Policy is not made in a vacuum. Disputes over tobacco control are fought within changing policy environments. This chapter explores key population-level factors that influence a national government’s decision to adopt tobacco control policy measures. These factors include social norms about smoking, the proportion of smokers in the population, societal support for tobacco control, and cultural values. These factors are interrelated in a specific way and to understand this, we will take a short detour into what is sometimes called “system thinking in tobacco control.” Ten years ago, the US National Cancer Institute (NCI) published a monograph on this topic (Best, Clark, Leichow, & Trochim, 2007), which acknowledged the complexity of tobacco control at the national level, involving as it does the interplay of factors over long periods of time, including feedback loops. According to experts from the NCI, a government’s willingness to acknowledge and address the smoking problem follows from its level of awareness that tobacco is a problem, and from the balance of lobbying forces that propose or hold back policy solutions. A government’s awareness of the problems associated with tobacco is further affected by specific population factors that are amenable to change. A country’s smoking rate is one of these: as long as the proportion of smokers is high, the government is more likely to be aware that there is a public health risk that needs to be addressed. Changes in the number of smokers also affect public support for tobacco control, which increases when adult smoking rates go down—a process which was believed to be mediated by social norms. Reduced smoking (people quitting or fewer people starting) shifts the balance between smokers and non-smokers, increasing the level of anti-smoking norms and altering public opinion. There is also evidence for the reverse effect, in that people quit smoking when social norms become less accommodating. Together, these population factors determine the context within which national tobacco control policymaking takes place. I have put the main factors together in a simplistic model, which I have called the flywheel model of tobacco control (Willemsen, 2011).

The flywheel model (Fig. 4.1) assumes that population-level factors interact in a circular feedback manner over long periods. The term “flywheel” reflects the notion that the process that moves a population in the direction of a smoke-free society is difficult to set in motion but once begun continues for some time on its own until it loses speed and eventually comes to a stop in the absence of a new impetus (i.e., new tobacco control interventions). The model assumes that as long as the wheel keeps turning, either through new policy input or because the process of denormalisation of tobacco use in society continues, smoking rates will go down. It predicts that policymakers are more willing to introduce tobacco control measures when they are supported by politicians, when the general public and civil society are supportive, when the public thinks more negatively about smoking, and when the prevalence of smoking is low. It assumes a gradual reduction of the proportion of smokers in the population, but eventually will reach its “destination” when smoking rates are at a level that is acceptable to society and government.

Fig. 4.1
figure 1

The flywheel model of Tobacco Control (TC)

Interest groups can influence each element of the flywheel model, with the exception of cultural values. Tobacco control proponents may give the wheel a spin by influencing any of its five sectors, while tobacco interest groups attempt to slow, stop, or reverse the wheel through the same access points. For example, the tobacco industry may develop campaigns to normalise smoking, apply strategies to lower the price of cigarettes in an attempt to offset the effect of tax increases, or present arguments that reduce political support for tobacco control.

The reader is invited to compare the flywheel model with the general conceptual framework in Chap. 1, which includes the same long-term feedback loop as the flywheel model. The flywheel is another way of conceptualising the dynamics of tobacco control, differing from the general conceptual framework in Fig. 1.1, in that it focuses on the population-level sociological factors that drive down smoking rates, resulting from the implementation of tobacco control measures, while ignoring the dynamics of the policymaking process itself.

The flywheel model starts with the implementation of tobacco control measures (TC). Depicted as one factor, in reality it consists of many possible policy solutions to the smoking problem. For example, tax increases make smoking less affordable, directly affecting tobacco consumption, while improvements in the smoking cessation infrastructure and smoking cessation campaigns build confidence in being able to quit and prompt smokers to quit. Tobacco control measures, when properly implemented, can have an impact on population smoking rates (Gravely et al., 2017; Ngo, Cheng, Chaloupka, & Shang, 2017). Some measures work indirectly through social norms (Rennen et al., 2014), particularly smoking bans (Betzner et al., 2012) and mass media campaigns that denormalise smoking (Durkin, Brennan, & Wakefield, 2012). A more detailed discussion of the effectiveness of different tobacco control measures appears in Chap. 7. In the current chapter I discuss the four factors of the flywheel that make up the sociological environment in which tobacco control policymaking takes place: cultural values, social norms, smoking rate, and public support for tobacco control.

Cultural Values

Some of the differences in acceptance and reaction to tobacco control between countries can be explained by cultural values (Hosking et al., 2009; Vogel, Kagan, & Kessler, 1993), which have been found to influence perceptions of tobacco products and smoking (Helweg-Larsen & Nielsen, 2009; Unger et al., 2003). Dominant cultural values and aspects of national culture are stable and relatively insensitive to outside influence, and are therefore located at the heart of the flywheel. Culture is revealed through a set of unique shared values and beliefs that exist for the majority of a population and distinguish it from other populations (Pasick, Onofrio, & Otero-Sabogal, 1996; Schwartz, 2006). According to S. Schwartz (2006), cultural values “shape and justify individual and group beliefs, actions, and goals. Institutional arrangements and policies, norms, and everyday practices express underlying cultural value emphases in societies.” In our context, cultural values determine whether specific tobacco control policy initiatives may fall on fertile soil.

According to a landmark study by Geert Hofstede, who analysed cultural values in more than 50 countries (Hofstede, 1980), Dutch national culture can be characterised as extremely individualistic. People value their freedom to make personal decisions. They expect people to look after themselves and be independent. Personal choice is highly valued. Dutch culture is also a typical example of a “feminine” cooperative culture, according to Hofstede’s research, meaning that negotiation and compromise are considered more appropriate than conflict. The combination of high individualism and high feminism has been proposed as an explanation for why the Dutch smoking ban in bars rested on the assumption that smoking customers would be cooperative, complying for the benefit of the employer who would be fined for non-compliance (feminine value orientation), while resistance to the ban reflected a high individualistic value orientation (Dechesne, Dignum, & Dignum, 2013).

Interestingly, the Dutch also score high on the dimension of indulgence, defined as “the extent to which people try to control their desires and impulses” (Hofstede Centre, 2015). A high score means that the Dutch recognise and respect other’s desires to enjoy life and have fun, which may help explain the recurring wish from policymakers to be considerate to smokers. When Health Minister Ab Klink opened a British American Tobacco (BAT)-financed smoking area in the Dutch parliament building in September 2008, he announced, “this smoking area is a symbol for our two-pronged policy. On the one hand, protection; on the other hand, we don’t want to go so far with regulating that we take away people’s pleasures.” The high score of the Dutch on individualism, femininity, and indulgence was confirmed by S. Schwartz (2006), who used similar orientations, albeit differently worded (respectively, intellectual autonomy, egalitarianism, and affective autonomy). The combination of these values is very alive today, illustrated by the still popular lines of the “alternative” national anthem Fifteen million people: “Fifteen million people/On that tiny strip of earth/You don’t patronize them with laws/You take them for what they are.”Footnote 1

The Dutch version of smoking bans further typify the libertarian and individualistic approach to smoking and the egalitarian, “feminine” value orientation. The bans are more smoker-friendly than in other countries. Many exemptions were included in early formulations, such as providing smoking rooms to accommodate smokers and setting up transitional regimes for sectors where smoking was considered more difficult to enforce. Smokers who disobeyed were not prosecuted, but instead the owner or administrator of the venue or property where the violation took place risked a fine. In one study we compared smoking bans in bars in the Netherlands, Germany, France, and Ireland (Nagelhout et al., 2011). After the implementation of the ban, reports of smoking remained fairly common in the Netherlands and Germany, two countries with lenient policies. In contrast, in Ireland and France where comprehensive bans were introduced with no exceptions and where fines for smokers were in place, smoking was reduced to almost zero, making these policies national successes.

Social Norms

At the core of comprehensive approaches to tobacco control are attempts to “denormalise” smoking rather than merely “controlling” it. By positioning social norms between cultural values and the other factors in the flywheel model, I want to express that societal norms regarding smoking are central to tobacco control. They reflect the deeply held cultural values, and in turn determine the preferences of groups of people for types of policy (Dechesne, Dignum, & Tan, 2011). Tobacco control evolves around how smoking is perceived in society. Normative factors explain why populations differ in their susceptibility to change (Chaiton, Cohen, & Frank, 2003). The famous British social epidemiologist Geoffrey Rose (1992) said, “Social norms rigidly constrain how we live. (…) We may think that our personal life-style represents our own free choice, but that belief is often mistaken. It is hard to be a non-smoker in a smoking milieu, or vice versa.” (p. 90) Smoking rates are indeed lower in constituencies that have an unfavourable “smoking climate” (Kim & Shanahan, 2003).

Changes in social norms have been found to be a driver of tobacco control (Hammond, Fong, Zanna, Thrasher, & Borland, 2006). Denormalisation of smoking reduces tobacco consumption (Alamar & Glantz, 2006; Baha & Le Faou, 2010; Biener, Hamilton, Siegel, & Sullivan, 2010; Hosking et al., 2009) and may result in more smokers quitting (Baha & Le Faou, 2010; Bosdriesz, Kunst, Muntaner, Willemsen, & O’Campo, 2017). Indeed, the second most frequently mentioned reason to quit smoking (after health concerns) is social concerns (McCaul et al., 2006). Best et al. (2007) proposed a feedback loop between social norms and smoking rates to acknowledge an independent process whereby when smoking becomes increasingly unpopular within the wider society, it leads to more people quitting and fewer young people starting, which make smoking even less popular.

The crucial role of social norms in tobacco control was already recognised in the 1970s. At the opening speech of the fourth World Conference on Smoking or Health in Stockholm in 1979, the director of WHO said that tobacco control advocates should try harder to reduce the social acceptance of smoking. Several tobacco industry representatives were present at the session. One of the industry observers wrote a memo, made public by the Norwegian Association on Smoking and Health, revealing that the centrality of social norms as mentioned by the WHO director was not new to the industry but was “just a confirmation of our own analysis that the social acceptability issue will be the central battleground on which our case in the long run will be lost or won” (Clairmonte, 1983, p. 85). More than a decade later, Philip Morris complained in an internal memo that smoking bans were not only hurting business, but that they had “a more important effect (…) on the social acceptability of smoking. Attempts to depict tobacco use as anti-social get a powerful boost when its use is banned in social settings. This impact on our business, whilst slower, is just as real” (Goldberg, 1999). Philip Morris’ PR firm formulated this in 1990 as follows: “ Social acceptability is ultimately the bedrock upon which the industry’s long run survival depends” (Burson-Marstellar, 1990). Indeed, in the 1990s the tobacco industry fought relentlessly with governments over the right to smoke in public.

The centrality of social norms in tobacco control has long been known to Dutch policymakers as well. In 1975 the Dutch Health Council advised that “activities against smoking must primarily aim at the creation of a psycho-social climate in which smoking is perceived negatively and a new attitude towards smoking emerges” (Beernink & Plokker, 1975). In 2000, Health Minister Borst defined her approach to tobacco control in parliament as follows:

At the core of the [tobacco control] policy is the objective that few young people start smoking and that smoking is increasingly seen as an abnormal behaviour. Tobacco needs “ denormalisation” in our society. This and other measures must lead to a social climate where non-smoking is the social norm and not starting or quitting smoking is the result.Footnote 2

It is one thing to recognise that social norms are important, but it is another thing to have good data on how a country’s social norms compare with those of other countries. For decades the tobacco industry had an information advantage. The big tobacco multinationals were able to organise worldwide comparison studies, cleverly exploiting the fact that they were present in a large number of countries across the world. In 1979 researchers who worked for tobacco manufacturer Philip Morris International analysed the social and political environment of that business in 27 countries (Unknown (Philip Morris), 1979). They remarked of the Dutch, “As the personal freedom concept is widely accepted and supported in Holland, the anti-smoking cause is not exceptionally strong. … Members of the medical profession and government appear to have highly individual opinions and the consensus is that smoking is a matter of personal choice.” Twenty years later, Philip Morris’ analysts remarked that “the Dutch resent government interference, [and] the public debate is more and more balanced,” and they characterised opinion as tolerant towards smoking (Philip Morris, 1996). Population survey data collected for Philip Morris further showed that social acceptance of smoking in the Netherlands was still high in 1997 (GfK Great Britain, 1998). Only 19% of non-smokers believed there was any element of risk about being around smokers in bars or pubs, or of “living with a smoker” or “working with a smoker.” This was one of the lowest results in Europe. The report also noted the low demand for government action against smoking among the Dutch.

It is only fairly recently that not only the tobacco industry but also the Dutch tobacco control community has come to realise that social norms regarding smoking are still more lenient than in other countries. Since 2008 the Netherlands participates in the International Tobacco Control (ITC) evaluation study. ITC data revealed that Dutch smokers were less often aware of societal disapproval of smoking than smokers in other high-income countries, at 63% compared to between 72% and 89% (ITC Project, 2015). Furthermore, only 22% of Dutch smokers often thought about the harm they did to themselves, while 33% of smokers in Germany, and between 43% and 56% in other high-income countries, did. The percentage of smokers who often thought about the harm they might inflict on others was extremely low in the Netherlands too (9%), and the Netherlands had the second-lowest percentage (21%) of smokers with a “negative” or “very negative” opinion of smoking, among the 13 ITC countries where this was measured (the German figure was 20%). All other countries scored between 45% and 62%.

How can the low concern among Dutch smokers and relatively tolerant norms towards smoking be explained? One explanation is the previously discussed “feminine” cooperative culture in the Netherlands that supports a tolerant approach towards smokers, which does not go well with confrontational media campaigns. For example, in the 1980s, the “Meinsma approach” (relentlessly hammering on health risks; more on Lenze Meinsma in Chap. 9) was replaced by a “more positive approach, where the advantages of non-smoking as part of an attractive lifestyle, are promoted” (WVC, 1984). In 1986, the government stated that awareness campaigns were important, but it wanted non-governmental organisations to run them, because the government said it was “handicapped,” hinting to the societal and political sensitivity of paternalistic lifestyle campaigns (WVC, 1986, p. 174). Since the 1950s, when the serious health consequences of smoking became clear, only one health risk campaign has ever been run in the Netherlands. This was part of a EU-funded project that made it possible to adapt the Canadian campaign Joanne for use in Dutch cinemas, featuring a young girl looking in a mirror, watching in horror as her face wrinkles and turns grey because of lifelong smoking. This campaign made a tremendous impression on smokers, who still recalled the campaign, many years later, when asked to give examples of anti-smoking campaigns.

The lack of hard-hitting media campaigns to deter smoking is a remarkable aspect of Dutch tobacco control, since many countries run confrontational anti-smoking media campaigns. They are an integral part of national tobacco control strategies and not regarded as particularly problematic in other countries. There is abundant evidence that, at least in the field of tobacco control, campaigns that make an emotional appeal can be effective (Biener et al., 2006; Borland & Balmford, 2003; Durkin et al., 2012; National Cancer Institute, 2008; Timmers & Van der Wijst, 2007; Wakefield, Loken, & Hornik, 2010). The Australian “Every cigarette is doing you damage” campaign targeted at adult smokers to quit is notorious in this regard. Campaign exposure has been associated with increased negative thoughts about smoking (Borland & Balmford, 2003) and contributed to reductions in smoking prevalence in Australia (Wakefield et al., 2008). This campaign was adapted for use in other countries as well, for example, in 2003 in Norway. Box 7.1 (Chap. 7) discusses the Dutch government’s reluctance to run such national media campaigns to deter smoking, which has to do with the belief that media campaigns offer expensive but ineffective ways to influence lifestyle.

Smoking Rates

When the ratio of smokers to non-smokers changes in favour of non-smokers, public support for policy restrictions increases. For example, when Health Minister Els Borst introduced her revision of the Tobacco Act in the parliament, she legitimised this by referring to the fact that already two-thirds of the adult population was non-smoking by then.Footnote 3 Countries with relatively few smokers, like Finland, the United Kingdom, and Australia, have the most comprehensive tobacco control policies. “Policy follows prevalence” said Kenneth Warner, an international tobacco control expert (quoted in D. J. Reid, Killoran, McNeill, and Chambers (1992)), and several ecological studies which used the country as a unit of analysis found that European countries with more stringent tobacco policies have fewer smokers—although the association is not very strong (Martinez-Sanchez et al., 2010; Willemsen, Kiselinova, Nagelhout, Joossens, & Knibbe, 2012). The direction of causality is not clear and obviously goes in two directions, as is captured in the flywheel model. In any case, within the problem stream of policymaking, data on the proportion of smokers among adults and adolescents constitute a crucial element in policymakers’ appreciation of the tobacco problem (see Chap. 10 on problem identification).

From an epidemiological perspective, tobacco’s worldwide spread is, as depicted by WHO, an epidemic (Roemer, 1982). The tobacco epidemic took many decades to unfold and will take even more time to resolve. Countries progress through the various stages of the epidemic in remarkably similar ways. A widely acclaimed model describes how the epidemic diffuses through populations (Lopez, Collishaw, & Piha, 1994; Thun, Peto, Boreham, & Lopez, 2012). In developed countries during the 1950s and 1960s, more than half of the population smoked, while in the higher strata of these societies almost all men smoked. High-income and male subpopulations were the first to become addicted to smoking, followed by lower income groups and females two decades later. After a period of gradual increase in smoking, prevalence reached a peak 40–50 years after onset, then slowly diminished. This pattern is clearly discernible in the Netherlands. In 1958 the cigarette epidemic had already reached its peak in the male population: 90% of men and 29% of women smoked. The proportion of male smokers in the Netherlands was exceptionally high: for example, in the United Kingdom 60% of the male population smoked in the 1950s and 1960s (Reid et al., 1992). Figure 4.2 shows how the proportion of smokers in the population has gone down since 1957.Footnote 4 The decline follows a similar pattern as in other developed countries, marked by a fast decline in the 1960s and 1970s and a slower decline from the 1980s onwards (OECD, 2014; Thun et al., 2012). In the Netherlands the percentage of women smoking peaked around 1970, followed by a reduction and convergence to male smoking levels in later decades. Internationally, the steep decline in smoking between 1960 and 1980 has been explained by widespread media attention to official governmental reports which showed that smoking causes death and disease (Farquhar, Magnus, & Maccoby, 1981; Reid et al., 1992). The Dutch public was confronted with similar messages in the media (see also Chap. 10) and many quit smoking despite the lack of governmental campaigns, while tobacco lost its aura of innocence and politicians called on the government to act.

Fig. 4.2
figure 2

Smoking prevalence in men and women since 1958. Sources: STIVORO (2012); Verdurmen, Monshouwer, and van Laar (2015). Note: whether one is a smoker or not is discerned by asking: “do you (ever) smoke or do you never smoke?”

In the 1980s smoking prevalence in the Netherlands was relatively high compared to other EU countries (European Commission, 1987). While the EU average was 37% in 1987, 44% of the Dutch population still smoked; only Denmark was higher with 46%. Although the decline in the general smoking rate followed roughly the same path in the Netherlands as in other developed countries (see Fig. 4.3 for a comparison with England and Canada), there was a noteworthy increase in the proportion of smokers between 1988 and 1996 (from 32% to 35%) in the Netherlands. Four years later, in 2000, the smoking rate was still higher than in 1988. In those years the Netherlands was very much a smoker’s country. With 2951 cigarettes consumed per adult per year, the Netherlands ranked third highest in the EU for consumption, comparable to Russia and Greece and much higher than neighbours Belgium, Germany, and the United Kingdom (Gallus, Schiaffino, La Vecchia, Townsend, & Fernandez, 2006). The proportion of male smokers in 2002 was still slightly higher than the EU-25 average, while female smoking was among the five highest (Zatoński, Przewoźniak, Sulkowska, West, & Wojtyła, 2012).

Fig. 4.3
figure 3

Trends in adult smoking prevalence (men and women combined) in the Netherlands, the United Kingdom, and Canada. Sources: United Kingdom: General Lifestyle Survey; Canada: Canadian Tobacco Use Monitoring Survey (CTUMS); Netherlands: Dutch Continuous Survey of Smoking Habits (COR). Note: The way smoking prevalence is measured differs slightly between countries. In the United Kingdom, all aged 16+ were asked, “Do you smoke cigarettes at all nowadays?” while in the Netherlands all aged 15+ were asked the question, “Do you ever smoke or do you never smoke?” In Canada all 15+ aged were asked, “At the present time do you smoke cigarettes every day, occasionally, or not at all?” Smokers are daily smokers and non-daily smokers combined

Figure 4.3 shows how smoking rates declined in the Netherlands, the United Kingdom, and Canada. While smoking rates were about the same around 1988, in later years the United Kingdom and Canada did better. The long-term background rate fell by less than 0.5% per year between 1990 and 2010 in the Netherlands (Willemsen, 2010), while prevalence fell by around 0.75% in Canada and 0.7% in the United Kingdom (Royal College of Physicians, 2016).

Some other countries (the United Kingdom, Australia, Canada, and the United States) also witnessed a faltering in the decline of smoking in the beginning to mid-1990s (Wakefield & Chaloupka, 1998), but not as distinct nor as prolonged as in the Netherlands, where it continued until 2000. It is no coincidence that this 12-year period of standstill coincided with a period of virtually no tobacco control interventions that might have impacted on smoking rates, while tobacco industry influence was strong both behind the scenes and in the media (Willemsen, 2017). Tobacco industry tolerance campaigns re-normalised smoking (see Chap. 8). The tobacco control flywheel had clearly lost its energy. The period of stagnation ended when 800,000 smokers made a quit attempt during a large-scale mass media quit campaign at the turn of the millennium. The revised Tobacco Act in 2002 led to a further reduction of the smoking rate from 31% in 2002 to 28% in 2004, after which it continued to fall. The reduction in the Netherlands between 1999 and 2009 (−18.7%) was almost the same as the Organisation for Economic Co-operation and Development (OECD) average reduction (−17.9%) (Bruggink, 2013). After the “correction” initiated by Minister Borst, the Netherlands was on a trajectory towards a non-smoking society at about the same pace as most other developed countries, but in more recent years smoking rates have seemed to rise again, from 24.5% in 2012 to 26.3% in 2015 (Van Laar & Van Ooyen-Houben, 2016).

Between 1992 and 1996 youth smoking increased (Fig. 4.4), despite a shift from daily smoking to less frequent smoking in this period (Willemsen, 2005). Since 1996, youth smoking rates declined almost uninterruptedly. “Regular smoking” in youth (10- to 19-year-olds) is defined as having smoked at least once in the past month.

Fig. 4.4
figure 4

Trends in youth smoking (10–19 year olds) between 1992 and 2013 in the Netherlands. Sources: STIVORO (2012); Verdurmen, Monshouwer, and Van Laar (2014)

The reduction in adolescent smoking seems to follow the general trend seen earlier in the adult population, including a period of stagnation in the 1990s. When adult smoking rates go down, youth rates follow; when adult smoking goes up, youth smoking again follows. Some have noted this as the most likely explanation for reductions in youth smoking (Chapman, 2007; Gielkens-Sijstermans et al., 2009; Hill, 1999). Young people are particularly sensitive to changes in what is regarded as “cool” in the wider society. Smoking has become less attractive since tobacco advertising was banned in 2002, and smokers were increasingly seen as social pariahs when smoking was banned in workplaces in that same year. Part of the explanation may also be that when parents quit smoking, fewer children are exposed to tobacco products at home and parents no longer are exemplars for smoking (Den Exter Blokland, Engels, Hale III, Meeus, & Willemsen, 2004).

Public Support for Tobacco Control

The flywheel model assumes that, at least in modern democracies, the adoption of policies reflects what the broader society wants. Tobacco policy generally reflects shifts in public opinion (Kagan & Nelson, 2001). Empirical evidence from US states has shown that when public opinion becomes more supportive of smoking bans, states are more likely to adopt them (Pacheco, 2012). Politicians know this and push for more stringent measures when they feel that society is ready. The battle is thus fought first in society, after which it moves to the political arena. When public opinion changes, politicians follow. As one former civil servant put it, “It is a very slow process to get societal support. Politics usually follow trends in society, because this assures that you remain in office and can come back again … it rarely happens that policy makers actively want to change public opinion.”Footnote 5

The need to have support from the general public is well known among Dutch politicians and policymakers. They routinely refer to public support when they defend or reject tobacco control policy proposals. For example, State Secretary for Health Joop van der Reijden (VVD) explicitly made decisions to intensify tobacco control contingent on the political discussion in parliament, which “shall made clear whether the climate is ripe for a really powerful policy” (WVC, 1986). In those days, the civil servants at the Ministry of Health who developed the first tobacco control policy measures complained that they did not feel supported by society or the medical sector.Footnote 6 Twenty years later, when Health Minister Borst defended her tobacco control bill in the senate, she said, “I believe that exceptions [to the ban] can only disappear when we have a totally different culture in the Netherlands, a culture in which nobody, exceptions granted, smokes. … We must not make things look nicer than they are. So we have to make exceptions for certain sectors.”Footnote 7 In 2005 the government looked back and remarked that the 1999 revised Tobacco Act “attempted to catch up with the changes in how society regards smoking” (VWS, 2005). State Secretary Martin van Rijn said in 2014, “A basic assumption [of my policy making] is that I want to carry a consistent and effective policy that has support from society” (Van Rijn, 2014). Even Dutch parliamentarians who are vehemently in favour of stricter tobacco control cannot easily propagate policies that lack general support from the public. Carla Dik-Faber, parliamentarian for the Christian Union (CU), reacted to the idea of banning smoking on terraces: “This topic is very much debated in society. At the moment there is insufficient political support. I can imagine that at one point terraces will become smoke-free. However, it is still too early for this. Political decision making must follow developments in society” (Van der Laan, 2015).

In 1987, when most EU countries had not yet adopted major tobacco policy measures, the European Commission (EC) wanted to know how supportive Europeans were of tobacco control measures (European Commission, 1987). The Dutch population answered somewhat below average on all measures, with a relatively low level of support for an advertising ban. The low support for tobacco control in the Netherlands has become more pronounced in later years. A Eurobarometer poll from 2005 showed that the Dutch were not very supportive of a smoking ban in bars (TNS Opinion & Social, 2006): only 46% of the total population was “somewhat or totally in favour”—one of the lowest ever levels of support in the EU. In 2009 when the EC conducted another poll (TNS Opinion & Social, 2010), the Netherlands emerged as a country with little general support for tobacco control measures, scoring the absolute lowest on plain packaging and on banning the sale of tobacco via the internet. The poll was repeated in 2012 (TNS Opinion & Social, 2012) and showed little change in the Dutch position: the Netherlands still had the least support for plain packaging of all 27 EU countries.

What might explain such low levels of support? According to the flywheel model, public support reflects dominant social norms in society. Social norms depend on knowledge about the problems associated with smoking (particularly from passive smoking) and level of acceptance of these problems by the public. Given that the level of concern is relatively low in the Netherlands, as was shown previously, one would expect that support for tobacco control is also low. This is indeed the case. I already referred to the striking data from the ITC project about the relative lack of concern among Dutch smokers about tobacco and health. I found a strong correlation in Dutch national survey data between believing that passive smoking is harmful and support for smoking bans (Willemsen, 2006). This association was consistent among both highly educated and lower educated groups—an association also found in other countries. With some colleagues I analysed Eurobarometer data from 2009, and what stood out was that smokers who lived in countries with comprehensive tobacco policies were more likely to support tobacco control measures, and such support was greater when they were more concerned about whether their smoking harms non-smokers (Willemsen, Kiselinova, et al., 2012). In an older study we had found that Dutch non-smoking employees were more likely to ask co-workers not to smoke when they had more negative beliefs about the health consequences of passive smoking (Willemsen & De Vries, 1996). Compliance with smoking bans is higher if smokers are more supportive of them and if they are more aware of the health consequences of passive smoking (G. E. Nagelhout, de Vries, et al., 2012).


Smoking rates have declined following patterns similar to those in other developed countries, typified by a fast decline in the 1960s and 1970s and a slower decline from the 1980s onwards. However, during the 1990s smoking rates stagnated then rose again. In these years the Netherlands was a smoker’s country, with more cigarettes consumed than in almost every other EU country. Smoking was socially well accepted, which might be partially attributed to the success of tobacco industry’s tolerance campaigns from 1970 until the end of the 1990s (see Chap. 8 for details about industry tolerance campaigns). These were exceptionally well received in the Netherlands, since the Dutch cherished the collective idea of being a tolerant people. There was a 12-year standstill in smoking rates (1988–2000), which coincided with virtually no action from the government to regulate tobacco. Only after the revised Tobacco Act was implemented in 2002 did people start to quit again. Smoking rates have continued to go down since then, and smoking among youth followed the example of the adults and also has gone down.

Societal support for most tobacco control measures has been relatively low compared with other EU countries. Support was lowest for confrontational tobacco control elements, which might be explained by “feminine” cooperative value orientations in Dutch national culture. The smoking bans implemented in 2004 were more smoker-friendly than those in other countries, reflecting such values. The Dutch government has been reluctant to run health awareness media campaigns, which are seen as ineffective, costly, and paternalistic. The lack of health awareness campaigns contributed to the relatively low levels of concern about smoking and the less than optimal support for tobacco control in even today’s society.