, and severity of poverty are greatest (National Statistical Office, 2005a; Prowse, 2007).

A variety of social protection instruments were then introduced. Initially mainly safety nets, these were closely related to agriculture and changes in agriculture policies as they attempted to address increased food insecurity and increasing vulnerability from, inter alia, declining holding sizes and soil fertility and the spread of HIV/AIDS among poor rural Malawians. Over time a wide variety of different social protection programmes and instruments were implemented (Slater and Tsoka, 2007), including targeted nutrition programmes, food transfers, public works programmes, school feeding programmes, credit transfers, and more recently cash transfers.

The agricultural synergies and conflicts of many of these programmes are well known: injections of cash and food into people’s livelihoods can make a critical contribution at lean times of year before harvest when labour is needed by people to work on their fields, these cash and food injections may allow them to work on their fields rather than seek work for cash or food (p.80) elsewhere. These injections can also generate local economic multipliers (Davies and Davey, 2008) and build up productive capacity and assets (Covarrubias et al., 2012). However cash or food for work programmes face a dilemma in that, if they are providing work and income at the time when people need it most, then this will take people from their fields and undermine their own production (Slater and Tsoka, 2007).These programmes also face wider problems regarding the extent and value of their contributions to rural assets and most importantly to the livelihoods of participants (Devereux, 2006). A tendency for programmes to lack long-term funding and consistency has also undermined the extent to which they can be relied upon by rural households (Slater and Tsoka, 2007).

4.5.3. Agricultural input provision programmes

As discussed earlier, recognition of the importance of agriculture for food security, of the need for fertilizers to raise yields for poor farmers with smallholdings and declining yields under continuous maize cropping, and of difficulties in accessing maize seed and inputs led to major political, economic, and developmental interests in policies and instruments aimed at increasing poor people’s access to inputs (seed and fertilizer) for maize production. Two different programmes and instruments concerned with input delivery to poor people are important for understanding the subsequent emergence of the agricultural input subsidy programme in 2005: ‘inputs for work’ and free input distribution (the latter under the ‘Starter Pack’ and ‘Targeted Inputs Programme’ or TIP). These programmes have operated at different scales and in different ways with different agricultural and social protection objectives of stakeholders in supporting different programmes: ambiguity and diversity in understandings of programme objectives have been widespread, and have had both benefits and costs.

‘Inputs for work’ describes the use of public works programmes aimed primarily at delivering social protection but, in contrast to food for work and cash for work programmes, participants are paid with agricultural inputs. ‘Inputs for work’ has only been implemented on a local scale by NGOs with donor funding. Payment with inputs is intended to overcome some of the difficulties with food and cash for work programmes by providing participants with work during the dry season, when there is little competition for labour with work on their fields, but benefits during the following cropping season (by easing labour and cash demands for households looking to purchase inputs) and/or by increasing subsequent maize harvests and food stocks. An evaluation of a pilot project in two districts of Malawi cited by Devereux (2006) concluded that the project was more popular with participants than food or cash for work, and yielded a very favourable return in the value of increased maize produced. (p.81)

Free input distribution has been a much more widely used approach to extending access to inputs across the country, with large-scale government distributions starting from 1993 in response to currency devaluation, the removal of fertilizer subsidies, the collapse of the credit system for maize inputs, and drought (Devereux, 2006). In 1998, the government implemented a universal ‘Starter Pack’ programme, under which every smallholder was provided with enough seeds and fertilizer to plant 0.1 hectares of land. This, with good weather, was a contributor to an estimated 67% increase in maize output, with maize production reaching 2.5 million tonnes (Levy, 2005).

The programme, funded by DFID, was continued in 1999 amid considerable controversy, rooted in different stakeholder interests in the programme, in the political context, and different perceptions of its objectives. As originally conceived, the Starter Pack was an agricultural development programme intended to promote farmer skills in more intensive maize production, diversification out of maize, and the growth of commercial input distribution systems in rural areas. It was intended to include maize fertilizers and legume and maize seed, to be accompanied by a strong extension programme, and to address the market and livelihood constraints discussed earlier. However the programme was actually funded and implemented more as a social protection programme, with a major emphasis on fertilizer provision to promote immediate food production, and less emphasis on agricultural education, provision of legume seed, or the development of commercial input delivery systems. The programme was highly politicized, being introduced just before the 1999 presidential elections, and was seen as particularly beneficial for the southern region, the ruling party’s power base.

Donors were concerned about the politicization of the programme, its high cost, its apparent emphasis on maize rather than on promoting diversification, its possible crowding out effects on input markets, and its efficiency as regards targeting and benefits to the poor. There was concern that large numbers of non-poor people were benefiting, and that receipt of inputs by such people was simply a transfer, with starter pack inputs displacing commercial purchases, although the extent of displacement is disputed. As a result DFID support of the programme was subsequently scaled back to the Targeted Input Programme (TIP).

Targeting, however, faced problems. There were considerable difficulties in the selection of beneficiaries and in the effectiveness of targeting. More fundamentally, however, Levy (2005) argues that the starter pack assisted poorer households in two ways: by increasing their own maize production and (by stimulating national maize production) reducing maize prices. The second benefit was lost when the programme was scaled back to a targeted programme. Dorward and Kydd (2005) simulate the effects of maize price and wage effects of the universal starter pack and compare this with effects under (p.82) a targeted programme, and argued that even if targeting could be achieved without exclusion and inclusion problems, and ignoring both the increased costs associated with targeting and displacement effects, the wage and maize price effects of a universal subsidy could be more cost-effective than a targeted programme in delivering welfare benefits to the target group. They were concerned, however, that by depressing maize prices, the universal programme ‘may undermine the important growth contributions of less poor households that engage in more intensive labour demanding maize-production’ (p. 274).

4.6. The 2005/6 Agricultural Input Subsidy Programme

High food prices and food shortages following poor harvests in 2000/1 and 2001/2 (after the scaling back of the starter pack programme), led to food security and fertilizer subsidies becoming major political issues in the period leading up to the 2004 presidential elections. Both the major parties and their candidates promised fertilizer subsidies, though the UDF and Bingu wa Mutharika, its presidential candidate, offered an extension of the rationed Starter Pack approach while the MCP and its candidate, John Tembo, offered a return to price subsidies through farmer groups. After the election the new government delayed the introduction of subsidies, perhaps due to the need for controlling government expenditure to qualify for debt relief (Chinsinga, 2006). Uncertainty about a subsidy led to delays in the decision to implement another targeted input programme, and also led to delays in fertilizer imports and to farmers delaying fertilizer purchases (Sahley et al., 2005). The result was another very poor season with subsequent food shortages, high prices, very expensive importation of maize, and considerable damage to people’s welfare and livelihoods and to the economy.

During 2005, therefore, the government decided to implement the Agricultural Input Subsidy Programme (AISP) as a rationed and targeted programme that would involve the partial subsidy of a much larger volume of inputs than were provided under the previous Starter Pack and Targeted Input Programme. While the impetus for this was the renewed concerns within Malawi over maize production, the decision provided an opportunity for President Bingu wa Mutharika to craft a programme that would be popular with the rural masses across the country, would provide potential resources for patronage, and was clearly different from previous subsidies or input supply programmes under Banda and Muluzi, but which also had a clear economic and welfare rationale that lay firmly in traditional domestic analysis of smallholder farmer needs. In this it was a critical strategic response to the political challenges facing Bingu wa Mutharika as he established his government without a majority in parliament. It also drew on new opportunities for (p.83) independent policy afforded by the introduction of large-scale budget support by Malawi’s major donors and (to a lesser extent) by ideological divisions within the donor community regarding the relative economic and welfare benefits of such a programme (with, for example, strong support from Jeffrey Sachs and the Millennium Programme).

The design, implementation, and impacts of the programme, and potential wider lessons from this, are the subject of the rest of this book.

4.7. Summary

This chapter has described the context of Malawi’s introduction of its Agricultural Input Subsidy Programme in 2005. An understanding of the economic context of smallholder agriculture and rural livelihoods in Malawi is critical for understanding technical arguments for the introduction of this programme, and its subsequent implementation and impacts. The importance of, and difficulties posed by, the ‘low maize productivity trap’ are particularly relevant. There are important interactions here with the arguments presented in Chapter 2 about subsidies’ objectives and modalities, as well as contrasts and similarities with other countries’ engagement with input subsidy programmes, as discussed in Chapter 3. However Malawi’s agriculture, rural livelihoods, and economy, and technical understanding of them, are themselves strongly influenced by its history of changing political circumstances and deep-rooted political influences and processes. Understanding these political issues is equally important for understanding the genesis, evolution, and impacts of agricultural input subsides in Malawi, and indeed in other countries. (p.84)


(1) From the mid 1990s there were major improvements in primary school enrolment and its gender balance (but not in the quality of primary education) and substantial falls in infant and under-five mortality (though these are still very high).

(2) From 2005 to 2009 there was a dramatic improvement in macro-economic management and consequent reduction of inflation and interest rates and much greater currency stability. Good weather and input subsidies also contributed to growth in food production, as will be discussed later. Macro-economic management declined, however, from 2009 to 2011.

(3) Profitability of and incentives for fertilizer use are commonly measured by the Value Cost Ratio (VCR) with a a general rule of thumb that it needs to be greater than 2 for smallholder investments in fertilizer use (Morris et al., 2007). School of Oriental and African Studies et al. (2008) note that with highly variable inter- and intra-seasonal maize prices and with rising nominal fertilizer prices, the VCR for maize varied markedly from the mid 1990s, with particular divergences between VCRs with peak pre-harvest and low post-harvest maize prices. In the latter case it was generally below 2, while in the former case it was generally but not always above 2. This suggests that profitability of fertilizer use on maize was a constraint to its use on maize grown for sale at or near harvest but not for maize is grown for own consumption. The divergence surplus and deficit maize producers’ VCRs are exacerbated if maize price risk considerations are allowed for, as these would lead to a lower (higher) subjective valuation of maize produced for sale (purchase).

(4) This summary is not a complete account of the many issues involved. Other causes for high dependency on maize include different crops’ calorific yields, dietary preferences, processing and storage considerations, farmers’ familiarity with the crop, and government policies. Poor macro-economic management also constrained wider growth before 2005.