Are agriculture and nutrition policies and practice coherent? Stakeholder evidence from Afghanistan

Abstract

Despite recent improvements in the national average, stunting levels in Afghanistan exceed 70% in some Provinces. Agriculture serves as the main source of livelihood for over half of the population and has the potential to be a strong driver of a reduction in under-nutrition. This article reports research conducted through interviews with stakeholders in agriculture and nutrition in the capital, Kabul, and four provinces of Afghanistan, to gain a better understanding of the institutional and political factors surrounding policy making and the nutrition-sensitivity of agriculture. Semi-structured interviews were conducted with a total of 46 stakeholders from central government and four provinces, including staff from international organizations, NGOs and universities. We found evidence of interdisciplinary communication at the central level and within Provinces, but little evidence of vertical coordination in policy formulation and implementation between the centre and Provinces. Policy formulation and decision making were largely sectoral, top-down, and poorly contextualised. The weaknesses identified in policy formulation, focus, knowledge management, and human and financial resources inhibit the orientation of national agricultural development strategies towards nutrition-sensitivity. Integrating agriculture and nutrition policies requires explicit leadership from the centre. However, effectiveness of a food-based approach to reducing nutrition insecurity will depend on decentralising policy ownership to the regions and provinces through stronger subnational governance. Security and humanitarian considerations point to the need to manage and integrate in a deliberate way the acute humanitarian care and long-term development needs, of which malnutrition is just one element.

1 Introduction

Public health nutrition is a major development priority and a pillar of the global sustainable development agenda (Horton and Lo 2013). Tackling malnutrition requires a combination of targeted nutrition-specific interventions to address the immediate causes of malnutrition as well as nutrition-sensitive interventions to address the underlying determinants. Improving the nutrition-sensitivity of agriculture is an approach to tackling malnutrition that exploits the multiple linkages between agriculture and nutrition such as food provision, income, employment and health (Hawkes and Ruel 2011; Ruel and Alderman 2013).

1.1 Agriculture and nutrition

The close relationship between agriculture and nutrition and wider contextual factors can be depicted by ordering the 17 Sustainable Development Goals as in Fig. 1. This highlights those Goals which immediately concern people’s health and welfare, with food and nutrition at the centre. In the second ring are those Goals proximally related to agriculture and nutrition which plausibly have a less direct but important local impact. Depicted outermost are the distal contextual and global issues, some implying governance, but most of which make a huge, sometimes less direct, contribution to agriculture and nutrition. This presentation emphasises how the agriculture and nutrition sectors are intertwined, and also how the wider environment of proximal and distal issues constitute significant enabling – or disabling – factors.

Fig. 1

Agriculture, nutrition and the Sustainable Development Goals (SDGs). Source: adapted from Waage and Yap (2015) and Poole (2017: 10)

Agriculture and nutrition are linked in obvious and less obvious ways. Diverse pathways were outlined by Gillespie et al. (2012) but the effectiveness of agricultural interventions upon nutritional outcomes has yet to be conclusively demonstrated. Reasons for this may be that studies assessing the effectiveness of agricultural interventions often have methodological weaknesses, flawed study designs, and fail to account for the multiple pathways, as well as the context and environment in which they operate (Masset et al. 2012). Thus, robust evidence of linkages between agricultural policy and nutrition is limited (Dangour et al. 2013).

1.2 The enabling environment

For agricultural interventions and policies to succeed, they must be embedded within a conducive ‘enabling environment’ (Gillespie et al. 2013; Ruel and Alderman 2013), meaning ‘the sociocultural, economic, political, institutional and policy contexts that govern the design and implementation of nutrition-relevant actions’ (van den Bold et al. 2015: 232). Gillespie et al. (2013) offer a framework for understanding the enabling environment for nutrition consisting of three key components. The political and governance component of the framework includes national governments, civil society, international and national development and research organisations, academia and the private sector firms which operate and interact with one another. It refers to the way that policies and programmes are created as a result of collaboration or conflict between sectors and organisations, the nature of hierarchical and horizontal governance structures and the availability and allocation of financial and other resources (Bryce et al. 2008). The knowledge and evidence component concerns how data, knowledge and information are created, distributed and utilised, capturing how nutrition is considered within a national or international agenda. The framing of an issue within the policy environment affects how data and evidence are collected and presented, and how they are communicated to governments and policy makers. The third component of the enabling environment highlights the resources and levels of human capacity within the policy environment including the capacity of policy makers, champions and implementers as well as the national and organisational capacity, finances and resource prioritisation.

These elements of an enabling environment focus on policy and capacity. However, they omit important contextual components suggested in Fig. 1, which recognises the SDGs as a comprehensive set of interrelated challenges. Therefore, in addition, we suggest that within the wider enabling environment there is, first, the natural resources base on which agriculture depends. This encompasses the soils, water and biodiversity, all of which in many places are threatened by climate change and rising temperatures, which phenomena in turn increase the risk of natural disasters (Blaney et al. 2009; Poole et al. 2016a). Second, security, justice and accountability are essential for productive activities, in rural and urban areas, and for household well-being (Barnett and Adger 2007; Lynch et al. 2013). Third, governance is the reach of the state and the non-state or private delivery of public goods and services. These include education, energy and infrastructure which are essential for development and improvements in the agriculture, nutrition and health sectors. These components are weak or absent in fragile states. This article concerns Afghanistan, where the delivery of public services is largely contracted to non-governmental organizations (Poole 2018). This public-NGO, or PUNGO, modality has advantages but also creates tensions. Finally, like capacity, political leadership by strong individuals has been identified as a common element of success in tackling under-nutrition (Nisbett et al. 2015; Nisbett and Barnett 2017).

The research reported here was part of the DFID-funded research programme Leveraging Agriculture for Nutrition in South Asia (LANSA). Analysis of the enabling environment and the nutrition-sensitivity of agricultural policies in South Asia has been central to LANSA. The research aims to identify opportunities for agriculture and food policies to increase impacts on nutrition in India, Pakistan, Bangladesh and Afghanistan. Among the research projects completed by the consortium are explorations of the political, institutional and policy challenges of creating nutrition-sensitive agricultural policies in India, Bangladesh and Pakistan (Gillespie et al. 2015); and analysing stakeholders’ knowledge, and the political and institutional context of, nutrition-sensitive agriculture (van den Bold et al. 2015).

In this paper we examine the extent to which an enabling environment for nutrition-sensitive agriculture exists within the fourth LANSA country, Afghanistan. The objectives of the research were not policy analysis per se, but understanding the perceptions and practices of policy makers and other stakeholders themselves, concerning sectoral and inter-sectoral policies and processes: to identify and evaluate the evidence base linking agriculture to nutrition, understand the perceptions of decision makers about policy making and implementation, and identify the capacities for improving nutrition through the agrifood system.

We draw lessons about the specific challenges of politics and governance, knowledge and evidence, as well as capacity and resources. In the next section (2) we review the Afghan context, and provide an account of the research methods (section 3), results (section 4) and discussion (section 4). Conclusions relate specifically to this case, but include lessons for the broader Asian context and for other conflict and post-conflict situations where agriculture, nutrition and health are compromised (section 5).

2 Afghanistan: agriculture, nutrition and the wider enabling environment

Afghanistan is primarily an agrarian country (MAIL 2015): 80% of the total population and 90% of the poor live in rural areas (World Bank 2014). Agriculture accounts for 40% of the labour market and about one-quarter of national gross domestic product (GDP), excluding the opium poppy economy (CSO 2014). GDP growth, driven principally by agriculture (World Bank Group 2017) was expected to be 2.6% in 2017. This lags the average annual population growth rate of 3% (World Bank 2017b).

Wheat is the principal food crop in Afghanistan, while fruits and nuts, and livestock products such as wool continue to play a large part in subsistence, in local marketed production, and some part in exports. On average, per capita consumption of wheat is 152 kg/year, among the highest in the world (Naseri undated).

Afghanistan is largely arid, but with productive agriculture in fertile well-watered valleys, and a significant pastoral economy: almost 70% of the population keep livestock of some sort (MAIL 2015). Dupree’s study during the period of relative peace and prosperity preceding the Soviet invasion in 1979, which precipitated the series of conflicts which have lasted almost four decades, illustrates the potential of Afghan agriculture and how much has been lost (Dupree 1977). She recorded a diverse and productive agriculture: besides major crops like wheat, maize, sesame, potatoes and rice, there were plums, apples and pears, apricots, figs, mulberries, cherries and almonds, walnuts and pistachios, figs and pomegranates, citrus, melons and watermelons. Other significant crops were sunflower, sesame, cotton and sugar cane. Raisins and fresh grapes constituted one of Afghanistan’s principal export sectors: ‘Afghanistan once produced 20% of the world’s raisins. In the 1970s, Afghan cut flowers graced tables from the Middle East to Europe’ (MAIL 2018). Livestock products were also important in home consumption as well as for local sale and export: meat, milk, cheese and fibres.

In a predominantly agricultural society, why is malnutrition so serious, and what can be done?

2.1 The current fragile context

Afghanistan is often described as a fragile or failed state, with serious negative consequences for human welfare, among other results. There is no single understanding of ‘fragility’ (Coburn and Larson 2014). The UK Department for International Development (DFID) proposes a somewhat narrow interpretation of failure in the political senses of lack of authority, failure to provide services and lack of legitimacy (Brown et al. 2009). Börzel and Risse (2015) use different concepts and terminology to analyse political economy that are useful for this research into agriculture and nutrition policies. They prefer to characterize as areas of ‘limited statehood’ situations where the formal state is dysfunctional or absent. In such contexts, governance through the provision of rules and regulations and the delivery of public services is sustained by institutions that are ‘functional equivalents’ of the formal state.

The enduring interventions in Afghanistan since 1978 overlay an historical fragility that frustrates governance and nationhood in multiple ways. Political fragmentation is enmeshed with cultural diversity, geographical extremes of temperature and altitude, military conflict, the drugs economy, weak public infrastructure and services, environmental vulnerability and migration. In the context of Afghanistan, we understand ‘fragility’ to encompass ongoing conflict, dependence on high levels of humanitarian aid, significant political instability, a weak capacity to carry out the basic state functions of governance and public services delivery, lacking cohesive and productive relations with society and a climate of natural disasters and environmental threats (Stewart and Brown 2009; FAO 2010; Gupte 2017; OECD 2017).

It is simplistic – but sufficient for our immediate purposes – to suggest that there are at least three major areas of governance in Afghanistan that could be regarded as Börzel and Risse’s functional equivalents of the state: first, the presence of international donor organisations representing overseas governments; second, the almost ubiquitous presence of the non-governmental organisation (NGO) sector, both national and international, which delivers public services such as the Basic Package of Health Services; and third, the diffuse social, tribal, political, economic, criminal and religious structures and relations which are the fundamental but often obscure fabric of society (Minoia et al. 2014; Mallett and Pain 2017; Cheng et al. 2018). These are all important components of the ‘wider enabling environment’.

The NGO sector is critical for service delivery particularly in health but also other areas of development and delivery of public services such as agriculture and education. Public services and project interventions are undertaken through partnerships between central government and NGO implementing organisations, often with the support of international organisations such as bilateral donors, UN agencies and the World Bank, a modality which redresses the limited capacities of the public sector in terms of human and financial resources and access to remote communities (Newbrander et al. 2014; Varkey et al. 2015; Poole 2018). For a resource-poor and conflict-ridden country, the PUNGO partnership approach is a workable solution. The complexity of the public sector-(I)NGO modality is signalled in section 2.2.4 in relation to food security and nutrition policy.

2.1.1 Poverty

In the United Nations Human Development Index for 2015, Afghanistan ranked 171 out of 188, the lowest of any country outside Sub-Saharan Africa (UNDP 2015). In 2007–2008, 36% of the population in Afghanistan was counted as poor (World Bank 2015). Most economic and growth indicators show very slow progress in the last decade (United Nations 2015), and slower than any other South Asian economy during 2014 and 2015 (Special Inspector General for Afghanistan Reconstruction 2016). The most recent data showed a worsening economic environment, with increases in unemployment and a rise in the poverty rate from 36% in 2011–12 to 39% in 2013–14, and an increase in rural poverty to 44% (World Bank 2017a).

2.1.2 Nutrition

A major food crisis in the early 2000s led to massive humanitarian intervention. The national rate of child malnutrition reduced from a catastrophic national level of more than 60% stunting recorded in 2003/04 to about 40% in 2011/12. High levels of hunger and micronutrient deficiencies persist, with stunting levels still exceeding 70% in some Provinces (UNICEF 2014). Food insecurity based on calorie consumption increased from 28.2% of the population to 30.1% between the National Risk and Vulnerability Assessment survey periods 2007/08 and 2011/12 (CSO 2014). Micronutrient deficiencies are strongly implicated in malnutrition among women and adolescent girls (Flores-Martínez et al. 2016), perpetuating generational consequences. While there are multiple causes of malnutrition, under-nutrition and lack of dietary diversity are significant causes and point to the need to target micronutrient deficiencies rather than generalised food insecurity (Levitt et al. 2009).

2.1.3 Agriculture

To improve nutritional status, a multipronged approach has been called for (Johnecheck and Holland 2007). Afghanistan has huge potential to re-orientate the whole agriculture sector towards the nutrition agenda but integrated, multi-sectoral policy making faces many challenges (Levitt et al. 2010; Levitt et al. 2011). Developing (licit) agriculture has been a recent focus (Islamic Republic of Afghanistan 2016; Poole et al. 2016b; Special Inspector General for Afghanistan Reconstruction 2016). However, sustainability of production systems is under threat. The threats to food and nutrition security are due, among other things, to ‘climate change, particularly the expected increase in the frequency and intensity of weather-related hazards; environmental conditions such as land degradation, deforestation, desertification and water scarcity’ (GoIRA 2012b: 59).

2.1.4 The wider enabling environment

Since 2001, Afghanistan has received hundreds of billions of dollars in foreign aid money – over $100 billion of which has come from the US alone (Lutz and Desai 2015). While such money is in theory earmarked for development, the primary aims have been security, nation-building and post-conflict reconstruction. Creating an enabling environment in Afghanistan requires overcoming impediments not present in most other development contexts. The withdrawal of International Security Assistance Forces during 2014 led to a collapse in the trade, construction and services sectors of the economy. The resurgence of Non-State Armed Groups (NSAGs) since 2015 has increased the fragility of markets, public services and regional integration. The massive financial inflows have fed corruption. Complex interrelationships exist between Afghanistan and neighbouring Iran and Pakistan that are not only geopolitical in nature but also concern agribusiness interests (Minoia et al. 2014). New humanitarian dynamics of emigration co-exist with increasing internal displacement and the return of refugees (see, for example, OCHA (2017)).

2.2 The policy and institutional architecture

The policy environment since 2001 has been characterised by a proliferation of initiatives, strategies, revisions and acronyms. Adoption and implementation of policies has been much less fruitful. Around the time of this study, more comprehensive and integrated approaches were in the process of formulation, but have not yet been implemented (AFSANA/AFSeN, below). It is not the intention to document all related policies, but below we summarise key elements at the time of writing.

The commitment to food and nutrition security is underpinned by Articles 13 and 14 of the Afghanistan constitution. The Afghanistan National Development Strategy (ANDS) constituted Afghanistan’s first multi-sectoral poverty reduction strategy, from 2008 to 2013 (MoE 2014). ANDS was complex, with 86 expected outcomes and 276 indicators including the agriculture and rural development and public health sectors. The latter was one of the more successful elements, focusing on reproductive and child health systems, notably with the introduction of the Basic Package of Health Services. Preventative strategies for public health and nutrition were not evident. Effectively, ANDS has been superseded by the Afghanistan National Peace and Development Framework (ANPDF) under the Ministry of Finance, now running from 2017 to 2021, with a focus on agriculture, extractive industries and trade as drivers of economic growth (MoF 2017). Agriculture is perceived therein as a source of employment, business generation and exports.

2.2.1 National priority programs (NPPs)

The National Priority Programmes (NPPs) are a set of 22 specific programmes and policies, grouped into six clusters that build upon the overall strategy outlined in ANDS, and now taken forward into the ANPDF. Ministries execute related programs. Within the NPPs, the Agriculture and Rural Development cluster contains nationwide programmes on water and natural resource management, agriculture, rural access and strengthening local institutions.

Food and nutrition objectives are linked to ten of the ongoing National Priority Programmes (MoF 2017). The latest National Comprehensive Agriculture Development Priority Program 2016–2021, including attention to food and nutrition (Strategic Priority 6: Food and Nutrition Security, and Resilience Building) is one of the National Priority Programmes.

Another NPP is the Women’s Economic Empowerment Program, an inter-ministerial initiative involving the Ministry of Labor, Social Affairs, Martyrs and Disabled (MOLSAMD) and the Ministry of Women Affairs (MOWA) (GoIRA 2016b). This is conceived primarily in economic terms and includes agriculture as a sector in which women are significantly active but from which nutrition is largely absent. Women’s affairs and capacity building for health are addressed under the Human Resource Development cluster.

2.2.2 The national solidarity programme (NSP) and the citizen’s charter

The National Solidarity Programme was managed by the Ministry of Rural Rehabilitation and Development (MRRD). The mechanism of NSP projects was to engage directly with communities, thus promoting community control over development initiatives. The MRRD contracted out activities in specific provinces and districts to implementing NGO partners. These implementing organisations facilitated the establishment of democratically-elected Community Development Councils (CDCs) who submitted proposals to and received funds from the MRRD. Projects were mostly infrastructure-based but touched upon pathways from agriculture to nutrition by increasing access to markets and raising household incomes. The NSP and its linkages to the population through CDCs is considered to have functioned well (Nijat et al. 2016).

This operational modality is the foundation of the new Citizens’ Charter (GoIRA 2016a), which is a partnership between the state and local communities. It aims to improve the delivery of a ‘basic package of services’ (modelled on the Basic Package of Health Services below) including education, basic rural infrastructure and agriculture services. CDCs will operate in both urban and rural community development programmes, with powers also devolved to city mayors and provincial governors in holding line ministries to account.

2.2.3 Agriculture

The ANDS had two main agricultural components: the Comprehensive Agriculture and Rural Development (CARD) strategy and the Agricultural and Rural Development Zone (ARDZ) initiative. The ANPDF has subsumed the CARD, becoming the National Comprehensive Agriculture Development Priority Program (MAIL 2016). Five of the seven Priority Programs are oriented towards increasing productivity of agricultural subsectors, and improved resources management. The Strategic Priority 6: Food and Nutrition Security, and Resilience Building is one of the National Priority Programmes and encompasses efforts directed towards agricultural development which enhances food availability, access and dietary diversity: ‘MAIL will collaborate closely with other Afghanistan Food Security and Nutrition Agenda (AFSANA) members to coordinate efforts through its Extension Workers and Home Economists to improve feeding and food preparation practices in a systematic and sustainable manner’ (MAIL 2016: 23).

A flagship activity is the Comprehensive Agriculture and Rural Development Facility (CARD-F), comprising investments through economic development packages in production and processing for selected value chains such as apiculture, cotton, pomegranates, poultry, saffron, and in the requisite public infrastructure to make agribusiness markets work. It focuses on commercialisation among a subset of farmers, aiming to diversify sources of income, improve farmers’ capacity through the provision of training and skills, and orientate agriculture towards markets. ARDZ, however, focused on large-scale agribusinesses and aimed to release state-owned land into the hands of private firms such as agro-processing enterprises. While neither strategy intentionally focused on agriculture-nutrition linkages, the policies do touch upon a number of pathways including raising incomes, lowering food prices, increased consumption from own production and post-harvesting processing.

2.2.4 Health

Considerable progress has been made in building a functional health service since 2001 (Salama and Alwan 2016). Several drivers of positive health sector outcomes have been identified that essentially concern the management and coordination of the system (Dalil et al. 2014). Fundamental has been the decision for the Ministry of Public Health (MoPH) to contract out to international and national non-government organisations the responsibility to implement the Basic Package of Health Services (BPHS). BPHS has been the cornerstone of public health interventions and is credited with having made significant improvements to the health status of the population (Newbrander et al. 2014).

The Public Nutrition Department (PND) of MoPH is responsible for the nutrition activities of the BPHS. Nutritional gains have been marginal, and more budget and better policies for improving nutrition are required (Varkey et al. 2015). Largely because of contextual issues, delivery is high cost and uneven across the country (Akseer et al. 2016), and much more remains to be done, particularly in the realm of nutrition (Varkey et al.; Salama and Alwan 2016). Data from the latest National Nutrition Survey (UNICEF 2014) suggest that health sector policies are focused on the treatment of severe clinical malnutrition and prevention of malnutrition through supplementation and food distribution programmes. Public sector ‘therapeutic’ approaches are supported by international agencies such as UNICEF who are involved in micronutrient distribution: for example, distribution of iron and folate to adolescent girls in schools. Food-based approaches have not been prominent hitherto.

Despite the improvements, a 2015 review of the health sector and MoPH identified a lack of specific strategic activities, goals, M&E frameworks, implementation and review plans, shortages of financial and human capacity and commitment, and insufficient inter-sectoral coordination (MoPH 2016). The new National Health Strategy 2016–2020 marks a shift towards specific objectives, implementation plans and monitoring, and an intention to decentralise and work within revised operating principles. The strategy to reduce acute and chronic malnutrition is through public health interventions, sustaining a curative approach, and in particular the treatment of micronutrient deficiency diseases. The closest the document gets to food-based approaches is Strategic Result 3.1.4: ‘Increased awareness about nutrition is achieved, as is adoption of healthy food practices among the general population’ (p. 37). Consideration of ‘agriculture’ is absent from the document. On the other hand, controlling the quality of imported food is a ‘top policy priority’ (p.7).

2.2.5 Afghanistan food security and nutrition agenda (AFSANA/AFSeN)

The Afghanistan Food Security and Nutrition Agenda (AFSANA, or latterly AFSeN) will be a key policy. It was initiated in 2012, delivered for review to the Ministry of Agriculture, Irrigation and Livestock (MAIL) in July 2015 and launched in October 2017 by the Chief Executive of Afghanistan.¹ On the same occasion Afghanistan joined the Scaling Up Nutrition (SUN) movement. AFSeN addresses directly and comprehensively the issues of food availability, access, food utilisation and nutrition. Horizontal and vertical links are envisaged among the principal ministries at the central level, and with international organisations and civil society. It will link to sub-national governance through provincial, district and community organisations and local councils, or ‘shuras’.²

The complexity of the policy and institutional architecture is illustrated by the number of government and other bodies envisaged at the outset:

‘They include central government ministries and agencies, decentralized and local government institutions as well as UN organizations, donor agencies, NGOs, civil society and the private sector organizations… The main players at the central government level are MAIL, MRRD, MoPH, MoLSAMD (Ministry of Labor, Social Affairs, Martyrs and Disabled) and ANDMA (Afghanistan National Disaster Management Authority). Others with direct or facilitating roles are MoCI (Ministry of Commerce and Industry), MoE (Ministry of Education), Ministry of Energy and Water (MoEW), CSO (Central Statistics Organization), Ministry of Women Affairs (MoWA), Ministry of Finance (MoF) and Ministry of Foreign Affairs (MoFA)’ (GoIRA 2012: 26).

Implementation and monitoring will be overseen by the Office of the Chief executive, and representatives from ten ministries, donors and international organisations.

3 Research questions and methods

3.1 Agri-nutrition impact pathways

In the current unpropitious context, the fundamental question underlying the LANSA research programme was to identify how agriculture and agrifood systems can be better designed through appropriate policies, interventions and strategies to improve the nutritional status of children and women. Six potential agriculture-nutrition impact pathways are as follows (Kadiyala et al. 2014):

• 1. Agricultural production as a source of food for own consumption

• 2. Agricultural sales and employment as a source of income for household expenditures on food and non-food goods

• 3. Agricultural and food policies which influence food access and availability through relative prices and affordability of specific foods, and foods in general

• 4. Women’s roles in agriculture, household decision making and resource allocation affecting intra-household allocations of, and expenditures on, food, health and care

• 5. Women’s employment in agriculture balanced with child care and feeding responsibilities

• 6. Women’s own nutrition and health status, affecting childcare and nutrition, which may be compromised by own agricultural labour

Two additional impact pathways are possible, which have potentially sector- and economy-wide effects:

• 7. Competition and entrepreneurism in the scale and diversity of agricultural production, and changes in productivity, which affect food access and availability at market, cf. household level

• 8. Changes in the agricultural sector and related industries as a whole which have a macro-level and medium-to-long term impact on gross domestic product and productivity

3.2 Research design

The research design was adapted from work in Bangladesh, India and Pakistan (van den Bold et al. 2015), which in turn was based on Gillespie et al. (2013). Changes were made to take into account the local context and regional differences, with particular emphasis on meeting provincial stakeholders. The work was undertaken between 2015 and 2017 by SOAS University of London staff of the DFID-funded LANSA research consortium in collaboration with the local partner, the Afghanistan Research and Evaluation Unit (AREU).

3.2.1 Data collection

The research was based on the work of Gillespie et al. (2013) whose framework for ‘creation of an enabling environment for accelerated under-nutrition reduction’ has guided the whole LANSA research programme. Data were collected through semi-structured interviews using a questionnaire with three sections covering stakeholders’ perceptions of the following:

• 1. The political and institutional context, governance systems, structures and processes

• 2. Knowledge and evidence concerning nutrition and agriculture policies

• 3. Stakeholder and policymaking capacity and resources

This framework has also been used to shape the results section and guide the analysis and conclusions.

Interviews were conducted between May and September 2015 in the capital, Kabul, and in the four provincial capitals of Badakhshan (Faizabad), Bamyan (Bamyan City), Kandahar (Kandahar City) and Nangarhar (Jalalabad). These provinces were selected to achieve a high degree of regional representation in terms of agriculture, remoteness and ethnicity; also, in these regions access could be facilitated by the Afghanistan Research and Evaluation Unit; and physical insecurity could be minimised (Fig. 2).

Fig. 2

Afghanistan showing interview locations. Source: author

Initial respondents were purposively-selected staff from public sector organisations, (I)NGOs ((international) non-governmental organisations), and universities who were willing to participate, and then approached through a ‘snowballing’ technique. Table 1 lists the respondents’ organisational affiliations, with females indicated by an asterisk. A total of 46 interviews were conducted.

Table 1 Stakeholder respondents: organizational type and geographical origin

Individual anonymity was assured, and therefore identifiable positions have not been reported. Included among the central government respondents was one Deputy Minister. Others were at Director-level and below. Provincial government staff interviewed included one Deputy Governor, Department Directors, programme managers and Provincial Council members. United Nations (UN) staff interviewed included programme and regional managers. NGO staff interviewed included programme and regional managers, heads of section and technical specialists. University staff were two lecturers and one university Dean. Some interviewees had a double affiliation, for example, university and NGO. Their primary affiliation has been used in reporting.

Six experienced national enumerators (five male, one female) were trained in the overall objectives of the research programme and the significant regional agriculture and nutrition issues and linkages. Enumerators tested the questionnaire among stakeholders in Kabul and appropriate changes were made to clarify questions. The interviews were conducted from April to September 2015, beginning in Kabul.

Interviews were conducted in the offices of the respondents in Dari and Pashto languages. Interviews lasted from 1 to 3 h, occasionally longer and over a two-day period. Sometimes, interviewees were accompanied by a junior colleague (in one case by a female colleague, not included in Table 1). Occasionally a senior public sector official handed over the interview to another colleague with more practical experience of agriculture and nutrition. Such interviews have been recorded as a single event with the first interviewee as the respondent. In one NGO interview, four staff were present, and views recorded acknowledged some differences of opinion.

3.2.2 Data analysis and validation

Interviews were recorded, and translated and transcribed into English. All transcripts were scrutinised and key concepts identified and related to the questionnaire objectives and questions. Key topics were coded using QSR NVivo and analysed using queries, text searches and word frequencies. The node structure was refined and data were re-coded during the analysis as new themes emerged. The report records the principal themes emerging from the stakeholder interviews.

Additional data from public sources and from the preliminary interviews in Kabul, Jalalabad, Nangarhar and in Kandahar have been clearly signalled in footnotes.

For the purposes of data validation, a knowledge-sharing event was held in Kabul in April 2016. This was a half-day dialogue with stakeholders including some of the central and provincial respondents. Comments from participants at this event have been used in this report and signalled as such in the text. Sources of quotations are indicated by the type of organisation and the Province, abbreviated as in Table 1. This manuscript draws on work previously published as a working paper (Poole et al. 2016b).

4 Results

We present the outcomes of interviews in terms of the three domains identified earlier which informed the data collection (Gillespie et al. 2013):

• framing, generation and communication of knowledge and evidence

• the political economy of stakeholders, ideas and interests

• individual, organisational and systemic capacity and resources

4.1 Framing, generation and communication of knowledge and evidence: policy and underlying narratives

This section concerns the approaches, issues and priorities for agriculture and nutrition that underlie policies, and how knowledge is collected and deployed, and the communication mechanisms. An account of the policies, political economy of stakeholders and policy making procedures follows in section 4.2, with questions of resources and capacity covered in section 4.3.

4.1.1 Limited perceptions of agri-nutrition impact pathways

Of the pathways through which agriculture can impact nutrition, the importance of food production, incomes from agricultural sales, and employment in agriculture were widely recognised by stakeholders (pathways 1–3 in section 3.1 above). There was negligible reference made to the multiple agro-economic and household roles of women set out in pathways 4–6: there was little awareness of the cross-cutting roles of women within the household, in agriculture, nutrition and economic development, and linkages between ministries on gender were not flagged in discussions. The exception was a respondent from MoWA who expressed strong aspirations about collaboration with other organisations:

‘We are the one Ministry to focus on women’s situation in Afghanistan and we should add gender issues in all policies, plans and strategies.’ CENTRAL GOVERNMENT, KBL.

Pathways 7–8 concerning meso and macro business and economic dimensions were implicit in some discussions but were not clearly articulated. Nevertheless, issues to do with food safety, agricultural trade, imports and exports and international agreements were raised in some interviews.

Most interviewees commented favourably on policies that promote nutrition through increasing the volume of food production and boosting agricultural incomes. Agriculture as a source of food security was discussed in terms of quantities and caloric content, but the micronutrient content of food and the importance of dietary diversity were less considered. Although respondents suggested that sectoral policies addressing agriculture and nutrition linkages may exist, few respondents could name specific policies. A lack of cross-sectoral policies was widely attributed to the absence of coordination between sectors at the national and provincial levels. All respondents agreed on the need for more collaboration:

‘I have worked in MoPH [Ministry of Public Health] for about 11 years; I haven’t seen anyone from the agriculture sector participate in the policy making workshops or meetings in here. To be honest, we don’t know about their policies and they don’t know about ours.’ NGO, KDH.

The concepts of food security and nutrition were prevalent in the language of NGOs. However, a comment made at the stakeholder consultation in Kabul in April 2016 implied the lack of a precise understanding of the actual challenges for food security and nutrition policies. With exceptions, the appearance was of a food security ‘bandwagon’ associated with securing external finance rather than a substantive and well-articulated commitment to improving nutrition:

‘Every NGO claims that they are working on food security.’ NGO, BMN.

‘In Afghanistan mostly the stakeholders and NGOs are working on food security but diversity of food is really much more important… For example, some people are producing eggs and while selling eggs they buy cake instead, and they have no knowledge of the nutrition.’ NGO, validation event, KBL.

4.1.2 Agricultural sector priorities: a strong production orientation

Rather than nutrition-sensitivity, approaches within the agricultural sector were perceived to be ‘productivist’, and focused upon three objectives: increasing agricultural production; providing jobs and economic growth; and addressing the need to reduce dependence on imported foods. An important consideration was to develop patterns of rural development that would reduce dependence on narcotics. Agriculture as a licit source of income and employment was the pathway most frequently discussed by informants.

Overall, there was a sound understanding of the importance of agriculture for the national economy as a whole, and for poverty reduction in particular.

‘Here in Kandahar our objective is to increase production so that people can easily get access to food products; when they have good access to food products, the problem of malnutrition will decrease.’ PROVINCIAL GOVERNMENT, KDH.

‘Our agriculture department… mainly focuses on increasing agricultural production. But the possible effects that increasing agricultural production have on improving nutrition… are given less consideration.’ NGO, BDK.

‘What they focus on in the agriculture sector is creating jobs for people and getting income from that. These issues have an indirect role for nutrition. When people cultivate their lands they just want to get income out of it.’ NGO, KDH.

4.1.3 Agricultural trade and food quality concerns

There was widespread unease about Afghanistan’s reliance on food imports which were alleged to be unfit for consumption, and the lack of control of imports by central government. Food safety and agro-industrial policies were said to be weak, and border controls deficient. Interviewees noted the negative implications for health and safety, as well as problems for developing the domestic agribusiness sector. (I)NGO and government officials alike commented:

‘Now all the products that are imported from other countries are all labelled ‘JUST FOR EXPORT’ and it shows that our neighbours are just sending poison for us to eat… we are dependent on our neighbour’. CENTRAL GOVERNMENT, KBL.

‘The chickens that are imported, they are expired to the point where even their bones have changed to the colour black.’ PROVINCIAL GOVERNMENT, BMN.

‘There are expired goods that came from Pakistan and Iran.’ NGO, BDK.

‘There is huge administrative corruption and we are really having lots of problems… The blame goes to Kabul, to the norm and standards directorate of MoPH because they should consider all the products but they don’t.’ CENTRAL GOVERNMENT, KBL.

4.1.4 Therapeutic approaches to treatment of malnutrition

Many interviewees acknowledged that MoPH is primarily concerned with clinical diagnosis, treatment and cure of malnutrition. This therapeutic approach was also evident within the departments of nutrition at the provincial level. Health sector NGOs focused on hunger, food security and the treatment of severe under-nutrition. One health professional made a distinction between nutrition and health, the former not being the responsibility of MoPH, even though the Public Nutrition Department (PND) is embedded within MoPH.

Interviewees commented on the delivery of treatments through existing state (although often NGO-run) infrastructure such as clinics and hospitals, with lower-severity malnutrition cases dealt with via in-home supplementary feeding programmes. This was the BPHS. Agriculture and food-based approaches were not widely acknowledged. No significant reference was made about the continuity of treatment or support after ‘curative treatment’ of acute malnutrition and short-term supplementary feeding interventions. There was no commitment to ongoing policies and support of households to prevent chronic malnutrition.

A series of significant quotations signals the gap between agriculture and nutrition, and nutrition and health:

‘All the NGOs which work for development haven’t evaluated nutrition based on health and no-one has focused on the quality of food yet. They are just working on increasing food quantity.’ NGO, KBL.

‘Through WFP [UN World Food Programme], we have a programme at district levels that distributes enriched flour and enriched oil to the families affected with malnutrition. They also distribute ready-made foods which are useful for the treatment of children who are affected by malnutrition. Those programmes were useful to cure malnutrition.’ NGO, BDK.

‘Nutrition is not our main purpose here; our main focus is on health. It is the responsibility of DoPH [Department of Public Health] to cure all malnutrition-affected patients. We admit those patients who have intensive malnutrition in hospitals and we cure them. PROVINCIAL GOVERNMENT, BDK.

‘According to the measurement of malnutrition that the children have, the doctors will decide which medicine they should give to the patient and for how long, which is for three months, six months or a one year period. They get the medicine from UNICEF [United Nations Children’s Fund] through a proposal, then UNICEF makes a contract with the other companies and distributes these products for them.’ UN AGENCY, BDK.

4.1.5 Humanitarian interventions

The therapeutic approach was evidenced by interviewees who reiterated the importance of food supplementation for at-risk groups. A lack of evidence of longer term food-based programmes strongly suggested that emergency and humanitarian interventions for nutrition were delinked from agriculture and longer term development interventions.

‘We deliver the food to people while WFP helps us by providing food. In this programme, we check and measure the arms of the women. We cover those pregnant women whose arms are less than 23 cm and lactating women who have children older than six months. In this project we deliver them complete healthy food; wheat, flour, salt, oil, and vitamin A.’ NGO, BDK.

‘The food products we distribute to them are standard and approved by both WFP and MoPH. Most of these products are milk and other foods, which are useful for malnutrition problems, and we cure the malnutrition patients by these food products.’ PROVINCIAL GOVERNMENT, BDK.

This therapeutic approach was criticised by a rare voice from an NGO for its short-termism and the common subsequent withdrawal of emergency services, without responding to the long-term needs:

‘WFP have done lots of work in the nutrition section in here… These sorts of works are not very helpful because their projects are short-term and after sometime they stop their aid. Instead of this type of help, they should implement more useful projects. We have lots of arable lands, water, and human resources. Instead of distributing food products they should improve the agriculture sector here and teach people how to produce food products; that type of project will be more effective.’ NGO, KDH.

4.1.6 Research methods and management

Responses concerning the role of data and evidence in contributing to policy formulation highlighted a contrast between the theoretical evidence-based policy making process and the reality expressed by most respondents throughout the interviews. On the one hand, a respondent from the Ministry of Women’s Affairs (MoWA) expressed confidence in the system, and supporting views were also found amongst NGOs. One DAIL (Department of Agriculture, Irrigation and Livestock) official explained the principles:

‘MAIL formulates all the policies based on the four principal programmes. They start formulating their policies from villages and districts. First they conduct several surveys and collect information about the problems and needs of the people. Based on the survey results, they make a proposal, they sort and prioritize the problems and then they make a complete plan or policy. Then they send the policy to the Ministry of Finance to get the budget for its implementation.’ PROVINCIAL GOVERNMENT, KDH.

However, research methods were reported to be weak, and data were either inaccurate or missing for many areas. Security concerns were cited as the main reason for lack of understanding of local realities. In fact, lack of security was said to affect every stage of the policy making process from the collection of survey data through the implementation of projects to the monitoring and evaluation of projects. Apart from national surveys — seen as infrequent, incomplete and unreliable — much of the data collection is done by NGOs and implementing partners when seeking contracts and tenders or evaluating programmes. As a result, there are concerns over conflicts of interest.

‘Sometimes the data are not accurate. They exaggerate and they do not tell the truth…They show weak data because they think that if they do so, they will have more funds.’ NGO, BAMYAN.

‘They try their best but there are some problems, maybe security problems, or the people who work with them can’t get the exact information. They can't go to some places because of the security problem and the people they have hired don’t give the right information to the policy makers.’PROVINCIAL GOVERNMENT, KDH.

Several times questions were raised about both the low quality of data collected when research is conducted, and the lack of oversight of the research process, and fabrication of data.³ The use of data was also an issue of contention, with many respondents frustrated that data collected do not get used to formulate policy. Stimulating the academic sector, especially technical agricultural research institutes, was seen as important. Some organisations were engaged in agricultural research but respondents from universities felt their research was not incorporated into the policy making process. The implications of weak research for policy making were expressed succinctly:

‘The main problem is that data are not accurate, surveyors don’t collect accurate data, and the reason for that is obvious, there are security issues and a lack of supervisory attention to control their surveyor and whether they actually go to the field or not.’ UNIVERSITY, KDH.

‘Why are their policies not effective? Because they don’t conduct any research and surveys and they don’t make policies based on research... In this country we don’t have accurate data.’ NGO, KDH.

4.1.7 Knowledge management

Knowledge management and communication were considered to be weak. Modern information and communications technology is used for internal management within many implementing organisations. Communication systems between organisations were said to consist of an ad hoc array of media: use of telephones prevails, especially cellphones; email contact was said to be common among NGOs but less so with government organisations. Face-to-face meetings in Kabul are rendered problematic by the insecurity in the city. One respondent emphasised that official business with government often adhered to use of hard copy and ordinary mail. An official of a UN organisation summarised the patterns of inter-organisational communication, with emphasis on how government entities depend on face-to-face meetings and letters. While use of ICT appears to be spreading within public sector organizations, there was no mention of web-based communication or file-sharing applications within and between organisations:

‘Mostly we communicate through emails and invitation letters; for example, when there is a monthly meeting in the health sector, the health department itself sends us an email, we participate in the meeting. After termination of the meeting, they send us the material of the meeting through the internet (email). NGOs and UN organisations usually communicate through software, especially email, and rarely through the telephone. Governmental organisations mostly communicate through documents, like their invitation letters are hard copies; they use hardware and most of their documents are hard; still there are pen, paper and documents in the government system.’ UN AGENCY, BDK.

Use of mass media to communicate with the public was also limited, particularly in rural areas where national television and radio offer limited coverage:

‘Actually 95 per cent of our people have TVs in the city while only 5 per cent of our people watch TV in the villages. Public awareness programmes via TV channels won’t help our rural people… Newspapers have gone from our system because most of our people are uneducated.’ UNIVERSITY, NGH.

4.2 Political economy of stakeholders, ideas, and interests: policy processes

This section concerns the perceived origins and contextualisation of policies and the horizontal and vertical processes through which policies are mediated.

4.2.1 Policy formulation in theory

The governance structure in Afghanistan ideally exploits the technical capabilities of central ministries and vertical consultation with provincial-level governments. In theory, data collected at the district level are aggregated at the provincial level and then fed into the policy making process at the central level. Policy is then formulated in Kabul by Government Ministries and cross-sectoral working groups and clusters — with input from Provincial Department heads. Policies are then disseminated in the form of plans and strategies to the provinces. Respondents cited few examples where this policy making process existed in reality.

4.2.2 Donor-driven policies

International donors were said to influence policy significantly because the Government of Afghanistan lacks technical capacity and financial independence. Over-reliance on foreign donors was a source of contention in many interviews. NGO staff and Provincial officials alike responded frequently:

‘The donors provide the funds; through their funding power they make policy makers accept their wishes and desires.’ NGO, BMN.

‘Generally, all the work done by MoPH is from foreign aid as the Government of Afghanistan doesn’t have its own budget.’ PROVINCIAL GOVERNMENT, BDK.

‘A foreigner comes in the country but is only allowed to stay here for 3-6 months. If he/she gets used to the environment at the first 3 months, he/she will end up wasting a lot of time. He/she then will leave and another person will replace him/her, someone who cannot take up his/her work. We have seen it many times.’ NGO, BMN.

4.2.3 Decontextualisation of policies

The majority of respondents considered that policy making was de facto significantly over-centralised. Lack of capacity was cited as a cause of the imperfections in policy making processes. Comments reflecting frustration with the lack of consultation were expressed, sometimes in strong negative terms:

‘… most of the time policies are made in Kabul behind the guarded walls.’ UNIVERSITY, KDH.

‘They should collect data from villages, districts and provinces and make a policy according to that. But the problem in here is the lack of capacity. If you go to a Ministry in Afghanistan they don’t even know how to write; then how do you expect them to make a policy?’ NGO, KDH.

‘Once the policy is made in Kabul, nobody has the right to either object or suggest anything. For example, if they say that the colour of milk is black and you write and say no it’s white, they will grab you and tell you that you are not obeying laws.’ PROVINCIAL GOVERNMENT, BMN.

Given the weak domestic knowledge environment, dependence on foreign advisors and donor organisations (discussed below) has led to formulation of national policies adopted or adapted from other countries that were inappropriate for the Afghan context and unsustainable.

‘The sources for all the health policies come from WHO [UN World Health Organization]... They conduct research all over the world, they bring the policies from other countries such as African countries to Afghanistan and implement them here… they make a guideline and give it to MoPH and they make their policies according to that. WHO makes different policies for all developed and underdeveloped countries.’ NGO, BDK.

‘… in Bamyan, we only have one growing season but in Jalalabad we have three. If the policy is made in Kabul and not contextualised well, then it will not work.’ NGO, BMN.

There was a widespread perception that central policy makers were remote and had had little understanding of the heterogeneous ecological and social contexts in provinces and districts. The remoteness of policy making from the provinces not only caused a lack of effectiveness but also mistrust among interviewees:

‘In every village we have different shuras. They should ask the people’s need from the shuras, and on that basis they should build their policies.’ CENTRAL GOVERNMENT, BMN.

‘Those who make policies, they come in bulletproof cars and they make policies behind the guarded walls. They don’t know about the situation of Helmand and about the agriculture and gardens there. Most policymakers come from abroad to the airport and from there, they go to their offices. How do you expect them to know about the situation of Afghanistan? I haven’t seen any policymakers go to the field to find out the problems of the people.’ NGO, KDH.

4.2.4 Engagement with the private sector

A range of private sector development activities related to both health and agriculture was reported. CARD-F (section 2.2.3) is the multi-institutional public sector-donor programme to develop commercial agriculture, comprising investment in commercial value chains and agribusiness infrastructure. Other donors have also invested in production enterprises and food storage firms. Another example is the United States Agency for International Development’s Kandahar Food Zone Program which encourages farmers to develop licit forms of agriculture, cf. opium poppy cultivation. Likewise, a respondent reported how MAIL, in collaboration with the UN Food and Agriculture Organization, provided support for the establishment of agricultural input firms producing and supplying improved wheat seeds in Badakhshan.

Private enterprises within the food sector such as salt manufacturers and grain millers have focused on micronutrients, especially iodine fortification of salt and mineral fortification of flour, in collaboration with government and international organizations. There was appreciation of agribusiness when it was reported:

‘The private sector has done lots of important work here. For example, they produce several vegetables like cucumber, tomato, etc., they built many greenhouses in here. Several dairy farms, aviculture farms are built by the private sector here.’ PROVINCIAL GOVERNMENT, KDH.

‘Our department had a project with UNICEF, and their aim was to include iodine into salt… We want all our projects to be like this.’ PROVINCIAL GOVERNMENT, BMN.

While there was government involvement in programmes aimed at stimulating private sector business to address economic development and public health, most projects were either independent or established through direct collaboration and support of NGOs. There was a widespread view that coherence across sectors required more government involvement and communication with the private sector. Generally, respondents were highly supportive of private sector involvement, but were aware of the need for further investment to reach into remote districts and achieve scale and effectiveness in implementation:

‘The majority of problems we face right now are because of the lack of a private sector… If we can improve our private sector it will have a positive impact on the improvement of food security and nutrition.’ NGO, NGH.

‘Unfortunately, we rarely have such dialogues [between government and the private sector]… I don’t think that the government has such meetings with these sorts of businesses to financially help these networks.’ NGO, BMN.

4.2.5 Civil society

Discussions about the role of civil society in promoting agriculture and nutrition were not fruitful. Respondents commented that civil society organisations are primarily involved in politics. In this context, respondents were not referring to village-level CDCs and Provincial Development Committees (PDCs) which are constituent parts of the sub-national governance architecture, but to national organisations. Ministry officials from Kabul also commented with some cynicism:

‘Civil society raises concerns as to why police die or why ISIS is here, and why our road is not constructed. In nutrition I do not think they have taken any action.’ UN AGENCY, BDK.

‘The civil society prioritises politics, not food security. They constantly talk about politics, they never talk about food security.’ CENTRAL GOVERNMENT, KBL.

4.3 Policy coherence, delivery, capacity and resources

This section focuses on policy coherence and coordination across sectors and between similar organisations vertically and horizontally. It also considers the ‘reach’ of policies and programmes, and the weaknesses in delivery.

4.3.1 Inter-sectoral coordination

There was significant evidence of horizontal coordination at the central government level in Kabul. The UN cluster system was frequently cited as an important coordination mechanism. Policy coordination among sectors is largely considered to occur among Ministries and within the sector clusters.

‘Within MoPH we have a Public Nutrition Department that makes a joint nutrition policy for all over Afghanistan. Organisations like UNDP [United Nations Development Programme], UNICEF and other organisations also review those policies. After every few years they make a guideline book and review those policies. The Minister of Public Health passes that policy and then they send those guidelines to all Public Nutrition Departments in the provinces.’ PROVINCIAL GOVERNMENT, NGH.

‘We have a cluster by the name of FSAC [Food Security & Agriculture Cluster], within this cluster are FAO and all the organisations that work in the agriculture sector; they indirectly work in the nutrition sector and jointly we work in this sector. In every project we implement, the MAIL, MRRD and other organisations cooperate with us...’ NGO, NGH.

‘We have the WASH [Water, Sanitation and Hygiene] cluster, nutrition cluster, protection cluster, and we have separate policies for these… the government is very weak and they don’t do a lot of work; it’s NGOs that do everything but still their programmes are in accordance with the government policies.’ NGO, NGH.

Respondents were asked about the linkages with sectors other than agriculture and health in which investment is necessary to enable development in agriculture and nutrition. Lack of investment in physical infrastructure was widely seen as a major challenge, with the lack of domestic energy infrastructure as a root cause of underdevelopment:

‘We don’t have any factory in Bamyan... and the reason is two things. First, people here are mostly poor. And secondly, we don’t have electricity in Bamyan.’ PROVINCIAL GOVERNMENT, BMN.

The Ministry of Rural Rehabilitation and Development (MRRD) were reported to be involved in infrastructure projects such as road building and irrigation, and their work was generally received positively. However, there was a common perception among respondents that projects were too few to meet the significant challenges, and that new technologies were not sufficiently utilised, particularly for energy. Roads and irrigation projects were seen as the most common investments in infrastructure. The need for greater investment in infrastructure supporting food value chains was recognised:

‘We particularly as MRRD have prioritised irrigation infrastructure as an area for us in the years to come in order to support the agriculture sector in a more sustainable manner that increases productivity... we are looking into the physical infrastructure in enhancing the value chain.’ CENTRAL GOVERNMENT, KBL.

4.3.2 Vertical coordination

Vertical coordination was much less apparent than horizontal coordination. As noted above, perceptions of the origins of policies shared the predominant view in all interviews that policies are made in Kabul, usually with the assistance of, or by, international donor organisations, sometimes copied from other countries, without knowledge or input from the provinces and districts for which they are designed. Vertical consultation on policies between the centre and the provinces was not the norm, although greater provincial participation was reported in project planning and implementation. Attitudes towards this centralisation were mixed. Some respondents from provincial governments and NGOs were satisfied with the level of engagement between provincial and central levels:

‘At the MoPH level, when they make a policy, they formulate it according to the national development policy of Afghanistan. For its implementation, they get the views of provincial officials, NGOs and other organisations that work in this sector.’ PROVINCIAL GOVERNMENT, KDH.

‘There was a guideline about malnutrition. They organised a 2-week workshop in Kabul regarding that. We participated in that workshop and now we are working to make a plan for its implementation.’ NGO, BDK.

Other provincial respondents differentiated between centralised policy making, and apparently more consultative processes for project planning and implementation:

‘People say drop by drop a river is created, but our mentality is that the entire river rains down on us. When we make policies, we should start from the smallest points and go to medium and then gather it altogether but that is not how they do it…They send plans for us to follow.’ PROVINCIAL GOVERNMENT, BMN.

‘They don’t invite people from the provinces to give their views in the policymaking process. But for implementation of these policies, they do ask … for our views and opinions.’ PROVINCIAL GOVERNMENT, KDH.

4.3.3 Horizontal coordination and ‘silos’ in planning, implementation and monitoring

There was evidence of horizontal coordination through meetings at the provincial level, and from the provinces upwards to the centre. NGO activities are linked to central government ministries and NGOs are granted permissions at the provincial level. Provincial governments therefore, in some circumstances, play the role of a coordinating body, subcontracting official Ministry policies and projects and preventing project duplication and conflict among NGOs. There was little evidence of integration of objectives and strategies across sectors, including among provincial departments, and little evidence of accountability by implementing organisations in practice:

‘The government plays a controller and moderator role here in Kandahar Province. All the projects are implemented by the national and international organisations, they are the ones who use resources and they do whatever they wish.’ PROVINCIAL GOVERNMENT, KDH.

‘Our main problem is that we don’t have coordination among DAIL and DoPH... There is a national nutrition committee at ministry level; MAIL, MoPH, MRRD, WFP, and UNICEF are in that committee… But at the provincial level we don’t have such committees or coordination meetings.’ PROVINCIAL GOVERNMENT, BMN.

Some large projects originate with donors and are centrally managed, characterised by good national level governance. CARD-F, (see section 2.2.5) is an agricultural project financed by DFID and DANIDA (Danish International Development Agency). It is an example of a sectoral programme, working in 14 provinces which is integrated horizontally into the activity of four Ministries (MAIL, MRRD, MCN and MoF), who manage the implementing agencies in an effective and coordinated structure.

However, most respondents reported that the implementation of plans and projects occurs in both sectoral and management ‘silos’. Projects often originate with donor agencies and bilateral organisations and are implemented by large international NGOs and consultancy firms which are ‘sectorised’, like CARD-F, and usually not locally accountable. Respondents reported that international organisations notify — but generally do not collaborate with — provincial departments. Rather, such organisations and agencies report to their head offices in Kabul and operate according to projects and policies designed in collaboration with central government Ministries.

A wide variety of NGOs had smaller agricultural projects that aimed to leverage agriculture for food security and nutrition. Agricultural policies and projects mainly focused around one of three activities: distribution of improved seeds and seed varieties; value chains and improving access to markets; and creating home gardens, vegetable patches and greenhouses. One NGO worker in Nangarhar stated:

‘We consider nutrition in most of our agriculture projects; food security is also included in it... For example, we have made vegetable demo-plots to increase cultivation of vegetables... Besides that, we build fishery farms and poultry farms… Also we have distributed 220 goats for poor and malnutrition-affected women. Besides these, we have distributed improved vegetable and corn seeds to poor farmers.’ NGO, NGH.

Thus, it appears that there is a form of parallel governance through which international NGOs and development agencies operate their own centralised policy making process. The effect is to sustain implementation and accountability ‘silos’: one where the implementing organisation takes policy direction from central offices in Kabul; and another where government Ministries issue strategic plans to provincial departments, implemented through smaller NGOs. There was said to be little interaction between the two. Government officials commented:

‘You know, either policy gets developed only on the government side, and non-government actors are not involved, or very often this happens to be on the non-government side and the government is not involved.’ CENTRAL GOVERNMENT, KBL.

‘They sometimes talk about parallel governance and in this case it is like that. The organisations (NGOs) have their own policies.’ PROVINCIAL GOVERNMENT, BDK.

4.3.4 Policy ‘reach’: sub-national governance

Some aspects of subnational governance work. However, effective ‘reach’ into the provinces is limited by many constraints. Despite the heavy top-down nature of policy making in Afghanistan, and excepting those projects implemented from Kabul through NGO ‘parallel governance’ channels, projects are often proposed at the provincial level, sent up to Kabul for approval, and then implemented at the local level within a complex bureaucratic system. As one DAIL official explained:

‘…now, plans are made on the provincial level with the coordination of the DOF, Governor’s office, and of the experts. The project is reviewed a few times in the PDC [Provincial Development Committee] to check whether this has already been done and that no other stakeholders should be working on the similar thing in the same area. Once finalised, the PDC signs it and sends it to the Governor but then again, it goes through the relevant Departments and afterwards to the relevant Ministries.’ PROVINCIAL GOVERNMENT, BMN.

Within this system the role of the Provincial Governor was said to be crucial. While each department has vertical line responsibility to its central Ministry, the Governor’s office was said to be the essential vertical link between diverse provincial level activities, and between the province and central government activities. The office is the source of legitimacy for the Provincial Development Committees [PDCs], and is the mechanism by which PDC decisions can be implemented.

NGOs frequently coordinate horizontally at the provincial level through Sector Working Groups (SWGs). SWGs are also where NGOs collaborate with the respective government Departments, who then liaise with Kabul via their line Ministry. Within-sector coordination also appears to be strong with government departments coordinating with the main NGO actors within their sector.

The PDCs are the most significant hubs at which communication and coordination between provincial-level sectors and development actors occur. Often meeting monthly, these Committees involve stakeholders in the development sector from NGOs, government Departments and donor organisations, and are generally multi-sectoral. Crucially for integrated agriculture and nutrition projects, this is the only regular and official point of contact between these two sectors at the provincial level. All respondents who attended PDC meetings held them in high regard as a communication and coordination hub, especially compared to the situation before PDCs operated. PDCs are forums to coordinate but also tend to focus on the biggest projects operating in the region such as large-scale development and construction projects.

‘We [the PDC] are directly in contact with Governor’s office. The Provincial [Governor’s] office has their own different sections like political, social, and executive sections. If the people face problems in any of these sections, we directly contact the provincial office and ask them to help solve our problems.’ PROVINCIAL GOVERNMENT, NGH.

4.3.5 Policy ‘reach’: sub-provincial coordination

Communication and cooperation activities at local and district levels operate through formal, informal and semiformal governance structures. Respondents commented on the formal structure of the National Solidarity Programme run by the Ministry of Rural Rehabilitation and Development, where elected CDCs are set up; each chairperson of such committees sits on an unelected District Development Assembly (DDA). CDCs were seen as powerful instruments as a result of their ability to direct development work, writing proposals and receiving funds for implementation. However, for this they are dependent on acceptance and approval of their proposals based upon policies developed centrally. As instruments of the National Solidarity Programme managed by MRRD and DRRDs (provincial Departments of Rural Rehabilitation and Development), CDCs also work with NGOs locally to propose projects for their local area. Proposals are written by CDCs with the assistance of the NGOs and then approved at the provincial level by government department members. NGOs appear to link closely to organizations at local and district levels. Nevertheless, limits were reported to constrain community involvement in local activities and decision making because of lack of information, and because initiatives seemed to trickle down from higher levels of government. This system of sub-national governance was re-envisioned by the World Bank in 2016 and is fundamental to the new Citizen’s Charter.

In addition to the CDCs, there are informal councils, the shuras, for considering specific issues. The degree to which these informal groups are consulted or involved in projects, however, is also limited. Shuras lack the power that comes as an official state institution. Nevertheless, shuras have local significance within traditional culture and structures. They may constitute both a bridge and a barrier between traditional eldership and the political/technocratic authorities. This creates a tension between traditional authorities and a younger generation of activists:

‘Now in every district and village we have councils; we have parents’ shuras, teachers’ shuras... In every village we have different shuras.’ PROVINCIAL GOVERNMENT, BMN.

‘If the policy makers want to work with people in the current time, there are two kinds of people in Afghanistan; the first kind is the elders whose views can be accepted, even while they don’t have knowledge, and the second kind is the educated people.’ UNIVERSITY, BMN.

4.3.6 Short-termism

Criticisms of policies and projects expressed by interviewees can be grouped into two broad areas related to the externally-driven agenda within a short-term framework, and the lack of accountability to the provinces and to the wider society.

Short-term projects were seen as a major problem because of lack of sustainability, linked to the over-dependence on ill-conceived central planning. Donor organisations were considered to be responsible for perpetuating a culture of short-termism. Insecurity leading to rapid rotation or replacement of staff accentuates this problem. To combat this, a shift from projects to longer-term programmes was seen as desirable, as was reducing dependence on funding from foreign donors. The following comment reflects the views of a number of respondents from NGO and government sectors:

‘These programmes should be long-term. The peoples’ capacity should be built so that good changes will come in terms of their nutrition and life. It would be better if NGOs work for 2-5 years because just like in business, the first year is an advantage while the next is loss… If they work for a long time, agriculture and livelihoods would be improved.’ NGO, BDK.

Given the scale of Afghanistan’s problems it is perhaps not surprising to find an emphasis on humanitarian relief programmes and policies. Within the nutrition sector this means distribution of emergency food rations and clinical treatment of severe malnutrition (section 4.1.5). While it is beyond question that such initiatives are highly important, there was evidence of a need to move towards a more strategic approach based upon long-term, sustained investment in the agricultural and nutrition sectors to complement the focus on emergency response:

‘For me, since it is a country coming from emergency, the emergency network has been really influential… We are thinking about slowly building capacity to move out from emergency to more protected and developed nutrition programmes.’ UN AGENCY, KBL.

4.3.7 Lack of accountability

Lack of accountability was linked also to a pervasive culture of corruption which not only siphoned money away from where it is needed but also distorted the programme priorities and the targeting of resources. Allegations of money wasted by organisations and by the government was also a frequent source of dissatisfaction. Criticisms from two provinces identified high operational costs within NGOs and aid agencies, as well as misguided, ineffective and symbolic projects without practical value.

‘NGOs have built many places to keep raisins and they spent hundreds of thousands of dollars on them, which is not very useful, and no one really uses them. The organisations and NGOs should spend the budgets on more useful projects.’ PROVINCIAL GOVERNMENT, NGH.

‘The things that have been done here are from the NGOs and not the government. The greenhouses that have been built by some of the NGOs are only symbolic and nothing else.’ PROVINCIAL GOVERNMENT, KDH.

The most frequent criticism of project implementation was the lack of monitoring and evaluation. The formal systems that were said to operate were explained by a member of DAIL:

‘Provincial Councils, District Councils, and civil society organisations always monitor the implementation of projects. Besides that, in the structure of MAIL there are departments of planning and policies; the Director of Agriculture along with head of planning and policy departments always monitor the implementation of projects.’ PROVINCIAL GOVERNMENT, KDH.

However, monitoring and evaluation were widely seen as ineffective. While major NGOs and international agencies were often seen to be accountable, the local project implementing partners were not viewed as accountable. Insecurity limits accountability: activities were rarely checked by people on the ground, and CDCs and shuras were said not to have an official mechanism of monitoring and evaluating programmes:

‘There is an accountability mechanism that exists but it is more on paper, not actual.’ PROVINCIAL GOVERNMENT, KDH.

‘Sometimes you have completed the project, which didn’t work, but it benefited the people implementing it and they want to repeat it because they will benefit from it. So you do something even though you are actually corrupting the whole system and you know it is not benefiting the people.’ UN AGENCY, KBL.

4.3.8 Limited technical capacity

We have noted the heavy dependence for the delivery of public services on non-state organisations. Nevertheless, there was evidence that some public services are offered through government departments at the provincial level, such as monitoring of food supplementation (salt). Also, respondents noted public agricultural services such as extension but these were mostly concerned the distribution of seeds and fertilizers. Programmes were typically scattered, irregular and were said not to reach remote regions.

‘Both DAIL and other organisations teach farmers how to use fertilizers for improvement of their land… These programmes are not regular, to allow farmers to practise these things and convey this information to all farmers.’ NGO, BMN.

Generally, government departments were seen by different stakeholders to be lacking capacity to handle projects and contribute to policy making. Insecurity is a significant factor in discouraging skilled staff from assuming front-line positions. Incompetence within government Ministries and Departments was also reported to be a serious problem by (I)NGO officials. Even where technology for knowledge management is available staff may not be trained to use it. There was a view that lack of government capacity was also due to insufficient salaries. Competent staff were recruited by NGOs or promoted to national level where they could receive higher salaries.

‘We are working in several districts but trust me, the [government] staff aren’t working at all… Some government staff are illiterate; they just sit in their chair and wait for a bribe.’ NGO, KBL.

‘For example, the Department of Education staff are still busy with pen, paper, documentation, and ruler, even though GIZ helped them with 18 computers last year and now maybe they have more than 50 computers. But they still use pen, paper and ruler.’ UN AGENCY, BDK.

‘There are a lot of professional staff in the (non-governmental) organisations because the salaries are much more than [those of] the government; everyone seeks a job in (non-governmental) organisations.’ NGO, KBL.

‘The staffing of governmental organisations is more than they need. For example an organisation which needs only 60 staff, now has about 180 staff but their work isn’t appreciable at all and their skills are very low.’ UN AGENCY, BDK.

4.3.9 Frail human capacities

Corruption is widely seen as a government problem, yet has knock-on effects for development projects: corrupt and nepotistic hiring practices were also said to be common within government departments. All respondents acknowledged that primacy should be attached to the appointment of people with technical skills, who will often be of the younger generation. Even where there were well-educated young staff, they were said to have to operate within the constraints of organisational and cultural norms. Different UN officials commented:

‘This is a big problem in the government, that they hire people based on relations and ties but not based on merit. For example a person who doesn’t know much about agriculture can be working in a very high position in the MAIL… Someone who has graduated from a law faculty is working as a director in the MAIL. People who work in the agriculture sector must be specialists in that field, and they must have good experience.’ NGO, BMN.

‘They should abolish the older staff slowly and they should bring the young generation to work, but regretfully they don’t hire the young generation.’ UN AGENCY, BDK.

Like security, physical remoteness was cited by a number of interviewees as a reason for the lack of skills at the provincial level, and ineffectiveness:

‘In Bamyan we rarely find individuals with a high level of capacity. Because Bamyan is a remote area, professional staff hardly agree to come here for work. Only people with less capacity come to work here.’ PROVINCIAL GOVERNMENT, BMN.

5 Discussion and conclusions

Emerging from this evidence are important themes concerning how agriculture can be leveraged for nutrition within Afghanistan. In a number of respects, the weaknesses identified resemble those identified by Levitt et al. (2011) during the period 2002–2007. Nevertheless, there have been changes in context and issues in the intervening decade, policies have proliferated, and new nuances have emerged in the policy agenda and in governance. This section addresses areas where changes are needed in policies and processes. Recommendations follow in section 6.

5.1 Gaps in responses

First, in relation to results, there were significant gaps in our expectations about the social and political mechanisms needed to reshape the food and nutrition environment:

• There was almost no articulation of the ways in which women’s participation in agriculture could help nutrition. The multi-sectoral linkages envisaged were largely absent in understandings of agriculture and nutrition policies. For example, there was no evidence of knowledge of MAIL’s recently formulated strategy (MAIL 2015). The only respondent who was able to explain the importance of women in agriculture was from MoWA. We should note that gender equity is now prioritised in a range of policy documents, but awareness is limited, and adoption and implementation will be considerable challenges.

• Apart from the huge role of NGOs as implementing organisations of public projects, there was no evidence that national civil society organisations were engaged in advocacy on technical agriculture and nutrition issues. While many organisations are active in development, a culture of awareness of sectoral and inter-sectoral development linkages and agriculture and nutrition was weak, and knowledge of shared experience was limited.

• No respondent cited nutrition education programmes or campaigns of advocacy and awareness that could influence both consumer behaviour and public policies. Policy and programme linkages to the education sector were equally deficient. Schooling is, of course, another challenge especially in respect of gender issues in attendance and attainment. Curriculum development that includes food-based approaches to better nutrition are essential for rural and urban situations alike in the context of the triple burden of malnutrition.

5.2 Policy framing and coherence

This section consider key elements of the enabling environment for agri-nutrition policies. The most trenchant criticisms concerned policy framing and policy-making processes. The framing of policies was said to be donor-driven, ill-designed through top-down processes, reflecting insufficient knowledge and awareness of local realities and unable to account for provincial social, economic, political and ecological heterogeneity. The knowledge base concerning the rural reality is shaky. Consequently, provincial stakeholders did not have confidence in the central policy makers. In practice they resented the remoteness of power, short-termism, overdependence on central and foreign interventions and initiatives that failed to address local needs or constraints and distorted strategies and operations.

At the time of writing, there were a number of ‘silos’ affecting policy formulation:

• agricultural development has focused on commercialising production, on employment, exports and strategies to combat illicit enterprises. Despite evidence of interventions designed to boost nutrition-sensitive agriculture such as kitchen gardens, fruit production and livestock, little attention hitherto has been given to food quality, dietary diversity and nutrition security of vulnerable populations;

• public health policies are largely curative and disarticulated from the prevention of malnutrition through agriculture and food-based approaches;

• humanitarian programmes and food supplementation policies to treat acute malnutrition due to micronutrient deficiencies were delinked from longer term strategies to achieve sustainable improvements in diets through nutrition-sensitive agricultural production and economic development pathways

In other respects, there were positive findings. The cluster system worked in Kabul, and at provincial level coordination was said to be fluid and efficient, supporting project implementation. National and international implementing organizations were a fundamental and largely successful element within the project and service delivery architecture. Local project implementation was considered to be more coherent than the policies themselves, partly due to the efficiency of implementing organizations even when they were often resourced from external sources. The office of the provincial governor and the PDCs were highlighted as potentially effective structures in local coordination.

Nevertheless, there was much evidence that knowledge sharing and communication between the hierarchy of national and subnational levels of government was poor. Research was said to be weak with data often fabricated, partly for fear of insecurity during field visits, and there was no systematic approach to using evidence in policy formulation. Information flows were fragmented within parallel and often ad hoc communication routes. There was little evidence of knowledge processes, either through communication from the communities, districts and provinces to the centre, or of precise policies from the centre to the provinces. Use of modern information management technology among government organisations apparently was insufficient. In separate discussions and through personal observations, the evidently widespread use of social media, at least in urban areas, was said to be confined to private and social purposes.

5.3 Perspectives for capacity and resources

Despite the long trajectory of the Afghanistan Food Security and Nutrition Agenda (AFSeN) and the evidence of horizontal coordination, we found a lack of integrated agriculture and nutrition policies at most levels of governance. Respondents were aware of broad policy approaches but were unable to cite objectives and strategies precisely. The approval of AFSeN offers the opportunity to integrate policies and sectors. The framing of the problems is sound, but the high level objectives still have to be translated into feasible and appropriate programmes in the provinces. In the first place, AFSeN will depend on successful management of a highly complex, multi-agency governance structure at the centre. Implementation will be constrained by limited capacities and resources. Precise strategies need to be formulated through consultations across sectors and throughout the vertical architecture of national, sub-national and sub-provincial governance. The Citizens’ Charter creates the needed opportunity and framework, but these are major challenges. Moreover, existing policies such as the National Strategy on Women in Agriculture (MAIL 2015) and emerging policies such as that concerning Zero Hunger should be integrated with AFSeN. Another requirement is to manage constructively the dependence on national and international NGOs for efficient delivery of public services.

In conclusion, major barriers remain to leveraging agriculture for nutrition. These are not confined to the specific sectors but are symptomatic of Afghanistan’s broader development challenges. Extreme dependence on external human and financial resources shapes policy and practice according to international expectations, but fails to deliver efficient and effective processes and outcomes. In particular, the lack of capacity and resources within government ministries and departments along with poor infrastructure remain major barriers to progress. Significant investment in Afghan human resources will require continued external assistance.

The apparently intractable insecurity and increasing humanitarian needs of internally displaced people, exacerbated by refugees returning from Pakistan and Iran, and entrenched corruption add further strain to the food and nutrition security system (OCHA 2018a, b). It goes without saying – almost – that the conflict must be resolved. Meanwhile, management of the humanitarian-development interface requires deliberate processes and appropriate resources in order to ensure coherence between approaches to Afghanistan’s acute and chronic problems through short-term interventions and long-term initiatives. Of these, agriculture and food and nutrition insecurity are major but not exclusive components.

Another challenge arises from the fragility of the natural environment: although interviewees only obliquely pointed to responses to anthropogenic and natural disasters, the agricultural capacity is highly dependent on climate-sensitive investment in seeds, irrigation and infrastructure, management of land and other natural resources.

6 Recommendations

6.1 Governance

Decentralisation of sectoral authority and responsibility for agriculture and nutrition to provincial levels is not a novel conception (MoPH 2016), but needs a cautious and comprehensive approach, with national political support and public commitment. Above all, it is essential for negotiation to take into account the complexities and destabilising effects of regional and ethnic differences and the presence of other non-state groups and coalitions. As noted, decentralisation is consistent with the new Citizen’s Charter employing distributed and inclusive forms of governance that embrace local and provincial structures, both formal and informal. It requires that the scale of operations strikes a balance between the availability of technical expertise and local capacity, and the ability for sectors and organisations to communicate effectively in PDCs and SWGs. Capacity at the provincial level is weak due to the recruitment of skilled individuals to work for (I)NGOs, UN agencies and central government for significantly higher salaries, and due to emigration. Increased investment is needed in departmental government capacity within the agriculture and health departments.

6.2 Context-awareness

Provincial-level governance potentially will result in policies that are context-specific, avoiding approaches developed centrally or copied inappropriately from other countries. Locating budgets and decision making at provincial level should facilitate the contextualisation of policies to local realities and enable more meaningful dialogue among local stakeholders, and across sectoral boundaries. Overall, it is necessary to formulate a ‘context-sensitive’ approach rather than a model of ‘one-size fits all’ (Nijat et al. 2016). And decentralisation processes must be accompanied by a comprehensive awareness of the security and peace challenges mentioned below.

6.3 Political ‘reach’

The prevailing public-NGO (PUNGO) partnership model for the delivery of services has merits, despite the political tensions (Newbrander et al. 2014; Poole 2017): at least, it is as sustainable as, and probably more effective than, any other form of delivery in terms of reach to remote areas under the circumstances, and fits well with decentralisation. Programme and project implementing organisations should be better linked, and accountability should be clearer, while recognising that it is they, where security permits, who have an effective presence and together constitute a ‘functional equivalent’ of the hitherto absent formal state. Tensions between the centre and the periphery, between the state and its functional equivalents, and informal political power brokers will have to be managed sensitively.

6.4 Knowledge and communication

All central and provincial stakeholders will benefit from improved information flows and knowledge management. This requires further moves towards e-governance: a publically-accessible knowledge-sharing architecture including up-to-date internet websites for public sector organisations, with appropriate links to donor and international organisations, including the UN. Increased use of provincial governance through both line ministries and the governors’ offices, and district level SWGs are two mechanisms to bridge the current poor communication flows between national and provincial levels. Team collaboration and document sharing software offer opportunities for knowledge management and sharing for better horizontal and vertical coordination.

Other potential improvements resulting from greater use of IT include enabling greater consultation in policy formulation at the central level and improved training in government policies at the provincial level. This is consistent with recommendations to strengthen the role and performance of traditional and modern civil society organisations at the lower levels of subnational governance (Nemat and Werner 2016). Even in this internet-enabled age, there is a role for national public mass media.

6.5 Regulation

Afghanistan’s membership of the World Trade Organization dates to July 2016. This should offer a platform for regularizing international trade within which domestic trade and investment policies for a competitive agribusiness sector can be developed. Regulatory frameworks will beneficially impact many current problems including but not limited to: undue reliance on foreign imports; allegations of inadequate food safety controls; lack of national micronutrient fortification of foods; and lack of agricultural incentives for farmers and food chain entrepreneurs. Successful implementation of good policies will depend, like many other interventions, on curbing corruption and conflict.

6.6 Investment

This article is hardly the place to reiterate the need for sustained support for Afghanistan from the international community which is discussed in other and higher-level policy fora. Nevertheless, it is pertinent to reiterate the need for investment in physical infrastructure for agriculture and nutrition, with particular focus on creating the facilitating conditions for successful domestic markets and enterprises (Flores-Martínez et al. 2016). Food chain organisation should also focus on creating processing and storage facilities and the essential services such as electricity and telecommunications upon which these depend. Roads, bridges and energy infrastructure that are resilient to natural disasters such as avalanches and flooding will link food chain stakeholders, enhance availability and access to stable food supplies in remote areas, facilitate the development of local agribusiness and expand the domestic agrifood market.

Above all, external support and investment in security is necessary to enable many other developments to take place. However, simple answers to conflict resolution will fail. Transition out of conflict will require sophisticated understandings of the political economy of war and peace, and the potential bargains that are necessary to ensure the peaceful inclusion of conflicted parties in governance processes that may not necessarily conform to existing expectations (Cheng et al. 2018). Meanwhile, agriculture and nutrition interventions with supporting economic developments are still needed to build the human health capital of the vulnerable Afghan populations.

6.7 Wider implications

It may be that Afghanistan is slipping from fragility towards failure, surpassed only by a few, mainly African, states. The situation commands attention, as the Economist magazine commented last year, ‘Few things matter more than fixing failed states’ (The Economist 2017). Afghanistan was one of four cases examined in the LANSA programme. The diverse Afghanistan contexts and challenges are not replicated exactly in Bangladesh, India and Pakistan, although challenges in some regions of the last are uncomfortably reminiscent of Afghanistan. Policy makers in the wider region and in parts of Sub-Saharan Africa and even Central America will recognise common challenges, for which the appropriate policies may resemble the recommendations included here.

Policy lessons about decentralising public services have a wider application than just failed states, as previous work on South Asia and East Africa shows (Gillespie et al. 2015; van den Bold et al. 2015). Lessons about better integrating agriculture and nutrition policies to address the burdens of malnutrition and achieve global development goals apply much more widely still. Many people live in areas of ‘limited statehood’ due either to ongoing conflict – the Middle East and Africa - or are struggling through the long haul from post-conflict status to stability – Colombia – or going in the other direction – maybe Nicaragua, Mexico and Venezuela. Managing the humanitarian-development interface is important, even under conditions of ‘mere’ natural disasters, let alone conflict. Many more people, both in South Asia, other developing regions and in advanced economies, suffer from the chronic effects of geographical and cultural remoteness from centralised policy making and an inefficient reach of public services. What currently works for agriculture and nutrition in Afghanistan is the cluster system of coordination at central level, inter-sectoral working groups at the provincial level, and extraordinarily dedicated people and organizations implementing national programmes in fragile areas. But there is no policy blueprint.

Overall, the weaknesses we have identified show that appropriate and inclusive policy formulation, multi-sectoral approaches to reducing malnutrition, and leveraging agriculture for nutrition do not work well in Afghanistan. Change needs the combined efforts of the national government, civil society and the international community to re-envision policy processes by adopting decentralized and territorial approaches to integrating agriculture and nutrition and wider development objectives (Poole 2017).

Glossary

ACBAR

Agency Coordinating Body for Afghan Relief

AFSANA/AFSeN

Afghanistan Food Security and Nutrition Agenda

AKF

Aga Khan Foundation

ANDMA

Afghanistan National Disaster Management Authority

ANDS

Afghanistan National Development Strategy

ANPDF

Afghanistan National Peace and Development Framework

APP

Agricultural Production and Productivity Programme

AREU

Afghanistan Research and Evaluation Unit

ARDZ

Agricultural and Rural Development Zone

ASGP

Afghanistan Subnational Governance Programme

BDK

Badakhshan Province

BMN

Bamyan Province

BPHS

Basic Package of Health Services

CARD-F

Comprehensive Agriculture and Rural Development Facility

CDC

Community Development Council

CSO

Central Statistics Organization

EDP

Economic Development Package

DCC

District Coordination Council

DDA

District Development Assembly

DFID

United Kingdom Department for International Development

DGO

District Governor’s Office

FAO

United Nations Food and Agriculture Organization

FSAC

Food Security and Agriculture Cluster

GoIRA

Government of the Islamic Republic of Afghanistan

HNSS

Health and Nutrition Sector Strategy

IDLG

Independent Directorate of Local Governance

IMC

Inter-Ministerial Committee

IOM

International Organization for Migration

IPC

Integrated Food Security Phase Classification

IRoA

Islamic Republic of Afghanistan.

IYCF

Infant and Young Child Feeding

KBL

Kabul

KDH

Kandahar Province

LANSA

Leveraging Agriculture for Nutrition in South Asia

MAIL/DAIL

Ministry/Department of Agriculture Irrigation and Livestock

MCN

Ministry of Counter Narcotics

MDGs

Millennium Development Goals

MoCI

Ministry of Commerce and Industry

MoE

Ministry of Education

MoEW

Ministry of Energy and Water

MoF

Ministry of Finance

MoFA

Ministry of Foreign Affairs

MoLSAMD

Ministry of Labor, Social Affairs, Martyrs and Disabled

MoPH/DoPH

Ministry/Department of Public Health

MoWA/DoWA

Ministry/Department of Women’s Affairs

MRRD/DRRD

Ministry/Department of Rural Rehabilitation and Development

NABDP

National Area-Based Development Programme

NADF

National Agricultural Development Framework

NAF

Nutrition Action Framework

(I)NGO

(International) Non-governmental Organisation

NGH

Nangarhar Province

NPP

National Priority Programme

NSP

National Solidarity Programme

OCHA

United Nations Office for the Coordination of Humanitarian Affairs

PC

Provincial Council

PDC

Provincial Development Committee

PGO

Provincial Governor’s Office

PND

Public Nutrition Department

SNG

Subnational Governance

SWG

Sector Working Group

UNDP

United Nations Development Programme

UNICEF

United Nations Children’s Fund

WASH

Water, Sanitation and Health

WFP

United Nations World Food Programme

WHO

United Nations World Health Organization