Neither sustainable nor inclusive: a political economy of agricultural policy and livelihoods in Malawi, Tanzania and Zambia

ABSTRACT

The food crisis of 2007/8, alongside rapid population growth, and the uncertainties of climate change propelled African agricultural transformation back into the development mainstream. New narratives of ‘climate-smart agriculture’ and ‘sustainable intensification’ underlie this contemporary transformation. We present a political economy analysis of agricultural policy and livelihoods in Malawi, Tanzania and Zambia, and use this to assess the challenge of achieving ‘sustainable and inclusive intensification’. We find little evidence that agricultural institutions have the capacity to deliver sustainable intensification in agriculture, or that agricultural policy drives changes in agricultural livelihoods that will make them either more sustainable or inclusive.

KEYWORDS:

• Agriculture
• sustainable intensification
• political economy
• inclusive agricultural livelihoods
• gender
• youth

1. Introduction

The 2012 revision to the FAO World Agriculture Report predicted that if the world adopted agricultural intensification, in the form of increasing crop production and higher cropping intensities, there could be a 90% increase in global food production and that ‘world agriculture should face no major constraints to producing all the food needed for the population of the future, provided that the research/investment/policy requirements and the objective of sustainable intensification continue to be priorities’ (Alexandratos and Bruinsma 2012, 20). Such predictions undoubtedly render sustainable agriculture intensification (SI) as an ‘organising principle’ through which global food and climate change problems can be solved (FAO 2009a). Moreover, whilst the idea of intensification in agriculture is certainly mainstreamed and is now entrenched in the push for a second green revolution (Fairbairn et al. 2014; Snyder and Cullen 2014; Tittonell 2014; Vanlauwe et al. 2014; Mdee et al. 2019), this can mean a narrow focus on technical interventions to increase production, and an unclear and contested relationship with the normative goal of ‘sustainability’.

This paper explores the gap between the compound and normative goal of ‘inclusive sustainable intensification’ and the actual practice of agricultural policy and livelihoods in Malawi, Tanzania and Zambia. Our argument is developed using a political economy analysis, to review relevant policies and recent research evidence from all three countries, triangulated through key informant interviews. Our starting point is the research question posed by the Policy for Equity in African Agriculture project, on sustainable intensification¹:

How can equity issues be best addressed in Sustainable Intensification approaches and policies to ensure the needs and interests of poorer smallholders, especially youth and women, are properly addressed?

This paper emerges from baseline research to support this donor-funded project in exploring how sustainable intensification interventions could also address concerns of equity and disadvantage. The term ‘inclusive’ is deliberately deployed in this paper to cover issues of equity and differentiation in agricultural policy and livelihoods, and we take an intersectional view of differentiation, rather than the more simplistic identity labelling (of ‘youth’ or ‘women’) implied in the question above (see Andersson Djurfeldt, Dzanku, and Isinika 2018; Doss et al. 2018). The paper has three main sections. Section one analyses the normative goal of inclusive sustainable intensification (ISI), noting that the concept of sustainable intensification itself is slippery and contested. Through a critical political economy analysis of agricultural policy and practice in sub-Saharan Africa (SSA), we argue that there is a significant gap between the normative aim of ISI and the actual dynamics and narratives at play in agricultural policy and livelihoods. We illustrate this with three areas of critical debate: control over land; rent-seeking and political capture of food production; and identity-based narratives of inclusion (just add women and youth). In section two, through a political economy analysis of agriculture policy and practice in Malawi, Tanzania and Zambia, we show that inclusive sustainable intensification is far from being mainstreamed and implemented. Our conclusions are based on literature review, a meta-analysis of previous data collection by the research team and supplemented with key informant interviews undertaken from November 2017 to February 2018. Finally, we assess the considerable barriers to the implementation of ISI in all three countries. Our findings highlight the substantial gap between normative donor-driven discourse on the nature of agricultural transformation and the actual dynamics of agricultural policy and livelihoods in all three countries.

2. The prospects for inclusive sustainable intensification

Sustainable intensification (SI) is a debated term, given the intensely slippery nature of ‘sustainability’ as a discursive and normative construct (Loos et al. 2014; Petersen and Snapp 2015). For Pretty, Toulmin, and Williams (2011, 7)

… .. it is defined as producing more output from the same area of land while reducing the negative environmental impacts and at the same time increasing contributions to natural capital and the flow of environmental services.

The Policy for Equity in African Agriculture research project focuses on analysing patterns of smallholder intensification with a focus on how women, youth and institutions can be incentivised to increase the intensity of agricultural production on existing land holdings using methods which do not cause environmental degradation, rather than the extensive cultivation of new land areas. Population growth and increase in land pressure are often characterised as directly proportional (Boserup 1965) and this argument underpins arguments for intensified production. However, we should remember the caution of Sara Berry (1993, 189) that:

Agricultural intensification has been neither inevitable nor continuous in African farming systems. In some areas, intensification was halted or reversed by changing environmental or political and economic conditions; in others, it has occurred not as an adaptive response to population growth or commercialisation, but in the face of growing labour shortages and declining commercial activity. Such cases underscore the importance of studying farming as a dynamic social process. As farmers contend with social as well as environmental conditions, changes occur not only in what is produced and how much, but also in when work is done and by whom. Thus, changes in cropping patterns and methods of cultivation are influenced by social factors which govern the timing as well as the amounts of labour devoted to farming, as well as the control of effort and output.

There is little agreement in either policy or academic literature on precisely what sustainable intensification consists of (Garnett et al. 2013; Wezel et al. 2015; Bernard and Lux 2017; Béné et al. 2019). Hence sustainable intensification can both denote efforts to improve agricultural output levels per unit area by increasing the use of more efficient external outputs and technological advances; and conversely the promotion of organic and conservation agricultural techniques with the reduction of external inputs. It remains unclear what SI is precisely, rather there is an open aspiration of ‘what it might look like’, hence it embraces different approaches that enable the fulfilment of the broader goal (Garnett and Godfray 2012; Garnett et al. 2013). Such approaches can include but are not limited to: new and conventional techniques to increase water, nutrient and other input efficiencies, development of new varieties or breeds of crops and livestock, conservation agriculture (reduced tillage/no till, soil cover and soil recovery strategies), integrated pest management (IPM), and agroecological approaches. Importantly, the breadth of SI discourse allows policymakers to integrate multiple goals and repackage existing policy and interventions under the concept (Loos et al. 2014). In this paper we don’t intend to add further to attempts to define SI, rather we are interested in the deployment of the concept in donor-funded agricultural interventions and how this sits at odds with the dominant narratives and concerns of policymakers, civil servants, farmers in Malawi, Tanzania and Zambia. Further, we seek to interrogate the notion that SI can also be inclusive.

The incentives, opportunities and obstacles for actors to promote (or discourage) agricultural intensification are related to both the political economy of development processes and a range of, often conflicting, national, regional and international policies and programmes. Institutional change processes and agricultural dynamics may lead to either inclusionary or exclusionary tendencies, e.g. small farmers may become landless if national policies favour agri-business investment (Vorley, Cotula, and Chan 2012; Tsikata 2016; Hall, Scoones, and Tsikata 2017; Chinsinga and Chasukwa 2018a; Manda, Dougill, and Tallontire 2018a, 2018b).

In addition, the adoption of inclusive sustainable intensification as a normative approach and outcome for agriculture in SSA would be subject to the agreement on such a vision by a multitude of different actors, with often competing interests and agendas. The barriers to this are obvious: donors remain confused about how to package, coordinate and deliver intervention to accelerate agriculture development (Mdee et al. 2019); and there are inconsistencies between global, regional and local policies on land governance and approaches to agriculture development (Chinsinga and Chasukwa 2012). Government budgetary spending and commitment to agriculture is rising but remains low in Africa (FAO 2009b, 2018); corporate investments are reproached for their unclear investment motives and destabilising relationship with smallholder farmers (Dubb, Scoones, and Woodhouse 2017) and smallholder farmers remain trapped in between the politics and dynamics of competing interests (Manda, Dougill, and Tallontire 2018a, 2018b). There is also little agreement from mainstream development partners on what sustainable agriculture looks like (Mdee et al. 2019). The contradictions among actors involved in agriculture development have historical origins hence it is important to understand the foundation of current agricultural policies in colonial regimes.

It is suggested that development partners and governments have a responsibility to create an enabling context of macro policy that could support inclusive and sustainable agricultural intensification (Pretty, Toulmin, and Williams 2011). In sub-Saharan African at a more meso-level, sub-national agricultural institutions and actors (including extension services, farmer organisations, traders, investors in contract farming, input suppliers and local authorities who control land tenure systems) are some of the most important mediators that determine the extent to which agricultural policies contribute to inclusivity and sustainability in existing agricultural dynamics. The resources and capabilities present at the meso-level have a critical role in enabling policy and strategy to be implemented and where resource and capability gaps are present, then the gap between stated policy and practice can be wide (Andrews, Pritchett, and Woolcock 2013; Mdee and Harrison 2019). At the micro level, farmers operate in a space in which they implement livelihood strategies in a context shaped by formal institutions, but also through more socially embedded customary arrangements (Harrison and Mdee 2017b).

It is apparent that the visibility of small-scale ‘peasant’ farmers in either colonial or post-colonial visions for agricultural transformation in Africa has been limited, presenting them as a problem to be solved, or a resource to be exploited (Birner and Resnick 2010). Birner and Resnick’s (2010) comparative political economy of agriculture policies in Africa and Asia argues that transformation of smallholder agriculture is not the focus of agricultural policy in SSA, despite their dominance in the agricultural sector, given the policy maker's preference for ‘modernised’ and commercial agriculture. However, large agriculture-dependent populations also need to be politically manipulated where democratic elections have been adopted (Bates and Block 2009) and where public spending on staple crops permeates political dynamics. This is pronounced in many subsidy programmes (Chirwa and Chinsinga 2012). We are conscious that there is an extensive literature on rural development and transformation, particularly relating to class formation and differentiation (for example Bernstein 2016), but focus our attention in this article on contemporary agricultural policy dynamics and capacity for implementation.

To what extent is it likely that this wide range of actors and interests will align on the normative goal of ISI? A critical examination of the literature on the political economy of agriculture in SSA suggests that this outcome is unlikely. In this section, we highlight three critical challenges to the normative goal of ISI: the scramble for ‘surplus’ land in Africa; the dominant narrative of agricultural modernisation/industrialisation and the politicisation of food crops; and the politics of identity in shaping the ‘inclusiveness’ of agricultural policy.

2.1. Challenge 1- the scramble for ‘surplus’ land

A narrative of agricultural intensification is frequently used as a justification to explain the motives behind large-scale land acquisition (LaSLA) in some sub-Saharan African countries by companies and external state interests (Engström and Hadju 2019). The argument is that such investment will ensure increased production, food security and wider agricultural transformation. Jayne, Chamberlin, and Headey (2014) suggest that the 2008 world food price crisis propelled a concerted effort to transfer land from customary tenure to the state or private individuals who, it was assumed, would more effectively realise the production potential of the land to meet national and global food objectives. This technical narrative of efficiency, often obfuscates the politics of agriculture and land access in order to frame answers to the question of ‘who will make the best use of Africa’s land?’ (Manjengwa, Hanlon, and Smart 2014).

A counternarrative suggests that these investments pose challenges for smallholder farmers and the food security (Dell’Angelo et al. 2017; Hall, Scoones, and Tsikata 2017). Modern ‘land-grabs’ have historical precedents. Africa has been the target of large land acquisition and investment, especially during colonial periods. Past and present waves of land reallocations/acquisitions focus on customary or ‘unused’ lands (Doss, Summerfield, and Tsikata 2014; Hall, Scoones, and Tsikata 2017; Kuusaana 2017). Allocations of land to incoming investors (either foreign or local) are often predicated on the assumption that there is spare and unproductive land, while land demarcations and designations are contested terrains. There is evidence in all our focus countries that estimates of spare land are exaggerated, and that almost all land is subject to claims of one kind or another, be they from the state, settlements, individual landowners or land users (see Chinsinga and Chasukwa 2012; Sitko and Chamberlin 2016; Bluwstein et al. 2018).

Where the state seeks to invest in agricultural transformation through ‘modernisation’ of agriculture then this can favour the commercial and corporate investor, and local elites, who use their power and resources to dispossess others of their access to resources (see, for example, Tsikata 2016; Kuusaana 2017; Bluwstein et al. 2018). This is reflected in the observed increasing differentiation of agricultural livelihoods across Africa (Hall, Scoones, and Tsikata 2015; Andersson Djurfeldt and Hillbom 2016; Andersson Djurfeldt, Dzanku, and Isinika 2018). Current patterns suggest that this process may be driving a new wave of marginalisation and dispossession.

2.2. Challenge 2- the politics of agricultural ‘modernisation’

Agricultural policy in sub-Saharan Africa is dominated by a historical and aspirational narrative of ‘modernisation’, whereby peasant agriculture is subject to transformation through the consolidation of lands, and the application of ‘modern’ technologies and production methods, organised around the potential to generate export revenue (Scott 2002; Bernstein 2010). As already noted above, this drives policy design, which often does not specifically tackle the challenges of small-scale agriculture (Birner and Resnick 2010; Mdee et al. 2019). A refreshed drive towards ‘industrialisation’ in African government policy further reinforces this dominant policy narrative (Verdier-Chouchane 2017).

A critical difference is that a post-second world war discourse on agricultural industrialisation does not fit the contemporary context, in which industrial agriculture is recognised as a significant contributor to carbon emissions, and environmental degradation. Despite donor-driven and international finance incentives to adopt a ‘green’ industrial strategy, this as yet shows little sign of coherent operationalisation, beyond limited policy statements on the adoption of ‘climate-smart agriculture’, ‘sustainable intensification’ and ‘conservation agriculture’ (Arslan et al. 2015; Mdee et al. 2019). As noted above, there is little agreement on what sustainable intensification in agriculture is, and critics have argued that it actually represents a significantly contested terrain, in which influential corporate actors and technology-driven philanthropists (e.g. Gates Foundation) offer technical solutions with green outcomes (GMOs, drip irrigation, drones and GPS controlled pesticide application), whilst others rooted in a more critical position on food sovereignty argue for a more radical reconfiguration of global food chains in the interests of both producers and the environment (Pretty et al. 2010; Morvaridi 2016; Mdee et al. 2019). There is a growing evidence base that genuine sustainable agriculture (low input, agro-ecological systems) can deliver intensification and improved livelihoods (Pretty 2001, 2002, 2003; Altieri and Toledo 2011; Altieri, Funes-Monzote, and Petersen 2012; Rosset et al.2011; Bennett and Franzel 2013; Nyantakyi-Frimpong et al. 2016; Khadse et al. 2018; Mdee et al. 2019). Yet, the barriers to agricultural policy being shaped by this evidence are significant (Isgren 2016). It runs counter to a deeply embedded and hegemonic policy rhetoric, where modernisation is equal to hybrid seeds, inorganic fertilisers, and large-scale mono-cropping (Engström and Hajdu 2019).

This policy rhetoric is built on a political economy of agriculture in which, the rural producers (as the backbone of the economy, and through democratisation as voters) can be embedded, controlled and exploited. In Zambia, for instance, Chapoto, Kabaghe, and Zulu-Mbata (2015) observe how heightened public expenditure on agricultural subsidies often coincides with general elections.

Taxation and price controls were common strategies that enabled elite control in the sector in both colonial and post-colonial eras. Thus, the prevailing basis of agriculture policies focuses on the dominance of political elites at the expense of small-scale agricultural development (Poulton 2012).

More overtly, subsidies and support provided for farmers can also act as an indirect source of funds for campaigns and electoral strategies for winning rural votes (Banful 2011) and serving as quick means to compensate for a lack of longer-term investment in rural infrastructure (Dorward et al. 2009). It benefits governments to keep the price of food low for noisy and politically active urban populations, whilst also subsidising production for the populous rural voters. Subsidies to agricultural production are frequently subject to local elite capture in their allocation and may limit sustainability of intensification through decreasing cropping diversification, promoting the use of inorganic inputs and ignoring soil fertility in the long term (Jayne and Rashid 2013; Chapoto, Kabaghe, and Zulu-Mbata 2015).

2.3. Challenge 3- the seductive simplifications of identity-based inclusion

Finally, addressing inequality in agricultural policy is framed predominantly through an identity-labelling approach (e.g. women or youth or vulnerable groups such as people living with HIV/AIDS). We see this reflected in the question which underpins this paper, and of course, this is built on substantial foundations in international development policy (Doss et al. 2018). This is consistent with neo-liberal understandings of poverty that frame it as an individual problem resulting from identity-based disadvantage, rather than a more structural product of class relations, resource control and power. The net effect of this is that in the international development industry, and in the policies that they influence we see a tendency to tackle inclusion through identity-based labelling. Therefore, it is argued that ‘women’ or ‘youth’ require special attention in order for them to overcome disadvantage (Hickel 2014). This may take the form of women or youth specific projects or targeted inclusion in decision making. However, this approach has significant weakness in addressing structural disadvantages.

Women, more than youth, have been labelled either as cardboard victims or heroines (Cornwall 2016). Ironically, at a time when it is argued that patriarchal decision-making power is declining in rural households, giving way for women and youth to gain more decision-making power and economic autonomy (Bryceson 2009), broad brush assumptions in relation to gender still dominate agricultural policies (Okali 2012). Doss et al. (2018) identify a number of dominant policy myths such as ‘women produce 60%–80% of food’. Whilst this statement is used to demonstrate the critical role women play in food production, they explain that these universalistic phrases, often based on patchy evidence, ignore the diversity of social and gender relations in agrarian households. Whilst such statements are widely accepted and propagated, they fail to create a meaningful change through the policies they inform (Okali 2012; Andersson Djurfeldt et al. 2018; Doss et al. 2018).

Without a more complex understanding of increasing differentiation within and between rural households, it is unlikely that policy will be able to respond to making agricultural intensification more inclusive (Doss, Summerfield, and Tsikata 2014; Tsikata 2016; Andersson Djurfeldt, Dzanku, and Isinika 2018).

This is not to say that gender and youth are not important factors in understanding inclusion and exclusion in agricultural livelihoods. However, they cannot be universalised and need to be understood as contributory and integrated into dynamics of differentiation, along with many other factors. Current evidence suggests a continued trend of increasing differentiation and class formation in agricultural dynamics (Jayne, Chamberlin, and Headey 2014). It is not enough to address this dynamic through identity-based intervention, rather state and political institutions have to address the consequences for those who are impacted negatively by current agricultural dynamics, such as those rendered landless (Snyder and Cullen 2014).

3. Assessing evidence for inclusive sustainable intensification in agricultural policy and livelihoods in Malawi, Tanzania and Zambia

Political economy analysis aims to provide a reasoned explanation for how a current situation comes to be as it is, in addition to elucidating different actor’s strategies and motivations (Jerven 2014; Nyame and Grant 2014). The nature of institutions and how they shape change is particularly key to understanding how change happens, who influences it and to what outcomes it has led.

Our analysis is based on the triangulation of multiple secondary and primary data sources. We began with a literature review of evidence on agricultural policy and donor intervention in the case study countries, with the purpose of understanding the broad dynamics of agriculture and how these align with the notion of ISI. This was substantially informed by the previous work of the authors. Anna Mdee has conducted extensive field research on agricultural livelihoods and natural resources governance in multiple locations in Tanzania since 1996 (Toner 2003; Cleaver and Toner 2006; Harrison and Mdee 2017a, 2017b; Mdee 2017; Harrison and Mdee 2018; Mdee et al. 2019; Mdee and Harrison 2019; Brockington et al. 2019); Michael Chasukwa has researched and published on the political economy of agriculture and on local government capacity in Malawi since 2005 (Chinsinga and Chasukwa 2012, 2015, 2018a, 2018b; Chinsinga, Chasukwa, and Zuka 2013), and Simon Manda has researched and published extensively on agriculture in Zambia, including a recent three year investigation of large scale land acquisition (Manda, Dougill, and Tallontire 2018a, 2018b, 2019). We also reviewed the AFRINT survey data for each of the countries and worked in close co-operation with senior researchers from AFRINT country teams.² Additional primary data collection was undertaken from November 2017 to February 2018 for the purposes of triangulation. This consisted of key informant elite interviews with government ministries, universities, NGOs and development partners both at national and district levels and primary data and focus group discussions with farmers in each country.³ Full details on the background fieldwork are published in a series of detailed working papers under the Policy for Equity in African Agriculture project: https://sairla-africa.org/what-we-do/research/policy-for-equity-in-african-agriculture-afrint-iv/.

3.1. Dynamics of agriculture in Malawi, Zambia and Tanzania

Notwithstanding the specificity of the detailed context of each of these three cases and variations in terms of politics, socio-economic conditions and agriculture policies, it is possible to identify a set of common trends and themes. This section should be read in conjunction with Table 1 which provides a synthesis of evidence on trends in agricultural policy and livelihoods.

Table 1. Dynamics of agricultural change in Malawi, Zambia and Tanzania.

Era	Pre-colonial	Colonial era	Independence – structural development	Structural adjustment and PRSP 1985–2000 – transition to multi-party democracy	MDGs 2000–2015 Debt relief Highly indebted poor countries	SDGs 2015
Agricultural Policies	None	Extensification Plantations Codification and taxation of customary lands Enforcement of patriarchal norms from UK e.g. on property ownership Modernisation of agriculture colonial extraction Malawi and Zambia – more settler agriculture and segregation of subsistence production	State led investment Nationalisation Agricultural subsidies State led co-operatives Modernisation of agriculture for state-building  - Tanzania- collectivisation  - Malawi and Zambia- elite accumulation Zambia – significant mining activity in economy, migration	Removal of subsidies Dismantling of parastatal and co-operatives Liberalisation- requiring agricultural modernisation for market-driven growth Re-emergence of subsidy (Zambia/Malawi) Agriculture Research heavily underfunded (market-led international research institutes take over e.g. seed companies)	Gradual freeing up of agricultural markets Greater access to inputs for those who can pay Liberalisation- stimulating agricultural modernisation for market-driven growth and attracting foreign investment Land reform (land grab?) Significant subsidy in Malawi and Zambia Public-Private Partnership New outgrower schemes	Intensification (new green revolution) Climate smart agriculture and sustainable intensification Missing potential from irrigation Industrialisation- underpinned by agricultural transformation
Key actors	Customary arrangements Arab traders (Tanzania) Early commercial interests	Colonial authorities Tribal authorities Commercial investors	Government donors, esp World Bank Tribal Authorities (but banned in Tanzania)	Government (but significantly reduced state investment) Donors Private sector and civil society growth Tribal Authorities (Zambia and Malawi)	Government – poverty reduction strategy Donors (with increasing influence from China) Private sector (increasing influence from South Africa) and civil society growth	Government – new industrialisation and climate change challenge Donors Private sector and civil society actors
Agricultural outcomes	Peasant and pastoral societies	Mainly peasant and pastoral mode with some large schemes – e.g. sisal and groundnuts, tobacco	Outcomes disputed – failure of some co-operatives Disruption of peasant mode in some places Peasant mode underpins food security Zambia – more urbanisation and migration; mining	Agricultural decline and soil degradation – food production keeps up with population expansion Diversification of livelihoods increasing off-farm income (deagrarianisation) Zambia – copper price decline	Agricultural growth – but limited Urban expansion- migration Continued increase of off-farm income Tz – high growth sectors: mining, construction Zambia – commodity price boom	No clear story – except of differentiation Extensification in some areas Decline in some areas Intensification in some areas
Inclusion of women and youth	Women and youth embedded in customary relations these are not uniform, unfixed or unchanging	Imposition of patriarchal laws from UK Taxation requires increased waged work Increasing home production burden on women and youth	Critical role of women within liberation movements Youth leagues Increasing access to education	Beijing Declaration on Rights of Women 1995 Special seats and women’s representation (Tanzania – 30% quotas) Increasing influence of NGOs working on women’s rights SAPs cause decline in access to education and other public services African Youth Charter 2006	Increased visibility of women in education and formal employment Legal reform on ownership giving women rights to land Civil society focus on women – access to loans	Educated Youth unemployment Increased youth interest in commercial agriculture (for those with resources) Increasing debt levels

All three countries share commonalities of colonial engagement, post-independence state-led agricultural investment, followed by structural adjustment and liberalisation, moving to a current concern with the ‘African green revolution’, large-scale commercial investment, and climate-smart agriculture. The discourse of agricultural modernisation is present in all three countries. Whilst these policy directions are heavily influenced by donors, the politics involved is complex. The aim of dominant political regimes of both colonial and post-colonial eras favours full-scale agricultural transformation, meaning the shift of labour from agriculture, and a concentration of land-holdings, through intensification and commercial investment; but this goal has remained elusive across the case study countries. Whereas the share of agriculture in GDP may be declining to variable degrees, small-scale agriculture remains highly significant in the livelihoods of most of the population (Manjengwa, Hanlon, and Smart 2014; Harrison and Mdee 2018). All three countries are experiencing large-scale land acquisition and increasing differentiation of rural wealth (Andersson Djurfeldt and Hillbom 2016; Chinsinga 2017; Chung 2017; Bluwstein et al. 2018). Whilst decentralisation shapes the governance agenda of all three governments, the process is incomplete, and local government support to agriculture is under-resourced and fragmented (O’Neil et al. 2014; Mdee, Tshomba, and Mushi 2017). Tanzania abolished ‘traditional authorities’, but they retain a central role in agricultural governance in Malawi and Zambia (Manda, Dougill, and Tallontire 2019).

Malawi, Zambia and Tanzania in this era of new green revolution have implemented state reforms through policy and donor-funded programmes geared towards accelerating modernisation and commercialisation of agriculture through Private-Public Partnership (PPP) approaches. In Tanzania, the Southern Agricultural Growth Corridor of Tanzania (SAGCOT) Programme, an investment strategy under Kilimo Kwanza (Agriculture First) aims to focus investment and commercialisation of agriculture in the central zone. However this has created space for local elites and companies to accumulate land and prosper at the expense of vulnerable groups and small producers (Chung 2017; Rasmussen and Lund 2018; Engström and Hajdu 2019). Similar criticisms are made of large-scale land acquisitions and PPPs such as the Green Belt initiative in Malawi (Chinsinga 2017), and the farm block initiative and out-grower schemes in Zambia (Sitko and Jayne 2014; Matenga 2017). Land policy in all three cases has similarities of duality with customary or collective tenure sitting alongside more formal systems of individual ownership. Again, there is existing broad literature on this that supports our analysis, and this is further confirmed in fieldwork and is explored in the next section. There is some evidence in all three countries that differentiation in land holdings is growing, and that exploitation of customary land is being enabled by traditional and other local leaders (Peters 2013; Andersson Djurfeldt and Hillbom 2016; Chinsinga 2017; Matenga and Hichaambwa 2017; Bluwstein et al. 2018: Manda, Dougill, and Tallontire 2019).

The shape of the economy as a whole is critical to the agricultural sector. The three countries in our study all have a large population base who derive a considerable proportion of their livelihoods from agriculture. However, there are differences relating to the size of land holdings and the shape of the wider economy. In Zambia, the mining sector underpins significantly more urbanisation than in Malawi or Tanzania. However, wider economic growth in mining, tourism and construction has also driven urbanisation in Tanzania. Malawi remains dependent on agricultural production for export revenue. The poorest households in all three countries depend on rain-fed agriculture, but strategies of migration and family network remittances are critical to their livelihoods (Andersson Djurfeldt, Dzanku, and Isinika 2018).

Malawi and Zambia both maintain high levels of state spending on input subsidies in the face of protracted donor opposition. These are politically important programmes and the focus of considerable debate, relating to their design, impact, and political capture (Sitko et al. 2017). For the past decade, FAO data has demonstrated increases, to different degrees, in central government expenditure in agriculture in all three countries (see Figure 1) which may partially relate to the adoption of the Comprehensive Africa Agriculture Development Programme (CAADP). Relative to government expenditures on sectors such as education and road infrastructure, spending on agriculture has not necessarily increased as a proportion of total government expenditure.

Figure 1. Trendline graph showing central government expenditure on agriculture in Malawi, Zambia and Tanzania (FAOSTAT data, 2018).

Gender and youth as categories of inclusion are present in the form of ‘policy noise’, workshops, committee quotas (very common in Tanzania) and groups. However, in all three cases, the greatest frontier of differentiation and exclusion appears to be class-based, within which gender and youth as categories also have some (but differentiated) significance.

In all three countries, donor and NGO intervention in agriculture is seen as confused, contradictory, driving elite capture, and supporting corruption. This view is repeatedly emphasised by elite interviewees in all sectors (government, NGOs and private). In Zambia, for example, this manifests itself in poor coordination of large-scale foreign investments in agriculture. This highlights a gulf between policy in theory and implementation capacity in practice. Analysis using the wider literature on local service delivery and governance underlines this case (e.g. Mdee, Tshomba, and Mushi 2017 on Tanzania). All three countries exhibit a critical lack of implementation capacity in local government support to agriculture and are forced to attempt ‘gap-filling’ through donor-funded projects. The outcome of this is an ad-hoc short term and fragmented approach that does not support longer term institutional capacity development. This is emphasised in elite interviews and reconfirmed in fieldwork at district levels. For example, in Malawi, a key informant at the local council made the observation: ‘we do not get enough resources from central government. The only way for us to be visible on the ground is to accept being involved in projects implemented by NGOs. We go by what they want us to do because it is a project’.⁴

Our research shows that irrigation development is seen as a critical component of agricultural modernisation. This again reflects government (with donor support) priorities of encouraging public-private investment in irrigation infrastructure – though state-led expansion remains slow (Manda, Dougill, and Tallontire 2019), farmer-led irrigation is expanding rapidly (Mdee and Harrison 2019). There is inadequate implementation capacity (even where there are comprehensive policy frameworks) for managing contested water use, and water scarcity is a growing concern (Harrison and Chiroro 2017; Harrison and Mdee 2018).

In Malawi and Zambia, state subsidy and market controls have significant impacts on driving maize production, although both countries are attempting to diversify their schemes (Hanjra and Culas 2011; Mpesi and Muriaas 2012; Mason, Jayne, and Mofya-Mukuka 2013). In Tanzania, the government subsidises maize through an input voucher, but on a much smaller scale (Gabagambi, Mkangwa, and Kadeng ‘uka 2015). It does intervene through the setting of tariffs and import/export bans on particular goods. These can be erratic and have significant impacts on markets in some cases, and we will explore this further below.

Evidence in all three countries suggests that access to reliable and fair markets remains a challenge for small farmers, even when they successfully increase their production (Chirwa and Chinsinga 2012; Matenga and Hichaambwa 2017; Harrison and Mdee 2017b).

From our analysis in Malawi, Tanzania and Zambia, we identify three key challenges that will need to be addressed if inclusive sustainable intensification is to be achieved: 1. Closing the gap between policy in theory and policy implementation in practice. 2. Addressing the role of maize and agricultural subsidies, and 3. Addressing the dynamics of elite capture in land and water acquisition.

3.2. The policy-practice gap

Malawi, Tanzania and Zambia have extensive policy frameworks relating to agriculture and related areas – such as irrigation, climate change, gender, etc. With donor focus on ‘good governance’-driven institutional change over the last decade, considerable resources and incentives have been in place to create policy and make ‘policy noise’. There is significant evidence of ‘isomorphic mimicry’ in creating a policy architecture that has the appearance of reform, without having the capability and capacity to implement extensive change (Andrews, Pritchett, and Woolcock 2013; Mdee and Harrison 2019).

How things work in practice is different from the neat intentions set out in policies. The reasons for this are multiple – relating to the creation of policy in vacuums in central government and elite levels, the hierarchical tendencies of government bureaucracy, the practical politics of power and patronage, and significant resource constraints (see Mdee, Tshomba, and Mushi 2017 for a detailed exploration of this in the Tanzanian context). In Zambia, recent evidence shows that ‘whilst possibilities for LaSLAs are created by state institutions, the state agencies seeking to administer land-based resources also limit their potential through competing authority and agendas’ (Manda, Dougill, and Tallontire 2019, 194).

An academic interviewee in Malawi also illustrates this issue:

The National Agricultural policy processes demonstrate that policies in Malawi involve multiple stakeholders, multiple arenas and multiple levels. This has significant implications for the level of coordination and coherence required to achieve the desired goals and objectives of any policy process. The level of complexity varies depending on the nature of the specific policy. The challenges demonstrate that policy and policy-making is conditioned and shaped by the political, social and economic context, as well as historical factors. It is also important to note that there is a lack of an internal sense of urgency to systematically address policy challenges within the government. This implies that the government is not proactive when it comes to initiating processes leading to the development of relevant policy documents. Almost always, policy processes are a reaction to international pressures and implemented on a project basis. Ministry or departmental led policy processes can be overshadowed by politically orientated policy pronouncements. There is often a sense of urgency to do something about these policy pronouncements since they may relate to political profiles for electoral purposes and are, usually, at the expense of other equally pressing policy processes and priorities. (Key informant Interview-Malawi 2018)

This view is consistently confirmed in interviews with representatives from government, NGOs, academics and donors (off record comments by local staff).

Aid is weakening everybody- we need to kick out the donors. There is so little capacity to implement the policies that are designed by them. In the end, the Districts just rely on NGOs to do anything. (Key informant interview, Malawi 2018)

For inclusive sustainable intensification in agriculture to be realised, policies will be required that start with agricultural conditions and resources as they are, and not as they are wished to be. This will require closing the gap between policies that exist in theory (only on paper) and those that can be actually implemented in the context of available resources. For example, Tanzania has a complex legislative framework for the management of water resources but is chronically lacking in capacity to implement them (Harrison and Mdee 2017b).

3.3. The politics of maize

In all three countries, maize remains a key staple food crop for both urban and rural dwellers, and thus is politically significant. The elevation of crops such as maize into a political food crop is predominant in Malawi and Zambia and presents itself as a form of social contract between the government and citizens (Banful 2011; Hanjra and Culas 2011; Mpesi and Muriaas 2012; Jayne and Rashid 2013; Chinsinga and Poulton 2014; Kato and Greeley 2016; Morgan et al. 2019). The dominant political narrative is that adequate maize availability translates into food security. These forms of social contract limit the government’s spending on other food crops and broader agricultural transformation initiatives and are fulfilled by the government in the form of inputs subsidy programmes and policies, consumer and producer price controls and marketing strategies, and these have significant impacts on farmers’ crop choice and production practice. Morgan et al. (2019) confirm our findings that such subsidies disincentivise production methods associated with sustainable intensification.

In Malawi, key informant interviews confirmed that not only are policy programmes like the Farmer Inputs Subsidy Program (FISP) problematic and subject to political manipulation, but they are also trapping smallholder farmers in maize production. Moreover, when the government establishes a parastatal agency that purchases a single food crop for farmers, it demonstrates the level of importance of the crop relative to other crops. In Zambia, the Food Reserve Agency (FRA) and FISP budgets dominate other agricultural expenditure in the country (Jayne and Rashid 2013). FRA purchases maize from farmers above the current market/privatized wholesale prices and deposits maize purchased into commercial mills. In Tanzania, the use of input subsidies is less extensive than Malawi and Zambia, however it is noted that the National Agriculture Input Voucher System (NAIVS) was subject to political demands for extension to certain strategic locations of the country (Gabagambi, Mkangwa, and Kadeng ‘uka 2015).

Hence, whilst providing input subsidies for maize might target crop productivity improvement, it also creates political and financial incentives for farmers to keep producing maize with inorganic inputs, (NPK fertilisers), for governments to manipulate rural voters, and for fertiliser importers and agro-business dealers to benefit from government contracts (Jayne and Rashid 2013). Additionally, there are potentially negative consequences of increasing the vulnerability of smallholder production to droughts and pests, and lowering the resilience of production systems (Morgan et al. 2019). In Zambia, the culture of defining agriculture in terms of maize production has been identified as stifling wider efforts towards diversification and market participation among local farmers (Manda, Dougill, and Tallontire 2019). Similarly, the narrative among citizens in Malawi that ‘food security is maize availability’⁵ leads political authorities to concentrate on promoting maize to resonate well with the electorate. It is evident, thus, that the these narrow subsidy programmes can frustrate efforts for crop diversification and climate change adaptation interventions (Chinsinga and Poulton 2014).

Key informants in all the three countries suggest that local agro-dealers deliberately do not stock diverse inputs because they are not profitable (confirmed in multiple interviews in 2018). An expert on agricultural subsidy in Malawi argued that the subsidy programme design assumes that agro-dealers should provide a comprehensive agricultural support system to farmers, including the provision of extension services and through supplying a range of agricultural inputs. However, in practice this has turned out to be impossible because of commercialisation of subsidy programmes and the temporary nature of agro-dealers – with most emerging opportunistically at the start of subsidy programme and closing the business at the end of the subsidy period. Furthermore, smallholder farmers in hard-to-reach areas are under-served because agro-dealers do not find any business justification to operate in less lucrative areas (see also Andersson Djurfeldt, Dzanku, and Isinika 2018). In these difficult-to-reach areas, people maybe systematically excluded because they are unable to walk long distances to buy farm inputs even if they have vouchers. At the District levels, agricultural extensions officers observed that it is also difficult for government to regulate agro-dealers and other suppliers because some are politically connected and represent interests of elites and political powers, and farmers are often at risk of purchasing fake inputs. Therefore whilst subsidy programmes may be nationwide in design, in practice they tend to focus on areas of commercial and political significance.

Without explicit recognition of the political incentives and economic interests that drive subsidy programmes it is difficult to see how governments will shift their focus from these programmes as their major form of intervention in agricultural livelihoods. It is conceivable that such programmes are redesigned to incentivise practices associated with ISI, but that will require political will.

3.4. Dynamics of differentiation

There is some evidence from our literature review, and supported through key informant interviews in all three countries, that differentiation in land holdings is growing, and that exploitation and appropriation of customary or collectively managed land is being enabled by ‘traditional’ and other local leaders (Peters 2013; Sitko and Jayne 2014; Sitko and Chamberlin 2016; Bluwstein et al. 2018; Rasmussen and Lund 2018; Chinsinga and Chasukwa 2018a). Moreover, there is evidence of elite capture in both land allocations and agricultural subsidy schemes. The formal systems (such as the distribution of vouchers or selection of recipients for subsidy) interact with the more informal systems of power. Elites should not be thought of only as the top political and business elite, but include the relatively wealthy and more powerful in a local setting. Elite capture contributes to class differentiation, as those with connections to powerholders access opportunities for subsidy, capital accumulation and investment (Mdee and Harrison 2019).

Initiatives seeking inclusion of ‘women’ or ‘youth’ should be understood against this background. It cannot be assumed that institutional arrangements are uniformly discriminatory in relation to gender and generation. Such social relations are dynamic and shifting, for example in relation to the shape of the wider economy, e.g. if there are significant waged labour opportunities that can enable men to migrate. The burden of those who take responsibility for domestic reproduction much greater where the state does not ensure the provision of basic services, and therefore gendering of labour must be understood in the wider context of the nature of the economy and class differentiation (Fakier and Cock 2018).

Debates on inclusivity in agriculture often focus on rights to land and accessibility of inputs for individuals. Many of the NGOs interviewed in this research are targeting input and productive interventions at women, using a rhetorical narrative of mainstreaming gender. Much of the research under this theme disaggregates agricultural production and practice based on the sex of the ‘household’ head or ‘plot manager’ (Andersson Djurfeldt et al. 2018). Surveys often assume that the household exists as a formation of one unit per dwelling. Rather, households are rather more diverse networks of kin and familial relations with various configurations of resource flows (Brockington et al. 2019). Farming is more usually a collective undertaking, with a wide range of diverse intra-familial arrangements for deciding how to go about this and how to share the benefits. In some parts of Malawi, land inheritance is patrilineal, in other areas matrilineal (see Peters 1997 for example) but such classifications also shift and change. Simply giving individual women title to land, or access, to inputs cannot be transformational to unequal gender relations (see also Andersson Djurfeldt et al. 2018). Historical literature suggests that colonial regimes introduced the gender and property norms of the Victorian United Kingdom along with the capitalist mode of production, and that these then overlaid a multiplicity of customary arrangements to shape the current landscape of social relations (Mandala 1984; Peters 1997). Therefore, gender relations and inclusion are far more complex than the standard government and donor discourse allows, as is asserted here in a key informant interview with an NGO manager:

My experience is that when we talk about mainstreaming youth and gender, we have to unpack these concepts. Representation does not become true mainstreaming. What do numbers tell us? We have to understand issues of access and control – can our monitoring really pick this up? Women will tell us in interviews that their husbands make the decisions, even if that is not the case. That is what they think that they should say (Interview, NGO Programme Officer, Malawi January 2018).

The concept of ISI cannot be confined to identity markers (such as gender or youth) as these mask economic and social relations that shape more complex and nuanced dynamics that drive inequality and differentiation. Within households and family farming units, decision-making and benefit sharing is differentiated and potentially disadvantages some and advantages others, and this may have elements of gendering (Andersson Djurfeldt et al. 2018).

However, key informants in all three countries suggested that the more powerful drivers of differentiation and inequality are to be found in patterns of land acquisitions, market access and unstable prices, unequal market inclusion leading to shortage of credit or debt traps and lack of alternative off-farm employment (Manda, Dougill, and Tallontire 2019).

4. Discussion and conclusion

Our analysis suggests that the normative goal of inclusive sustainable intensification in agriculture is somewhat divorced from current realities of agricultural policy and practice in Tanzania, Malawi and Zambia.

Firstly, the capacity of governments and local institutions to deliver coherent and multiple policies that ISI will require is highly doubtful. There is a very large gap between policy commitments on paper and the capacity of local institutions to deliver. There is no clear agreement on what characterises inclusive and sustainable agricultural production, and little evidence of coherent or sustained policy implementation to that affect, outside of donor or NGO funded projects. Agricultural policy remains dominated by a narrative of technology-driven modernisation and commercialisation, but evidence of a sustained advance towards that outcome is also limited. Agricultural subsidies are critical components of this landscape and currently create incentives for maize production in particular, and disincentivise the adoption of sustainable practices in agriculture (see Morgan et al. 2019). They also link symbiotically to the dynamics of rural politics and elections.

Secondly, our strongest conclusion is that agriculture is becoming less inclusive. This is being driven by the possibilities for elite capture in state subsidies, an increasing ‘middle’ class and elite accumulation of land, as well as donor-corporate-government alliances favouring large-scale land acquisitions. A renewed policy rhetoric on industrialisation might assume that the small farmer becomes a labour reserve for new industry, but there is little sign of new industries to fulfil this role.

There appears to be a failure of policy at multiple scales: at the global level policies of sustainable intensification are framed by a vague understanding of environment and agricultural production in a way which few can question on normative grounds. Who can possibly be opposed to a policy of sustainable, inclusive, intensification? This policy rhetoric then interacts with national and local political dynamics which, coupled with limited implementation capacity, produces an environment that favours local elite resource accumulation, and at the same time also encourages the entry of a multitude of donor-funded actors to compensate for the shortfall in state implementation capacity, creating a fragmented mess of donor-driven interventions.

Where agricultural intervention does exist, then elite capture at all levels is a frequent issue, as is illustrated in the example of the Farm Input Subsidy Programme (FISP) of Zambia. Traditional Authorities and resource allocation committees are frequent sites of elite capture and potential exploitation. Elite capture may have a gender dimension but requires explicit consideration of class. This also applies fundamentally to land policy, where land titling and formalisation may have made it easier for the poorest farmers to be dispossessed of land. A poorer man is very much more disadvantaged than a wealthier woman in this regard.

Policy frameworks are dominated by an aid-driven donor discourse. State investment in agriculture remains limited. Private finance is unaffordable to the small farmer and out-grower schemes have disappointed many of those involved. Elite and commercial interests are favoured in legal frameworks and in the normal business of institutional actors. Markets remain exploitative, hard to access, and unreliable for the small farmer. Exploitation and dispossession of resources are the dominant trends. These trends run entirely counter to the notion of inclusive sustainable intensification, and significant political will be required to reverse them. The dynamics of agriculture in Malawi, Tanzania and Zambia are currently neither sustainable nor inclusive.

Acknowledgements

Thanks to Agnes Andersson Djurfeldt for her insightful comments during the preparation of this paper, and to the two anonymous reviews for their guidance in shaping the final paper. We would also like to thank Gasper Mdee, Darlen Dzimwe and Royd Malisase for their assistance and insights during the fieldwork. This paper draws on a set of working papers produced for the Policies for Equity in African Agriculture Project, led by Lund University and funded by the UK Department for International Development. However this paper represents the views of the authors alone.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

‘The Equity and Institutions in Sustainable Africa Intensification’ [grant number S-69] research project was funded by UK Government and supported through the Sustainable Intensification Research and Learning in Africa (SAIRLA) programme. The views expressed in this article are those of the authors and do not necessarily reflect the UK government’s official policies.

Notes on contributors

Anna Mdee

Anna Mdee is Professor in the Politics of Global Development at the University of Leeds. She is an anthropologist who has worked on agricultural livelihoods and local governance in Tanzania since 1996. She is also Associate Director of water@leeds and has a particular interest in the politics of irrigation in the transformation of agriculture in Tanzania.

Alesia Ofori

Alesia Ofori Dedaa is currently a PhD student at the School of Politics and International Studies at the University of Leeds. Her PhD research seeks to excavate some of the complex political issues around the nexus between mineral exploration and community water supply, access and distribution in sub-Saharan Africa, particularly Ghana. Alesia’s research interests encompass rural governance and development, agrarian change and the political ecology of waterscapes.

Michael Chasukwa

Michael Chasukwa is a Senior Lecturer and Head of the Department of Political and Administrative Studies, Chancellor College, University of Malawi. He obtained his doctorate in development studies from the University of Leeds, United Kingdom. He has published in various journals including the International Journal of Public Administration, Agrarian South: the Journal of Political Economy, Insights on Africa, Journal of Development Effectiveness, Africa Review, and Journal of Asian and African Studies.

Simon Manda

Simon Manda is a Lecturer at the University of Zambia Department of Development Studies. He holds a PhD in Environment and Development and an MA in Global Development from University of Leeds, UK. He also holds a Bachelor of Arts Degree in Development Studies from the University of Zambia. His current research focuses on labour migration in agriculture.

Notes

1 The project is funded by the UK Department for International Development under the programme Sustainable Agricultural Intensification Research and Learning in Africa (SAIRLA) https://sairla-africa.org/. The Policy for Equity in African Agriculture project is led by Lund University under the AFRINT project https://www.keg.lu.se/en/research/research-projects/current-research-projects/afrint.

2 AFRINT is a multi-country longitudal study of agricultural dynamics in 8 African countries, based on household surveys: https://www.keg.lu.se/en/research/research-projects/current-research-projects/afrint.

3 20–30 key informant interviews were conducted in each country – purposively sampled to cover the perspectives of government ministries, agricultural extension workers, local, nation and international NGOs, development partners (inc World Bank and faith-based organisations). Numbers of focus groups varied by country and in relation to existing data sets. Focus groups differentiated participants by gender and were used to do a cross check on the themes emerging from key informant interviews.

4 Key informant interview, Dedza, Malawi.

5 Key informant, Mulanje District Council.