Introduction

Fifteen years ago, the Framework Convention on Tobacco Control (FCTC)1 became the first global health treaty negotiated under the auspices of WHO, following 4 years of intense negotiations. The development of the FCTC was unusual for its high level of Member State engagement, particularly that of low-income and middle-income countries2 whose influence was seen as a key factor in the strength of the resulting Convention.3 In the context of global governance—where the interests of poorer countries are often overshadowed by those of powerful elites4–6—such an achievement is to be highly prized.

While the FCTC is justly hailed as a public health triumph,7 8 its success is reliant on the strength and coherence of the international movement that gave rise to its existence and that sustains and protects its realisation in the face of persisting challenges.2 This is widely acknowledged with reference to the need for accelerated implementation of core FCTC measures at national level. There is growing evidence that such implementation accelerates tobacco control measures and reduces smoking prevalence.7 9 Yet while the Convention has now been ratified by 180 countries and the European Union,10 only an estimated 10% of the world’s population are protected by best practice in tobacco taxation, with comprehensive advertising bans and smoke-free environments covering just 15% and 20%, respectively.11 Progress in many areas is hampered by competing priorities,12 limited national capacity13 and the active opposition of the industry.14 15

The salience of such political dynamics illustrates that there is nothing inevitable about the success of the FCTC process, highlighting the importance of key governance attributes. Of particular significance here is the Conference of the Parties (COP), the governing body charged to ‘keep under regular review the implementation of the Convention and take the decisions necessary to promote its effective implementation.’1 The COP met for the first time in Geneva in February 2006; subsequent iterations were held in Thailand (COP2, 2007), South Africa (COP3, 2008), Uruguay (COP4, 2010), South Korea (COP5, 2012), Russia (COP6, 2014) and India (COP7, 2016), with COP8 returning to Geneva in October 2018. Among multiple challenges confronting the COP,16 its attributes of participation and representation have become both politicised and increasingly vulnerable in two key respects.

First, the issue of travel support to enable low/middle-income country participation in the COP became highly contested following COP4.16 17 The active leadership role of low-income and lower-middle-income countries, widely seen as critical to the success of FCTC negotiations,2 3 has been underpinned by their receipt of travel support to enable attendance. The implementation of a COP4 decision to adopt a far more restrictive policy (for ‘least developed’ countries) was delayed during COP5-COP7 via successive ad hoc funds, with the Head of the Convention Secretariat warning of ‘a risk that the COP would no longer be quorate’ (p30)18 in their absence.19

Alongside anxiety about a widening democratic deficit in the FCTC, discussions at COP6 and COP7 were characterised by increasing concern about country delegations with tobacco industry affiliations. Potential responses to this issue included consideration of requiring delegates to declare any such links20 21 while civil society participants at COP7 noted an apparent evolution in tobacco industry strategy, with some country delegations obstructively deploying procedural or legalistic arguments to delay discussions.22 The significance afforded by the tobacco industry in seeking to influence the work of the COP was subsequently demonstrated by an internal Philip Morris International (PMI) document obtained by Reuters.23 This framed the FCTC as ‘a regulatory runaway train’ in which health concerns dominate over other interests; the document set out PMI’s aim to ‘(p)ush for a balanced delegation’, by encouraging greater representation of finance, foreign affairs, customs and commerce ministries in national delegations.24

In this context, it becomes increasingly important to understand the dynamics of the COP and to discuss their implications for tobacco control policy. We have therefore built on an earlier analysis of participation at COPs 1–416 by analysing data from COPs 1–7 and combining this with data on the strength of parties’ national tobacco control policies. The aim of this paper is to consider developments in countries’ participation in the COP, including by income group; to examine changes in the composition of delegations (with reference to the proportion of health officials); and to assess the relationship between participation in the COP and the strength of key tobacco control measures at national level.

Methods

Results

Participation in the FCTC COP

While party participation in COP meetings remains extensive, it has declined over time (figure 1). The first COP in 2006 enjoyed very high levels of participation, with over 95% of all eligible countries represented at that meeting. The following 10 years saw a broad decline in party ‘turn-out’ which had dropped to around 75% at COP7 in 2016.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Participation at first seven COP meetings by all parties* and by parties from low-income and lower-middle-income groups†, 2006–2016. *That is, Member States that were parties to the FCTC at the time of the relevant COP meeting. †World Bank income categories exclude some very small states (eg, Cook Islands, Niue) which are therefore not included in calculating participation by income group. COP, Conference of the Parties; FCTC, Framework Convention on Tobacco Control.

Participation in the COP has consistently been higher among low-income and lower-middle-income countries, which typically make up around half the parties represented at the meetings (online supplementary table 2). At COP7, for example, attendance among eligible low-income and lower-middle-income countries was 82%, while among high-income and upper-middle-income countries it was only 73%. Overall, low-income and lower-middle-income countries are more likely to have attended some or all of the seven COP meetings (online supplementary table 3) with only 5.4% never attending or attending only once, compared with the 9% of high-income and middle-income countries that have attended either once or not at all.

While low-income and lower-middle-income countries are more likely to attend the COP, they tend to have much smaller delegations than high-income or upper-middle-income countries. Thus, delegates from low-income and lower-middle-income countries have historically comprised only a quarter to a third of all country delegates (online supplementary table 4), despite making up half the parties represented at the meeting. The proportion of delegates from low-income and lower-middle-income countries has slightly increased over time and was particularly high at the last COP in India (a lower-middle-income country), although all other COP meetings were numerically dominated by delegations from wealthier countries.

We also found considerable variation in the composition of delegations by income group and over time. Delegations from lower income countries are more likely to be drawn from national ministries of health and tobacco control focal points, which together comprise over 60% of delegates from low-income countries; while delegates from non-health departments such as foreign affairs and finance make up a greater proportion of delegations from higher income countries, with health departments making up just 42% of all delegates (figure 2). The proportion of COP delegates drawn from ministries of health (including tobacco control focal points) was particularly low at COP1 in Geneva, reflecting the tendency for many countries to use their standing diplomatic representatives to the United Nations. While the proportion of health-focused delegates has been strikingly higher outside of Geneva, this has tended to decrease over time (table 2); from a high of almost two-thirds of delegates at COP3 (in 2008), to only a third of all meeting delegates at COP7 (in 2016).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Institutional affiliation* of national delegates attending COP meetings (2006–2016) from low-income and high-income countries. *‘Health’, Ministry of Health or equivalent; ‘TC focal point’, national tobacco control focal point (usually within Ministry of Health); ‘Foreign Affairs’, Ministries of Foreign Affairs, Diplomatic Missions or equivalent; ‘Finance’, Ministry of Finance, Treasury or equivalent; ‘Agriculture’, Ministry of Agriculture or equivalent; ‘Trade’, Ministry of Trade, Commerce or equivalent; ‘Other’, all other categories including other government departments (eg, Justice), delegates described by role (‘scientist’, ‘expert’, etc), delegates from non-governmental institutions and delegates with no institutional descriptor. COP, Conference of the Parties.

View this table:

• View inline
• View popup

Table 2

Proportion of delegates from health ministries (including tobacco control focal points) by Conference of the Parties (COP) session and income group

Relationship between participation in COP and strength of tobacco control

We found evidence of a correlation between parties’ attendance at COP meetings and the strength of their national tobacco control policies, as indicated by our three selected measures. The proportion of parties with comprehensive implementation of policies (for tobacco taxation, smoke-free environments and advertising bans) increased with increasing frequency of attendance at the COP meetings (figure 3). This relationship is clearest for taxation and advertising bans (gamma coefficient=0.35 for both), and weakest for smoke-free policies (gamma coefficient=0.29).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Strength of parties’ tobacco control policies (% with minimal, partial and comprehensive implementation) by frequency of Conference of the Parties attendance.

Attendance at each additional COP session (from 1 to 7) was associated with an increased likelihood of parties being in the most comprehensive category of tobacco control, although this was statistically significant only for taxation. For each additional COP meeting attended, the OR for comprehensive implementation was 1.28 (95% CI 1.08 to 1.52) for taxation, 1.19 (95% CI 1.00 to 1.41) for smoke-free policies and 1.24 (95% CI 1.00 to 1.53) for advertising bans.

The strength of this relationship appears to be mediated by income group as well as the specific tobacco control measure examined (online supplementary table 5). When we stratified parties by income, we found that low-income countries showed a particularly strong correlation between COP attendance and implementation of advertising bans (gamma coefficient=0.65). High-income and lower-middle-income countries showed a moderate correlation between COP attendance and implementation of tobacco taxes (coefficient=0.38 and 0.37, respectively), while upper-middle-income countries showed a moderate correlation between COP attendance and implementation of smoke-free environments (coefficient=0.31). We were unable to perform logistic regression by income subgroup as small numbers in some cells meant the model was unstable.

Discussion

Our findings reveal some important trends in relation to FCTC parties’ participation in COP meetings. First, attendance at the COP has declined steadily over time, from 96% at the first COP to 75% at the seventh. Second, participation in COP meetings is consistently higher among low-income and lower-middle-income countries, although these parties tend to have much smaller delegations and are therefore less numerically dominant at COP meetings than high-income and upper-middle-income countries. Third, a greater proportion of delegates from low-income and lower-middle-income countries come from national ministries of health (including national tobacco control focal points), in part reflecting the status of these ministries as recipients of travel support. Overall, the smaller size of delegations from low-income and lower-middle-income countries means that the composition of COP meetings is reflective of more wide-ranging high-income and middle-income delegations, which tend to include substantially more representatives from non-health ministries (particularly foreign affairs and finance).

Participation in the COP is important for maintaining representation and accountability in both formal decision-making terms and with respect to the legitimacy and efficacy of the COP as a negotiating forum. A turn-out rate of 66% is required for the COP to be able to take decisions.35 While attendance from COP1 to COP7 has remained above this level, the general downward trend raises the possibility that turn-out at a future COP meeting could drop below the required threshold for the session to be able to perform its crucial governance functions. In addition to jeopardising the formal requirement for quoracy, the experience of other global governance mechanisms4 suggests that declining attendance could trigger diverse problems including reduced buy-in from disenfranchised Parties, weakened legitimacy and accountability,6 and—ultimately— stalling of progress on implementation. This highlights the scale of potential challenges to the viability of the COP, underpinning the risk for tobacco control of sleepwalking towards chaos in the absence of support that can ensure effective participation.

In relation to participation by low-income and lower-middle-income countries, their strong attendance record highlights their importance in maintaining adequate representation at COP meetings. Participation rates for high-income and upper-middle-income parties have fallen below 66% at three of the seven COP meetings (online supplementary table 2), meaning these meetings have been crucially reliant on much higher levels of attendance among low-income and lower-middle-income countries. The decision to substantially reduce travel support to the majority of these countries28 could therefore have serious consequences, not only for their representation and involvement in decision-making, but for the viability of the entire COP process.

Beyond the basic issues of numbers, it will be important to monitor both participation and discussions emerging from the eighth COP in Geneva in order to address key questions regarding the governance dynamics of tobacco control and global health. It seems almost certain that the proportion of COP delegates drawn from ministries of health will continue to decline, given the understandable pressures on financially straitened governments to rely on diplomatic missions to provide representation. Additionally, the fact that the Illicit Trade Protocol came into effect in 2018 means that COP8 and subsequent COP sessions will be immediately followed by the Meeting of the Parties to the Protocol (MOP). While interested high-income countries will be able to send complex and extensive delegations drawing on expertise from across multiple ministries, countries struggling to send a minimalist delegation might be expected to send an official from, for example, customs and excise rather than health.

It is, of course, highly desirable that the COP (and MOP) should be able to engage with and work across multiple ministries and policy spheres. Yet such a shift also clearly aligns, however inadvertently, with PMI’s desire to ‘(m)ove tobacco issues away from (health)’ by ‘push(ing) for a balanced delegation’, that is, one with much stronger representation from ministries of finance, commerce and industry.24 The tobacco industry’s desire to undermine the pre-eminence of health officials at the COP serves as a reminder that the wider primacy of health goals in addressing tobacco issues within the UN system and global governance is neither inevitable nor unchallenged. WHO did not assume lead status on tobacco issues within the UN until the development of the FCTC2; some agencies, notably the International Labour Organization, have been strikingly reluctant to manage risks of tobacco industry interference in policy36; and it cannot be assumed that governance instruments established to promote health will not be subverted to advance other agendas. The potential for the tobacco industry to appropriate core aspects of the ITP, including by countries adopting the industry-designed product pack marker system CODENTIFY37–39 (now Inexto Suite), emphasises the need to buttress the strength of health-oriented officials in delegations for FCTC negotiations.

This case is reinforced by our analysis demonstrating a broad correlation between parties’ attendance at the COP and the strength of their national tobacco control policies, as indicated by legislation on tobacco taxation, smoke-free public places and bans on advertising, promotion and sponsorship. This finding is perhaps unsurprising, since attendance at the COP and implementation of key FCTC measures will both be linked to a country’s commitment to and capacity for advancing tobacco control16 40; while, less positively, low levels of COP attendance and weak implementation of tobacco control may reflect low political commitment or be linked with higher levels of industry involvement in national policy.15 Yet this correlation also speaks to the value of countries’ involvement in global health governance in terms of democratic participation, policy engagement and capacity building. COP meetings are a forum for exchanging information and experience, promoting best practice and increasing awareness of both threats and opportunities to effective tobacco control,16 and participation can be expected to both strengthen and be strengthened by delegates’ commitment to reducing the future health burden of tobacco.41 This relationship may be particularly important for low-income countries, in whom we found the strongest correlation between COP attendance and implementation of advertising bans (often seen as one of the first steps in a national tobacco control strategy42). Such dynamics have long been viewed as central to the wider process impacts of the FCTC, with information exchange seen as crucial to accelerating policy diffusion.43 In turn this suggests that there is a strong case for treating strong investment in travel support, which may be viewed as something of a bureaucratic burden, as a core investment in global tobacco control; maximising parties’ participation in the COP should be framed as an important mechanism for advancing implementation of the FCTC.

Our study has some limitations, most notably its reliance on official records for details of parties’ participation—which indicate numbers registered as attending, but not the level or quality of their participation. Our classification of delegates’ institutional affiliation is also imperfect, since some delegates are listed in official records without an institutional affiliation, a smaller number are affiliated with non-government agencies such as research institutes, and non-governmental personnel (eg, from non-govermental organisations or academia) occasionally sit in meetings as part of national delegations. Our reliance on World Bank categories of national income meant we were unable to classify some very small states (such as the Cook Islands and Niue) that are not included in this classification. Finally, our assessment of the strength of national tobacco control policy focuses on the presence (or absence) of specific legislation—which will not capture more nuanced aspects such as the extent of enforcement or budgetary commitments to tobacco control. We have not sought to assess other aspects of FCTC implementation in which there may be important distinctions between income groups, including countries’ contributions to treaty funding and their support for establishing a financing mechanism such as a global fund.

More broadly, while our data demonstrate important trends in COP participation they do not speak directly to the reasons for parties’ attendance or non-attendance (eg, how decisions regarding attendance might be affected by the location of a COP, or whether the development of consensus positions at preceding regional meetings might reduce the perceived importance of attending the COP itself). Our results point to the need for more in-depth qualitative research to understand factors influencing parties’ participation in the COP (and Protocol’s MOP) and to identify potential barriers and facilitators. Developing such understanding is critical to realising the potential of the COP as a beacon of effective, democratic global health governance.

In conclusion, our study highlights the importance of maintaining structural support to protect parties’ participation in the COP, particularly that of low-income and lower-middle-income countries. The FCTC has rightly been hailed as a rare example of effective global governance in which low-income and middle-income countries have played a significant role,3 44 but its ongoing success and effectiveness are not a certainty. In many respects, the FCTC remains surprisingly fragile; its financial basis and support structure still require further development.16 Maintaining high levels of representation and participation at the COP requires sustained attention on the policy agenda alongside maximising transparency in delegations and developing a strategic framework for FCTC implementation.