Comparing global alcohol and tobacco control efforts: network formation and evolution in international health governance

Abstract

Smoking and drinking constitute two risk factors contributing to the rising burden of non-communicable diseases in low- and middle-income countries. Both issues have gained increased international attention, but tobacco control has made more sustained progress in terms of international and domestic policy commitments, resources dedicated to reducing harm, and reduction of tobacco use in many high-income countries. The research presented here offers insights into why risk factors with comparable levels of harm experience different trajectories of global attention. The analysis focuses particular attention on the role of dedicated global health networks composed of individuals and organizations producing research and engaging in advocacy on a given health problem. Variation in issue characteristics and the policy environment shape the opportunities and challenges of global health networks focused on reducing the burden of disease. What sets the tobacco case apart was the ability of tobacco control advocates to create and maintain a consensus on policy solutions, expand their reach in low- and middle-income countries and combine evidence-based research with advocacy reaching beyond the public health-centered focus of the core network. In contrast, a similar network in the alcohol case struggled with expanding its reach and has yet to overcome divisions based on competing problem definitions and solutions to alcohol harm. The tobacco control network evolved from a group of dedicated individuals to a global coalition of membership-based organizations, whereas the alcohol control network remains at the stage of a collection of dedicated and like-minded individuals.

KEY MESSAGES.

• Tobacco and alcohol are two risk factors contributing to the rising burden of non-communicable diseases (NCDs) in low- and middle-income countries.

• Global attention to both risk factors has increased during the past decades, but tobacco control has made more significant gains with regard to policy formation, resource acquisition and reducing harm, especially across a number of industrialized countries.

• These differences in attention and progress in addressing harm are driven by interactions between issue characteristics, global health networks and the policy environment. The tobacco control network has been more effective in creating and maintaining wide-spread consensus about effective policies to harm reduction, expanding its reach in low- and middle-income countries, and combining evidence-based research with effective advocacy at the highest levels of the World Health Organization.

Introduction

Tobacco and alcohol use, alongside unhealthy diet and physical inactivity, represent two risk factors contributing to the rapidly rising burden of non-communicable diseases (NCDs) in low- and middle-income countries (World Economic Forum and Harvard School of Public Health 2011; Institute for Health Metrics and Evaluation 2013). The rise of NCDs not only increases demands on health care systems but also negatively affects economic development and growth. Effective responses to these mounting health challenges rely on the generation of knowledge about harm caused, policies that address those health issues and the ways in which such policies are generated and diffused globally and domestically.

During the past decades, more rapid progress has been made in addressing harm caused by smoking, while efforts to address alcohol harm have lagged behind. At the global level, the 2003 Framework Convention on Tobacco Control (FCTC) (Yach 2014) establishes legally binding obligations on its signatories, while it took until 2010 for the adoption of a non-binding Global Strategy to reduce harmful use of alcohol. In most industrialized countries, smoking prevalence has decreased sharply since the 1960s, but alcohol consumption remains steady or increased across most regions of the world (World Health Organization 2011: 8). Most importantly, decades ago both smoking and drinking were socially accepted in developed countries, industry interests were powerful, and few to no resources were dedicated to addressing the significant harm caused by both risk factors. Today, the tobacco industry stands out as a pariah in the commercial sector, while the alcohol industry is still viewed by many as a legitimate stakeholder in shaping domestic and international policies designed to reduce harm caused by its products.

Dedicated global health networks play a central role in raising awareness about these health issues, but we still have limited knowledge about their emergence, evolution and strategies. Comparing two networks dedicated to reducing harm caused by smoking and drinking through a public health approach offers important insights into their relative ability to shape global health policy over time. This public health approach highlights the importance of surveillance to collect reliable data about the problem (1), research focused on the causes and contributing factors (2), community interventions addressing the root causes of the condition (3) and regular monitoring and evaluation to ensure efficiency and effectiveness (4).

The comparison presented here explores in what ways differences in issue characteristics and the policy environment interact with network structures and strategies to produce specific outcomes expressed in the adoption of policy solutions and the amount of financial resources dedicated to the health problem. It provides an initial set of propositions about how global health networks can enhance their effectiveness by focusing attention on the political process of expanding their reach through coalition-building and gaining crucial support outside of the core group sharing a public health understanding of a problem and its root causes.

Conceptual framework

This study is part of the Global Health Advocacy and Policy Project (GHAPP), a research initiative examining networks that have mobilized to address six global health problems: tuberculosis, pneumonia, tobacco use, alcohol use, neonatal mortality and maternal mortality. Its aim is to understand how networks crystallize around health issues and why some are better able to influence policy and public health outcomes. GHAPP studies draw on a common conceptual framework grounded in theories on collective action from political science, sociology and economics (Snow etal. 1986; Stone 1989; Powell 1990; Kingdon 1995; Kahler 2009). The introductory paper to this supplement presents the framework in detail (Shiffman etal. 2016).

The GHAPP studies specifically examine network outputs, policy consequences and impact. Outputs are the immediate products of network activity, such as guidance on intervention strategy, research and international meetings. Policy consequences may include the adoption of international resolutions or treaties, increased funding, national policy adoption and the scale-up of interventions. Impact refers to the ultimate objective of improvement in population health.

The framework consists of three categories of factors (Shiffman etal. 2016: Figure 1). One category, network and actor features, concerns factors internal to the network involving attributes of the actors that created and constitute the network and its strategies. This category covers characteristics of individuals and organizations that shape network capacity to act and influence their environment. A second category, the policy environment, concern factors external to the network that shape both its nature and the effects the network hopes to produce. The third category, issue characteristics, concerns features of the problem the network seeks to address. GHAPP studies begin with the presumption that no single category of factors takes precedence. Instead, analysis focuses on how factors within each category interact with one another to produce policy and public health effects.

Several factors in each category may be particularly influential. Among network and actor features, the existence of effective leaders may be one reason networks crystallize in the first place, and why, once they appear, they are able to achieve their objectives. The quality of governance may also matter, in particular the institutional forms adopted by network members to pursue collective goals (Buse and Walt 2000). A third factor is composition. Diverse networks that link scientists, advocates, policymakers and others from both high- and low-income countries may achieve better outcomes because diversity improves collective understanding and problem solving capacities (Page 2007). But heterogeneity can sometimes also be problematic, as it may cause internal divisions. The fourth factor is framing strategies (Snow etal. 1986) or the activities network actors display in publicly position an issue.

Several factors in the policy environment may be particularly influential. Among these are potential allies and opponents. If there are many groups whose interests align with a network’s goals, that network is more likely to expand and be effective than one that faces a dearth of potential allies. Opponents, such as the alcohol and tobacco industries, may both hinder and facilitate network outcomes: they seek to discredit the network or co-opt members, but may also inspire mobilization. Substantial funding may enable a network to flourish; however, a network set up at the behest of donors may be perceived as less legitimate than those that emerge from grassroots activism. Norms—standards of appropriate behaviour for a particular group of actors—may also be influential. Important examples of influential norms in global health are those that the health-related Millennium Development Goals advance (Fukuda-Parr and Hulme 2011). These goals have raised expectations that states, intergovernmental organizations and other global actors act to reduce burden from that subset of global health problems selected for inclusion.

Among issue characteristics, severity, tractability and the nature of affected groups may be particularly relevant. Robust networks may be more likely to emerge around issues with high mortality and morbidity rates or social disruption. Also, individuals and organizations may be more likely to act on problems perceived to be solvable (Stone 1989). In addition, affected populations that inspire sympathy, such as children, may be more likely to lead to network mobilization (Schneider and Ingram 1993) than those that do not. Also, positive network results may be more likely if affected populations are able to mobilize on their own behalf, as some people living with HIV/AIDS have done.

Methodology

This study combined a process-tracing methodology involving in-depth examination of social and political processes with a paired comparison. Process-tracing is used to uncover causal mechanisms linking specific explanatory factors to policy outcomes (Yin 2008; Bennett 2010). The paired comparison allows for retaining the in-depth qualitative analysis, while also making some inferences about the relative importance of factors highlighted in single-case studies across other cases (Brady and Collier 2004; Tarrow 2010). GHAPP researchers used the same methodology, began with the same basic set of questions, and were in regular communication in order to share insights as the studies unfolded. The alcohol–tobacco comparison presented here relies on an analysis of events over time that identifies the key events relevant for understanding the evolution and effectiveness of the two global health networks. This descriptive inference (Collier 2011) is used to ascertain the relative importance of factors identified in the conceptual framework. For example, we used the interviews to inquire about the relevance of prior policy efforts such as prohibition or how exactly network members participated in international policy negotiations around their health issue.

The comparison of the alcohol and tobacco case commenced with a careful study of documents and archival materials, followed by semi-structured interviews with experts focused on the emergence and evolution of each network. Documents and archival materials consisted of several hundred scholarly articles produced by network members, policy submissions to national and international bodies, editorials, press releases and World Health Organization (WHO) background documents. Semi-structured interviews were conducted with a total of 55 individuals familiar with the actions of the global health networks. Interviewees were offered anonymity to enhance the overall reliability of the data collected regarding network features and strategies. Supplementary Appendix I contains a list of interviewees who gave consent to recording and transcribing their answers. Some of these individuals were interviewed more than once and have provided feedback throughout the development of the case studies. Supplementary Appendix II provides an outline of the themes covered in the interviews.

Additional interviews and background conversations were conducted after the completion of initial draft versions of the individual case studies. We conducted a member checking process by inviting seven key informants to provide detailed written feedback on the case studies. Other interviewees and experts provided feedback at network conferences where initial results were presented. We received extensive feedback on our results at meetings of these global health networks and other experts, including the 2012 American Public Health Association meetings in San Francisco, the 2013 Global Alcohol Policy Conference in Seoul, South Korea and the 2013 Alcohol Policy 16 conference in Arlington, Virginia.

The results reported offer initial evidence about how issue characteristics, network features and the policy environment shape the effectiveness of global health networks. Three core limitations emerged during the research process and suggest the need for additional research on this topic. First, data and information about the early stages of network emergence and evolution are less reliable because of very scant written documentation about the inner workings of the networks as well as interviewees’ limited capacity of recollection. Second, we initially identified interviewees through a literature review and later using snowball sampling. This may have generated a selection bias and limited the diversity of viewpoints collected. Third, the comparison concerns two networks at different stages of their development. What this study cannot answer is if the alcohol control network will experience the same level of success witnessed in the tobacco case over time, or if its challenges identified here will limit its effectiveness in the long-term.

Results

We organize the results section along the three categories of issue characteristics, network and actor features and policy environment. This presentation serves two main purposes. First, we highlight the factors that the paired comparison identifies as most relevant to explaining why the tobacco control network was more effective than its alcohol counterpart. While the global health network on tobacco control has already been studied in some detail (Roemer etal. 2005; Mamudu etal. 2011), the first systematic review of the global health network on alcohol is included in this issue (Schmitz 2016). Second, we address how factors explaining relative network effectiveness are causally linked to each other.

We define network effectiveness as the ability to set agendas of international institutions, to prompt international and domestic policy adoption in line with network objectives, to raise funds from public and private donors and to ultimately reduce harm and improve population health. In relative terms, the tobacco control network has been more effective with regard to global policy adoption, resource mobilization and harm reduction. At the global level, both networks were able to increase awareness about their issue and succeeded in getting the WHO and its member states to adopt policies designed to reduce harm. But in the alcohol case, this commitment remains non-binding, whereas the FCTC imposes more well-defined legal obligations on its signatories. In addition, major philanthropic donors, including the Bloomberg Philanthropies and Bill and Melinda Gates Foundation, pledged in the mid-2000s more than $600 million to tobacco control (Bloomberg Philanthropies 2011). Both initiatives share a focus on the promotion of tobacco control policies and laws in low- and middle-income countries, but have developed different emphases in their grant programs. In sharp contrast, in the four years after the adoption of the Global Strategy no significant financial commitments from public or private sources have been forthcoming in support of global alcohol control. Although the WHO currently projects an increase in smoking deaths from 6 to 8 million by 2030, tobacco use has significantly dropped in most industrialized countries where the most stringent public health measures have been in place for some time. In the alcohol case, we observe considerably less progress, including no significant funding dedicated to harm reduction and virtually unchanged global consumption levels in the past decades (World Health Organization 2011: 8).

How do we explain the increase in global attention to both risk factors, but their diverging subsequent trajectory? We argue that the respective global health networks advancing public health solutions play a critical role in conjunction with issue characteristics and the policy environment. Issue characteristics and historical experiences with reducing harm caused by smoking and drinking have shaped the conditions under which both networks emerged and evolved. Most importantly, the prevalence of a range of problem definitions in the post-Prohibition era explains why the global health network on alcohol faced much greater challenges to coalition-building and effectiveness from the very start. Perceptions of failed prohibition led to the establishment of competing problem definitions and solutions, including alcoholism as a treatable medical condition, the recovery approach represented by Alcoholics Anonymous (AA), the activism of social movements against drunk driving and the public health approach focused on population-level policies and the role of industry as a ‘vector of disease’ (Gilmore etal. 2011). At its inception, the tobacco control network had to insure that it created and maintained a consensus about how to address harm beyond simply identifying the industry as the main enemy. This positive consensus required agreement not only about specific interventions, but also on strategic choices about how to implement those policies, including the push for the FCTC.

As the networks developed and matured, their intrinsic features (leadership, governance, composition and framing strategies) become increasingly important in explaining differences in their respective ability to expand membership and recruit allies. Both networks are effective in using evidence-based research in advancing agenda-setting and their own policy solutions, but the tobacco network grew more quickly, its leadership had higher-level access in the WHO hierarchy, and its expanding presence in low- and middle-income countries provided it with greater legitimacy when compared with the alcohol network. The policy environment as the third and final category of factors began to matter when both networks are established and actively agitate for their causes. In the incipient stages of both networks, the policy environment in the industrialized world was dominated by powerful tobacco and alcohol industries, and social acceptance of smoking and drinking was high. As the networks began to mobilize and evolve, responses in the policy environment differed. Most importantly, the tobacco industry’s persistent denial of harm provided anti-smoking advocates with a strategic advantage once evidence about the link between smoking and cancer became irrefutable. In contrast, the alcohol industry effectively combined a denial of harm to the vast majority of the population with an acknowledgement that ‘heavy drinking’ is a problem that the industry is effective in addressing through self-regulation. As a result, the well-funded efforts of the alcohol industry to establish their definition of harm and solutions reinforce the existing divisions among activists with divergent sets of problem definitions.

Issue characteristics: severity, tractability and affected groups

Alcohol and tobacco use represent behavioural risk factors with significant similarities, including their addictive nature, strong scientific evidence linking use to personal harm, the power of commercial interests, the disproportionately negative effects on low-income groups and the social harm caused by second-hand smoke and drinking. Fourteen % of the world’s population smoke, whereas 42% drink alcohol (World Health Organization 2011). Although mortality rates for smoking are higher than those for drinking, the differences in severity are comparable when considering rates of morbidity and overall social harm caused by alcohol consumption. Smoking is currently responsible for ∼8% (5.4 million) of all deaths and 3.9% of the global burden of disease and injury (measured in disability-adjusted life years). In comparison, alcohol use caused 5.9% (3.3 million) of all deaths and 5.1% of the global burden of disease and injury in 2012 (World Health Organization 2014). Alcohol’s much greater prevalence causes significant social harm, including loss in economic productivity and injury to others. Low- and middle-income countries and their economic development are particularly threatened since alcohol is the leading risk factor for death and disability for the economically productive 15–59 age group (World Health Organization 2014: 57).

Considering the high burden of harm caused by both tobacco and alcohol, severity alone cannot explain the observed variation in global responses to both risk factors. Instead, the following comparative summary of issue characteristics focuses first on differences in severity, then on how variation in tractability is shaped by historical experiences in reducing harm, and finally on the significance of affected groups.

While the health risks attributed to alcohol and tobacco use are well researched, there are important differences with regard to our current understanding of harm. Only after decades of research and activism, there is a consensus today that tobacco use kills half of all smokers using the product as intended. This evidence offers a solid base for mobilizing around the issue, policy solutions (e.g. raising taxes and banning marketing) and shifting responsibility away from individuals to the activities of industry. In contrast, alcohol use is currently linked to 200 diseases (Room 2013), but many adverse effects remain under-researched, including links to domestic violence or how alcohol use undermines recovery from other health problems. In addition, some health benefits of moderate alcohol consumption at age 45 or older also present greater framing challenges for public health advocates. In the tobacco case, the gap between objective harm established by scientific research and the public perception of harm has narrowed to a much larger degree than in the alcohol case.

Differences in the perceived complexity of adverse health effects have become the basis for diverging approaches to addressing alcohol harm. The diversity of alcohol-related harm, including immediate (e.g. drunk driving, domestic violence) as well as more long-term effects creates demand for a broad variety of solutions. While the global health network addressing tobacco harm succeeded over time to develop and maintain a general consensus about how to reduce smoking rates, it has proven to be more challenging to create such a consensus in the alcohol case. In the tobacco case, coalition-building played a crucial role in expanding the network and in acquiring policy expertise to engage in the broader politics debates of the issue at domestic and international levels. More intensive collaboration reinforced unity and allowed individual groups to benefit from the capacities of allies (Weishaar etal. 2015). In the alcohol case, such a broader network bringing together all parties interested in reducing harm (except for the industry) remains to be established.

A second characteristic that sets alcohol apart from tobacco is the Prohibition era and its legacy (Schrad 2010), which have shaped contemporary debates about problem definitions and policy solutions (Interviews A1, A3, A4, A9). The widely perceived failure of Prohibition established in the early 20th century in the USA and some European countries (Okrent 2010) has profoundly shaped subsequent policy responses to rising harm caused by alcohol at domestic and international levels. While the temperance movement became one of the most powerful social movements of the late 19th and early 20th centuries (Schrad 2010), efforts to control smoking were much less relevant at the time primarily because tobacco consumption was low and links between smoking and cancer had yet to be established. The end of Prohibition in the 1930s was less relevant for tobacco control efforts later on, but splintered the surviving alcohol control movement during the 1930s and again after World War II into several separate movements with distinct and often incompatible approaches to addressing alcohol harm (Interviews A1, A5, A9). Today, three distinct approaches to alcohol harm—public health, individual (moral) responsibility and medical treatment—are advanced by separate groups disagreeing about what constitutes the problem and how to address it. AA, founded in 1935, advances an individualistic approach that explicitly rejects public policy engagement. Similarly, the medical treatment and recovery community also focuses exclusively on the small subset of people identified as addicted drinkers and emphasizes individual approaches to recovery (Beauchamp 1980; Roizen 1991). Advocates against drunk driving do seek changes of public policy, but focus exclusively on one particular harm and rarely endorse population-based measures (Lerner 2011). It is the perceived failure of prohibition that played an important role in giving rise to competing approaches to alcohol harm (Interviews A1, A5). Divergence in how different groups define the problem and issue characteristics are not merely a reflection of the objective harm caused, but a result of competing normative claims about the place of alcohol in society (Gusfield 1981).

In contrast, tobacco as a regulatory issue only gained momentum when scientific evidence about harm became widely shared knowledge among scientists and the medical community. While agreement about harm is important, networks are particularly relevant in forging consensus on appropriate solutions. Here, the anti-smoking movement early on focused on establishing a common public health approach focused on prevention, protection and smoking cessation. While tobacco control represents a policy field that encompasses a diverse set of interventions (e.g. non-smoker protection, demand reduction, product regulation), these dimensions are all part of a public health approach, which advances government-led tobacco control intervention as effective solutions. Tobacco control debates between reducing harm and preventing all forms of tobacco use do represent two competing end goals, but these disagreements did not undermine the broader consensus maintained as the network expanded its activities and began to focus on the WHO and the FCTC.

A final issue characteristic setting tobacco apart from alcohol is the constitution and perceptions of affected groups. Because there are no safe levels of tobacco use, the policy focus is simply on all users and highlights cessation as the single strategy of choice. Protecting non-smokers from harmful effects of tobacco entails also population-based approaches that effectively reduce not only exposure, but also overall consumption. In contrast, different forms of alcohol use have given rise to a diverse set of policy proposals focused on only addressing drunk driving, protecting the unborn from fetal alcohol syndrome, or reducing incidents of cancer. Unlike the tobacco case, each of the affected groups has given rise to separate and often competing activism to reduce a particular harm. Even more importantly, the availability of functioning health care systems in many industrialized countries with the highest levels of alcohol consumption diminishes perceptions of severity through the medical treatment of harm. Finally, almost half of the global population consumes alcohol and many of those claim to enjoy it and are relatively well-off. When compared with tobacco use, the group at risk is not only larger, but also more diverse, undercutting perceptions of urgency to address the substantial harm caused by this drug. As tobacco use in developed countries dropped more quickly among high-income than lower income populations (Centers for Disease Control and Prevention 2009), raising taxes and other interventions faced less resistance at the elite levels. In contrast, alcohol control has yet to reach a tipping point where reduced prevalence combined with policy interventions would create a reinforcing virtuous cycle.

Network and actor features: composition, framing strategies and leadership

Issue characteristics and historical precursors set up very different starting conditions for the two global health networks emerging in the 1970s and 1980s. While the strengthening of tobacco control advocacy tracks closely the growing scientific evidence on harm caused by smoking, the global health network on alcohol control emerged amidst disagreements about how to best address drinking as a global health issue. The emerging networks differed with regards to their composition, framing and leadership, equipping them with different abilities to influence their respective policy environments. In particular, the tobacco control network was more effective in expanding its reach into low- and middle-income countries and its leadership was able to get access to the top-level leadership at the WHO to advance its goal of an international treaty. The anti-smoking network’s ability to expand its support base globally combined with its leadership explains why it was able to successfully push for a legally binding treaty as a key step towards attracting more funding and setting in motion a process of implementing meaningful tobacco control measures in WHO member states. In contrast, the alcohol control network remains much smaller, less diverse and features more limited leadership skills needed to move from an agenda-setting role to an ability to shape policy formation and implementation.

Following the official acceptance of scientific evidence linking smoking and cancer both in the United States and Europe, network activity around global tobacco control started to emerge during the late 1950s and early 1960s although it would take several decades for an effective network for global tobacco control to emerge. Until the 1980s, the primary venue for international collaboration was the Conference on Tobacco or Health, which allowed scientists and activists to exchange research and discuss policy responses mainly at domestic levels. When evidence about the harm of second-hand smoking finally became overwhelming in the early 1980s, the issue began to draw wider interest among organizations representing non-smokers. The 1985 International Summit on Smoking Control, organized by the American Cancer Society, was the first meeting explicitly focused on the need to coordinate global action (Interview T2).

The emerging network around tobacco control resembled an epistemic community including scientists and advocates marked by a high level of cohesion around the framing of tobacco control as public health issue, the industry as vector of disease and population-based policies as effective solutions (Mamudu etal. 2011). As network members constructed a shared understanding of tobacco control as public health issue based on scientific evidence, advocacy leadership around global tobacco control also started to crystallize. In 1993, a small group of activists proposed for the first time the idea of an international treaty. This group, led by Roemer, Taylor and Mackay combined legal and medical expertise to exercise norm entrepreneurship in setting the course for how to proceed collectively at the international level. In their view, an international treaty represented a crucial step in legitimizing and strengthening tobacco control initiatives across countries.

The year 1999 represented the transition from the incipient stage of the network to its rapid growth as the WHO started to negotiate an international treaty on tobacco control—the FCTC. The same year, the Framework Convention Alliance (FCA), which would represent the central coordinating network of tobacco control advocates and scientists during the FCTC negotiations, was formally founded. The creation of the FCA allowed the network to broaden its reach into low- and middle-income countries, helped to further strengthen ties between network members and augmented the capacity and expertise of network members across the globe (Interviews T5, T15). During the past decade, the network has further institutionalized its presence and reached into domestic policy contexts around the world through the creation of regional network organizations, such as the Southeast Asian Tobacco Control Alliance (SEATCA) or the African Tobacco Control Alliance (ATCA). Between 2000 and 2003, membership grew quickly from 25 to 195 members, and it stands today at ∼350 non-governmental groups (Interview T14).

In the alcohol case, the emergence of a distinct global health network did not grow out of a better understanding of harm, although evidence about the rising mortality and morbidity rates in low- and middle-income countries would contribute in the 1990s and 2000s to the slow globalization of the network. Instead, scientists and activists concerned about rising global alcohol harm split from existing groups concerned with the issue based on developing a different understanding of the problem as well as the appropriate solutions. In 1986, former members of the International Council on Alcohol and Addictions (ICAA) launched the Kettil Bruun Society (KBS). ICAA had long served as a host for a wide range of approaches to reducing alcohol harm and allowed its members to collaborate with industry. In contrast, KBS members emphasized the role of social conditions in shaping population and health, including an emphasis on the alcohol industry as a major contributor to the problem.

In 2000, a conference of public health-focused efforts to reduce alcohol harm resulted in the creation of the Global Alcohol Policy Alliance (GAPA). The explicit goal of establishing a global body was to counter the growing marketing efforts of industry targeting low- and middle-income countries (Hesse 2015), and to expand the existing network beyond the transatlantic context (Interviews A3, A6). Subsequently, GAPA members established regional bodies in Asia (Indian Alcohol Policy Alliance, 2004), the Pacific (Asia Pacific Alcohol Policy Alliance, 2005) and Africa (East African Alcohol Policy Alliance, 2009/Southern African Alcohol Policy Alliance, 2012), but the speed and depth of globalizing the network has lagged behind the tobacco case.

A final factor distinguishing both networks is variation in the type of leadership exercised during crucial phases of network creation and evolution. In the tobacco case, evidence about harm increased pressure for action, but it was the idea of creating an international agreement introduced by legal experts Roemer and Taylor that provided a viable strategy integrating the network. In the alcohol case, in contrast, GAPA and its predecessors emerged from an explicit desire to abandon previous approaches to addressing alcohol harm. When Gro Harlem Brundtland became Director-General of the WHO in 1998, she embraced tobacco control as a top priority and this additional leadership support legitimized and broadened the network further and sustained the issue until the adoption of the FCTC in 2003. In contrast, the alcohol control network never received this kind of top-level leadership support at the WHO.

The tobacco network developed relatively quickly into a diverse and globalized movement that cohered around a single approach of addressing harm through reducing smoking. During the late-1990s, the crucial element of leadership exercised by Brundtland and the rest of the WHO put the issue on track for the crucial step of a legal agreement, setting up new global norms and funding opportunities. In contrast, the alcohol network globalized to a much lesser degree, had greater difficulties in diversifying support for its distinct public health approach to reducing harm, and has yet to gain the leadership support visible in the tobacco case. While the FCA as a key umbrella group for global tobacco control can point to hundreds of membership organizations from around the world, GAPA and its allies still represent primarily a collection of dedicated individuals, and not yet an alliance of organizations with a crucial capacity to mobilize. These differences with regard to network composition, framing and leadership explain why tobacco control has received more funding, has given rise to more powerful international norms has and more effectively excluded commercial interests from the policymaking process internationally and domestically.

Policy environment: norms, funding and allies/opponents

Issue characteristics shape network emergence and evolution, but networks and their members also seek to proactively ascribe new or different meanings to these characteristics. For example, one of the key outputs of health networks is the production of scientific knowledge which shapes public perceptions of severity and tractability. These activities are primarily targeted at the policy environment that consists of potential allies and opponents, offers funding opportunities and is the arena where activists pursue policy change. Comparing the alcohol and tobacco cases reveals important differences along all three factors, including a stronger international legal framework represented by the FCTC, substantially more funding for tobacco control and broader alliances and less influential industry opponents when comparing the tobacco and alcohol cases.

Today’s significant differences regarding the policy environment show that global health networks matter because these differences did not exist at earlier stages of network development. Decades ago smoking and alcohol use were both socially accepted in developed countries, industry interests were overwhelmingly powerful, and there was very limited funding available for advancing the goals of the respective networks. In addition, the spread of free trade policies increased the availability of tobacco and alcohol products. Without recognition of global health networks as significant actors it is impossible to understand why these similarities have given way to today’s quite different policy environments for both cases.

Tobacco control advocates succeeded in their efforts to get a strong, legally binding global treaty adopted in 2003, while their counterparts in the alcohol case successfully lobbied WHO member states to adopt a non-binding agreement to reduce alcohol harm in 2010. The leadership and personal relationships of individual tobacco control advocates proved useful in facilitating their success. The idea of a treaty for tobacco control first emerged during conversations between Roemer, Taylor and Mackay and was subsequently introduced at the 1994 Conference on Tobacco or Health (Mackay 2003: 551). Roemer, Taylor and Mackay combined their respective international law expertise and personal ties within the WHO. Backed by the International Non Governmental Coalition Against Tobacco (INGCAT) and its broad support across many developed and developing countries, these activists gained the support of Jean Lariviere, a Canadian World Health Assembly (WHA) delegate who successfully lobbied other WHA members to request a study on the feasibility of an international instrument (Roemer etal. 2005). However, a majority of member states raised objections to a legally binding agreement, and other proposals, including non-binding instruments, were introduced into the debate. This changed in 1998 with Brundtland’s appointment to Director-General, which gave Mackay direct access to her as a member of the transition team. This direct lobbying contributed to Brundtland’s decision to include tobacco control into her key cabinet projects and led to the creation of the Tobacco Free Initiative, a significant elevation of the issue within the WHO bureaucracy.

The FCTC negotiations marked the point where the demands of the global health network became part of the international health agenda and the interactions between the network and its policy environment had significant impact on levels of global attention and policy formulation. By the early 2000s, INGCAT represented more than 1,000 member organizations from 150 countries and was able to mobilize this representation to regularly call for a strong global treaty in support of tobacco control. Founded by the International Union Against Cancer, the International Union Against Tuberculosis and Lung Disease and the World Heart Federation, INGCAT was more than a collection of dedicated researchers and could claim broad support across a wide range of membership-based groups focused on different diseases. The adoption of the FCTC in 2003 institutionalized global tobacco control within an inter-governmental governance structure and established a set of policy prescriptions and government obligations for effective public policymaking around tobacco control. The FCA institutionalized its advocacy efforts by creating formal network organizations at regional and domestic levels. While FCTC member states’ resource contributions have remained limited, private funding sources, such as the Bloomberg and Gates Foundations, have added more than $600 million to global tobacco control efforts.

In the alcohol case, the creation of a global normative framework within the WHO context was only possible after the completion of the FCTC negotiations. During the 1990s, members of the global health network focused on alcohol harm had contributed research on the severity and tractability of the issue and pushed the WHO to pay greater attention to the topic. By 1999, the WHO published the first Global Status Report on Alcohol (World Health Organization 1999), followed by additional reports released in the early 2000s (World Health Organization 2001, 2004a,b). A key focus of these publications was to raise awareness about low- and middle-income countries as ‘long-neglected areas where alcohol problems are likely to increase at an alarming rate in the future’ (Le Galès-Camus 2004).

Nordic countries then took the lead in putting alcohol back on the WHO agenda, arguing that evidence about rising harm necessitated a global response (Bull 2005). In 2005, the WHA adopted its first resolution in favour of a non-binding Global Strategy to Reduce the Harmful Use of Alcohol. Subsequently, countries opposed to WHO action on alcohol sought to slow down the process. For example, a Cuban foreign policy official expressed doubts in a Swedish newspaper questioning ‘why push the alcohol question so hard when people lie dying of AIDS, tuberculosis and malaria’ (cited in: Grimm 2008: 863). Actual negotiations on the substance of the Global Strategy commenced in 2009 in Geneva and offered members of the global health network and other civil society groups to consult and engage in lobbying activities aimed at strengthening the public health language of the proposed agreement (Interview A12). The adoption of the Global Strategy required overcoming strong resistance by the United States government and other countries as well as increasing support by low- and middle-income countries, including Kenya, Rwanda and Thailand whose representatives increasingly took over leadership on the issue from Nordic countries (Interview A14).

Despite the success in establishing a global norm to reduce alcohol harm, this issue continues to lag behind tobacco to a degree that cannot be fully explained by objective measures of severity. The adoption of the Global Strategy did not include any financial commitments on the part of governments aimed at reducing harm, and private foundations have been largely absent from this issue. A recent study on public and private donor commitments concluded that in 2007 less than 3% of global health spending was dedicated to addressing NCDs (Nugent and Feigl 2010). While a significant part of this spending on NCDs is dedicated to tobacco control, only $4 million was identified as being explicitly targeted at alcohol. Scholars have highlighted a persistent neglect of funding for alcohol control even after the adoption of the Global Strategy (Zeigler and Babor 2011), which stands in sharp contrast to post-2006 decisions by the Bloomberg Philanthropies and the Bill and Melinda Gates Foundation to dedicate significant resources to tobacco control. The highly unequal funding streams put alcohol control advocates at a significant disadvantage as they compete for talent and attention with other NCD risk factors, including obesity and physical inactivity (Interviews A8, A12). The comparison shows that expanded funding is particularly crucial for maintaining network momentum after the adoption of global policies such as the FCTC or the Global Strategy when global health networks can base their legitimacy on the public health approach now supported by WHO member states.

Addressing alcohol and tobacco harm involves facing industry interests that have been identified by members of the global health networks as ‘vectors of disease’ (Jahiel and Babor 2007; Gilmore etal. 2011). Since both tobacco and alcohol industries are important economic actors, comparing changes in their involvement in policymaking at international and domestic levels represents a key indicator of network effectiveness and future likelihood of domestic adoption of effective policies. The tobacco industry was not only formally excluded from the FCTC process and subsequent policy negotiations but is formally identified as an obstacle for global tobacco control within the FCTC (Article 5.3). Due to the exclusion from the policymaking arenas, the tobacco industry has been forced to resort to confrontational strategies to counteract the global policy momentum for tobacco control within individual countries (Interviews T11, T16, T21). Interfering strategies include the refutation of projected tobacco control policy outcomes, the activation of front groups to protest tobacco control policies and the use of global, regional and bilateral trade and investment agreements as strategic avenues to combat tobacco control policy diffusion (Fooks 2011; Brandt 2012).

In contrast, the alcohol industry managed to retain its position within the emerging policy field of global alcohol control by supporting counter frames against the public health perspective on alcohol control. Learning from the experience of the tobacco industry (Bond etal. 2009), the alcohol industry does not categorically deny harmful effects, but insists on an exclusive focus on excessive alcohol use and voluntary efforts. While the industry has limited direct access at the WHO, it exerts power domestically by disseminating its own policy templates and cultivating key relationships with domestic policymakers, thereby effectively shaping member state preferences in policy negotiations regarding alcohol policies (Interviews A10, A15). At the global level, the alcohol industry has become increasingly concentrated into larger corporations that can exert greater influence both at the international and domestic levels (Jernigan 2009). In recent global negotiations about addressing NCDs, the industry has managed to be classified with the food industry as a possible stakeholder while the same document recognized a ‘fundamental conflict between the tobacco industry and public health’ (United Nations General Assembly 2012: para 37 and 38).

Although the alcohol industry has not been entirely excluded from policymaking processes at the international level, there is increasing consensus that commercial interests should not be at the table when discussing policies designed to reduce alcohol harm (Babor etal. 2013). The Director General of the WHO, Margaret Chan, has endorsed such a conflict of interest policy (Chan 2013). As both the tobacco and alcohol industries remain powerful industries with extensive lobbying efforts both globally and domestically, differences in their status at the policy table matter greatly for public health outcomes.

Discussion

This article provided evidence about the role of global health networks in contributing to effective policy solutions in reducing harm caused by alcohol and tobacco use. In both cases, increased global levels of harm, documented by network members, led to network crystallization during the 1960s and 1970s. But beyond initial agenda-setting, the two networks diverge sharply with regard to the degree of consensus built among activists about how to reduce harm. In the tobacco case, activists united early on against the tobacco industry and came to a consensus focused on prevention (targeting youths), protection (from second-hand smoke) and cessation (targeting current smokers). In the alcohol case, a very similar public health focused network formed, but never overcame differences in problem definition and policy solutions that separated it from other approaches to alcohol harm, including groups focused only on drunk driving (Lerner 2011), addiction treatment (Beauchamp 1980; Hester and Miller 2002) and the self-help approach represented by groups including AA.

Why then is the alcohol case characterized by persistent disagreements with regard to problem definition and policy solutions? We argue that specific issue characteristics and the policy environment play important roles. With regard to issue characteristics, the legacy of the Prohibition era and perceptions of more limited harm caused by alcohol create greater challenges for the global health network to expand and spread consensus. As the network struggles to move beyond initial agenda-setting, commercial interests contesting alcohol control in the larger policy environment also actively foster disagreements about problem definition and solutions. Learning from ‘big tobacco’ and its failed strategy of denial, the alcohol industry has adopted a proactive stance focused on narrowing the problem to ‘excessive drinking’ only. This allows the industry to exploit existing divisions among activists while also projecting an impression of socially responsible behaviour (Table 1).

Table 1.

Tracking differences between the tobacco and the alcohol cases

Issue characteristics: severity, tractability and affected groups		Network and actor features: leadership, composition, governance and framing strategies		Policy environment: allies/opponents, funding and norms
Tobacco: Growing research evidence enables consensus about preferred policy solutions	Alcohol: Legacy of Prohibition era gives rise to wide range of competing understandings of harm (1), preferred solutions (2), and definition of affected groups (3)	Tobacco: Network grows globally and diversifies to enhance its legitimacy; leaders effectively advocate for WHO access and a focus on the adoption of the FCTC	Alcohol: Network grows slowly, but remains dominated by individual researchers with limited advocacy expertise; differences with other groups interested in reducing alcohol harm are not overcome	Tobacco: Industry power diminishes as its representatives are excluded from international policymaking processes; FCTC leads to an expansion of the network and justifies push for domestic implementation of tobacco control measures	Alcohol: Industry learns from tobacco case and proactively claims efforts of self-regulation; network remains still weak and struggles to attract new funding sources or allies

The alcohol–tobacco comparison reveals that choices and strategies of global health networks shape the trajectory of their causes, in particular their ability to maintain consensus while expanding their influence beyond the narrow health field. While the tobacco control network grew along with the mounting evidence about harm and a consensus about effective interventions, the alcohol network competed from the very start with established approaches focused on treatment and individual responsibility. Both networks were effective in producing scientific evidence to raise awareness, but disagreements about what constituted harm and effective remedies remained more prevalent in the alcohol case. The policies of increased taxation, restrictions on marketing, and excluding industry from policymaking are today widely accepted in reducing tobacco harm, but continue to have much more limited support in the alcohol case.

As the networks developed and matured, their intrinsic features (leadership, governance, composition and framing strategies) became increasingly important in explaining differences in their respective ability to expand membership and recruit allies. The tobacco control network evolved from a small group of individuals to a broad civil society coalition supported by membership-based organizations. This enabled the network to expand its capacities beyond knowledge generation to acquire also greater advocacy and policy expertise. In contrast, the alcohol control network remained a smaller coalition of dedicated and like-minded individuals with limited capacities to engage in the political struggles associated with the adoption of their population-based interventions (e.g. taxation, marketing bans) to reduce alcohol harm. As a result, the INGCAT and the FCA with their broader networks of regional and domestic organizations were able to make more credible claims regarding representation and legitimacy (Gneiting 2016).

Finally, the policy environment gains in relevance as both networks evolve and actively seek to change the world around them. The industry as the main opponent to both networks is first to respond and shape further outcomes. The tobacco industry’s choice to deny harm over decades contrasts with the alcohol industry’s more proactive stance of accepting some responsibility and seeking to define the problem as limited to excessive drinking. In the tobacco case, the industry and its denial fostered network cohesion, while in the alcohol case the activities of industry reinforced existing divisions through the creation of its own civil society groups and funding to researchers. Differences in network strength then explain why the tobacco control network successfully pushed for a legally binding global treaty, while the alcohol network had to settle for a non-binding agreement. In turn, this difference precipitated a virtuous cycle of attracting increased funding to tobacco control, an increased focus on domestic implementation and greater legitimacy for tobacco control activists able to point to enforceable state commitments. Both issues are likely to benefit from their inclusion in the third goal of the Sustainable Development Goals promoting healthy living and well-being (United Nations 2015).

Conclusions

The results of this comparison offer broader lessons about the effectiveness of global health networks, especially in the context of NCDs. An initial implication highlights how perceptions of issue characteristics, including severity and tractability, are shaped by historical legacies and prior policy efforts. These legacies establish an important context for the emergence and evolution of global health networks. The comparison provides some initial evidence that the legacy of Prohibition represented a formidable challenge for alcohol control efforts and needs to be more explicitly addressed in how the network advances its policy interventions.

This insight leads to a second implication about the power of specific strategies and arguments. While global health networks primarily focus on research documenting harm and evaluating effective solutions, they are often less well equipped to engage in the political struggles that emerge when they seek broader support for their preferred interventions. In the alcohol case, policies of raising taxes or limiting marketing require engagement with a broader range of actors beyond the narrow health field. Arguments that increased taxes lower alcohol harm work well within the public health community, but they do not necessarily respond to counter-arguments by industry interests that aim at portraying alcohol control as prohibitionist and a limit to personal freedoms. To win the public debate outside of the health arena, global health networks have to rely on political skills required to attract allies and maintain cross-sectoral coalitions (Morley 2015). In the alcohol case, such efforts were successful at the domestic level in the United States during the 1970s when the Center for Science in the Public Interest (CSPI) led a coalition of activists with different problem definitions in jointly lobbying Congress (Schmitz 2016). At the global level, such a broad coalition emerged in the tobacco case, but not yet in the alcohol case.

A second promising strategy to overcome these adverse effects of prior policy efforts is to broaden the legitimacy of the network by expanding into low- and middle-income countries. First, such efforts would bring in more network members from regions where harm is increasing but where no legacy of Prohibition creates potential rifts among activists. Second, it would allow the network to claim greater representativeness. Being able to claim global representation was a crucial ingredient of tobacco control advocates when gaining access to high-level policy negotiations in the 1990s. When the WHO leadership changed in 1998, the tobacco control network was ready to take advantage of this opportunity and succeeded within 5 years in getting the FCTC adopted. This policy victory then served as a key stepping-stone to reaching new goals, including attracting new allies and financial resources dedicated to the broader issue of tobacco control.

A third implication highlights the temporary nature of success as defined here by the extent of global attention and policy responses. Tobacco control advocates can point to the FCTC as a key accomplishment, but also face the problem of increased tobacco use in many middle-income countries as well as threats to their consensus approach by new technologies, including electronic cigarettes (Grana etal. 2014). At the same time, the alcohol case considered here has gained global attention much later than tobacco control. As a result, the impact of its activities may not yet be as visible as the well-documented success of the tobacco control network. Ongoing changes in issue characteristics and the policy environment require global health networks to focus continually on maintaining consensus and reaching out to allies that may or may not share the basic tenants of the public health approach.

Conflict of interest statement. None declared.

Ethical approval

We cleared the study protocol through the Institutional Review Boards of American University, Syracuse University and the University of New Mexico, which granted the study exempt status, as it focused on public policy and was deemed to pose minimal risk to informants.

Supplementary data

Supplementary data are available at HEAPOL online.

Funding

This research was supported by the by the Bill and Melinda Gates Foundation [OPPGH4831].