1 Introduction

South Africa, along with many other low and middle income countries, is experiencing a rising burden of diet-related Non Communicable Diseases (NCDs) while still struggling to address persisting household food insecurity and undernutritionFootnote 1 (Muzigaba et al. 2016). Over the past 40 years, the prevalence of stunting among children in South Africa has remained at around 25% (Said-Mohamed et al. 2015). In the 2012 South African National Health and Nutrition Examination Study (SANHANES) study, 54% of the South African population reported being food insecure, and 28% were at risk of hunger (Muzigaba et al. 2016). More recently, the prevalence of obesity has risen to 39% among women and 11% among men, and diabetes in the adult population to 10% (Shisana et al. 2014).

Addressing this double burden of malnutrition will require a comprehensive policy approach, which supports both demand for healthy food and its supply (Food and Agriculture Organization of the United Nations and World Health Organization 2014). In this paper, we focus on supply side interventions – and particularly, the need for policy across sectors to support availability of affordable healthy food (Republic of South Africa Department of Health 2013; Government of South Africa 2014). Global evidence shows that government action to promote a healthy food supply can be in tension with government objectives to pursue economic growth, particularly through economic liberalization (Hawkes 2005; Mihalache-O’keef and Li 2011; Popkin et al. 2012; Margulis 2013; Baker et al. 2014; Thow et al. 2015a). This tension between policy objectives of different sectors can result in policy incoherence. In contrast, policy coherence refers to ‘the systematic promotion of mutually reinforcing policies across government departments to create synergies towards achieving agreed objectives and to avoid or minimize negative spillovers in other policy areas’ (OECD 2016). Policy coherence is prioritized in Sustainable Development Goal 17 (United Nations 2015).

There are three key facets of incoherence between economic policies and food security and nutrition policies that have been observed globally. First, economic policies focused on liberalization – particularly of trade and investment – can have negative impacts on nutrition and food security. For example, increased competition and economies of scale associated with trade and investment liberalization, particularly for corporate and multinational food processers, manufacturers and retailers, have helped to decrease the price and increase the availability of highly processed foods, contributing to diet-related NCDs (Baker et al. 2014; Thow and McGrady 2014; Schram et al. 2015; Thow et al. 2015a; Thow et al. 2015b; Timmer 2016). In addition, poorer households may experience increased food insecurity through volatility of global food prices and negative impacts on employment as a result of trade liberalization (Brooks and Matthews 2015). For example, during the global food crisis of 2007–2009, shocks such as speculative behaviour in food commodity markets and the diversion of food crops to fuel production led to sudden increases in the prices of staple foods (De Schutter 2009). Second, nutrition-related policies that aim to reduce the availability and affordability of unhealthy, highly processed (and often highly profitable) foods can be at odds with economic policies that aim to attract or incentivize trade and investment in food processing, service and retail. This can create tensions for governments due to the political power of investors with significant investments at multiple points in supply chains (Thow and McGrady 2014), because nutrition interventions may adversely affect the profitability of investments in food processing or agriculture. For example, initiatives such as a product tax or labeling measures to reduce highly processed food consumption. A potentially concerning result of this is the possibility for measures to be challenged under investor protection clauses within Investment Agreements (Thow and McGrady 2014; Woolfrey 2014). Third, policy incoherence can result from supply chain policies (including development, trade, finance, industrial and (some aspects of) agricultural production policies) that focus on objectives related to economic growth, but give little consideration to nutrition and food security objectives related to increasing access to affordable healthy food. As a result, nutrition and food security policy objectives can be undermined by economic policy action (Walls et al. 2015; Ruckert et al. 2016).

Previous research has indicated that these tensions appear to exist in South Africa, where the food supply is subject to binding agreements regarding trade in goods and services, and Foreign Direct Investment (FDI) (Greenberg 2017). There have been significant increases in food trade and investment, particularly related to processed foods, in the past two decades (Igumbor et al. 2012; Schram et al. 2013; Thow et al. 2015a). In this paper, we investigate, systematically, if and how the policy objectives articulated in economic, food security and nutrition related sectors across the Government of South Africa are coherent; identify if there are tensions among the policy objectives; and explore the underlying actor beliefs and frames that shape the current policy context. We then consider opportunities to improve policy coherence, and, particularly, outcomes for food security and nutrition.

2 Methods

The aim of the study was to identify 1) instances of policy incoherence and 2) opportunities to improve policy coherence among sectors with responsibilities for food supply policy related to food security and nutrition in South Africa. Our research questions were:

  • What are the main current food supply policy objectives and actions related to food security and nutrition in South Africa?

  • What are the political dynamics and actor beliefs that underlie food supply policy related to food security and nutrition?

  • How could policy coherence be improved in relation to food security and nutrition?

This qualitative study used two methods for policy analysis: document analysis of existing policies and strategies, and interviews with actors engaged in the South African food policy space. Our analysis focused on South African national government sectors with policy responsibilities related to the food supply. These include Agriculture (food production and marketing), Investment (food production and processing), Commerce and Industry (food processing, marketing and distribution), Trade (food distribution), and Health (food and health-related legislation). To underpin our analysis, we drew on Sabatier’s conceptualization of coalitions of actors as influential in shaping policy outcomes for a given policy area. The selection of this framework for analysis was informed by the observation during data collection that the incoherence evident in the policy content appeared to reflect obvious divergences in actor beliefs. In other words, the policy incoherence within the food policy subsystem appeared to not simply reflect different policy objectives across sectors, but also reflected different beliefs about food security and nutrition as a policy issue.

2.1 Policy content review

We searched government websites for relevant policies using each of these sectors as search terms, together with the words ‘policy’, ‘strategy’ and ‘action plan’, and then identified further policies through cross-references in policy documents. Based on the literature and global recommendations regarding 1) best-practice food security and nutrition policy and 2) the implications of international economic agreements and nutrition, our objective was to identify policy content that fostered positive incentives for food security and nutrition within the food supply, or indicated points of incoherence. We developed a framework for policy content analysis that enabled us to compare food policy objectives and content across sectors. We first extracted content relevant to the food supply from food security and nutrition policy documents. We then analysed policy content with respect to stated policy objectives, informed by the OECD policy coherence framework (OECD 2016), to identify the ways in which policy objectives and activities in the relevant economic policy documents supported or undermined food security and nutrition policy objectives.

In the case of trade, investment and industry policy, we augmented the assessment of policy coherence related to objectives and content with a review of the literature that identifies impacts from these economic policy sectors on food security and nutrition outcomes and policy space. This was to aid identification of policy incoherence, since these economic policies tend not to explicitly mention food security and/or nutrition. The policy provisions identified are presented in Box 1.

Box 1 Summary of provisions in trade and investment agreements and related economic policies with implications for food security and nutrition

2.2 Interviews

The interviews were designed to explore the nature of food policy incoherence in relation to food security and nutrition. In particular, policymaker beliefs and frames used to inform the development of the policies and which might help explain the policy (in)coherence between food security and nutrition policy objectives and actions related to the food supply on the one hand, and economic policy objectives and actions on the other. Interview schedules were based on policy analysis frameworks (Bennett and Howlett 1992; Shiffman and Smith 2007; Reich and Balarajan 2012) and the OECD policy coherence framework (OECD 2016), and asked about: influential actors; policy processes; policy priorities; policy context; framing of nutrition; and opportunities to improve coherence.

AMT, SG and MH conducted 14 semi-structured interviews, each 1–1.5 h in length, with 22 actors engaged in the South African food policy space in September 2016, in Cape Town, Pretoria and Johannesburg. Participants included 12 national-level government food policy actors from agriculture (n = 6), economic policy (n = 3), and health (n = 3); 2 academics, 2 independent food policy consultants, and 6 food industry stakeholders. Participants were recruited through formal letters of invitation to the heads of relevant agencies, and through snowball sampling. We also requested interviews with three Investment Banks, as the largest source of investment in the food supply in South Africa, but the opportunity was declined. Interviews were all jointly conducted by three of the authors (with expertise in fisheries, agriculture and nutrition). All interviewers took detailed notes during the interviews, and each of these was combined into a single interview summary.

The first author conducted the analysis using NVIVO™, and coding and validity of the themes were then reviewed by two other authors. NVIVO is software that supports qualitative and mixed methods research by helping researchers to organize and analyze data (QSR International 2018). Themes were informed by Paul Sabatier’s Advocacy Coalition Framework (Jenkins-Smith et al. 2014), as an established framework for understanding policy dynamics and opportunities for policy change. This framework identifies the role of actor coalitions (both inside and outside of government, bound together by beliefs) as core in shaping policy outcomes within a policy subsystem – in this case, the food and nutrition security policy subsystem. Themes examined included:

  • Actor coalitions with interests in the food policy space in South Africa, and the beliefs and resources available to these coalitions

  • Institutions and forums relevant to multi-sectoral policy making for nutrition and food security

  • Framing/beliefs about food security and nutrition by different types of actors in the food policy space

  • Context – particularly relevant characteristics of the political system and broad policy priorities

  • Perceived policy opportunities that could further support the production and consumption of healthy foods

Findings from the document review were synthesized in relation to policy incoherence, and then – also drawing on the interview data – we identified three coalitions within the policy subsystem. We first focus on evident policy coherence and incoherence in policy content related to food security and nutrition in the food policy subsystem – and then describe the three coalitions identified, based on the Advocacy Coalition analytical framework.

This study was granted ethical approval by the Humanities and Social Sciences Research Ethics Committee of the University of the Western Cape.

3 Results

This analysis of food supply policy identified a number of tensions and points of incoherence between economic perspectives on food supply policy goals, production-oriented perspectives on food security, and health-focused perspectives on nutrition. These points of incoherence primarily reflected instances in which nutrition and food security policy objectives were not supported – or were undermined – by food supply policies in the economic sector.

3.1 Review of policy content

We identified 40 policy documents and related government initiatives relevant to food security and nutrition, including those relevant to the food supply more broadly (Tables 1, 2, 3, and 4). We start here by reviewing nutrition and food security policy priorities relevant to the food supply, and then examine the economic policies that govern the food supply.

Table 1 Nutrition/Food policy priorities in South Africa
Table 2 Agricultural policy priorities in South Africa
Table 3 Overview of South African trade and investment policy documents for potential nutrition implications (including all existing/terminated BITs with English language text available)
Table 4 Economic incentives for investors, potentially relevant to nutrition

3.1.1 Nutrition and food security policies

The Government of South Africa has identified specific policy objectives to improve nutritional health (Table 1). These include prevention of NCDs and promotion of health and wellness (Strategic Plan for Prevention and Control of NCDs, “NCD Strategic Plan”, 2013–2017), and, in line with the Government’s commitment to the Right to Food, ensuring “availability, accessibility and affordability of safe and nutritious food” (National Policy on Food and Nutrition Security, 2014). These policies both reflect global recommendations for using food supply policy to improve nutrition, with interventions targeting increased access to affordable healthy food, an explicit activity. The NCD Strategic Plan mandates engagement (by health) with relevant government departments, including agriculture, trade and industry, and treasury to achieve this. The National Policy on Food and Nutrition Security identifies the need to increase access to production inputs, leverage government procurement, use market interventions and trade measures for food security, and address land tenure.

3.1.2 Food security and agricultural policies

Food security has been identified repeatedly as a national priority, including in the National Development Plan (2012), which mandated the preparation of the Department of Agriculture, Fisheries and Forestry’s Integrated Growth and Development Plan (the national agricultural policy). The Government of South Africa’s agricultural policies are the Integrated Growth and Development Plan (2012) issued by the Department of Agriculture, Fisheries and Forestry and the Agricultural Policy Action Plan (2015–2019) (Table 2). National agricultural policies in South Africa include strong support for food security, although there is no mention of nutrition (Table 2).

The objectives that are emphasized in agricultural policies are those relating to economic growth, employment creation and rural development, and the dominant frame through which attainment of food security is articulated is economic and aggregate production oriented (rather than distribution oriented). In contrast, the text of the Food and Nutrition Security Policy is framed in the context of the right to food, and access to safe and nutritious food for households (Table 1). However, overall there is an implicit focus in all the food security related policies on the issue of quantity of food, and little consideration of nutritional quality.

3.1.3 Economic policies relevant to the food supply

South Africa’s economic policies that affect the food supply have clear objectives: to increase economic productivity and employment through agriculture, food processing and food retail. Trade and investment policy commitments include reducing barriers to trade and investment with respect to goods and services (including food), and protecting intellectual property rights and investors (including in the food system). As part of this, specific measures to promote agri-food processing are highlighted as a growth area from an economic perspective (Box 1). Food processing – which is an issue of concern from a nutrition and health perspective – is a priority in the National Development Plan and Trade Policy and Strategy Framework, and investment incentives are provided in the Industrial Policy (Tables 3 and 4).

These economic food supply policy objectives are supported at a whole-of-government level by the National Development Plan, which focusses on economic and social development (National Planning Commission 2012). Key priorities relevant to the food supply include increased employment, poverty reduction and improved agricultural production – all of which would generally have positive spillover impacts for food and nutrition security. However, food supply policy objectives contained in this broader government policy agenda focus primarily on food as an economic commodity (for example, as a source of income and employment). The National Development Plan does include nutrition as a priority, but only in terms of direct (health sector) interventions for maternal and child undernutrition, with no mention of food supply intervention. (National Planning Commission 2012) Food security objectives in the National Development Plan focus on increasing economic access to food and decreasing the cost of food, but not on the nutritional quality of that food. Somewhat contradictory definitions of food security appear in the National Development Plan, which further explain this lack of focus on the nutritional quality of the food supply. Although the Plan does acknowledge the definition by the United Nations Food and Agriculture Organization, which states that ‘everyone has access to sufficient, nutritious and safe food at all times’ (p.230), the focus overall is on household access dimensions of food security (for example, ‘Household food security is determined by the ability to access food rather than its availability’, p230). There is no mention of health, nutrition or food security in the preamble or objectives of trade or investment agreements (Box 1, Table 3). The only trade agreement with any exceptions for public health (implicitly including nutrition) is the Trade, Development and Cooperation Agreement between South Africa and the European Union. In addition, there is no mention of nutrition in relation to NCD prevention in Trade, Investment, Industry, or Agricultural policy documents, or of the food supply policy actions identified in the National Strategic plan for NCDs. There are also no provisions that explicitly protect food security where economic interests might be in conflict – for example, to ensure that expanding protection of intellectual property rights does not interfere with smallholder access to seeds.

However, trade policy directions and priorities in South Africa have evolved over the past decade to have more of a focus on equitable development (Box 1). A review of investment policy in South Africa was undertaken between 2007 and 2010, in part in response to an international investment dispute regarding the Broad-Based Black Economic Empowerment Act in 2007 under the Belgium/Luxembourg - South Africa Bilateral Investment Treaty (Mossallam 2015). This review led to the termination of several ‘first generation’ Bilateral Investment Treaties, and the new Protection of Investment Act 2015 (Mashigo 2014; Adeleke 2015), designed to maintain a level of investor protection while bringing current agreements into line with the priority given to non-economic (particularly social, sustainable development and equality focused) policy objectives (Government of South Africa 2010; Mossallam 2015).

The 2010 Trade Policy and Strategy Framework explicitly identifies the need for trade policy commitments to support broader national development objectives (Table 3). In addition, the new Promotion of Investment Act 2015 limits the scope for the food industry to contest food security and nutrition policy measures that might impact on the value of investments. The termination of existing investment agreements with very ambiguous definitions of key terms such as Fair and Equitable Treatment, and no broad development objectives in their preambles, opens a potential opportunity for policy space to protect and promote food security and nutrition (Box 1).

3.1.4 Summary of policy document analysis: points of incoherence in food supply policy

Overall, nutrition and food security policy objectives articulated in health and agriculture policies – particularly those that relate to food supply change that promote availability of affordable nutritious foods – are not explicitly supported by the economic policy sectors. Economic policies relating to the food supply do not include explicit consideration of nutrition or food security policy space, but focus on food as an economic commodity.

3.2 Interviews: Actors and coalitions

Analysis of the interview data identified three key coalitions relating to food and nutrition policy, which reflected the points of policy incoherence identified in the policy content analysis. The dominant subsystem we termed ‘Economic Growth’, due to its framing of the role of food systems as contributors to economic outcomes and employment. The second was termed, ‘Food Security and Agriculture Production’ ('Food Security'), due to its emphasis on production aspects of food security. The third we termed the ‘Health’ coalition, due to its emphasis on food as a nutrition and health issue. These coalitions were mutually exclusive, in terms of the interviewees who articulated the main tenets of each coalition, but there was some overlap observed in beliefs between the Food Security and Health coalitions, and in some of the framing of food security between the Economic Growth and Food Security coalitions, as indicated below. Most of the actors in each coalition interacted formally in various forums and were thus linked not only in their frames and beliefs but also in formal policy processes. However, they did not refer to each other personally during the interviews, although many referred to colleagues’ institutions as influential.

3.2.1 Economic growth coalition

The Economic Growth coalition focused on the role of economic growth and employment in delivering improved food security and nutrition outcomes. Key actors were the Department of Trade and Industry and other economic policy departments, Agricultural trade within the Department of Agriculture, Fisheries and Forestry, the food industry, and agricultural producers/traders.

The Economic Growth coalition appeared to be a dominant coalition, as it had the most resonance with the priorities and frames of overarching government policy objectives: in particular, the resonance between the beliefs of this coalition, regarding food as primarily an economic commodity, and the primacy of economic growth (employment, balance of trade and other indicators) within the priorities of the National Development Plan. The framing and beliefs of this coalition were evident in the focus on economic growth and employment that permeated policy documents relating to the food supply, described above. In line with this policy content focus on economic growth, the interviews identified a strong bias towards economic interests (both government and private sector) in the formal government forums convened to inform policy making. Indeed, the dominance of this coalition was further indicated through the direct access that the food industry has to government, formalized through their participation in a range of forums, including the National Agroprocessing Forum, the Agricultural Trade Forum and Value Chain Roundtables on key commodities.

[We] have walked a long journey building relationships with government, so we can engage robustly [Interview 4, Food Industry]

Within government, cross-sectoral Clusters are convened at a high level (DG/Ministerial) to deal with policy coordination and cross cutting issues in government. These include Economic, Social, Trade and Foreign Policy Clusters. It was notable that interviewees did not identify an obvious place of responsibility in either the roundtables or these cross-cutting forums for food security and nutrition to be discussed.

Framing of food security and nutrition

The Economic Growth coalition framed hunger and undernutrition as the priority issue to be addressed by food security and nutrition policy. Nutrition in the context of NCD prevention was perceived as something that would resolve itself with economic growth (i.e. consumers becoming wealthier), and trade in particular was seen as critical for ensuring food security. The Economic growth coalition framed the causes of nutrition and food security problems as a lack of access to (healthy) food – related to income – among vulnerable sectors of the population. This was closely linked to employment opportunities.

40% unemployment is at the bottom of the issue… South Africa can’t have food security when people don’t have incomes … [Interview 6, Academic, Food Science]

In the Economic Growth coalition there was acknowledgement of dietary change and a shift towards processed foods, with negative implications for nutrition. However, these trends were framed as the result of individual preferences for fat, salt and sugar, related to taste and palatability. Consumer decisions not to purchase healthy foods were also framed as personal decisions based on preference and consumer desires for ‘status’ foods. In line with this, solutions for NCDs were framed as addressing personal factors through improving education. This was seen as an avenue to improve consumption decisions and also as a mechanism to improve the food supply, since industry was seen as responding to consumer demands.

However, there was also some acknowledgement of the current food and nutrition problems as the result of systemic issues. One actor in the Economic Growth coalition identified the need for policy entrepreneurs and advocates as ‘visionaries’. Such actors could use systems-thinking approaches to identify and articulate long-term economic consequences (including with respect to health) of policy making.

There may be unintended consequences [of liberalizing trade] for diets… South Africa has potential to pick this up earlier than the wealthy countries… We needed visionaries in negotiations back then… Thinking in systems would change policy making. [Interview 7, Agricultural trade].

Beliefs: economic solutions to achieve nutrition and food security

Actors aligned with the Economic Growth coalition held an evident belief that food security and nutrition were positive by-products of economic growth. In contrast to the Health coalition (described below), there was little perceived tension between the goals of economic policy and nutrition/food security policy. Integration into global value chains was seen as the main opportunity for agricultural industries, which was a key point of difference with the Food Security coalition (described below). This reflected the priority given to employment and perceived indicators of economic development, such as maintaining a positive balance of payments.

In this coalition, there was an evident belief that industry was a (if not the) key stakeholder in achieving food policy goals. Food industry actors were portrayed as highly knowledgeable stakeholders, and the avenue through which policy objectives would be achieved. A strong, formalized, competitive local industry was seen as critical to achieving development goals. The food industry supported this view, and articulated a belief that they were key to achieving food security and nutrition policy goals. As part of this, there was a belief that Government and Industry could be mutually supportive in achieving food security and nutrition goals.

It was evident within this coalition that there was little perceived tension between achieving goals of economic growth and food security/nutrition. There was a belief that the market would resolve any perceived tension:

If industry doesn’t have a healthy market then they are not going to be economically viable, so they have vested interest in maintaining a healthy market … It is possible to reconcile health/nutrition and profit motives… at the end of the day, the food supply is mostly consumer driven and a high level of competition means it is too risky to be unethical in marketing… [Interview 2, Food Industry].

Resources: high level political will and industry support

Alignment with the core objectives of the National Development Plan and priority government economic goals gave the Economic Growth coalition a high level of political support. For example, agriculture and the food industry were identified as a means to achieve the government’s stated goal of creating 11 million jobs by 2030.

Industry and industry associations positioned themselves as key resources to achieve not only these economic objectives but also food security and nutrition policy goals. In particular, as the main holder of technical expertise, as evidenced by their assistance to government in setting food standards; as a source of innovation in food and nutrition; and as experts in logistics, essential for meeting food needs (e.g. fish, staples). Industry actors portrayed themselves as direct contributors to food security and nutrition, through general food production and their Corporate Social Responsibility activities. For example:

[Regulators] grossly underestimate [our] contribution to food security. [Interview 1, Food Industry]

3.2.2 Food security and agricultural production coalition

The Food and Nutrition Security Policy was the product of what we termed the (minority) ‘Food Security and Agricultural Production’ coalition (hereafter, the Food Security coalition). It was developed by the Food Security Division of the Department of Agriculture, Fisheries and Forestry in consultation with (primarily) Education, Health and Social Development.

Framing of food security and nutrition

In the Food Security coalition, food insecurity (i.e. lack of sufficient food) was framed as a major problem requiring a policy response, and there was little consideration of other issues of nutrition, such as NCD prevention and micronutrient deficiencies. In contrast to the Economic Growth coalition, the cause of the problem was framed primarily as one of increasing food prices and insufficient production, in a context of poverty, rather than a lack of access to income/employment. Although these are two sides of the same issue (food affordability), the difference in emphasis (production rather than employment) influenced perceptions of appropriate policy solutions.

Drought is a big problem … Not enough food [was] produced so prices went up… the whole food basket [is] affected [Interview 13, Agriculture]

Framing of food supply problems tended to focus on the lack of support for local production of diverse food crops. The policy agenda of the Food Security coalition was perceived by other actors as being almost too focused on primary production [Interview 7, Agricultural trade]. Similarly, while public procurement was identified as a strategy to increase availability of fresh healthy food, provide stable income for local farmers, and support rural development, this was embraced more cautiously on the side of the Economic Growth coalition. Public procurement is a key issue in trade agreements, which tend to minimize the potential for preferential local public procurement.

However, there was some overlap between the Food Security and Economic Growth coalitions. For example, there were a few crops mentioned across coalitions as high economic value, and feasible for smaller scale production, such as berries [Interview 5 & 7, Agriculture]. Related to this are opportunities to improve demand for fresh produce through supporting local markets that link farmers more directly with consumers. This would increase accessibility of healthy food for consumers, support local farmers in a way that was inclusive of smallholder farmers, and support economic growth in rural areas (an economic policy priority) – and potentially reduce costs incurred by long supply chains. This strategy was also seen as feasible from the economic growth coalition because creating local markets and encouraging diversification would benefit small scale farmers economically.

Beliefs: food is a local, social good, and not an economic policy issue

The Food Security coalition held an evident belief that food security was a social rather than economic issue. Actors highlighted the policy tension between food as an economic commodity, and as a social good.

Food security is a social issue… [it] will always be at opposite end to economics… The Government is trying to bring these together but it is not possible…. Economists will tell you that economic growth brings spin offs, but social issues are marginalized [Interview 13, Agriculture].

In line with this, the Food and Nutrition Security Policy was developed with limited input from the economic sector, despite the economic sector (Trade, Investment, Commerce etc) having a significant policy influence on the food supply. The key actors involved in the development of the policy were the Department of Education – seen as particularly relevant with growing interest in school feeding (“nutrition is a side issue for them but they are interested because of what children eat” [Interview 13, Agriculture]) – the Department of Health, because of their expertise in nutrition, and the Department of Social Development, which has early childhood development centres and provides social grants.

The Food Security coalition was also characterized by a belief that local markets (production and consumption) would strengthen food security, through a focus on providing access to consumers and also supporting poor farmers. Social grants were seen as helpful contributions at the household level, but the primary need was framed as affordable, accessible food.

The local food processing industry also positioned itself as contributing to the policy objectives of this coalition, particularly processing companies. They framed their supply chain expertise and preference for local primary produce as expanding production capacities and increasing local production:

[We are] supporting local farmers, and try to source locally … we prefer not to import due to cost…and purchase 2 million tonnes of agricultural commodities per year, of which 2/3 is local [Interview 4, Food Industry].

Resources: high level political will but limited Civil Society engagement

The Government of South Africa has prioritised food security as part of national and international commitments, including the Constitution and the Sustainable Development Goals (Table 4).

However, although this political commitment exists, the use of the term ‘food security’ in the National Development Plan appears to align more with the Economic Growth coalition’s frame than the Food Security frame, with more of a focus on increasing household employment and national measures of food security. In addition, while the Food and Nutrition Security Policy was developed by the Ministry of Agriculture, the locus of implementation of the Food Security and Nutrition Policy is with the National Government in the planning department. This may indicate future challenges in maintaining the conceptualization of food security used by the Food Security Coalition.

Actors in this coalition reported interest from Civil Society Organizations (CSOs) in contributing to food security policy, but CSOs had limited involvement in policy making – and notably, no formal avenues for input. In particular, they weren’t included in the development of the Food Security and Nutrition policy. In the next policy stage (implementation) CSOs will be included in the high-level council on Food Security. However, an actor from Government expressed caution about engagement with CSOs, voicing a perception that small informal producers have limited representation.

Civil Society Organizations wanted to be consulted on policy … But it is not clear who they represent. It is unlikely to be the rural poor… [and] farmer associations don’t have local roots. CSOs are not organized… [they] need to be properly represented and organized [Interview 13, Agriculture].

3.2.3 Health coalition

The actors in what we termed the Health coalition evidenced beliefs regarding the importance of food supply policy in creating healthy food environments for good nutrition for health (including aspects of food security, but more health focused). For example:

[Health and nutrition is] not just about education, because nutritious food/healthy convenience food is not affordable, even when you are not poor [Interview 11, Public Health]

The main actor was perceived as the Department of Health, as the focal ministry for nutrition related policy, and also CSOs. However, there was also recognition that achieving nutrition policy goals would require action in economic sectors.

The Food Legislation Advisory Group is the multi-stakeholder forum related to food and health, which sits with the Department of Health, and acts as a platform for Government to engage with industry associations, academics, and other government departments on food regulation. Although interviewees from a range of coalitions identified this as the key multisectoral forum relevant to food security and nutrition, it is not constituted to address either of these issues, and its main focus is on food safety.

Framing of food security and nutrition

The Health coalition framed the main nutritional problem as the coexistence of multiple forms of malnutrition (undernutrition, micronutrient deficiencies and diet-related NCDs/obesity). In particular, they emphasized that these different forms of malnutrition are affecting common (not different) populations, and thus needed to be addressed in a coordinated manner.

In the Health coalition the problem of poor diets was framed as a response to food environments, rather than an issue of personal choice, in contrast to the Economic Growth coalition – although there was recognition that limited household finances also played a role in skewing consumption to cheap, unhealthy foods. They identified the relative inexpensiveness of unhealthy foods, as well as industry efforts in the marketing and advertising of such foods as key factors driving dietary change.

Energy dense, low nutrient foods are what is commonly consumed… This is very cheap and tastes nice – e.g. chips [Interview 11, Public health]

In line with this, the solution was framed as a need for systemic change – to increase access to healthy affordable foods, such as fruit and vegetables. However, the food system was seen as very difficult to change in terms of reorienting to healthy food production.

Beliefs: solutions require food policy change, without industry influence

Actors in the Health coalition articulated a belief that considerations of health and nutrition were marginalized in food supply policy. This was perceived as being due to the prevailing focus on “bringing investment, not on the impact of investment” [Interview 14, Public Health]. The Department of Health also had limited participation in decision-making regarding economic policy relating to the food supply.

Nutrition is not really considered [in the Interministerial Committee on Investment]… it is the domain of the Department of Health [Interview 8, Economic policy].

There were overlaps in beliefs about food security and nutrition policy between the Food Security and Health coalition, and also some evidence of collaboration between the Ministries of Health and Agriculture, for example on the development of orange-fleshed sweet potato. This was also supported by the consideration of micronutrient content of crops in the national agricultural policy (Table 2). However, a key difference appeared to be the focus of the Health coalition on the outcomes of agricultural production for the health of consumers, which was not perceived as a core issue for consideration by agricultural policy makers, who were seen by the health sector as more concerned with ensuring the welfare of farmers. This was seen as limiting the scope for more significant collaboration on nutrition.

In contrast to the Economic growth coalition, the Health coalition held a strong belief regarding the need to limit industry involvement in food security and nutrition policy making due to conflicts of interest:

Policy space [for nutrition and food security] needs to be protected – industry should not be involved in the policy space … Health must engage with industry but needs to have very clear rules of engagement and guidelines [Interview 10, Public Health].

Resources: limited influence and capacity for enforcement

Part of the marginalization of health was perceived as due to the lack of Civil Society activity in the nutrition policy space, which is needed to help raise concerns to the attention of policy makers. However, there appeared to be no formal mechanism for their engagement (in contrast to industry actors in the economic policy space. The advantage to the health sector of having Civil Society actors calling for strong food security and nutrition policy, was that it would circumvent the need for the health sector to directly lobby against government policies that they perceived as favouring industry over health. It would also give access to broader expertise than contained within the government.

Health actors need to consider how to strengthen Civil Society, in terms of capacity and education, and to create organized lobbying. Improving food security and nutrition needs multidisciplinary expertise… [Interview 10, Public Health]

The Health coalition also faced a significant challenge in the form of lack of resources. This meant that policies weren’t implemented. Two areas of resource imbalance were highlighted in the interviews, which made it difficult for the Health coalition to successfully shape policy agendas. One was the lack of resources for government nutrition promotion, compared to the advertising budgets of industry, resulting in poor quality health promotion interventions.

[There are] huge disparities in resources – the Department of Health doesn’t even have 100 million rand/year for prevention… but industry spends 100 million rand on just one ad. [Interview 14, Public Health].

The marginalization of nutrition interests was compounded by an imbalance in resources and influence for lobbying. The resources available to industry to fight policy (for example, in relation to the proposed [at the time] soft drink tax) were much more significant than that available to the Health Coalition actors – primarily government and civil society. The direct access that industry had to economic forums was also seen as giving them a preferential position in food supply policy making.

4 Discussion

4.1 Current policy agendas

Food security and nutrition policy is a political and contested policy space. This study identified three different policy coalitions contributing to policy incoherence regarding food supply and food security and nutrition in South Africa. Drawing on Sabatier’s conceptualization of coalitions of actors as influential in shaping policy outcomes in a given policy subsystem, we analysed the framing of food security and nutrition by different actors, the resonance of these frames with policy content, and the evident beliefs and resources that characterized each coalition. Overall, we found recognition across all the coalitions that the government is trying to balance competing agendas in the food security and nutrition policy space. One of the key challenges to policy coherence identified was the very different framing of food and nutrition between the Food Security and Health coalitions, with the problem narrowly (coalition-based) defined as: hunger and economic access to calories; or rising consumption of unhealthy foods; or lack of diversity in diets based on staple foods. There is an implicit incoherence between economic/agricultural policy emphasis on value-adding, which is primarily an avenue for job creation, and health policy emphasis on fresh, unprocessed (healthier) foods.

The dominant policy coalition, whose beliefs we see most clearly reflected in policy documents governing the food supply, we termed the Economic Growth coalition. Actors in this coalition frame food insecurity and malnutrition as primarily the result of a lack of income and a lack of knowledge about healthy eating. This understanding of the problem as primarily deriving from individual level factors, such as being poor (i.e. lack of economic access to food) or personal preference (e.g. for foods high in fat, salt and sugar), is reflected in the focus on personal education and economic growth (to provide employment and income) as core components of the solution. The core beliefs of this coalition are that employment and economic growth, within a neoliberal economic dispensation, are the primary mechanism to improve nutrition and food security, and that industry is therefore key to achieving food security and nutrition in the long term. This coalition is supported by high level political will – these beliefs are reflected in the National Development Plan and other core food supply policy documents. They are also heavily supported by the food industry, which is represented as a resource to achieve food security and nutrition policy goals. Industry has several formal mechanisms to input into policy making; their role is framed as both technical experts in food systems, and their significant contribution to economic growth.

Support for economic growth within a neoliberal, unregulated framework have been documented elsewhere as dominant in food policy making (Pinstrup-Andersen 2013). The food industry has been heavily engaged with developing policy solutions that focus on individual responsibility and portray the food industry as a key part of the solution (Jenkin et al. 2011; Scott et al. 2017), and in framing food as primarily an economic good and the food industry as a significant contributor to GDP (Friel et al. 2016). The heavy involvement of industry in policy forums in South Africa raises concerns about conflicts of interest in nutrition policy making. The World Health Organization has unequivocally recommended that nutrition policy processes be protected from the influence of vested interests (WHO 2013). However, it is unclear how this can be operationalized when policies shaping the food supply are both nutrition and economic policies.

The second policy coalition is focused on food security, with frames and beliefs resonant with the National Food and Nutrition Security Policy. Actors in this coalition frame food insecurity as primarily a problem of production and accessibility of food. This framing regarding production is reflected in solutions focused on increasing production for local populations, such as through increased investment in local markets. Food security is a political priority, and the Right to Food is enshrined in South African legislation; the planned National Food and Nutrition Security council will be situated under the President’s Office. However, there is ambiguity in the use of the term food security in high level policy documents – for example, the National Development Plan and national agricultural policy reflect much more of the framing of the Economic Growth coalition, in contrast to the national food and nutrition security policy, which is much more in line with the framing and beliefs of the Food Security coalition. A key opportunity to increase the resources available to this coalition is the civil society interest in this framing of food security. However, they have had limited participation in policy development to date.

The concept of food as a social good is embedded in a social perspective on food security (Riches 2016), and reflects aspects of the food sovereignty discourse in its focus on smallholder production and the right to food (Jarosz 2014). However, in this context this seems to be core to the marginalization of the Food Security coalition. Food trade and industry-led growth is a tenet of the dominant framing of food security by the Economic Growth coalition, and a focus on smallholder farmers and local markets is marginalized by the privileging of large-scale production and seen as unable to meet overarching policy objectives for economic development. This tension is reflected in recent calls to ‘revision’ of agricultural and food systems with respect to nutrition, which highlight the need to identify opportunities to achieve both economic and nutritional policy goals through agricultural production and distribution (Jones and Ejeta 2015; McDermott et al. 2015; Pingali 2015). As in this study, recommendations include strengthening incentives for diversification to nutrient-rich crops and strengthening markets. However, the potential of promoting small scale agricultural production of vegetables, fruit and small livestock to both supply more accessible nutritious food and create livelihoods remains marginalised in policy discourse in South Africa.

The third coalition identified we termed the Health coalition, which frames food security and nutrition from the perspective of malnutrition as a health outcome. Actors in the Health coalition frame malnutrition (undernutrition, micronutrient deficiency and diet-related non-communicable diseases) as primarily the result of an unhealthy food environment, in which unhealthy foods are among the most affordable and heavily marketed. In contrast to the Economic Growth coalition, the solution is thus framed as primarily the responsibility of the food system to deliver healthy affordable foods. A core belief of the Health coalition is the need for food supply policy to support nutrition objectives. Another key belief – particularly in relation to NCD prevention – is that the influence of the food industry on food-related (nutrition related) policy making should be circumscribed. However, the influence of this coalition does not appear to extend far beyond health policy documents and it is characterized by limited resources: in particular, limited engagement by civil society organizations, a low capacity for enforcement, and limited financial resources for raising awareness and exerting influence on policy (particularly in contrast to industry).

The challenges faced by the health coalition in translating their core beliefs into policy action have been observed elsewhere (Roberto et al. 2015). For example, a marked difference in the beliefs and frames between actors in public health and trade/agriculture has also been observed in the EU (Walls et al. 2016). This has often been attributed to resource constraints, including lack of political will and human and organizational capacities, which have resulted in limited policy action on malnutrition in other low and middle income countries (Pelletier et al. 2012). The lack of civil society engagement observed here has also been identified as a barrier to policy action for nutrition globally (Timotijevic et al. 2010; Huang et al. 2016). Recent research has identified strategies to build public support for nutrition policy action such as: improving public information; population-specific framing; strengthening media advocacy; and cultivation of change agents within government (Huang et al. 2016). One argument that has been adopted globally by nutrition policy advocates, but had little presence in the data we collected, was on the economic cost of poor nutrition and NCDs (Batura et al. 2015; Shekar et al. 2016).

4.2 Improving policy coherence

Evident in the understanding of the problem of food insecurity and malnutrition and the solutions identified by these policy coalitions is a tension between overarching policy objectives, as the Government of South Africa seeks to reconcile priorities of economic growth and productivity with health, social transformation, and the right to food. The renegotiation of investment agreements by the Government of South Africa, and the explicit policy priority for achieving social and development goals in the context of trade agreements indicates that the economic policy paradigm may be changing. This presents a potential policy window for inclusion of public health / nutrition considerations into trade and investment policy, such that policy space for current and future nutrition policy interventions is protected. This changing investment policy space in South Africa and the Southern African Development Community (SADC) reflects wider concerns regarding the potential for investment agreements, including Bilateral Investment Treaties, to constrain national policy space for achieving social, health and other objectives. The United Nations Conference on Trade and Development has recently concluded that “Today, the question is not whether or not to reform [International Investment Agreements], but about the what, how and the extent of such reform” (UNCTAD 2016).

Policy coherence in this context would mean that food security and nutrition policy objectives are not undermined – and ideally are supported – by economic policy that relates to the food supply. Change in the economic paradigm that has resulted from the social policy subsystem disruption (which led to the significant changes in investment policy described in Findings) might create scope for more positive economic policy in the food system. Leveraging this opportunity to improve food security and nutrition will require acknowledgement of broader development, food security, nutrition and health objectives in economic policy objectives (including economic development, trade, investment, industrial and agricultural policies). It will also require food security and nutrition to be perceived as a domestic policy priority, to be pursued in the protected policy space. With the current dominance of the Economic Growth coalition in framing the issues, it is not clear whether nutrition and food security would be prioritized, even with increased policy space to do so. Previous research in South Africa has documented the tendency for relatively minor policy change with respect to the food supply. Even in response to the significant food security and nutrition crisis engendered by the global food price increases of 2007–09, South African food policy focused on household food access rather than changes to the food supply (Kirsten 2012; Watson 2017). This likely reflects the political expediency of maintaining the status quo, as well as the power of food system actors that benefit from the current policy structure. For example, the strong influence of business interests on South African policy priorities that has been documented elsewhere (Kirsten 2012).

This raises the question of what might incentivize increased policy priority for food security and nutrition with respect to the food supply. The research presented here clearly indicated the influence of whole-of-government policy objectives on policies across sectors, and particularly the dominance lent to the economic growth coalition by its clear alignment to the overarching government priorities for employment and economic growth. It is possible that agencies with a whole-of-government mandate, such as the Department of Monitoring, Planning and Evaluation, or the 5-yearly African National Conference, which shapes the programme of work for government (Kirsten 2012), may have an increasing interest in improving policy coherence for food security and nutrition with respect to the food supply. In particular, due to the high social costs of food insecurity and malnutrition and the ineffectiveness of current approaches to address these (evidenced by high rates of food insecurity, undernutrition and NCDs), as well as growing global consensus that food supply-oriented policy is part of an effective response. The recent adoption by the Government of South Africa of a soft drink tax, in the face of strong industry opposition, indicates growing recognition of the importance of – and willingness to take policy action for – a food supply that supports good nutrition.

This analysis indicates that the forums for stakeholder engagement in the food policy subsystem heavily favour industry, suggesting that formal mechanisms for capacitating civil society and promoting its engagement might help to improve policy coherence. Interviewees from the Food Security and Health coalitions identified the need for CSOs to engage in more strategic advocacy for consideration of social, environmental and health issues in food security and nutrition policy making. Improving outcomes for food security and nutrition through increased civil society engagement will require increased capacity for CSO lobbying and communication in the food security and nutrition policy space. The capacity of civil society to support public interest, engage with policy issues, and bring key issues to the attention of policy makers has been identified as a significant facilitator of nutrition policy action globally (Roberto et al. 2015; Huang et al. 2016; Ruckert et al. 2016). Further research is needed to investigate the opportunities and challenges to increasing capacity of civil society actors to support more coherent food security and nutrition policy in South Africa.

Addressing the double burden of malnutrition will require a policy focus on rendering foods of high nutritional quality accessible geographically and financially to consumers across the income spectrum, to complement the current focus on poverty reduction. The dominant framing and beliefs in the Economic Growth and Food Security policy coalitions focus on production of (and access to) sufficient food, but not on nutritional quality. This is a global challenge; there have been repeated calls for food systems that deliver nutritional quality and not simply calories, such that they would achieve food security and nutritional objectives while not discounting other (economically oriented) food supply policy priorities (McDermott et al. 2015). Taking a nutritional quality and food security lens to the existing policy priorities, content and interests regarding the food supply in South Africa in this study identified four specific opportunities to strengthen policy for food security and nutrition. First, the opportunity for specific changes to economic policy relating to the food supply that achieve both food security/nutrition and economic objectives, such as incentivizing small scale producers to produce foods of high nutritional value, that also create employment through their high economic value (such as fruit and vegetables). Second, creating links between producers and consumers, through markets and fiscal incentives, that make healthy and fresh foods more accessible and affordable. Third, increasing formal avenues for engagement by Civil Society in nutrition and food security policy making, as well as avenues for food security and nutrition policy makers to engage in economic policy forums that affect the food supply. Fourth, to include consideration of the nutritional quality of the food supply in policy objectives across sectors, to create a framework for policy coherence across sectors relating to the food supply.

South Africa is a co-chair of the global Sustainable Food Systems Programme, which does not currently address nutrition, but this might afford an opportunity to open a broader dialogue about relevant and appropriate policy objectives to address the pervasive nutrition challenges that South Africa faces. Another opportunity may be strategic use of public procurement. In Brazil, local public procurement for schools has played an important role in promoting food security as well as rural development, including through reduction of the costs associated with long supply chains with multiple actors (Sidaner et al. 2013).

4.3 Limitations of the study

The main limitation of this study is the limited number of interviews conducted. The selection of stakeholders is likely to have shaped the coalitions identified. However, complementing interviews with a systematic review of policy content is likely to have balanced out this risk. Further research in this space would be strengthened by including interviews with cross-cutting agencies such as the Department of Monitoring, Planning and Evaluation and the Competition Commission, with retailers, and with civil society actors. Another limitation is our focus only on national level policy. It is also likely that these kinds of policy incoherence also manifest at the provincial and local level.