Global governance and the broader determinants of health: A comparative case study of UNDP’s and WTO’s engagement with global health

ABSTRACT This comparative case study investigated how two intergovernmental organisations without formal health mandates – the United Nations Development Programme (UNDP) and the World Trade Organization (WTO) – have engaged with global health issues. Triangulating insights from key institutional documents, ten semi-structured interviews with senior officials, and scholarly books tracing the history of both organisations, the study identified an evolving and broadened engagement with global health issues in UNDP and WTO. Within WTO, the dominant view was that enhancing international trade is instrumental to improving global health, although the need to resolve tensions between public health objectives and WTO agreements was recognised. For UNDP, interviewees reported that the agency gained prominence in global health for its response to HIV/AIDS in the 1990s and early 2000s. Learning from that experience, the agency has evolved and expanded its role in two respects: it has increasingly facilitated processes to provide global normative direction for global health issues such as HIV/AIDS and access to medicines, and it has expanded its focus beyond HIV/AIDS. Overall, the study findings suggest the need for seeking greater integration among international institutions, closing key global institutional gaps, and establishing a shared global institutional space for promoting action on the broader determinants of health.


Introduction
The World Health Organization (WHO) holds the mandate as the 'directing and coordinating authority on international health work' (WHO, 1946). At the level of global governance, efforts to improve global health have primarily focused on strengthening WHO, especially its role in coordinating other actors with mandates focused on global health (Cassels, Kickbusch, Told, & Ghiga, 2014).
Yet WHO is not the only international institution with an impact on global health. A wide range of global actors, both inside and outside the United Nations (UN) system, and the policies and programmes these implement, play an influential role. Working in parallel with WHO, the World Trade Organization (WTO) and the United Nations Development Programme (UNDP) lead global efforts in two other sectors with substantial health impact: those of trade and development, respectively. After WTO's creation in 1995, scholars have shown many connections between public health objectives, and WTO trade law and policy (Andrew, Bouchard, Labonté, & Runnels, 2010). Notable examples include the links between the WTO Agreement on the Trade-Related Aspects of Intellectual Property Rights (TRIPS) and access to essential medicines (Hoen, Berger, Calmy, & Moon, 2011); and the ongoing tensions between WTO law and States' regulatory autonomy to respond to the increasing burden of non-communicable diseases through measures such taxes, restrictions on marketing, product regulation, and labelling (McGrady, 2011).
Governance of the development sector is more diffuse. Many non-governmental and multilateral organisations are active in international development workbroadly understood as efforts to reduce poverty, improve people's well-being, and increase the capacity of economic, political, and social systems to sustain human well-being over the long-term (Soubbotina & Sheram, 2000). At the global level, the UNDP and the World Bank stand out as the global agencies with the greatest influence and the largest portfolio of development programmes, many of which are directed at the broader determinants of health.
The adoption of the UN Sustainable Development Goals (SDGs) further underlines the importance of addressing global health as an intersectoral issue with an emphasis on addressing the broader determinants of healththe economic, environmental, social, and political conditions affecting population health (Buse & Hawkes, 2015;Rasanathan et al., 2015). An area where further research is needed is examining whether and how international actors in other sectors and without explicit health mandates integrate health into their institutional activities (Faid, 2012). This study aimed to address this research question by investigating how UNDP and the WTO have conceptualised their roles in global health, exploring how these institutions' relationship with global health have evolved over time, and identifying barriers and opportunities for collaboration across institutions.

Study design, data collection, and limitations
This study was designed as an exploratory multiple case study (Yin, 2009) where each organisation's engagement with global health was defined as a discrete case study and initially analysed independently before contrasting and synthesising insights from across the cases. Data collection consisted of a review of key institutional documents and a series of semi-structured, in-depth interviews. The document review was completed through a two-staged process. The first stage involved reviewing organisational documents to inform the development of a common interview guide, and to provide background information about both organisations to help probe the interview responses in greater detail. Here the findings from one organisation informed the other so as to formulate questions broad enough to be relevant to both organisations. The second stage traced, analysed, and contrasted organisational views on global health through their official statements and publications, including views and activities (such as statements by the leadership, the production of an independent or joint report, or establishment of a dedicated programme) on specific global health issues. The limitations of using official organisational documents are that these tend to contain insufficient detail to inform the research questions, be biased towards presenting the work of the organisation in positive light, and incompletely reflect the views and contributions of the organisation on the issues studied (Bowen, 2009). The full description of our approach is available in supplementary file S1. Overviews of documents reviewed and the main findings from these are available in supplementary files S2, S3, and S4.
An interview guide with open-ended questions (Table 1) was designed to elicit the experiences and perspectives of UNDP and WTO interviewees. This study did not intend to assess views held by the majority of UNDP and WTO staff. Rather, we wanted to recruit interviewees providing unique experiences and perspectives from the higher levels of leadership in these organisations. Accordingly, potential interviewees were purposively sampled on the basis of their leadership position. The UNDP interviewees were coordinators and directors of thematic clusters within the major technical bureaus of the organisation. The WTO interviewees were staff at the senior counsellor rank and above. 10 interviews were conducted; four with WTO leaders and six with UNDP leaders. The UNDP interviewees had (including country office experience) on average 5.7 years of experience in the organisation (range: 1-11 years), and the WTO interviewees had on average 12.5 years of experience in the organisation (range: 3-28 years). The recruitment process balanced the need to obtain a broad range of perspectives with constraints of time, resources, and access. All invited leaders agreed to participate in this study and interviews were conducted at various times between October 2012 and June 2014. To protect the interviewees from any harm, and to enable them to express the fullest range of views and opinions without constraint, directly identifiable information was eliminated from all audio records and transcriptions.

Data analysis and interpretation
Two investigators (UG and AL) independently analysed the interview transcripts, compared their coding and interpretation, and discussed any reasons for variation. Interview transcripts were analysed using an inductive approach inspired by grounded theory (Yin, 2011). The initial coding was further processed and refined by iterative categorisation of the codes as part of identifying major categories and themes responding to our research questions (Yin, 2011). Finally, a cross-case synthesis was conducted, comparing and contrasting the themes identified in the qualitative data from the UNDP and WTO interviews. The interpretation of interview data was informed by previous work on agency in international organisations, including studies of bureaucratic autonomy of international organisations and applications of principal-agent theory (Bauer & Ege, 2016;Eckhard & Ege, 2016;Elsig, 2011).
The list of codes (including the number of interviewees who raised the issue represented by the code), categories, and themes responding to our research questions are available in supplementary   Table 1. Interview guide used for the semi-structured interviews with senior leaders. Introduction 1. Briefly outline some of the global governance challenges that you encounter in your work. Do these have an impact on global health? How so? Framing Global Health 2. In the day-to-day affairs of your organisation, how is the advancement of global health conceptualised and framed in relation to the organisation's stated primary objectives? [Can you provide any specific examples?] 3. What are the persuading factors that might lead your organisation to consider and engage with health in its day-to-day affairs? Advancing Global Health through Global Governance 4. What steps are taken by your organisation to ensure that health is protected and promoted within its deliberations, policies, and activities? Where and when in the planning, deliberation, and implementation processes of your organisation is health considered? 5. What are the major barriers to collaboration between your organisation and a global governance institution from a different sector, towards the advancement of global health? 6. What are the major enablers to collaboration between your organisation and a global governance institution from a different sector, towards the advancement of global health? 7. Given the major barriers and enablers, and the conceptual and procedural contexts discussed, are there any proposals or solutions you would like to see? file S5. The insight from the interviews and documents were interpreted together with scholarly books that have examined the history of each institution (Murphy, 2006;Skard, 2007;VanGrasstek, 2013). These provided a reference point for situating the narrative of each organisation's relationship with global health issues within a broader historical context. Constant comparison, searching for negative cases, and rival thinking were strategies implemented to minimise biased interpretation of the qualitative data (Yin, 2011).

Ethics, consent, and permissions
Prior to initiating the study, approval for conducting the interviews, and recording and storage of audio was obtained from the Data Protection Official for Research under the Norwegian Centre for Research Data (project no. 31093). An informed consent form was signed by each participant prior to conducting the interviews.

Main findings and discussion
We begin our discussion of the findings with a narration of each organisation's engagement with global health. We then examine the access to medicines issue as a key example to compare and contrast the organisations' experiences with a specific global health issue. Finally, we examine views and perceptions about the state of the global architecture for health. Together, these insights allow for proposals about what international institutions collectively should do to more effectively address the broader determinants of health.

WTO: A battleground between trade policy and public health objectives
The WTO was established on 1 January 1995 as the major reform to the multilateral trading system resulting from the 1986-1994 Uruguay Round of multilateral trade negotiations. The major objective of the Uruguay Round was to broaden the scope of multilateral trade rules to include agriculture, intellectual property, investments, and services. The WTO and a series of other agreements built upon and formally replaced the General Agreement on Trade and Services 1947 (GATT), which had functioned as the set of rules regulating international trade since 1947.
Tensions between WTO's agenda and public health objectives have been under scrutiny for many years (Labonte & Sanger, 2006). When asked to reflect on the relationship between public health and their agency's institutional mandate, WTO interviewees described that 'one of the principle factors in achieving public health is reduction of poverty, and more equitable participation in economic systems' (WTO 1) and that 'the perspective of people in the WTO is we're trying to do things that are ultimately positive and good for public health because it's a part of the development process, part of economic progress … ' (WTO 3). While all WTO interviewees recognised tensions between public health objectives and the WTO's trade liberalisation agenda, it was also expressed that 'since the trade rules were created there has always been an overriding exception for purposes of health protection' (WTO 2) and that 'governments have wide latitude to adopt measures that are necessary to protect public health, even if they create trade barriers' (WTO 3).
Interviewees raised non-discrimination as a fundamental WTO principle for governments to consider when designing public health measures, for example for tobacco control, which has been a prominent public health issue in WTO committees and disputes (Lencucha, Drope, & Labonte, 2016). For example, it was argued that 'overall it should be possible for countries to do anything they want to restrict smoking, to restrict tobacco products, as long as they do it on a non-discriminatory basis' (WTO 3). The principle of non-discrimination is reflected in a number of WTO provisions, including the most-favoured-nation treatment and national treatment clauses that in different forms exist in GATT 1994 and other WTO agreements (GATT, 1994). Most-favoured-nation treatment entails not discriminating between trading partners (for example lowering tariffs for the same goods from one country, but not another), while national treatment involves treating imported and locally produced goods, services, and intellectual property equally.
The case brought by the US against Thailand under GATT 1947 over Thailand's tobacco import restrictions was the first international trade dispute involving tobacco control. It is considered 'a landmark dispute at the intersection of trade and health' (Drope & Lencucha, 2014). While this specific dispute preceded the establishment of WTO, it sheds light on several aspects of the interface between health and trade, including the application of the principle of non-discrimination in a trade dispute involving a public health issue, what role WHO and the use of scientific evidence supporting health norms can play in WTO proceedings, and how the WTO and WHO secretariat have interacted about the lessons learned from the dispute.
In this specific case, the United States Trade Representative threatened trade sanctions, protesting that Thailand's anti-tobacco policy discriminated between domestic and foreign products by permitting domestic cigarettes produced under the Thailand Tobacco Monopoly. The dispute was ultimately referred under the GATT for arbitration in 1990, where the GATT panel ultimately judged that Thailand's import ban violated the principle of non-discrimination and could not be justified on health grounds, since less trade-restrictive measures for tobacco control were available. An interesting feature of the case was the role played by WHO during the proceedings. At Thailand's request, WHO submitted its own views on the issue. WHO supported a number of Thailand's arguments in favour of an import banincluding that a 'major difference between manufacturers of American cigarettes and of Thai cigarettes was that the former designed special brands aimed at the female market', and that opening Thailand's market to multinational companies would likely increase tobacco consumption. The US challenged WHO on this latter point, arguing WHO not to be specially competent to address the 'health consequences of the opening of the market for cigarettes' (WTO, 1990). However, the substantial input by WHO was referenced in multiple sections of the final GATT report, and is considered among the first examples of an attempt to integrate public health norms within an international trade policy context (Drope & Lencucha, 2014).
Another example of how tensions between trade and health interests have emerged in the WTO system is when industry or governments promoting industry interests have attempted to use the WTO and its rules to exercise influence over intergovernmental negotiations on global health issues. At the time when WHO's Tobacco-Free Initiative facilitated the drafting of the Framework Convention on Tobacco Control, British American Tobacco sent letters to then WTO Director-General Mike Moore. This was an attempt to lobby the WTO to exercise resistance to any efforts by WHO to make tobacco products exempt from benefits of the multilateral trade agreements (Mamudu, Hammond, & Glantz, 2011;Shears, 1999). In this particular case, Moore expressed that the WTO Secretariat had briefed members of WHO's Inter-Agency Task Force on Tobacco Control, but that the 'WTO Secretariat has no role in the negotiation of the FCTC, which is a matter for WHO member states' (Moore, 2000).
This presents an interesting question, namely the role played by the WTO Secretariat amidst these tensions. Interviewees argued that it is primarily up to members to resolve tensions between international trade policy and public health, and described the WTO as 'a contractual arrangement among sovereign states' (WTO 3) without the authority nor the capacity to advise states about how to take into account public health considerations: If you're talking about a process of trade negotiations, that is driven directly by the governments concerned, and so it's essentially their collective choice as to how to take into account public health considerations. (WTO 1) We can't do that … .We don't have the capacity, we don't have the knowledge to do a health impact assessment. We always describe the WTO as a member driven organisation. We cannot as a secretariat propose that we set up even recommendations for members to do a health assessment of different measures. (WTO 2) From this perspective, the main role of the WTO Secretariat is to manage and inform states about the inter-governmental process, but not provide states with independent advice shaping their decisionmaking. Accordingly, WTO's role should be understood within a principal-agent framework, where members as a group of principals mandate WTO to fulfil a limited set of tasks and ensure that the Secretariat does not over-step its mandate (Elsig, 2011).
Yet studies on the bureaucratic autonomy of international organisations have suggested that the WTO Secretariat may indeed exercise some level of autonomy (Bauer & Ege, 2016;Jinnah, 2010;Trondal, 2016). This may particularly be true when the actions are consistent with the values embedded in the organisational culture and the ideology guiding the organisation's objectives, which in WTO is dominated by neoliberal free trade theory (Chorev & Babb, 2009;Lera St Clair, 2004). For example, Bøås and Vevatne (2004) have described how WTO adapted the interpretation of sustainable development so as to reinforce trade liberalisation as an important policy instrument for achieving sustainable development.
With respect to global health, one example where the WTO Secretariat appears to have exercised its autonomy to gain influence is the first joint WHO/WTO Study on Trade and Public Health. This report assessed relevant WTO agreements and the ways in which these influenced public health and health policies (WHO & WTO, 2002). The tobacco dispute between Thailand and the US described above was among the cases reviewed. The joint WHO/WTO study acknowledged that market entry of multinational tobacco producers initially may increase tobacco consumption, but credited the GATT decision for supporting and encouraging domestic tobacco control policies, such as advertising bans, as long as these were implemented in a non-discriminatory manner. However, the joint study failed to mention WHO's support for Thailand's argument that multinational tobacco companies were equipped with techniques to circumvent such advertising bans. For this and other reasons, critics have argued that the report reflected a compromise between WHO and WTO (Lee, Sridhar, & Patel, 2009), with Howse (2004 going so far as to describe the report 'a co-option of the WHO by the WTO, rather than being the product of genuine dialogue between the two organisations'. Overall, specific trade concerns raised in response to public health measures undertaken by governments are increasing in the WTO system (Lencucha et al., 2016;McGrady & Jones, 2013;Thow, Jones, Hawkes, Ali, & Labonté, 2017). For example, in 2014, the WTO Committee on Technical Barriers to Trade reported health protection and labelling for food and drink to be 'emerging as a dominant theme in many of the "specific trade concerns" raised by members' (WTO, 2014). As states increasingly adopt stronger tobacco control measures and additionally use lessons learned from tobacco control to address other risk factors for NCDs, such as diet and alcohol, the need to address tensions between public health measures and WTO obligations is expected to continue. Interestingly, studies indicate that official WTO bodies too (e.g. the WTO dispute panel and Appellate Body) are increasingly considering prevailing public health norms during judicial interpretation of WTO law (Drope & Lencucha, 2014;McGrady & Jones, 2013). For example, scholars note how the WTO panel in a more recent dispute between the U.S. and Indonesia over the U.S. ban on clove cigarettes in its report drew heavily on the WHO FCTC Partial Guidelines on Articles 9 and 10 as 'a growing consensus within the international community to strengthen tobacco-control policies through regulation of the content of tobacco products' (Drope & Lencucha, 2014;McGrady & Jones, 2013;WTO, 2017a).
UNDP: Emergence, evolution, and expansion as a global health actor UNDP was established in 1966 as part of a merger between two UN programmes dedicated to international developmentthe Expanded Programme of Technical Assistance (EPTA) and the UN Special Fund. The major goal was to establish a 'central funding and coordinating organisation for technical assistance from the UN' (Skard, 2007). UNDP is a programme within the UN system (similar to UNICEF, UNAIDS and UNFPA), and not a specialised agency such as the WHO. The main distinction has been that programmes are financed by voluntary rather than assessed contributions; however, this distinction has become less clear in recent times, as voluntary contributions make up a greater share of the budgets of most specialised UN agencies.
Historically, UNDP's engagement with health issues was focused on supporting disease-specific programmes, such as those against river blindness and guinea worm in the 1980s (Murphy, 2006). The HIV/AIDS crisis during the late 1980s and 1990s broadened the organisation's engagement with global health. WHO's Global Programme on AIDS (WHO-GPA), under the leadership of its first director Dr. Jonathan Mann, had initially been credited with adopting a broad, human rights approach to the HIV/AIDS crisis, and recognising the social, economic, demographic, cultural, and political nature of the challenge (Hilts, 1990). In 1990, Mann resigned as director of WHO-GPA. Among the major reasons were disagreements with then WHO Director-General Dr. Hiroshi Nakajima about 'the importance and strategy of the fight against AIDS' (Hilts, 1990), and that Nakajima's predominant biomedical approach to AIDS conflicted with Mann's broader approach (Leon, 2011;Merson & Inrig, 2018). Over the next years, concerns were increasingly raised about WHO-GPA 'taking too medical an approach to the AIDS epidemic' (Knight, 2008). UNDP was a prominent critic of WHO's response to HIV/AIDS, and the intense confrontation between UNDP and WHO over the response to the HIV/AIDS epidemic has been studied in greater detail by others (Balter, 1994;Leon, 2011). Against this background, a defining moment in UNDP's engagement with global health was the founding of UNDP's own programme on HIV and Development to address HIV/ AIDS as a multi-sectoral, human rights issue under the leadership of Elizabeth Reid. Shortly after, the establishment of UNAIDS signalled that WHO-GPA represented an insufficient response to the HIV/AIDS crisis, and that there was a need for an inter-agency entity that 'had the ownership of a broad set of UN agencies' (Knight, 2008). UNDP was later designated the lead organisation for 'HIV/AIDS, development, governance and mainstreaming, including instruments such as PRSPs, and enabling legislation, human rights and gender' (UNAIDS, 2005). By becoming a focal point for technical support to governments on these areas, UNDP's role was further reinforced. Interviewees underscored UNDP as leader of a broader responsegoing beyond medical interventionsto address the HIV/AIDS crisis: … our contributions have been particular to what UNDP brings to the table, particularly when it comes to capacity development going beyond health sector interventions, looking at integration across sectors, and the governance dimension and the human rights areas. So, we've helped ground the HIV/AIDS response in a broader policy setting, which is a precondition for it to be effective … . (UNDP 1) All UNDP interviewees framed the organisation's response to global health issues in terms of the broader determinants of health. One interviewee explicitly contrasted UNDP's approach against a health care-focused approach to health by expressing that 'our approach is around social determinants which is really anything that doctors and nurses don't do' (UNDP 2).
It can be argued that UNDP, through lessons from its work with HIV/AIDS, has broadened its engagement with global health in two respects. First, in addition to its on-the-ground support to countries, UNDP is increasingly facilitating processes to provide global normative direction for addressing global health issues. One prominent example is the Global Commission on HIV and the Law, convened by UNDP in 2010 to 'examine the relationship between the legal responses, human rights, and HIV' (El Feki et al., 2014). While the Commission's report officially is independent from UNDP, statements from UNDP's previous Administrator Helen Clark and publications from UNDP staff suggest strong support for the report's recommendations (Clark, 2013b;El Feki et al., 2014). This support was also reflected in the way one interviewee described the work of the Commission: … we made a significant contribution along those lines with this Commission on HIV and the Law … we basically convened the best expertise in the world and policy makers around the agenda of HIV and the law in order to really get to the issues of discrimination and exclusion … when we bring our evidence to the table, others bring their evidence and then out of this comes a new political high level policy agenda on how to move forward on these issue. And it has implications hopefully to global governance on a normative level … . (UNDP 1) Second, UNDP has begun to broaden its focus to issues beyond HIV/AIDS. At the time of the interviews, this intent was communicated by reports and policy documents discussing the importance of tobacco control and addressing the social determinants of NCDs (UNDP, 2013(UNDP, , 2014. Interviewees confirmed that UNDP was amidst a process drawing lessons from their HIV/AIDS response in order to 'move forward better in terms of dealing with the broader challenges when it comes to health' (UNDP 1). As one interviewee explained: So we're now embarking on a process where it's not one particular retrovirus which is the only issue on the planet … it's how do we incorporate a lot of the lessons learned from the HIV response into broader health governance issues … we've done a lot of work on the legal environments and HIV which we are going to look at other diseases as well. We're looking a lot at emerging pandemics like non-communicable diseases in particular and looking at the governance of some of the risk factors. (UNDP 2) These ideas are encapsulated in UNDP's most recent strategy on HIV, health and development (UNDP, 2016), which consolidates various strands of UNDP's work on global health, focusing on three areas: (1) reducing inequalities and the social exclusion that drives HIV and poor health; (2) promoting effective and inclusive governance for health; and (3) building resilient and sustainable systems for health. Overall, UNDP's own framing of global health in terms of the broader determinants of health is consistent with the organisation's human development approach to international development which has dominated the organisation since the 1990s (Lera St Clair, 2004).

Contrasting experiences and ideological differences: Access to anti-retroviral medicines
The HIV/AIDS epidemic became a pivotal global health issue for both UNDP's and WTO's engagement with global health. As described above, HIV/AIDS became an entry point for UNDP's broader engagement in global health. For WTO, the implementation of the TRIPS agreement coincided with the emergence of novel HIV/AIDS drugs in 1996. At the turn of the millennium, the majority of the estimated 24.5 million people living with HIV in sub-Saharan Africa did not have access to anti-retroviral drugs (UNAIDS, 2000). Lack of access was largely attributed to the TRIPS agreement and originator companies who priced the drugs out of reach for patients, governments, and non-governmental organisations in low-and middle-income countries (Hoen et al., 2011). One WTO director described access to anti-retroviral drugs as an issue that defined how the WTO was perceived externally: In some ways I think the WTO kind of took the hit, people were looking for somebody to blame for this huge epidemic and you know, the IP providers were kind of an easy target. I think they [pharmaceutical companies] opened themselves up to it by being incredibly narrow minded about how they dealt with it. (WTO 3) One example of the response by the WTO Secretariat during the height of these tensions was former WTO Director-General Mike Moore's article 'Yes, Drugs for the Poorand Patents as Well'. While acknowledging that high costs contributed to the lack of access to HIV/AIDS medicines, Moore mainly defended the TRIPS agreement, arguing that 'were it not for a patent system that rewards companies for risking millions on research, anti-AIDS drugs would not exist', and that 'the TRIPS agreement tries to strike a healthy balance between the short-term need to make vital drugs available to those who need them and the long-term, equally vital need to encourage research into new drugs' (Moore, 2001).
WTO's major institutional response was in 2001, when its members adopted the Doha Declaration and affirmed the right to prioritise public health interests and increase access through compulsory generic licensing of patented AIDS drugs if needed (WTO, 2001). In 2005, WTO members unanimously adopted a protocol amending the TRIPS agreement so as to enable developing and least-developed countries without the capacity to produce generic drugs to get access to medicines from third country producers under compulsory licensing arrangements. This amendment, which entered into force in 2017, is the first amendment of WTO accords since the organisation was established in 1995 (WTO, 2017b). The on-the-ground impact of the Doha Declaration and the TRIPS amendment is a topic of much debate (Beall, Kuhn, & Ford, 2012;Correa & Matthews, 2011). However, one WTO interviewee cited the declaration and the amendment as a reflection of the WTO system's willingness to give priority to public health interests: This was a conscious choice by trade ministers to articulate public health as a priority and to do more than say it's a desirable policy goal, to actually require specific steps, specific legal readings of treaties to serve that objective … . (WTO 1) This in turn may have enabled the WTO Secretariat to exercise greater autonomy and discretion than in other domains, engaging more actively with the access to medicines issue. To this end, it was expressed that the relationship between intellectual property and access to medicines remains an issue where involvement from WTO is essential: the relationship between trade and health and intellectual property protection, that's a unique area where I think clearly the WTO has gotten involved and it has to be involved. We have to deal with that trade off and we can't really duck it. (WTO 3) In comparison, UNDP has from the beginning of its engagement advanced a pro-access stancefocusing on how global intellectual property rules can be managed to maximise access to patented drugsand coming down firmly in favour of a human rights approach: I think the way we approach the issue is that the right to health is a right that is enshrined in many treaties … A core component of the right to health is the ability to obtain affordable and accessible essential medicines. And, one of the factors that determines the availability of medicines is price. One of the factors, probably one of the key factors that impacts price is intellectual property. (UNDP 3) Consistent with broadening its scope beyond HIV/AIDS, UNDP has issued several discussion papers focusing on the use of TRIPS flexibilities to increase access to drugs for NCDs (Correa & Matthews, 2011;UNDP, 2013). UNDP's increasing engagement and normative leadership on this issue was also seen in initiating and hosting the secretariat for the recent High-Level Panel on Access to Medicines (UN, 2016). Comparing itself to WTO and other international institutions, UNDP interviewees argued that access to medicines is being approached from very different normative frameworks: There are downright ideological differences on the role of IP. So, you know, you have WTO and WIPO on the one extreme I'd say, WHO somewhere in between and then more towards the centre you probably have UNCTAD and slightly to the left of that you have UNAIDS and I'd like to put us left of the UNAIDS. (UNDP 3) They don't sometimes understand. Dealing with IMF for example, when we say that there has to be policies, maybe in terms of subsidies in a particular country. So that people can have access to those drugs, whether it is HIV/AIDS, whether it is NCDs, but IMF sometimes takes the position "no", because that would be interfering with the market. (UNDP 4) Yet claiming that WTO is completely disengaged from the human rights framework with respect to access to medicines may not be accurate. A simple comparison suggests that this relationship is one that is gradually evolving. In 2002, the first joint WHO/WTO study on WTO agreements and public health made no specific references to human rights. In contrast, the trilateral study published by WHO, WTO, and WIPO in 2013 presented the human rights framework as an 'important mechanism to further the public health policy goals of ensuring and improving access to medicines for those who are most in need' (WHO, WTO, & WIPO, 2013). This development caused one NGO to declare that 'the trilateral report's references to the obligations to safeguard the right to health in the context of public health and trade rules would have been unthinkable 10 years ago' (KEI, 2013). Consistent with this view, one WTO interviewee expressed that the 'WTO system is evolving more and more toward fitting free trade into a more mature social framework, framework of kind of social responsibility, social rights, social protections', and that this development is 'sort of inevitable in the future, but it's going to take time' (WTO 3).

Specialised mandates and sectoral interests continue to represent barriers to effective intersectoral collaboration
Interviewees from both organisations described the need to improve their own sectors' links with other sectors. However, interviewees raised that narrowly defined mandates and sectoral interests operate as barriers to effective intersectoral collaboration. These barriers could partially be traced historically to the establishment of these international organisations: You got an architecture of the United Nations which differentiates responsibilities and this was [established] during a period of time when differentiation and responsibility was the ultimate definition of efficiency. Over time that doesn't work because when you've only got verticalisation and non-horizontal, you miss out on what we now euphemistically call synergies. (UNDP 2) In general, WTO interviewees were comfortable engaging with global health through collaboration with WHO, and argued that effective communication should be able to resolve most challenges: I think the system that does exist of UN and other specialised agencies with their distinct mandates is probably [the] most effective as long as you have very good collaboration and communication between them. When you don'tthere's where there's problems. (WTO 2) In contrast, UNDP interviewees were more vocal about the need for improved space and mechanisms for institutions outside of WHO to engage with global health issues. One interviewee noted that the struggle we have is that we have an intellectual discussion which recognises the roles that culture, economy, politics and poverty, and all these things play in the wellbeing of individuals, but we don't have an architecture that will allow that discussion to have a meaningful operational life. (UNDP 2) While respecting WHO's mandate in global health, interviewees expressed that the forums convened by WHO were alone insufficient to bring all the needed actors and sectors together. One interviewee expressed that the current global architecture for health did not allow for a conversation about broader factors affecting health issues, thereby also neglecting the potentially meaningful engagement on global health by other international institutions than WHO: … we don't have mechanisms for that, there is no authoritative place for these kinds of discussions today, except in the WHO, but that is the medical profession basically running it, I mean health ministries and medical people. So we lose out tremendously there. (UNDP 1) Accordingly, the interviewee called for an 'an effective mechanism for leadership on the health agenda in the future' (UNDP 1). At the same time, concerns were raised about whether UNDP's expanded engagement with global health could be seen as encroaching upon WHO's institutional mandate, thereby risking unhelpful competition where the institution aim to ensure that their own expertise gain prominence over that of others: there is a need for a broader perspective and so WHO needs to think beyond its own remit and not kind of think that they have to own this and dominate this, and feel insecure if other actors come in on the health issue. But, likewise, we need to also lower our guard and be willing to work with others, not thinking first and foremost about what's in it for UNDP. (UNDP 1) Overarching conclusions: Towards integration and closing institutional gaps?
Similar to climate change, food security, and other issues where organisations have partially overlapping mandates and membership, global health, too, is in short of an integrated, comprehensive regime to govern its affairs. Instead, the institutions that have a direct and indirect impact on global health interact within a regime complex, which entails that different institutions are loosely coupled and work in parallel without a hierarchical architecture to set rules and structure their interaction (Fidler, 2007;Keohane & Victor, 2011;Margulis, 2013). In recent years a number of scholars have proposed mechanisms for strengthening collaboration on global health across sectors, including establishing more formal linkages between existing institutions (Lee et al., 2009), or creating new institutions (Nordström, 2014). Our study informs this debate by exploring how two international institutions without a formal institutional mandate on health -WTO and UNDPhave engaged with global health issues.
Overall, an evolving and broadened engagement with global health issues can be observed in both WTO and UNDP. Key differences include the scope and activeness of the Secretariat's engagement, and the extent to which their interaction with global health rely on WHO. In WTO's context, specific trade concerns raised over public health measures have been increasing over the past 5-10 years (McGrady, 2011;Thow et al., 2017;WTO, 2014), and the need for resolving tensions between public health measures and trade commitments is expected to continue. Our study indicates support among leaders in the WTO Secretariat for considering international trade commitments in a broader policy context. While the WTO Secretariat has no formal influence over the decision-making processes of WTO's official bodies, it can through institutional memory, knowledge-brokering, networking with governments and NGOs, and its role as administrators of WTO processes (Jinnah, 2010;Nordström, 2005) frame the discourse on trade and health in WTO, and support efforts to resolve tensions between health norms and trade policy objectives. However, the Secretariat perceives a limited role for itself on these matters except for increasing cooperation with WHO. Accordingly, and as argued previously by other scholars (Lee et al., 2009;Liberman & Mitchell, 2010), more effective participation by WHO in trade issues and stronger inter-institutional cooperation between WTO and WHO might be strategies for deeper internalisation of health-related norms and standards in WTO including through joint reports, workshops or inter-agency taskforces, or more formal joint committees on trade and health (Liberman & Mitchell, 2010). To this end, WHO has taken its own steps to strengthen capacity on these issues by, among other things, employing trade law experts and facilitating workshops to examine issues at the nexus of trade and health (SPC, C-POND, UNDP Pacific Centre, & WHO, 2013). Finally, with respect to making progress on promoting a regulatory space for governments to pursue public health objectives, it is important to pay attention to the shift away from WTO to regional and bilateral trade and investment treaties. Much concern has been raised over the investor-state dispute settlement provisions associated with these agreements, enabling foreign investors to challenge legislation and regulation by governments to promote environmental and health objectives (Lencucha, 2017;McNeill et al., 2017).
In comparison to WTO, UNDP has taken a more independent approach with the aim of closing what it perceives to be an institutional gap in global health. The organisation has been vocal about the valuable role it can play (Clark, 2013a(Clark, , 2013b(Clark, , 2015, with the recent UNDP health strategy reinforcing the organisation's emphasis on 'addressing the social, economic and environmental determinants of health' (UNDP, 2016). One explanatory factor for the broadening of UNDP's role can be what is claimed to be WHO's seemingly insufficient engagement with the broader determinants of health (Gopinathan et al., 2015;Hoffman & Røttingen, 2014;Sridhar & Gostin, 2011). A key question for the future is whether the relationship between UNDP's expanding role will complement or compete against the authority of WHO, which too has the social, economic and environmental determinants of health as a core mandate (WHO, 2017). Could there be the potential for both institutions to engage, leveraging their comparative advantages, without fragmentation or duplication, particularly when it comes to normative positions and technical advice following on from those? To answer this question, a priority for the new WHO Director-General Tedros Adhanom Ghebreyesus should be to look beyond the historically siloed architecture of the UN and other specialised agencies, and engage the many actors beyond the health sector to reexamine what the WHO mandate as the 'directing and coordinating authority on international health work' means in the era of the sustainable development agenda. Reviewing the experiences from various UN interagency arrangements on health issues, such as the Interagency Task Forces on tobacco control and the prevention and control of NCDs, can be a starting point for examining the scope for more systematic intersectoral collaboration on global health issues (WHO, 2018a(WHO, , 2018b. The SDGs call for an integrated approach, which suggests the need for increasing engagement among global institutions beyond WHO when addressing issues of significant importance for global health (Buse & Hawkes, 2015). Yet there currently exists no permanent global institutional space to more systematically address global health issues that cut across sectors, and to consider tensions and synergies between global health objectives (including SDGs), and the objectives and interests of other sectors. Seeking greater integration among international institutions, closing key global institutional gaps, and establishing a shared global institutional space for addressing the broader determinants of health is needed to support states in achieving the objectives set by SDG3 on health.