Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Twenty years ago I worked on an advisory report on the effectiveness of various tobacco control policy measures, commissioned by the Dutch Ministry of HealthFootnote 1 as part of the process of presenting a revised Tobacco Act to the parliament (Willemsen, De Zwart, & Mooy, 1998). Soon after the report was finished I attended the World Conference on Tobacco or Health in Bei**g, where I spoke with a civil servant from the Dutch Ministry of Health. I asked him what would happen with the report and was shocked when he told me that many of the conclusions were “not politically feasible” and could not be taken up.

This was the first time that I was confronted with the concept of “political feasibility.” In hindsight this was rather naïve of me, but students and researchers who invest time and effort in understanding better ways of hel** people to overcome tobacco addiction sooner or later come to realise that the tobacco problem has political roots. To do something about it on a societal level, one has to acknowledge what many people would describe as “nasty” politics. Many scientists shy away from this, just as I did then, because they believe that science is independent and politics-free or because they are intimidated by what they perceive as complexity, unpredictability, and irrationality in politics for which they are not prepared, being used to working from within an evidence-based science paradigm. A more effective strategy to address the tobacco problem on a societal level is to try to understand the policy process and why this process at times appears so irrational.

Since my professorship in tobacco control research, journalists, scientific colleagues from other countries, and students have asked me the same question: why is the Dutch government not doing more to control tobacco? This book is my attempt to formulate an answer. As an introductory text to the field, it seeks to provide an understanding of the full complexities of tobacco control policy. It further aims to offer a broad framework for thinking about tobacco control policymaking. Many of the understandings in the book can be applied to other public health areas, and lessons drawn from the analysis of the Dutch case may be of interest to other countries, particularly those with similar multi-party parliamentary democracies.

How does the trajectory of Dutch tobacco control compare with other developed countries? In the case of the Netherlands, there were nine years between the time that national data on the dangers of smoking were presented (1948) and the time that the government admitted there was a problem and the public should be informed (1957). It took another 25 years for the first regulative measure (health warnings on cigarette packs in 1982). Another six years passed before the Netherlands had a Tobacco Act (1988), and a further 14 years before effective measures such as an advertising ban and a workplace smoking ban were implemented (through a revision of the Tobacco Act in 2002).

This is an enormous period of time. Why was there such a delay between recognition of the tobacco problem and the policy response? Some might say this is a moralistic starting point for a book because it assumes that the government could have reacted sooner, faster, or more decisively. This is undoubtedly true—at least in theory, the government might just as well have reacted later, slower, and less decisively. In fact, the Netherlands has not done particularly badly in comparison with many other countries: the Netherlands is in some periods a laggard, and most of the time just struggles to keep up with the mainstream, but everywhere in the world there has been a wide gap between realising that there is a problem with tobacco and actually implementing effective solutions. Major tobacco control measures have had to be won in hard and long-running political battles, because tobacco control is a highly contested and politically sensitive topic. Even countries leading the way in tobacco control such as the United Kingdom, Australia, and Canada needed 20–30 years to come up with a comprehensive policy response (Cairney, Studlar, & Mamudu, 2012).

Idealists in the tobacco control field may expect that the presentation of scientific facts will automatically result in rational policy decisions, and when this does not happen a common explanation is that the tobacco industry has been successful in casting doubt on the evidence and in lobbying to delay regulation (Larsen, 2008). The industry is certainly well known for casting doubt on science, misleading politicians, and opposing or delaying tobacco control (Baba, Cook, McGarity, & Bero, 2005; Bornhauser, McCarthy, & Glantz, 2006; Costa, Gilmore, Peeters, McKee, & Stuckler, 2014; Lie, Willemsen, De Vries, & Fooks, 2016; Tobacco Free Initiative, 2008), but pointing to the tobacco industry as the sole reason for why governments do not take action is a gross simplification. Although efforts by the tobacco industry to prevent and delay tobacco policy making are important factors—and I will present many details on how this was done in the Netherlands—there are many other factors that one must take into consideration if one wants to understand the nuances and complexities of tobacco control policymaking.

With this book I move beyond the mainstream tobacco control literature which often assumes that knowledge on smoking risks leads, or should lead, to tobacco regulation. I want to explore what can be learned from insights from public policy research. I have already found a superficial glance at this rich literature rewarding, as it offers many insights that are immediately applicable to the tobacco control field. It can teach us, among other things, that public policymaking is not a rational, linear process starting with the identification of a problem, followed by selecting the best solution, finalised by adopting, implementing, and evaluating. Such models of knowledge transfer do not do justice to what happens in the real world. It is not so much that “knowledge plays no part in tobacco control, but that it is just one factor among many other policy determinants, and one that needs a political interpretation to have a policy effect” (Larsen, 2008, p. 764). Indeed, progress in tobacco control is a function of the internal dynamics of the policy process itself, and it has been argued that a more profound understanding of the political dimensions of health policy will help “to better anticipate opportunities and constraints on governmental action and design more effective policies and programs” (Oliver, 2006). Understanding these dimensions is crucial for those who want to contribute to more effective policies, including to so-called endgame strategies that may eventually eradicate the sale and consumption of tobacco products (Cairney & Mamudu, 2014; McDaniel, Smith, & Malone, 2016).

Understanding the Policy Process

A central tenet of public policy studies is that the relative influence of actors such as politicians, bureaucrats, and lobbyists on policy formation differs according to the policy sector (John, 2012, p. 5), so that a distinct sector such as tobacco control should be studied in its own right. However, to date there is not much understanding of politics in tobacco control. In 2014, I conducted a study where we searched scientific literature databases and counted the number of scientific publications in 31 European countries between 2000 and 2012 that had nicotine or tobacco as their main research topic (Willemsen & Nagelhout, 2016). Of the almost 15,000 papers identified, the proportion that had either “policy” or “politics” in their title was 0.9% and only half of these dealt with the determinants or impact of policies, leaving less than 0.5% of research that had the policy process as its main focus (unpublished data).

Despite some recent studies that have drawn on political sciences, most studies on tobacco control have paid little attention to policy processes. For example, one recent study tried to explain why smoking rates increased between 2005 and 2010 in France, which was “an unusual occurrence in countries in the ‘mature stage’ of the smoking epidemic” (McNeill, Guignard, Beck, Marteau, & Marteau, 2015). The research was a case study, comparing France with the United Kingdom, where smoking rates continued to decline in the same period. The main explanation was that in France there had been no tobacco price increases in that period, “stemming from the lack of a robust and coordinated tobacco control strategy.” Furthermore, the French government had continued to financially compensate tobacconists (small tobacco shops) with more money than was spent on tobacco control, and was too permissive regarding tobacco control, resulting in violations of the French Tobacco Act. The researchers did not say why the French government, the first country in the European Union (EU) to ratify the World Health Organization’s (WHO) Framework Convention on Tobacco Control (FCTC) in 2004, had no strong tobacco control strategy in the years 2005–2010, although they referred to a “lack of clear and consistent political will.” When we look at the Netherlands, only one scientific publication has tried to explain Dutch tobacco control policy. It is a case study of the implementation of the smoking ban in the hospitality sector (Gonzalez & Glantz, 2013), but does not really answer the question of why a smoke-free policy failed in the Dutch context (De Leeuw, 2013).

The tobacco control field has come to realise that we need to know more about the policymaking process. In 2007 the US National Cancer Institute published the monograph “Greater than the sum: Systems thinking in tobacco control” (Best, Clark, Leichow, & Trochim, 2007) which concluded that slow progress in tobacco control “is likely due to many complex and overlap** factors that must be better understood if more effective action is to be taken.” Fortunately, a small but growing body of literature emerging on the politics of tobacco control is slowly gaining attention (Cohen et al., 2000). In the past ten years or so, some useful attempts have been made to understand how and why specific tobacco policies have emerged in specific social, cultural, and political contexts (Albæk, Green-Pedersen, & Nielsen, 2007; Bryan-Jones & Chapman, 2008; Feldman & Bayer, 2004; Grüning, Strünk, & Gilmore, 2008; Kurzer & Cooper, 2016; Nathanson, 2005; Reid, 2005; Studlar, 2002, 2007a, 2007b; Young, Borland, & Coghill, 2010). To give one example, Nathanson (2005) examined differences in countries’ political systems and cultures, and how these evolve over time. She explained the diversity in trajectories by pointing to differences in how policymaking is organised and structured (e.g., whether there is a federal or centralised government and how much executive power the government has), the resources and access to policymakers that anti- and pro-tobacco groups have, and the dominant ideologies regarding tobacco use and the role of the state versus individual responsibility.

The two theories that I find most useful in understanding tobacco control policy are the Advocacy Coalition Framework (ACF) (Sabatier, 1998, 2007; Sabatier & Weible, 2007) and the Multiple Streams Approach (MSA) (Kingdon, 2003; Zahariadis, 2007). ACF is the more ambitious of the two because it aims to describe the dynamics of the policy process within changing environments, using competition between two or more advocacy coalitions in a specific policy subsystem (in our case the subsystem of all people and organisations involved in tobacco control) as a starting point. External events, such as a new government, may shock the subsystem, and such shocks result in policy change when one of the coalitions is better at exploiting the opportunity to reinforce its position—usually by demonstrating that its belief system can solve the policy problem better than opposing coalitions can (Cairney, 2013). Whether this is successful depends on the coalition’s resources and how good it is in framing its preferred solution, in exploiting public opinion, and in generating societal and political support.

What are “coalitions” in ACF theory? A network can be called an advocacy coalition when it is composed of people who share beliefs about the causes and solutions of a policy problem and have common core values. They must also engage in a “nontrivial degree of coordination” (Weible, Sabatier, & McQueen, 2009) and the stronger the coordination, the more efficient and successful their lobbying power can be. People in a pro- or anti tobacco coalition may have a variety of positions and may include interest group leaders, politicians, government officials, experts, researchers, and journalists. Shared beliefs act as the glue that binds them. The motivation to align with others in the same coalition is strengthened by what is called in ACF theory the “devil shift” (Sabatier & Weible, 2007): the tendency to perceive actors in opposing coalitions as more threatening and more powerful than they usually are. Some scholars make a distinction between purposive coalitions and material coalitions to acknowledge the fact that coalitions such as the tobacco industry coalition mainly exists because its members share material (economic) interests (Sabatier & Weible, 2007, p. 197). ACF theory assumes that, in most policy domains, two or more coalitions of policy actors can be identified (Sabatier & Weible, 2007). Often there is a dominant coalition of people who share a particular belief system, challenged by one or more competing coalitions.

Participants in a pro- or anti tobacco coalition share what ACF calls “policy core beliefs” about solutions to the problem. Such deeply ingrained beliefs remain stable within coalitions for long periods of time, explaining the resistance to change. Once a conservative coalition dominates the policy subsystem, policy change is unlikely. When one coalition’s ideas about and framing of an issue resonate better with policymakers than with a competing coalition, there is a shift in the balance of power in relation to the new “core ideas.” Subsystems contain “policy brokers,” persons such as civil servants who communicate and deal with both sides. The ACF has been used to analyse the process of tobacco control policymaking at the national level: for example, in Japan (Sato, 1999) and Canada (Breton, Richard, Gagnon, Jacques, & Bergeron, 2008); in the EU (Smith, Fooks, Gilmore, Collin, & Weishaar, 2015); and at the global level (Farquharson, 2003).

The MSA is useful in identifying the defining moment when policy solutions have become accepted by policymakers as the answer to a policy problem. While the ACF can tell us what drives change and helps us understand how actors produce policy change over longer periods of time, the MSA tells us when change is most likely to happen. It distinguishes three policy domains, each having its own dynamics and actors (Kingdon, 2003; Zahariadis, 2007): the domain of problems (is tobacco seen as a problem that needs a governmental response?), the domain of policy (which solutions are available for the problem?), and the domain of politics (is the government willing and able to act?). Substantial policy change is more likely to occur when problem appreciation, policy solution, and political opportunity align. Kingdon (2003) assigns a central role to “policy entrepreneurs” (lobbyists, activists, politicians, civil servants), who develop policy alternatives and couple these with problems at the right time (when ideological and political opportunities are favourable). In the political stream, a major electoral change may present an opportunity to advance or roll back government regulation. Changes in the perception of problems can also open up a window of opportunity: this might happen when activists point out that other countries have more advanced tobacco control policies and are more effective in tackling the smoking problem. It may then become clear to policymakers that there still exists a problem that can and should be tackled. Important changes in the policy stream can also open windows of opportunity. This may happen when new information is revealed about the feasibility and effectiveness of existing and novel policy options.

Kingdon (2003) developed his ideas after interviews with people involved with policymaking in the context of US health-care reform, regulation of transportation in the United States, and US tax changes, but not in tobacco control. The theory is nevertheless applicable and relevant, and is becoming more popular as a simple theoretical framework to understand tobacco policy (Asbridge, 2004; Barnsley, Walters, & Wood-Baker, 2015; Blackman, 2005; Bosdriesz, Willemsen, Stronks, & Kunst, 2014; Cairney, 2009; Mamudu et al., 2014; Schwartz & Johnson, 2010).

Almost ten years ago, Larsen (2008) noted that there was remarkably little interaction between mainstream tobacco control literature and public policy literature. This is slowly changing. For example, in a recent special issue of the scientific journal Tobacco Control, out of 20 contributions on the topic of “the tobacco endgame,”Footnote 2 most were from experts from the medical, public health or behavioural sciences, or from tobacco control advocates, but two were from public policy scholars. One discussed the political feasibility of various tobacco elimination endgame scenarios (Rabe, 2013), the other provided advice on how endgame solutions could be implemented and organised (Isett, 2013). Such interaction between political sciences and public health in tobacco control is welcome, but still rare. With this book I hope to contribute to this emerging literature, using the Netherlands as a case study.

A Multi-Lens Approach

Advanced tobacco control policies thrive within “policy environments” that are favourable to the implementation of tobacco control measures. Theories can inform the characteristics of such environments (Cairney & Mamudu, 2014), but a general problem in public policy is that there are almost as many models and approaches to the complex reality of policy formation as there are scholars. Each scholar takes a specific perspective, such as emphasising the context, or focusing on institutional factors, or emphasising the role of lobbyists as in the MSA, or starting with changes in policy core beliefs such as in the ACF. Discussions of the various approaches and accompanying theories can be found in textbooks on public policy (Birkland, 2011; Buse, Mays, & Walt, 2012; Cairney, 2012; John, 2012; Sabatier, 2007). Each theory offers a valuable but incomplete account of the policy process.

A complementary approach, using insights from multiple theoretical approaches, has the most chance to explain the complexity of policymaking. According to John (2012), “the approaches are not rivals; they can complement each other and be part of the explanation” (p. 14). Instead of selecting the most suitable theory for each policy problem or using multiple theories and determining which describes the data and the observations best, my book follows an approach advocated by Donley Studlar and Paul Cairney (Cairney, 2007; Cairney et al., 2012; Studlar, 2007b, 2015), who identified the core constructs from diverse theories that complement each other. They assumed that much might be gained from looking at the same policy case several times, each time from a different perspective and applying a different analytical lens.

Cairney et al. (2012) differentiated five fundamental ways of approaching tobacco control policy change, roughly coinciding with the major strands of thought within public policy science: looking at the context, at institutions, at the diffusion of ideas, at networks, and at agenda setting. John (2012) distinguished the same five elements and explained that policy change emerges from their interaction. These five ways of approaching the problem can be conceptualised as lenses through which the policy process can be analysed, and I will use these lenses in different chapters of this book. I interpret the five analytical lenses to mean the following:

  1. 1.

    Context refers to the social, cultural, and economic environment in which tobacco policymaking occurs. The social environment consists of factors such as public knowledge and concern about smoking, and public support for policy measures. Cultural values lie at the core of ideological preferences and societal rejection of tobacco control measures. The economic aspect has to do with the economic importance of tobacco for the national economy, which relates to the tobacco industry’s leverage to influence policy making.

  2. 2.

    Institutions refers to how policy is shaped by a country’s specific policymaking system. What is the dominant policy system and what are its formal and informal rules? What is the role of bureaucracy and parliament, and what are the opportunities for and constraints on lobbying? Which part of government is responsible for tobacco policy? Which level of governance is responsible for tobacco: the sub-national, national, or supranational, or some combination of these?

  3. 3.

    Diffusion of ideas refers to the role of medical and scientific knowledge. To what extent is policymaking influenced by the transfer of information, knowledge, and ideas from within and from abroad? What is the level of knowledge that has accumulated within the political system, and how is this important when making decisions on tobacco?

  4. 4.

    Networks refers to the balance of power between policy “entrepreneurs” who are typically organised in policy coalitions. Networks become coalitions when network members share a common set of core beliefs and when there is a certain level of coordination. What changes can we see over time in how the tobacco control and the tobacco industry coalitions are organised, and how effective has their lobbying been? Has there been a shift in the balance of power?

  5. 5.

    Agenda setting refers to the process by which tobacco policy appears higher on the political agenda. This involves identifying at some point that there is a problem and that the government needs to do something about it. Ideological factors have to be taken into account: is it a responsibility for the government, or should the government leave it to citizens or to the free market to tackle the problem and find solutions? Problem and solutions need to be “sold” to politicians and policymakers.

To understand how the five elements fit together conceptually, and how they may either inhibit or promote tobacco control policy, I developed a conceptual framework (Fig. 1.1) that depicts the relationships between the five elements. They are congruent with the findings from the research for this book, and are consistent with theoretical conceptualisations of the policy process. The distinction between relatively stable and relatively dynamic contextual factors is borrowed from the ACF (Sabatier & Weible, 2007; Weible et al., 2009) and, like the ACF, I award a central place in the framework to the competition between pro- and anti-tobacco coalitions and how effective they are at building supportive networks and setting up an effective lobbying apparatus. The model further reflects the basic idea from the MSA (Kingdon, 2003) that policy change is more likely to happen when advocates from coalitions succeed in bringing their conceptualisations of the problem and their preferred policy solution to the attention of politicians and policymakers. A topic’s position on the policy agenda will be higher when both problem and policy solutions align with political opportunities. Opportunities result from changes in the policy environment, such as increased social support for tobacco control, new demands from the EU or the WHO, or a change of government with another dominant ideology. Ideology is particularly important in understanding tobacco policy (Cohen et al., 2000; K. Smith, 2013; Tesh, 1988). According to political scientist Silvia Tesh, “More powerful than vested interests, more subtle than science, political ideology has, in the end, the greatest influence on disease prevention policy” (Tesh, 1988, p. 155).

Fig. 1.1
figure 1

Conceptual framework for understanding tobacco control making

Full size image

At the heart of the struggle for tobacco control is the almost universal fight between the economy, public health, and ideology. The model illustrates how new scientific information from domestic or international sources, about both problem and solutions, may feed into a coalition’s repertoire, strengthening or broadening its lobbying capabilities. However, since statistical facts rarely speak for themselves and the making of facts may be commissioned by advocacy coalitions, the arrow between ideas and networks points in both directions. To have an effect, research findings need to be translated and “sold” to decision makers (Warner, 2005). The framework assumes that (at least in parliamentary democracies such as the Netherlands) a government’s decision to adopt tobacco control measures depends on the presence of sufficient political support (majority positions in the parliament and cabinet), which is subject to the lobbying activities of the pro- and anti-tobacco control advocacy groups. In the background are the more stable and enduring contextual factors (notably cultural values), and institutional policymaking structures which directly or indirectly reinforce or inhibit the extent to which opposing coalitions can take advantage of new opportunities. Coalitions that are the best at taking advantage of opportunities arising from changes in the policy environment will be the most successful.

The framework provides for one large feedback loop. Tobacco control measures that are implemented by the government contribute to reductions in tobacco consumption at the population level. When there are fewer smokers and smoking is less visible, smoking becomes less popular and public opinion changes (smoking de-normalises), leading to more public support for further tobacco control measures and eventually contributing to new opportunities for the tobacco control coalition to advance its agenda. The empirical evidence for this loop is discussed in more detail in Chap. 4.

Outline

The next chapter is a detailed narrative of the events that shaped Dutch tobacco control policy: from when the government began to take the problem of smoking seriously in the 1960s until around 2014, when an era ended with the closure of the Stichting Volksgezondheid en Roken (Dutch Smoking or Health Foundation) (STIVORO), the national expert centre on tobacco control. Particular attention is given to the interaction between decision makers (ministers and state secretaries) and politicians in the parliament, to reveal what was done by governments in those years to tackle the smoking problem, and the struggles and politics involved. Chapter 2 structures the steps taken by the Dutch government to control tobacco, applying the idea of policy cycles, before comparing the pace of tobacco control policymaking in the Netherlands with other European countries. It will show that the Netherlands started relatively late, but caught up with mainstream Europe at the beginning of this century. Chapter 3 is the first chapter to apply one of the various analytical lenses, positioned so that we first look at policymaking from afar and gradually approach until we examine the internal dynamics of policymaking. The explanatory factors that are most distant from the actual policymaking process are external and relatively stable parameters, and serve as contextual structures that set the boundaries within which policymaking occurs. Chapter 4 first looks at the social and cultural environment. The government’s willingness to consider tobacco control measures is influenced by social norms and societal support, and by the balance between the numbers of smokers and non-smokers. There is also a feedback loop, since these factors are also affected by the adoption of a new tobacco control policy. The way these factors influence each other over long periods of time at the population level is captured in the “flywheel model” of tobacco control. Chapter 5 examines the institutional structures and the “rules of the game” that make policymaking possible but also constrain it. Chapter 6 looks at how national tobacco control policy is increasingly determined by international institutions, particularly the EU and WHO. Chapter 7 examines the role of science and the diffusion of new ideas and knowledge about what works best in tobacco control at the national level. Chapters 8 and 9 respectively discuss pro- and anti-tobacco coalitions within the broader context of the tobacco control policy arena, and describe the failures and successes of advocacy efforts. The diminishing importance of the tobacco production and trade sector to the Dutch economy is also discussed in Chap. 8. Chapter 10 takes us to the core of public policymaking, which is problem definition and agenda setting by advocacy groups. The final chapter attempts to come to a synthesis of the main findings, answering the question of how tobacco control policy comes about in the Dutch context.

Research

As part of the research for this book, I examined several data sources. In addition to the scientific literature, I made extensive use of the database of parliamentary documents and proceedings of public debates on tobacco policy in both chambers of the Dutch Parliament, using NVivo software to facilitate the process of data ordering. This was a tedious task, given the large number of documents (more than 400), many of which were minutes of lengthy debates, but the documents proved most useful in presenting a detailed historical account of the policy process given in Chap. 3. The book further benefitted from interviews with key stakeholders and informants. I conducted 22 in-depth interviews with informants from the government, health organisations, and tobacco industry, focused on the five elements of public policymaking (context, institutions, agenda setting, ideas, and networks). Finally I examined documents made public through two freedom of information requests by investigative journalist Joop Bouma (Bouma, 2001). My research team added the Bouma documents to the Truth Tobacco Industry Documents database with help of the Maastricht University Library and the Dutch Cancer Society, so that the documents are now accessible for research.Footnote 3 They were most relevant in describing the lobbying practices of the Dutch tobacco industry network (Chap. 8). For the chapter on the tobacco control network (Chap. 9), I accessed documents from the archive of STIVORO to supplement what was learned in the interviews.