Diet-related noncommunicable diseases (NCDs) are responsible for the largest proportion of premature mortality globally [1]. Industries that profit from the sales of health harming commodities drive this NCD crisis [2]. The products and practices of the food industry exert a significant influence on population health and health equity [2]. While the effectiveness of mandatory regulation (e.g., taxation of sugar sweetened beverages) is well established [3, 4], non-binding ‘soft’ modes of regulation, particularly multi-stakeholder platforms and public-private partnerships, have emerged as widespread approaches at multiple levels of governance [5]. Within global governance, the United Nations Sustainable Development Goal 17 calls for the establishment of multistakeholder partnerships [6], with national-level governments launching multistakeholder platforms to tackle a range of climate and health challenges.

While multistakeholderism portrays and promotes collaboration between public and private actors as necessary and transformative in tackling NCDs, the legitimacy and effectiveness of this approach is contested [7, 8]. Although platforms engaging with industry to address issues such as food fortification have been implemented successfully, the wider literature points to stalled progress against the backdrop of increasing NCDs, competing interests and political contestation, particularly where proposed policies that address NCDs threaten core industry interests [2, 7, 8].

At least four factors have been put forward to explain why policy engagement to address diet-related NCDs with the food industry is often conflictual. First, nutrition as a field is multisectoral, including various government policy sectors from industry and trade to health and education, civil society, the food industry (and other associated industries, such as transport or public relations) and sometimes academic organisations [8, 9]. Moreover, it is multidisciplinary: it spreads over disciplines of food science and manufacturing, agriculture, public health, trade, education, economy, etc. [8, 9]. Such diversity inherently increases the likelihood of conflict as the actors often have contrasting worldviews, advance competing interests and promote different approaches to lowering the prevalence of diet-related NCDs [8, 9]. Second, reconciling conflicting interests of public and private actors can prove extremely challenging, given that the core business model of many food companies depends on maximizing profits from ultra-processed foods that are harmful to health [2, 8]. Third, the complexity of global and national nutrition governance makes reaching consensus among a variety of global actors and mechanisms challenging [8, 9]. Four, major power imbalances between private, public, and civil sector actors can render prioritisation of health interests difficult, resulting in a consensus that favours corporate interests and diminishes the health benefits of the collaboration [8, 10, 11]. Hence, policy engagement with the food industry can potentially serve to exacerbate and institutionalise existing power asymmetries [8, 10].

Despite the challenges posed by public private engagement to health equity, there are few indications that this paradigm of multistakeholderism will be challenged, with governments, international organisations, and industry actors remaining committed to collaboration. This reflects the prevailing idea that “the food industry is part of the solution” to solve the NCDs crisis [12]. This might be explained by the influence of large corporations on governments and multilateral organisations [13], the paradigm of ‘soft’ governance that has come to dominate transnational governance [14], and the notion that partnership and engagement between public and private actors can address market inefficiencies [15]. As a result, the value of collaborative approaches is often taken for granted by policy makers as they are perceived to be more effective, less costly and more straightforward to operationalize compared to mandatory regulation [5, 16].

Public private engagement (PPE) – the formal and informal governance processes through which state and non-state actors collectively work to design, promote, maintain and implement regulation – institutionalises the power stakeholders exercise over policy making. According to Moon’s [17] typology, policy actors may draw their power to influence governance from eight sources: economic power, through using material resources; institutional power, through established rules and processes in governance and decision-making; expert power, through being recognised for knowledge or skills; discursive power, by being able to shape the way others think; moral power, by being able to shape others’ moral principles; network power, through being able to harness the collective power of others; structural power, enabled the structures of society (e.g., governments); and physical power, through the use of force. Drawing on Moon’s typology [17], the institutional power granted by a PPE to private actors that may already hold considerable power from other sources (such as the economic power of transnational food manufacturers or the network power enjoyed by industry associations) can solidify these actors’ influence over public policies, which can be used to jeopardise public interests. However, the terms (i.e., conditions and forms) and governance arrangements of PPE will determine the level of institutional power formally granted to stakeholders and the extent to which that power can be exercised to undermine public policy goals in any given PPE [18].

Policy makers working within the prevailing paradigm of multistakeholderism are often confronted by institutional ambiguity about the appropriate terms of engagement with food industry actors. While the idea of partnership is synonymous with engaging the private sector, partnership itself may be used to describe varying types of engagement and does not necessarily reveal much about the involved parties’ respective responsibilities and decision-making authority (i.e., institutional power) [9]. Hawkes and Buse [9] propose to distinguish partnerships by the direction of funding (from private to public, from public to private, or joint funding), and whether decision-making authority is shared among participants or sectors. Others differentiate partnerships by the relationship between public and private actors, suggesting that their processes and outputs may be philanthropic, transactional, or transformational [19]. While this typology can help to understand the inter-organisational dynamics of partnerships, this approach may be less useful practically when proactively designing PPE. These typologies highlight the different forms PPE can take but could be complemented by more direct consideration of governance arrangements that can be embedded in policies from the start to enable government officials to make healthier public policy within multistakeholder approaches.

Tools have been developed to assist governments in considering whether to engage with particular food industry actors in advancing public health objectives. These include the draft World Health Organization approach on the prevention and management of conflicts of interest in the policy development and implementation of nutrition programmes at country level [20], complemented by a simplified triage tool [21]. While these resources suggest conducting a risk assessment for each individual actor, the public health community tends to discuss engagement with the food industry as if industry were a homogenous group of players, without differentiating the range of actors in this space [4]. Although the same groups of industry actors are often found across multistakeholder food platforms (e.g., trade associations, large transnational manufacturers), recognising the heterogeneity among industry actors is important in considering public health outcomes, as different types of actors (e.g. local fruit producers compared to industrial processors) are likely to have different products (and hence externalities), interests, resources, activities, sources and amount power, and thus might aim or able to influence population diet and food regulation differently [4]. This has led to a growing recognition among researchers that more analytical definitions of food industry actor types would help in conceptualising PPEs, which would be of use to policy makers and advocates [4].

In view of this, this paper seeks to conceptualise the forms that public sector engagement with the food industry can take, with particular focus on governance arrangements, taking into consideration the power resources at the disposal of and exercised by different stakeholders in these platforms. It does so by proposing the Governance Typology for Public Private Engagement in the Nutrition Sector, a typology for government-led PPEs established with industry actors to tackle diet-related NCDs. In this way, it seeks to inform policy makers and support them to consider a range of options for engaging on health policy matters with industry under the guise of “partnerships”.

The governance typology for public private engagement in the nutrition sector

The consideration of three domains may help to better conceptualise PPE with the food industry to lower the prevalence of diet-related NCDs: (i) the form of food industry and civil society actor engagement, which shapes the ways institutional power can be exercised by stakeholders; (ii) the type of food industry actors engaged, reflecting pre-existing power attributes and interests; and (iii) the substantive policy focus of engagement, which offers additional considerations about the suitability of different stakeholders in PPEs. These domains are discussed in detail below and summarised in Table 1.

Table 1 The governance typology for public private engagement in the nutrition sector

A recent systematic review [7] assessed a range of PPEs with the food industry to improve population health. Based on this review, we applied the Governance Typology for Public Private Engagement in the Nutrition Sector (Governance Typology of Public Private Engagement, henceforth) to six PPEs that aim to address diet-related NCDs and were ranked as having high quality evaluations published. The domains of the Governance Typology for Public Private Engagement in relation to this select sample of PPEs is discussed below and summarised in Table 2.

Table 2 Analysis of PPEs based on the governance typology for public private engagement in the nutrition sector (MSMEs – micro, small, and medium sized enterprises)Form of food industry and civil society engagement

Existing typologies of PPE concentrate on organisational input and policy output and neglect the processes and procedures of decision-making and policy deliberation inherent in collaboration [19]. We outline two procedural dimensions to classify government-led PPEs based on the form of food industry and civil society engagement: (i) the degree of participation in formal decision-making, indicating the institutional power granted to stakeholders; and (ii) the stage(s) of the policy cycle they are engaged in, defining where and when the granted institutional power can be exercised.

The role of civil society actors is often to promote and safeguard public health interests and/or advocate for other policy outcomes [5, 22]. Considering the common power imbalance between food industry and civil society actors [8, 10], PPE may deepen this imbalance or potentially level the playing field by granting more institutional power to civil society and allowing them to exercise it more widely than industry actors in the policy making process. Therefore, in the design of PPE arrangements it is important to consider the rules on exercising institutional power for both industry and civil society actors.

Degree of participation in formal decision-making

Defining the industry and civil society actors’ degree of participation in formal decision-making helps understand the extent of institutional power the PPE will grant to these stakeholders. We drew from the International Association for Public Participation (IAP2) Spectrum of Public Participation [23] that was developed to assess civic participation in public policy making [23]. Applying the IAP2 framework to PPE is justified in our view as states and international organisations may ‘orchestrate’ governance arrangements with non-state actors, initiating and steering engagement with an industry actor, as the regulated party (or as co-regulator) [5, 21, 24]. The following differentiation of PPEs based on the stakeholders’ degree of participation in formal decision-making expands on Hawkes and Buse’s [9] framework.

According to the Governance Typology of Public Private Engagement, government agencies may engage with food industry and civil society actors in any of the following five ways: (i) information provision (provide information to enable compliance), (ii) consultation (obtain information and feedback on policy options, without the promise of those being consistently considered, and policy makers make all decisions); (iii) direct involvement (stakeholders are directly engaged in discussions with policy makers and their “concerns and aspirations are consistently understood and considered” [23] and systematically incorporated into the policy design, but policy makers make all decisions); (iv) collaboration (stakeholders and policy makers make decisions together); and (v) empowerment (government agencies allow stakeholders to make the decisions on aspects of PPE). At the risk of over-simplifying what may in practice constitute a continuum, in partnerships, the public sector typically offers collaboration or empowers private and civil society actors, while in public policy making, the public sector tends to inform, consult and directly involve other stakeholders.

We can use the above-identified categories to characterise PPEs (Table 2). For example, in the HOME GR/OWN initiative of the City of Milwaukee, United States, micro-, small- and medium-sized enterprises (MSMEs) specialised in urban farming and civil society organisations were directly involved in program development [25]. In contrast, in the case of the United Kingdom’s (UK) Public Health Responsibility Deal, the government and food industry actors collaborated in developing voluntary measures to improve population diet [26, 27]. The food industry actors had significant influence over the scope and content of the Deal due to being granted high levels of institutional power, while civil society actors reportedly had less influence as they received less institutional power by only being directly involved in the development of the voluntary measures [26, 27]. In the case of Argentina’s Less Salt, More Life PPE, food industry actors were empowered as they were granted autonomy to decide which food products to include in product reformulation [28]. As with the Public Health Responsibility Deal, civil society actors wielded less institutional power by being directly involved but not empowered in the way the food industry was [28]. These examples reveal that actors might be engaged in different ways within the same governance arrangement in a PPE due to underlying inter-organisational dynamics. These PPEs also highlight the importance of defining when, where and how non-governmental stakeholders are allowed to exercise institutional power.

Stage of the policy cycle

The terms and governance arrangements of PPE define what part of the policy making process stakeholders can influence through the exercise of their powers. The Governance Typology for Public Private Engagement suggests differentiating the form of food industry engagement according to the stages of the policy cycle where the engagement takes place [29]: (i) policy formulation (i.e., for the purposes of this paper, goal, rule and standard setting, or outlining specific measures for implementation); (ii) implementation; (iii)monitoring; and (iv) evaluation. Involvement in policy formulation entails food industry and civil society actors contributing to the development of the policy content. Taking part in implementation involves these actors having responsibility in executing aspects of the policy. Food industry actors’ participation in monitoring often includes the provision of quantitative or qualitative data on inputs, process and or outputs and outcomes. Civil society may be assigned a role of services delivery or take the opportunity to observe and monitor industry actors’ activities. Undertaking evaluation typically involves industry and/or civil society actors appraising the collaborative efforts in relation to the attainment of the PPE goals (as well as meeting their own interests). The policy formulation phase of PPEs is where the governance arrangements are developed, and this is the stage during which our proposed typology might be best applied. While the policy cycle approach may not reflect the messy realities of policy making, it offers an analytical lens that can help make sense of complex processes and in this case guide policy makers towards a well-considered public health enhancing PPE design.

Food industry interference in agenda setting, policy development and decision-making has been a major barrier in adopting mandatory food regulations in jurisdictions around the globe [2]. Voluntary, collaborative arrangements, such as PPEs, are often advocated for by industry as alternatives to mandatory measures in the agenda setting phase of the policy [7]. Thus, it is likely the case that by the time policy makers decide to engage with industry actors, the influence of these actors has already been exercised. We have not incorporated agenda setting into the typology, as its addition as a separate analytical stage would provide limited value given the aims of the typology – which are to aid in considering different PPE governance arrangements (i.e., after the idea for a PPE is already on the agenda). However, if the Governance Typology for Public Private Engagement was to be developed into a risk assessment tool, the incorporation of the agenda setting phase could raise its analytical value.

Australia’s Food and Health Dialogue provides a good example where food industry and civil society actors were involved in formulating voluntary measures on product reformulation [30, 31]. Industry actors were also responsible for implementation and participating in monitoring by providing self-reports, but they have had no role in evaluating the PPE to date [30, 31]. The UK’s Food Standards Agency sodium reduction strategy was government led, and food industry and civil society actors were consulted as part of policy formulation on technical issues related to the establishment of targets; monitoring combined both industry self-reporting and independent review [32]. Argentina’s Less Salt, More Life and the Milwaukee HOME GR/OWN initiative involved food industry actors in the policy formulation and implementation phases, but not in monitoring or evaluation, while the civil society had a role in policy formulation and monitoring [25, 28]. These examples demonstrate that engaging stakeholders at different stages of policy cycle may limit these actors’ ability to exert institutional power to jeopardise public health goals in PPEs (Table 2). Additionally, these examples reflect the need to consider stakeholders’ level and sources of power before enhancing their influence over PPEs.

Type of food industry actors engaged

Conceptualising the differences between the food industry actor types may help guide consideration about the form of engagement, by gaining a better understanding of how the governance arrangements of PPE enhance or limit pre-existing power structures. Earlier typologies of PPE do not typically incorporate a distinction between industry actor types [9, 19]. Following a pragmatic approach, the Governance Typology for Public Private Engagement suggests classifying food industry actors based on three factors: (i) product portfolios for profit, indicating their interests; (ii) size, indicating economic power; (iii) and function, potentially indicating other sources of power.

Product portfolios for profit

Considering the engaged stakeholders’ interests is important to predict whether the institutional power granted through the PPE might be used to undermine public health goals. The Governance Typology for Public Private Engagement recognises that food industry actors that profit from manufacturing and selling products that are not part of a healthy diet, such as ultra-processed foods and drinks, are likely to have different motivations and interests in participating in a PPE compared to those that profit primarily from the sales of healthy food items. Considering a food industry actor’s product portfolio is vital to any assessment of engagement, as it can forecast the likelihood of consensus on PPE goals and the potential for conflicts of interests.

For example, the HOME GR/OWN initiative engaged MSMEs that profit from selling fresh fruits and vegetables (Table 2) [25]. The aim of the PPE was to increase the consumption of these foods in the local community; therefore, the interests of these MSMEs were aligned with those of the PPE [25]. Thus, the government did not risk public health goals by granting institutional power to these otherwise less powerful food industry actors.

Size

The size of a company may indicate the extent of its economic power (e.g., through their assets and influence over markets and governments). The Governance Typology for Public Private Engagement categorises industry actors’ size based on number of employees and annual revenue, and whether the company pursues its primary activities in more than two countries. Industry actors can be categorised as micro-, small-, and medium-sized enterprises (MSMEs), large local enterprises, and transnational businesses [33]. Small-sized companies have fewer than 50 employees and a turnover of less than USD$10 million [33]. Medium-sized enterprises employ fewer than 250 workers with a turnover of up to USD$52 million, and large companies exceed these numbers [