“Private and personal”: Corporate political activity, informal governance, and the undermining of marketing regulation in Brazil

ABSTRACT Few studies have investigated corporate political activity by unhealthy commodity industries in low- and middle-income countries, and the significance of social and political context has been largely neglected. This study aimed to explore the stalled development of marketing restriction policies in Brazil with an analysis of strategies used to undermine the Legal Framework for Early Childhood. Using a constructivist approach based on a typology of corporate political strategies, decision-making processes were assessed to understand interference by food companies in the Legal Framework, and how this was perceived by policy actors. Semi-structured interviews were conducted with public health advocates, academics and legislature officials. Three broad strategies that contributed to the stalling of marketing restrictions in the Legal Framework were identified: relational approaches to policy influence; collective participation in formal decision-making; and specific strategy choices (information and financial incentives). Key opportunities for policy influence through informal social networks in the ‘backstage’ of policy making are found to privilege commercial sector actors. Informal policy making may have a critical function in obstructing the development of health-focused regulation in Brazil. This highlights the need for a better understanding of non-codified and hidden corporate efforts to shape the policy environment.


Introduction
Brazil has been internationally recognised for the significance of its food and nutrition security policies (FAO, 2019;Lutter et al., 2011). The 2014 Brazilian Dietary Guidelines (BDG) have been widely acknowledged as an internationally significant innovation, representing a blueprint for other countries (Monteiro et al., 2015;Swinburn et al., 2019). These were the first such guidelines that used the NOVA classification system, which groups food products according to the degree of industrial processing, ranging from minimally to ultra-processed food and drink products (Monteiro et al., 2015). Yet while there has been considerable progress in reducing stunting, obesity rates are increasingly problematic (FAO, 2019) and require enhanced and expanded regulatory measures, such as marketing regulation.
In 2010, a regulation was published requiring warnings to accompany advertisements for foods and beverages high in fats, sugars, or salt. The Brazilian Health Regulatory Agency (Agência Nacional de Vigilância Sanitária -Anvisa)'s Resolution of the Collegiate Board (Resolução da Diretoria Colegiada -RDC) no. 24 sought to regulate advertising, information and other related practices in the commercial promotion and advertising of such foods and beverages (RDC 24, 2010). This regulation led to Brazil being identified as 'the most advanced country in Latin America with regard to government led approaches' of marketing regulation (PAHO, 2011). However, since 2013, this resolution has been suspended following a judicial decision in response to a request by the Brazilian Food and Industry Association (Associação Brasileira da Indústria de Alimentos -ABIA). ABIA alleged that Anvisa lacked competence to regulate on this issue (Baird, 2016;Santarelli et al., 2018).
Since Anvisa RDC no 24, no other relevant legislation has been approved by Brazil's federal government. Multiple bills have been initiated within the legislature, from both the Chamber of Deputies and Federal Senate. For example, Bill no 5.921 from 2001 was one of the first about children's advertising presented in the Congress and has still not been approved. This illustrates the protracted nature of debates in this policy sphere. According to the Alana Institute, a non-profit organisation that has sought to prohibit marketing communication directed at children since 2005 (Child and Consumption Programme -Programa Criança e Consumo), pressure groups have been acting against regulation or any kind of restriction on children's advertising and have actively participated in shaping the policy process (Instituto ALANA, 2016).
The importance of mechanisms to shield the development of nutrition policy from conflicting commercial interests has received increasing attention (Swinburn et al., 2015;WCRF, 2020). The Lancet Report on the Global Syndemic of Obesity, Undernutrition, and Climate Change attributed international policy inertia to the combined effects of inadequate political leadership and governance, strong opposition to those policies by powerful commercial interests, and a lack of public demand for policy action (Swinburn et al., 2019). In this context, the Lancet report presented the BDG as reflecting a particularly democratic mode of nutrition governance that has 'kept vested interests at bay' (Swinburn et al., 2019). The inability of public health advocates to secure effective marketing regulations, however, suggests the need to re-examine such claims and highlights the need for further research into corporate political activity (CPA). CPA is defined here as 'corporate attempts to shape government policy in ways favourable to the firm' (Hillman et al., 2004).
Hillman and Hitt's taxonomy of CPA is one of the most widely used frameworks for mapping the strategic action of unhealthy commodity industries (Savell et al., 2014). This framework is based on three categories of political activities describing strategies to influence government policy: constituency building, information and financial incentives (Hillman & Hitt, 1999). Information management focuses on influencing policy through the production and dissemination of favourable information (for example, industry-sponsored research, reputation management, and contestation of public health evidence), while financial incentives offer direct or indirect monetary support to political decision makers. Third, constituency building describes attempts to build public support for preferred policies through corporate social responsibility initiatives, building coalitions with civil society actors, and public relations (Hillman & Hitt, 1999). In addition, the authors note that, while corporate actors may pursue a transactional issue-specific approach to influencing government policy, a relational approach to political strategy that builds long-term relationships with policy makers and publics is often preferred, so that 'when public policy issues arise that affect their operations, the contacts and resources needed to influence this policy are already in place' (Hillman & Hitt, 1999).
Public health analyses of CPA have typically itemised diverse tactics used by particular industries, often categorised into broader strategies following or building upon Hillman and Hitt's taxonomy (Mialon & Gomes, 2019;Savell et al., 2014;Ulucanlar et al., 2016). In practice, such accounts often focus on short-term, issue-specific transactional politics in which commercial actors develop strategies that are reactive and seek to counter emergent public health initiatives.
Public health researchers have highlighted the comparative dearth of analyses of CPA of the ultra-processed food products (UPP) industry in Latin America and the Caribbean (LAC) (Mialon & Gomes, 2019), echoing wider limitations within the business studies literature regarding developing and transitional country markets (Gorostidi-Martinez & Zhao, 2017;Londoño-Correa & Garzón, 2016). Brazil constitutes an important context for such analysis, given its status as both a global leader in food and nutrition policies and as much the largest LAC market for the UPP industry (Euromonitor, 2019). Similarly, a focus on marketing is appropriate given its significance to both health policy and to the global food industry, while the fractured and confused context of marketing regulation in Brazil suggests the need to explore the relevance of industry opposition. On one hand, there are rules to kerb the practice of marketing targeted at children, such as the Statute of Children and Adolescents (Estatuto da Criança e do Adolescente -ECA), the Resolution no. 163 of the National Council on the Rights of Children and Adolescents (Conselho Nacional dos Direitos da Criança e do Adolescente -Conanda), the Law no. 11265 that regulates the marketing of food for infants and young children, and provisions in articles of the Brazilian Federal Constitution and the Consumer Defence Code (Código de Defesa do Consumidor -CDC) (Código de Defesa Do Consumidor, 2017;Lei 11.265, 2006;Lei 8.069, 1990;Resolução 163, 2014). Health campaigners have, however, highlighted the ongoing need for a comprehensive law to regulate advertising and promotion and establish norms that would protect children (Instituto Alana, 2016). In this context, the Legal Framework for Early Childhood (LFEC) (Law no. 13,257/2016) constitutes a key initiative that initially included an article that would restrict advertising, an article that was contested and diluted across the approval process (Cedes, 2016). This paper seeks to examine this case to explore political strategies used by commercial actors in undermining this proposed regulatory provision.
The public health literature on CPA (Mialon & Gomes, 2019;Savell et al., 2014;Ulucanlar et al., 2016) has documented specific examples of tactics used by corporations and their allies across geographical contexts and industry sectors and can bring invaluable contributions to developing public health responses to the political influence of unhealthy commodity industries. The emphasis of such work on developing a transferable taxonomy of tactics and on recording examples of reported interference or misconduct can, however, risk neglecting the significance of specific contextual factors in understanding the drivers and outcomes of CPA. Hillman and Hitt acknowledge that 'as institutional arrangement vary by country, so will firms' political actions' (Hillman & Hitt, 1999), yet their taxonomy says relatively little about how the institutional contexts of policy making shape such political activity.
This article addresses this gap in examining the informal politics of legislative decision-making in Brazil and the strategies used by food industry actors to influence policy making. Drawing on institutionalist theories from political science, informal decision-making refers to non-codified rules that 'make available additional channels of influence for political participation beyond formal institutions' (Lauth, 2000, p. 26). In contrast to formal rules and procedures that require formal institutions to be open and transparent, informal decision-making is characterised by a shift into more hidden (or 'backstage') areas (Reh et al., 2013), where 'social relationships and webs of influence play crucial roles' (Harsh et al., 2012, p. 489). As a result, informal politics is structured around 'socially shared rules, unusually unwritten, that are created and enforced outside of officially sanctioned channels' (Helmke & Levitsky, 2004, p. 727), limiting participation to a restricted subset of actors (Reh, 2012). In Brazil, broader analyses of social policy highlight the importance of social networks in understanding policy change (Sugiyama, 2008), with the policy influence of commercial actors seen as shaped by 'systems of informal relationships' characterised by 'the exchange of personal favours' (Bazuchi et al., 2013). This paper therefore aims to examine how policy debates regarding marketing regulation have been shaped by social influence operating outside and between formal negotiations, seeking to contribute to understandings of CPA via an explicit focus on informal governance. It analyses the defeat of a legislative proposal intended to heavily circumscribe advertising to children in Brazil and situates the CPA taxonomy within a context in which policy making is defined by informal relationships and is heavily shaped by social capital.

Methods
This study entailed a policy analysis that combined bibliographical and documentary research with semi-structured interviews. Secondary data included official documentation of the legislative process in the National Congress, such sources included audio records of the sessions, texts of the different versions of the bill (initial, final and modified versions), amendments, rapporteur's opinions, special committees' statements and news about the legal process (Brasil, 2014b(Brasil, , 2014aBrasil, n.d.). Other public sources consulted included official websites of relevant institutions (including the Maria Cecília Souto Vidigal Foundation (FMCSV) and Alana Institute), of politicians (including authors of the legislative drafts, rapporteurs and members of the Congress involved in the discussions) and of key food industry associations (National Industry Confederation (Confederação Nacional da Indústria -CNI) and ABIA) (ABIA, 2014(ABIA, , 2015Cedes, 2016;Fundação Maria Cecilia Souto Vidigal, n.d.;Instituto Alana, 2019;Perondi, 2014;Rede Nacional Primeira Infância, n.d.).
The study is part of a wider research project that explored different aspects of the Brazilian food system and related public policies to address obstacles to healthy eating (Dos Passos et al., 2020;Duran et al., 2019;Jaime et al., 2018;Leite et al., 2020). Interviewees were selected according to their relevance in the policy process and identified via the documentary research and suggestions from other participants. This paper draws particularly on eight interviews: four with civil society representatives, two with advisers from the executive and legislature, and two with academic researchers. While this appears a relatively small sample, it represents high levels of access to the restricted number of individuals participating in the formal and informal politics of legislative decision-making with first-hand knowledge of the LFEC case. Interviewees comprised advocates from civil society organisations that were involved in the case; advisers who worked with prominent parliamentarians; while two interviewees had previously worked with relevant private sector actors.
The interviews were semi-structured, conducted by the lead author in 2017 and 2018. Interviews followed a themed interview schedule, which was elaborated to collect information around this particular case and was focused on private sector involvement in food and nutrition policy. The length of interviews ranged according to the participants availability and varied between 13 and 96 min. When it was necessary to optimise the time of interviews, questions related to the case in question were prioritised. All interviewee signed a written consent form which allowed the interview to be digitally recorded.
The data collected were analysed in line with Hillman and Hitt (1999) (Hillman & Hitt, 1999), seeking to capture data relating to processes of informal governance approach. The research project was approved by the research ethics committee of the Faculty of Public Health of the University of São Paulo (USP), opinion no. 2,338,071.

Results
The legislative proposal in question is first described and then the taxonomy of constituency building, informational and financial strategies within the Brazilian context is situated.
The legal framework for early childhood and the primacy of relational politics in Brazil In December 2013, deputy Osmar Terra, presented bill 6998/2013, which aimed at inserting provisions on the Statute of Children and Adolescents. A Special Committee on Early Childhood was quickly created and tasked with its assessment. In April 2014, Resolution no. 163 of the National Council on the Rights of Children and Adolescents (Conselho Nacional dos Direitos da Criança e do Adolescente -Conanda) was published, under which any marketing communication aimed at children under 12 years of age would be considered abusive (Resolução 163, 2014). Following a series of seminars and public hearings, the bill was broadly approved in December at the 15th meeting of the Special Committee on Early Childhood in the Chamber of Deputies. Importantly, however, the Special Committee rejected three provisions of the bill, with the broad prohibition on advertising to children being jettisoned alongside a proposed increase of maternity leave and a requirement for companies to establish breastfeeding facilities. The rapporteur's analysis of 25 proposed amendments was discussed in a long session, with a break that allowed deputies to discuss key issues including advertising (Cedes, 2016). The rapporteur rejected two amendments relating to the article on advertising, including one which sought to withdraw it entirely. Most of the deputies did approve a final text whereby proposed regulation of advertising was substantially diluted.
In explaining this outcome, interviewees emphasised the strategic actions of commercial actors in building relational, long-term interactions with deputies. One interviewee from civil society, explained that the process of voting was shaped by 'an establishment of a prior relationship, because as it is all very fast there, you have little time to convince a deputy that you have a proper argument'. This account depicted the idea of rational decision-making as a fiction, as often what 'happens there at the time of voting is only a reinforcement'. Rather, the voting behaviour of deputies is seen as shaped by existing relationships created via complex social networks, with decisions to support a particular position being 'given in an emotional and relational way'. This suggests that relationship building by commercial actors both served to secure physical access to key decision-makers and reflected an understanding of unwritten, socially shared rules that could enable policy influence. One civil society advocate emphasised the significance of such relationships in the context of often rapid decision-making at key moments in the legislative process: 'People have difficulty understanding this, that in a space of 20 min or at maximum one hour, everything happens there. (…) The rite of the House is not a rite that gives transparency in the process. It is a rite that is done just for people, only for those who are familiarized, who can quickly react. It is very difficult to understand the process'. (civil society representative) In similar vein, another interviewee situated the relevance of such relationships within links across Brazilian political, economic and social elites. The ability to leverage social networks for political influence is presented as an extension of socialisation from an early age cross schools, universities and professional contexts, since the key actors 'are people who have attended the same places for their whole life'. Importantly, the strategic significance of such relationships is seen as reinforcing inequities and marginalising civil society voices: 'When you look at networks of relationships, of influence of who is there where the information circulates, who has the ability to indicate representatives, there you do not find anyone from civil society, so there will always be a huge discrepancy of power, of possibility of influence in this'. (academic representative)

Categorising specific strategies in shaping informal governance
Recognition of the primacy of the relational approach of commercial sector actors and the significance of informal governance in the case of the LFEC shifts attention towards understanding more long-term and indirect routes to policy influence. While for heuristic purposes our account is organised around Hillman and Hitt's taxonomy, what they present as 'three theoretically distinct strategies' is better understood here as intertwined.

Constituency building
Policy negotiations for the LFEC occurred within a context designed to broaden participation in governance, epitomised by the 2011 creation of the 'Early Childhood Parliamentary Front' in the Chamber of Deputies. This Front included more than 200 parliamentarians from different parties, and was part of a wider coalition, including governmental bodies, corporate foundations, corporations and civil society organisations (Rede Nacional Primeira Infância, n.d.).
Official records of the 15th meeting of the Special Committee on Early Childhood in the Chamber of Deputies only recorded the 22 participating deputies, though it was also attended by representatives from companies, industry associations and civil society organisations (Brasil, 2014b). Hence, there is no record of which actors entered these sessions, nor of the reports, materials or policy briefs distributed to parliamentarians. This lack of transparency about attendees other than deputies is a feature of the sessions in the Chamber of Deputies.
Interview data suggests that participants included representatives of business umbrella groups, such as the National Industry Foundation (Confederação Nacional das Indústrias -CNI), leading producers of UPP (Coca-Cola, Ambev, Nestlé, Mondelēz), and the Brazilian Association of Radio and Television Broadcasters (Associação Brasileira de Emissoras de Rádio e TV -ABERT). Such active engagement by key economic interests had been expected from the outset. According to one parliamentary adviser, 'we would have a tremendous difficulty in getting this bill passed in the Chamber', given the presence of a 'pressure mechanism' and the challenges of conducting debates in plenary when 'this involves important sectors of society'.
While civil society organisations were involved in the policy process, interviewees viewed their influence and effectiveness as limited. The FMCSV was viewed as the most important civil society actor involved, but according to one interviewee, FMCSV 'never got into this fight' to protect ambitious provision for advertising regulation. Created in 1965 by a banking family, FMCSV had focused on early childhood development since 2007 and was a key financial funder of the Parliamentary Front. While presented as a public interest non-governmental organisation (NGO), FMCSV collaborates closely with the commercial sector. Its partners include business interest NGOs and foundations including Danone Early Life Nutrition, FEMSA Foundation (closely related to Coca-Cola) and Lemann Foundation (founded by the owner of AMBEV). FMCSV was described by one participant as having the operational mode of a classic insider protagonist rather than of an advocacy organisation; 'The Foundation has no activist DNA, it has a negotiation DNA, of being a negotiator, sitting at the table and discussing … Dialogue is always the first option.' This preference was seen as typifying Brazilian organisations 'coming from rich families', with a reluctance to confront opposing interests seen as limiting its effectiveness in managing negotiations around advertising regulation.

Information strategy
Predictably, the development of legislative proposals to restrict advertising stimulated a response by the food and beverage industry. Countering the prospect of more stringent regulation of advertising is evident, for example, as a priority in the ABIA's annual reports for 2014 and 2015 (ABIA, 2014(ABIA, , 2015. The LFEC initiative is specifically mentioned in the 2015 report, with its president noting that 'advertising, labelling, nutritional profile and tax issues are increasingly associated with processed food', and ABIA is described as 'strongly engaged in following the discussions on children's advertising' in the National Congress (ABIA, 2015).
More broadly, interview data highlights how ways of targeting policymakers with information were adaptive to the context and practices of informal governance. One interviewee noted limited evidence of lobbying by commercial sector actors during the formal, open aspects of the negotiations, commenting that 'you don't often see lobbyists during public hearings'. Interviewees who attended the 15th meeting of the Special Committee instead emphasised the significance of informal discussions between deputies and industry representatives, operating outside the formal sessions, and parliamentarians' interactions with the commercial sector were seen as less transparent than with civil society. There was a clear sense that detailed, logical or technical arguments were much less significant in shaping outcomes than social capital or political calculation; prior relationships outweighed the specifics of arguments.

Financial incentive strategies
Available data do not suggest that these negotiations were strongly shaped via the direct provision of financial incentives to policymakers. However, official funding declarations indicate that the Deputy who wrote the amendment to withdraw provisions relating to advertising received electoral campaign contributions from a company largely owned by Ambev, the regionally dominant brewer, from Rexam Beverage (a global beverage can maker), and from a pharmaceutical industry association (Brasil, 2010). According to official records of the session, this Deputy made a public speech advocating withdrawal of the article (Brasil, 2014a).
Events in the aftermath of the LFEC decision indicate the significance of building and maintaining reliable long-term relationships with policymakers, and of the relevance of financial resources in doing so. Dozens of the participants in the Parliamentary Front attended an 'Executive Leadership Programme in Early Childhood' held at the Centre on the Developing Child at Harvard University (Cedes, 2016). This programme is a capacity building initiative in which parliamentarians are invited to develop expertise in approaches to advancing policies on early childhood and is run with the support of the Nucleus Science for Childhood, a partnership involving FMCSV alongside INSPER and the University of São Paulo (USP). INSPER is a non-profit teaching and research institution focusing on administration, economics, law and engineering and aiming to develop innovative leaders and professionals (INSPER, n.d.). One of its on-government relationships is directed specifically to private sector representatives. In it, the informality of such work is presented as one of the biggest challenges in Brazil: 'most relationships are private and personal, professionals are not accredited' (INSPER, 2015). Participants in the course are instructed in how to develop a 'permanent dialogue' with the government (INSPER, 2015).

Discussion and conclusions
While Brazil's achievements in food and nutrition policies and contributions to global developments have been notable, landmark initiatives such as the BDG (Monteiro et al., 2015;Swinburn et al., 2019) should not distract attention from ongoing challenges in developing comprehensive and effective nutrition governance. Though the negotiation of the LFEC advanced the prominence of early childhood development on Brazil's policy agenda, the dilution of its innovative and ambitious rights-based approach to marketing regulation provides a powerful lens through which to explore such challenges.
From a food policy perspective, the interest of this case lies in the attempt to pursue health objectives beyond the confines of health policy. The attempt at a wide-ranging approach to protecting children's rights entailed engaging with diverse actors and communities beyond the health sector. In seeking to address political and economic drivers of child health, the case highlights multiple challenges confronting nutrition policy (Balarajan & Reich, 2016). These include difficulties in promoting effective co-ordination across multiple stakeholders with varying priorities and conflicting interests; limitations in leadership and accountability; and highlight the need to invest in developing policy champions and training them to manage the policy process (Balarajan & Reich, 2016).
This initiative is notable for the ambition of its intended comprehensive rights-based approach to nutrition. Such approaches have attracted increasing academic and policy interest, with reports from the Lancet Commission on Obesity (Swinburn et al., 2019) and from the United Nations Children's Fund (UNICEF) (UNICEF, 2018) identifying normative, legal and strategic advantages of engaging with rights conventions and frameworks. While the UNICEF report and the WHO recommendations (WHO, 2012) see comprehensive approaches as encompassing restrictions on marketing to children 'of foods that are high in saturated fats, trans-fatty acids, free sugars or salt' (UNICEF, 2018), advocates for the LFEC sought to regulate marketing to young children as intrinsically and necessarily infringing on the rights of young children. If successful, such an approach could have had implications for health beyond nutrition, including for developing coherent approaches to regulation across unhealthy commodity industries (Collin et al., 2017;Moodie et al., 2013). Importantly, the undermining of the LFEC's provisions highlights the vulnerability of rights-based approaches to challenge when they are seen to impact on powerful commercial interests.
This case study suggests that nutrition governance in Brazil may not have protected against industry interference as effectively as existing accounts suggest. Brazil's formal participatory structures have been presented as facilitating democracy and promoting accountability in nutrition policy, helping to ensure that 'governments and civil society, acting on behalf of public interest, outweigh the interests of the private sector' (Swinburn et al., 2015). While such structures are clearly valuable, awareness of the relational dynamics of informal governance highlights facets of policy making which in practice privilege private sector actors.
The case of the LFEC illustrates the significance of the high levels of social capital and informal mechanisms of access and influence leveraged by the UPP industry, its representatives and allies. In contrast with accounts that emphasise civil society's role in shaping nutrition policy in Brazil (Swinburn et al., 2015;Swinburn et al., 2019), here its impact appears to have been limited. Participation via formal structures in the front stage of policy making can serve to distract attention from the exercise of power and influence that takes place at the hidden margins of those processes. Private sector actors appear much better placed to leverage contacts, networks and influence in the decisive backstage moments, undermining health protections and shaping the outcomes of a protracted policy process. This partly reflects intrinsic political advantages long enjoyed by Brazil's social and economic elites, but inadequate development of formal accountability mechanisms exacerbates the disadvantage experienced by health advocates. In particular, the official record of the Special Committee's deliberations is strikingly limited regarding details of actors attending and in the omission of information, reports and lobbying materials presented to Committee members. Alongside participant accounts that suggest industry representation heavily outnumbers civil society, failures of transparency impede the ability of advocacy actors to exercise key accountability functions (Swinburn et al., 2015), further privileging private sector actors.
This study does have a number of limitations, including those intrinsic to studies of a single case (Merriam, 2009;Yin, 2009). Our focus on the negotiation of the LFEC as an instance of contested policy making has the potential to focus attention on specific political activities undertaken by commercial sector actors, arguably understating the significance of their broader influence as political and economic actors and risking privileging agency over structural power (Farnsworth & Holden, 2006). Equally, our focus on relational approaches to exercising influence may risk neglecting the significance of transactional approaches. This emphasis, however, clearly reflects the evidence provided by interviewees, it highlights important aspects of policy making and influence in Brazil and contributes to the public health literature on CPA that has focused mainly on transactional aspects. Our account of this case is somewhat constrained by a small number of interviewees, though we would contend that our access has been strategic given that the key stages of decision-making involved comparatively few participants, limited documentation and high levels of political sensitivity. The shift of decision-making to informal channels involves intrinsic challenges for research, which we feel have been successfully managed given the range of key actors with whom we secured interviews.
The preceding analysis suggests that a focus on informal governance can enhance understanding of CPA and inform the development of effective public health policy. From a research perspective, this case study highlights the difficulties involved in analysing strategic actions that advance corporate interests behind the scenes, with very limited visibility or transparency, and where companies themselves do not directly perform identifiable actions. Understanding the significance of backstage contexts may be particularly important in the context of governance mechanisms that have strong performative or legitimating functions. This can be particularly important for multi-stakeholder platforms and public-private partnerships, where formal procedures may create the illusion of accountability and legitimacy, downplaying informal processes of decision-making (Ralston, 2018). Understanding the importance of a relational approach to policy influence also has important implications for supporting skills acquisition in advocacy for nutrition (Balarajan & Reich, 2016) and for public health more broadly (Blenner et al., 2017;Gagnon et al., 2017). If lengthy policy making processes can be decisively shaped in a few minutes, outside of formal decision-making, by leveraging social capital and networks, it is important that training in how to develop and utilise such relationships is not confined to the commercial sector.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding statement
This paper is a product of the project 'Improving food systems in Brazil: overcoming obstacles for healthy and sustainable diets' that was funded by the International Development Research Centre: (grant no 108166-001). IDRC had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.