Making Tobacco Control Policy Work: Rules of the Game

Tobacco policy is made by actors who operate in an institutional environment with specific characteristics (Scharpf, 1997), and differences in these characteristics can explain much of the variation in tobacco control between countries. Governments and parliaments, and their bureaucracies, have formal and informal “rules of the game.” Informal rules include conventions, unwritten procedures, and expectations. Formal rules are official and legal procedures. While the contextual factors discussed in Chap. 4 are relatively dynamic and amenable to change by tobacco control interest groups, the rules of the game that are the subject of the current chapter are more static. In terms of the Advocacy Coalition Framework, they are the constitutional structural factors (Breton, Richard, Gagnon, Jacques, & Bergeron, 2008; Sabatier, 2007) which determine how countries differ from each other—less about how countries change over time, although institutional factors and conditions can gradually change. Knowledge of these factors is not only key to understanding why tobacco policymaking is most of the time a tedious and slow process; it also helps to understand how tobacco interest groups may influence policymaking and why some groups are more successful at this than others.

Making Tobacco Control Policy Work: Rules of the Game Tobacco policy is made by actors who operate in an institutional environment with specific characteristics (Scharpf, 1997), and differences in these characteristics can explain much of the variation in tobacco control between countries. Governments and parliaments, and their bureaucracies, have formal and informal "rules of the game." Informal rules include conventions, unwritten procedures, and expectations. Formal rules are official and legal procedures. While the contextual factors discussed in Chap. 4 are relatively dynamic and amenable to change by tobacco control interest groups, the rules of the game that are the subject of the current chapter are more static. In terms of the Advocacy Coalition Framework, they are the constitutional structural factors (Breton, Richard, Gagnon, Jacques, & Bergeron, 2008;Sabatier, 2007) which determine how countries differ from each other-less about how countries change over time, although institutional factors and conditions can gradually change.
Knowledge of these factors is not only key to understanding why tobacco policymaking is most of the time a tedious and slow process; it also helps to understand how tobacco interest groups may influence policymaking and why some groups are more successful at this than others. Groups of countries that are politically and culturally similar and have comparable institutional arrangements are sometimes referred to as "families of nations". Policies in countries that are similar with respect to culture, institutional make-up, and economic development tend to converge (Lenschow, Liefferink, & Veenman, 2005). Convergence between countries within the same family has been shown to occur for tobacco control, with similar countries choosing similar policies and adopting policies at roughly the same tempo (Studlar, 2007). The Netherlands is considered part of the continental (Western European) family, consisting of countries like Belgium, Germany, and Austria (Castles & Obinger, 2008;Obinger & Wagschal, 2001). Other families are an Anglo-Saxon (English-speaking) group (Australia, Canada, New Zealand, the United Kingdom, the United States), a cluster of Scandinavian countries, and a peripheral group of European countries (Italy, Spain, Portugal, Greece). The continental family, to which the Netherlands belongs, is characterised by the central position of Christian Democrats, a "politics of the middle way" (combining social democratic and liberal principles), and political cultures that are oriented towards bargaining and compromise seeking. This chapter describes this system. It starts with a discussion of the two most important aspects of Dutch policymaking: corporatism (pressure groups incorporated into the policymaking process) and the continuous need to seek consensus in a multi-party political system. The reader's attention is directed to the tension between party ideology and tobacco control policymaking. Separate sections deal with the process of drafting new legislation and the opportunities available for advocates to influence this process through lobbying, either directly (targeting civil servants of the ministry), or through parliament. Finally, our attention turns to two long-term processes that gradually undermined the position of the central government as the main producer of tobacco control: deregulation and decentralisation.

Corporatism
Dutch pragmatic policymaking is sometimes called interactive governance, but most often it is referred to as corporatism (Van Tulder, 1999). The Netherlands has a conservative-corporatist model of the welfare state (Fenger, 2007) and is a prime example of a corporatist country, where policymakers incorporate interests of legitimate stakeholders into the decision-making process. In a review of 23 studies that assessed the level of corporatism in developed countries, the Netherlands ranked fourth on corporatism after Austria, Norway, and Sweden (Siaroff, 1999).
Corporatism has been defined as "institutionalized and privileged integration of organized interests in the preparation and/or implementation of public policies" (Christiansen et al., 2010). In a corporatist structure, government and interest groups cooperate rather than compete (Andeweg & Irwin, 2009). Pressure groups do not need to influence the government from the outside since they are incorporated in the policymaking process-hence the name "corporatism." If the government formulates a new policy, parties such as trade unions, business representative organisations, and societal organisations-with an obvious interest in the policyare routinely invited into the policymaking process. Some groups obtain a privileged monopoly over a policy topic, and these chosen few enjoy close relationships with state representatives. The government bargains and seeks compromise with these groups, trusting that this will result in "better" policy outcomes-more equal, flexible, or more effective. In corporatist systems a government is most likely to consult the tobacco industry on new tobacco control initiatives. This contrasts with pluralistic democracies, where interest groups freely interact with each other and compete for power (a "may the strongest party win" approach) and influence policy from the outside. Sometimes the term "lobbyism" is used for pluralist systems (Rommetvedt, Thesen, Christiansen, & Nørgaard, 2012). Outsider anti-tobacco lobbying groups thrive best in open and competitive pluralist environments (Studlar, 2007). Countries with strong pluralistic characteristics are New Zealand, Australia, the United Kingdom, Italy, Ireland, Canada, and the United States (Siaroff, 1999).
A feature of the corporatist system is that interest groups are more or less routinely invited by the government to comment on policy proposals. In the Netherlands, involvement of stakeholders is historically organised in an informal manner. Consultation is confidential and open only to invited stakeholders. This practice has benefitted the tobacco industry tremendously in the past (see Chap. 8 for examples), but is currently changing. In 2011, the cabinet announced that it wanted to use open internet consultation rounds more often, in addition to the traditional practice of invitation-only consultation-although the ministries are still only committed to submit 10% of their regulatory proposals for internet consultation. Recent consultations on tobacco covered the implementation of the European Union (EU) tobacco product directive, adaptations of tobacco taxation, and an intended ban on the display of tobacco products. 1 In the Netherlands corporatism is grounded in Christian Democratic values and associated political principles, two of which are particularly relevant. The first is subsidiarity: the principle that decision-making should take place at the lowest possible (administrative) level. It means that regulation by the state is only appropriate if it is not possible to solve matters at regional or local levels. The second principle is the idea that het maatschappelijke middenveld (civil society) is needed for political support and to have effective policies (Andeweg & Irwin, 2009, pp. 177-178). This goes back to the nineteenth-century Calvinist doctrine of soevereiniteit in eigen kring (sovereignty in one's own circle), the idea that organisations determine their own fate independent of government. Christian Democrats believe that organisations such as churches, unions, welfare institutions, and voluntary organisations must not become too dependent on government support but must retain responsibility and have their own decision-making powers, and that this will benefit social cohesion. These historical principles were re-introduced by Prime Minister Balkenendean influential ideologist in the Christian Democratic Party's think tank-. His idea that government must work together in partnership with business and society was a guiding principle in the 2002-2010 cabinets. In 2007 Health Minister Ab Klink, former director of the think tank of the Christian Democratic Party and co-author with Balkenende on CDA position papers, introduced it as the cornerstone of his prevention policy. During the debate on the Tobacco Act on 31 May 2001, Siem Buijs of the Christian Democratic Party (CDA) debated with Rob Oudkerk (Labour Party) on the need for an advertising ban instead of continued selfregulation. Oudkerk attacked Buijs for hiding behind self-regulation and being naïve about the real intentions of the industry: "Keep on dreaming, CDA, this is not how it works," Buijs replied: You keep on dreaming, Mister Oudkerk, that we can influence behaviour in this country through laws and regulations. Forget it! We see this in many places. This is not what the CDA wants with shared responsibility. … We want behaviour to be influenced from the bottom up. It should not be controlled from the top down through laws, after which you can relax because everything is settled. I believe that is a sham. … I want to say this to you: "I have a dream." It is a nice dream and I want to make it true. It assumes behavioural change and attitude change and not an iron rod and a stick.
In corporatist arrangements, if an interest group is not recognised by the government as a legitimate negotiation partner, it can be a voice in the wilderness. This is what happened to the Cancer Society's director Dr Meinsma in the 1970s, and explains why the confrontational, more aggressive lobbying strategy that was adopted around 2008 by the Dutch tobacco control coalition was ultimately ineffective and eventually contributed to the downfall of STIVORO (see Chap. 9). The national corporatist character explains why the Netherlands does not have a strong tradition of the type of grass roots activist movements that were crucial to advancing smoke-free legislation in the United States and Australia. This is not to say that civil lawsuits and citizens' initiatives play a marginal role in the Dutch system. They have an important role in voicing societal concerns and may occasionally succeed in putting tobacco on the political agenda, especially when they use the weapon of legal action (see the section on venue shopping, later in this chapter).

Consensus seeking
The Netherlands has long been one of the purest examples of a consensus democracy (Lijphart, 1999). This means that the political landscape is so fragmented that no single party is likely to achieve a majority position in parliament. This sets the Netherlands apart from majoritarian democracies such as the United Kingdom and the United States, where governments can impose policy in a top-down manner as they do not need to obtain consensus from rival political parties (although this has changed somewhat in the United Kingdom in more recent years). The fragmented Dutch political landscape has its origins dating back to times when the Netherlands was divided into the four religious or ideological "pillars": Catholics, Protestant-Christians, Socialists, and liberals or neutrals. Because no group was large enough for a majority, they needed to cooperate to maintain a stable and viable society. This was done by a process of consultation between spokespersons from the four groups. Specific but mainly informal "rules of the game" were developed and abided by to safeguard democratic stability (Lijphart, 1999).
The four pillars no longer exist, but the political landscape is still fragmented. After the Second World War the lower chamber of parliament consisted of six political parties roughly conforming to the traditional pillars, but the number of parties had increased to 13 by 2017. Since governments prefer to have a majority in both chambers of parliament, they usually need two or three parties, sometimes even four, to form a government.
Minority cabinets are rare in the Netherlands. Coalitions are constructed in such a way that the government has an obedient majority and a relatively powerless minority in parliament. This can work in two ways. If the government presents a coalition agreement that includes a strong tobacco control paragraph, parliament can only delay its passage as the paragraph is accepted by the parties that make up the majority. Parliament cannot vote against it without risking the fall of the cabinet. On average, less than one government bill is defeated by parliament per year (Andeweg & Irwin, 2009). However, if a strong tobacco control paragraph is missing from the coalition agreement and the government is unwilling to advance tobacco control, the power of the parliament to push for tobacco control is extremely limited because it has little chance to obtain sufficient votes to reach a majority.
Because of its multi-party nature, the cabinet must seek approval from a coalition of parties for important decisions, especially about sensitive matters. Because of the continuous need to secure support from parties that have different ideologies and interests, yes or no decisions are avoided and politicians go to great lengths to avoid having identifiable winners and losers, since this could create resentment and would cause problems for future decisions where one might need support from the opponents of the day (Koopmans, 2011). The Dutch version of consensus-based policymaking became widely known as the polder model during the late 1990s, when employer organisations, trade unions, and the government made compromises regarding wages and social security, and the astonishing positive effect on the national economy attracted international attention. Nowadays polderen is used more broadly, referring to the process of finding compromises between political parties or between government and civil society. The first policy reflex to a new social problem is still to find a solution that does right by as many stakeholders as possible.
The Christian Democratic ideals of subsidiarity and the reliance on support from civil society and the associated tradition of polderen mean that informal self-restraint and self-regulatory agreements with industry and other interested parties are preferred. The government feels most comfortable with voluntary agreements by the sector as the first policy option, before eventually proposing and trying legislation. This has happened time and again in tobacco control policy making, and continues to be the first reflex of Dutch politicians, many of whom believe it has advantages. These were summarised in 1996 by the minister of economic affairs in a debate in parliament over tobacco policy: Self-regulation certainly has advantages: it makes it possible to have tailormade solutions for each sector, it creates flexibility, and changes in agreements can be made relatively quickly because it is not necessary to come to formal changes in law. A disadvantage is that the legal status of agreements often is unclear. 2 In a parliamentary debate in 2001 about the Tobacco Act, Minister of Health Els Borst called self-regulation "the royal way" for government. 3 The preference for self-regulation is held by political parties in the centre and on the right flank of the Dutch parliament (who usually hold the majority of seats), while parties from the left question it as they are more suspicious of the intentions of businesses.

Consensus Seeking in the Cabinet
Dutch ministers wear two hats, one as a minister who has tasks, responsibilities, and challenges unique to his or her department, the other ideological, resulting from representing a political party. After the breakdown of the ideal of pillars, consensus seeking became concentrated more and more within the cabinet itself, where departmental ministers compete and negotiate with their ministerial colleagues and political ideologies are taken into consideration. From the 1960s on, the cabinet became increasingly politicised (Andeweg & Irwin, 2009, pp. 142-145). A state secretary and a minister in the same department are often from different political parties, so that the coalition parties are represented in a balanced manner in the cabinet. Health ministers were most often from liberal or Christian parties (Table 5.1). Few have had a background in health: from 1967, this has been the case in only 6 of 15 instances, and none has had a medical profile since 2002 (Table 5.1). Another point to note is that until 1994 tobacco control was handled by state secretaries (junior ministers). This gave tobacco control a none too strong negotiating position, because state secretaries are not present at the weekly cabinet meetings and do not have the right to vote. They attend only when a topic in their area is being discussed.
The cabinet is further politicised through the practice of holding party meetings of political parties' ministers and state secretaries with their leaders and party chairpersons in the second and first chamber the day before each weekly cabinet meeting to prepare for next day's agenda. Party discipline may be enforced so that ministers vote along party lines.

Coalition agreements
After an election, the political leaders of the parties of a new ruling coalition draft a set of policy intentions. As in other multi-party countries, such coalition agreements are the result of a sometimes long and difficult negotiation between party leaders and are subject to extensive lobbying from interests groups. The outcome is "a register of policies that coalition parties wish ministers to implement" (Moury, 2011). About two-thirds of pledges formulated in Dutch coalition agreements are transferred into government decisions, according to one study that examined this for two cabinets (Lubbers III, 1989-1994Kok II, 1998 (Moury, 2011). Coalition partners tend to negotiate policy intentions for the next four years at a rather high level of detail so that there is little room for the new government to decide on new policy except for unexpected issues and responses to crises. If we disregard decisions about things such as crises, Dutch ministers make less than 20% of policy decisions spontaneously. They must promise to adhere to the coalition agreement before they are sworn in, and they sometimes refer to it as their "Holy Bible" (Andeweg & Irwin, 2009). 4 Jeroen Dijsselbloem, Labour Party politician and former Minister of Finance, described such a coalition agreement as "a sad thing. It lists many far-reaching policy intentions, intentions that are devised and negotiated by a small group of political insiders behind closed doors. Civil society and science are not taken into consideration, so that it has no nonpolitical checks and balances" (Slob & Staman, 2012). Tobacco control was mentioned three times in the coalition agreements of the 15 cabinets since 1972 (Elsevier, 2010; Van den Braak & Van den Berg, 2017). The first was in the second cabinet of Labour Party Prime Minister Kok (1998Kok ( -2002. Tobacco policy was called a priority: tobacco control was to be intensified and it was announced that when the code of conduct regarding tobacco advertising ended (in May 1999), the European directive on tobacco advertising would be implemented. This gave Health Minister Els Borst a strong mandate to realise her Tobacco Act, since it meant that the Liberal-Conservative Volkspartij voor Vrijheid en Democratie (People's Party for Freedom and Democracy) (VVD), part of the ruling coalition, was committed to the agreement and could not put up much of a fight. The second time that tobacco was mentioned was in the fourth cabinet of Christian Democratic Prime Minister Balkenende (2007)(2008)(2009)(2010), in which the Labour Party also participated. The tobacco control lobby succeeded in getting a sentence about a smoke-free hospitality sector in the coalition agreement that said that, in collaboration with the hospitality sector, the government would work towards a smoke-free hospitality sector. The next coalition between VVD and CDA (Rutte I) announced in 2010 in its coalition agreement that the smoking ban for bars would be relaxed because "in many small pubs, where there is no personnel employed, there is no need for a smoking ban." Rutte explained it in his public statement about the coalition agreement: "We give responsibility back to the people in this country. No patronising, no untenable smoking bans in small cafés, and furthermore no unnecessary regulations" (Elsevier, 2010, p. 176). This was the outcome of a deal between the VVD and the CDA. The CDA would give up its wish for smoke-free bars in exchange for an assurance that it would be compensated in other dossiers. 5 seCuring long-term tobaCCo Control poliCy Dutch tobacco control policymaking is a drawn-out process. The Ministry of Health pushes for a stricter and more effective tobacco policy, supported by tobacco control organisations. This is most effective when there is sufficient collective memory and capacity within the bureaucracy to support long-term policymaking, independent of the whim of the day and the ever-changing ideologies of Dutch governments. However, this capacity is restricted. In the past, institutional policy continuity was supported by the fact that high-level officials remained at their posts when governments changed, so that policy knowledge was preserved. However, since the emergence of the Algemene Bestuursdienst (senior civil service) at the end of the 1990s, top-level bureaucrats (secretaries-general, directors-general, and inspectors-general) are supposed to change position every four or five years in order to reduce compartmentalisation. This has led to the emergence of career civil servants: managers with little affinity for the subject matter. According to one person, "that's the big story of these top levels who rotate faster and faster so there's less and less collective memory. This plays into the hands of the industry." 6 In more recent years, institutional memory within the tobacco control unit of the Ministry of Health has been further hampered when civil servants from lower levels also changed positions. Between 2008 and 2013 there was much turnover, with civil servants "doing" tobacco control for short periods of time. 7 One way of overcoming problems of such institutional amnesia is to rely on organisations and individuals outside the governmental bureaucracy to provide continuity and the preservation of knowledge (Smith, 2013). In the Netherlands, STIVORO traditionally fulfilled this role, and new civil servants in the Ministry of Health could quickly become familiar with the tobacco control "dossier" after a few meetings with experts from this organisation. 8 One mechanism that supports policy continuity from one administration to the next is the practice of introductory dossiers, presented to a new minister on the first day in office. They are written by high-level civil servants and are a list of policy intentions, unresolved issues, and pressing matters that need to be solved in the short-and mid-term. Introductory dossiers are also an attempt by the bureaucratic system to assure that a minister stays as close as possible to mid-and long-term policy issues.

Compromises between ministries
While tobacco control policy is subject to compromises between political parties and between the government and interest groups, compromises are also sought between different ministries: particularly between the Ministry of Economic Affairs, which protects business interests, and the Ministry of Health, which protects public health. Until the mid-1990s tobacco policy was jointly determined by these two departments, after which it became the prime responsibility of the Ministry of Health. Health Minister Borst's tobacco policy document in 1996 was the last one cosigned by both ministers. This signified an important transformation from an economic to a public health-dominated perspective on the tobacco problem. A similar handing over of responsibilities from trade to health has occurred in other high-income countries (Cairney, Studlar, & Mamudu, 2012): in the United Kingdom, for instance, such a shift took place in 2003 , six years later than in the Netherlands. In the Netherlands, despite this shift in formal responsibilities, the trade ministry continues to have a say in tobacco control policy (see Chap. 8). For one thing, the health minister, like every other minister, has to have support for plans and budget proposals from the full cabinet.
Proposals for new tobacco policy are usually made after frequent consultation between civil servants from the Ministry of Health and their counterparts from the Ministry of Economic Affairs, but may also involve the Ministry of Social Affairs (in the case of smoking bans), the Ministry of Education (e.g., regarding youth prevention programmes and smokefree schools), and the Ministry of Finance (for taxation of tobacco and requests for a larger budget). Tobacco taxation, which is the most effective tobacco control measure, is still firmly in the hands of the Ministry of Finance. Ministers do not want issues to be discussed in the cabinet while they are in their infancy, so civil servants from ministries negotiate with each other until a compromise is reached. By the time an issue is brought to the cabinet meeting for approval, it has been thoroughly discussed and reworked by the bureaucracy.
Disputes that cannot be solved by civil servants at the highest level are negotiated in cabinet meetings. The practice of consultation with other ministries encourages slow decision-making and compromises. Minister Els Borst reflected on the long time it took to realise a new tobacco policy: There are 15 different ministers in this cabinet, with different opinions, often based on personal beliefs about tobacco and alcohol policy. They further have a specific political stance and these three political parties [which form the government] have different views on the matter. It is not the case that the minister of health and the minister of economic affairs can decide on tobacco policy quickly on their own. These discussions on tobacco control policy (…) have taken much time within the full cabinet. 9 When debating in parliament, the minister is supposed to speak on behalf of the cabinet. Ministers mask disputes with colleague ministers. When Borst was trying to get her proposal for revision of the Tobacco Act adopted, she was in constant conflict with Minister of Economic Affairs Annemarie Jorritsma, who vigorously defended tobacco industry interests and sent Borst so-called blue letters 10 urging her to tone down her policy intentions. In 2000 MP Jan Marijnissen (Socialist Party) initiated a debate in a fruitless attempt to get to the bottom of this. 11 lobbying the bureauCraCy of the ministry After a minister or state secretary has instructed the bureaucracy to draft a bill, civil servants go to work on it. The start of the creation of new legislation "is the most relevant part of decision-making, where most influence can be exerted" (Scheltema Beduin & Ter Weele, 2015). The process usually involves talks with stakeholders. Lobbying at this stage, before any formal public consultation may be organised, is regarded as the most important and opportune moment to influence decision-making, and this lobby is completely unchecked in the Netherlands (Scheltema Beduin & Ter Weele, 2015)-although this improved for tobacco control since the implementation of Article 5.3 of the FCTC (more on this in Chap. 6). As one Dutch tobacco industry lobbyist put it, "You need to sit with the person who has the white paper sheet in front of him" (Van der Poel & Gutter, 2011). In 2000 Socialist Party parliamentarian Jan Marijnissen questioned the fact that civil servants from the Ministry of Economic Affairs gave suggestions about the wording and content of the proposal for the new Tobacco Act after being prepared by officials from the Ministry of Health-that had led to textual changes. 12 Marijnissen suggested that the tobacco industry had influenced the policymaking process through the trade ministry. The health minister replied that "there are constantly contacts about tobacco prevention policy between the Ministry of Health, Economic Affairs, sometimes Finance and sometimes Social Affairs. (…) These are normal, common contacts." However, it is clear that these interdepartmental contacts gave the tobacco industry ample opportunity to influence tobacco control policy via contacts with civil servants.
Approval from other ministries is also sought at this stage. When the civil servants have written a draft legislation, it gets "in the line" for approval and amendments by ever higher echelons of bureaucracy until it receives a paragraph of approval from the minister. It is then discussed in a small committee with the involved ministers. After that, it is discussed in the cabinet before being sent for advice to the Council of State and to a commission that checks whether the proposal imposes administrative burdens to society and businesses. Only when it has passed these hurdles is it sent to the second chamber of parliament. It is then discussed in an expert committee of the parliament with the minister, after which it is subject to a plenary debate in parliament. Only then does the bill become public. The parliament can, and most often does, propose further amendments. Amendments need a simple majority vote. This process of amending and rewriting may take several years in the case of politically controversial issues such as tobacco policy, after which it is sent back to the second chamber for a final vote. After approval it is sent for scrutiny to the first chamber (the senate).
In the Netherlands about 250 bills are introduced each year, and they take on average 14 months to reach the adoption stage in the senate, but there have been cases where the process took more than 20 years (Andeweg & Irwin, 2009). The first Tobacco Act took four years (1984)(1985)(1986)(1987)(1988) from presentation of the first draft to the second chamber and approval by the first chamber, but the preparatory work had started already around 1981. The second (amended) Tobacco Act took two years (1999)(2000)(2001), but the political and bureaucratic process had already started in 1996 when the government presented its tobacco control policy intentions for scrutiny and debate in parliament.
It is a peculiarity of the Dutch legislative process that bills do not die with a change of government, as is the case in some countries. Despite the fact that drafting a piece of legislation does not stop when a cabinet resigns, ministers are keen to complete the legislation process while they are still in power, since the survival of a bill is uncertain when a new majority coalition takes over. It was crucial to the realisation of the second Tobacco Act that Minister of Health Els Borst remained in office for two full cabinet periods (1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002). In the last year she and her bureaucracy put much pressure on the process, supported by the tobacco control network, and she managed to get the act through the senate just before the cabinet resigned.
After legislation has been approved, the government can issue regulations for the implementation phase. Governmental decrees (orders-incouncil) 13 and ministerial decrees 14 are part of a higher-order legislative act, and are used for fine-tuning legislation during implementation. Formal approval by vote from the parliament is not necessary. This socalled delegated legislation was extensively used while working on the Tobacco Act and postponed the most controversial elements of the legislation to a later date. Decrees are often used since they make legislation possible while at the same time functioning as a big stick to put pressure on self-regulatory trajectories, for example, with the various temporary exceptions to the workplace smoking ban under the revised Tobacco Act.

lobbying parliament
The drafting of legislation is the joint constitutional responsibility of government and parliament. The second chamber of parliament (lower house, house of representatives) may take the initiative to draft a law and has the right to amend pieces of legislation that are proposed by the government, while the first chamber (upper house, senate) can only adopt or block proposals. It does not have the right to amend laws, although it can force the minister to reconsider a law, withdraw it, or send a revised version to the senate. In practice, the second chamber of parliament rarely uses its right of initiative (Van Outeren & Pergrim, 2015). New legislation stems mainly from the governmental bureaucracy, but this may be, to various degrees, adjusted and amended in parliament.
There is a strict party discipline in parliament. Votes on a proposal are by show of hands. Normally only the hands of the leaders of the parliamentary factions are counted, since it is assumed that party members vote uniformly. A recent count revealed that in a period of four years (2008)(2009)(2010)(2011)(2012), there were 11,000 cases where parliamentarians had to vote (Okhuijsen, 2012). Only in 64 cases was there a roll-call vote, where individuals were counted, and in only 25 of the 11,000 cases did members of parliament cross the floor. In tobacco control policymaking, only once was there a situation where a member of parliament went against party discipline: when the lower house voted on the amendments to the Tobacco Act in 2002 by show of hands. Erica Terpstra broke VVD ranks and voted in favour of the bill.
Despite its relatively weak power position, opposition in the parliament has influenced tobacco control policy considerably. The main way parliament may influence legislation is to adopt resolutions (motions) during debates. In order to be tabled, draft resolutions have to be seconded by at least four other MP's. The purpose is to urge a minister to act or to change intended legislation, for example, to come up with a new or additional proposal. Resolutions are not binding, but when a parliamentary majority approves them, ministers are expected to carry them out. Most resolutions introduced in the Dutch parliament to strengthen tobacco legislation have been defeated because of majority support for the government's position. However, in some cases, resolutions received backing from both ruling and opposing parties, and were influential. For example, parliament strengthened tobacco control by including the worksite smoking ban in the amended Tobacco Act. On the other hand, it diluted or delayed tobacco control by opting for an age key system instead of a ban on tobacco vending machines, by setting the age limit for the sale of tobacco at 16 instead of 18 years, by demanding exemptions to the smoking ban and by rejecting new policy ideas such as graphic health warnings in 2006 (discussed in Chap. 2). Box 5.1 presents more examples of resolutions that influenced tobacco control, illustrating how the parliament is an important lobbying venue for both sides of tobacco control. On balance, the industry lobby seems to have been most successful.

Box 5.1 Important pro-and anti-tobacco control parliamentary resolutions
In 1996 Marijnissen (Socialist Party) received a majority vote with backing from the Christian Democrats. 15 The resolution requested that the government come up with a plan to restrict the sale of tobacco to specialty shops. To this day the government has not executed the motion, but it is still occasionally referred to and in that manner continues to plague the government such as in 2013 when State Secretary Martin van Rijn was challenged by parliament to come up with a proposal to reduce the number of points of sale. 16 As in other democracies, Dutch parliamentarians have the right to ask oral or written questions to ministers, who are obliged to answer within three weeks to written questions. Questioning is therefore an important additional means for parliamentarians to control government and, through friendly members of parliament, for lobbyists to put pressure on the government. See Chap. 10 for a further discussion of parliamentary questions.

parliamentarians: targets for lobbyists
In the Netherlands lobbying is part of day-to-day politics. It is less regulated than in many other countries in Europe (Scheltema Beduin & Ter Weele, 2015). At best, Dutch parliamentarians have only one personal staff member to help them with their complex tasks. They are among the least equipped parliamentarians in Europe, making them particularly vulnerable to lobbyists (Korteweg & Huisman, 2016, p. 17). Parliamentarians often rely on interest groups to help them control the government or draft In 1997 the VVD (with support from D66 and CDA) was successful in weakening the advertising and promotion ban by exempting the brand names or trademarks of tobacco product already in use for non-tobacco merchandise before the law went into effect. 17 The industry could continue circumventing the ban by promoting its brand through clothes, shoes, and the like.
In 2001 the Labour Party, together with the Socialist Party, managed to round up a majority for a resolution demanding that the government make effective smoking cessation support eligible for financial reimbursement through the national health insurance plan. 18 This eventually contributed to the adoption of the current reimbursement system, although it took ten years to be implemented.
In 2002 CDA Senator Werner received full support for a resolution that requested that the government find the necessary budget for media campaigns to help smokers quit smoking as an accompaniment to the workplace smoking ban. 19 It was contingent on this condition that the senate was prepared to approve the Tobacco Act.
In 2005 conservative parties, with support from the Socialist and the Green-left parties, forced the government to consider ventilation as an alternative to smoking bans. 20 initiative laws, as they lack both the specialised knowledge and the time to gather all the facts and figures they need. How they process information they receive from lobbyists is at their own discretion-they do not have to declare or make this public.
There is increasing demand for lobbying transparency in Dutch society and the media. In 2012 the Algemene Rekenkamer (Court of Audit) noted that the information possessed by the tobacco sector is better than that of specialised parliamentarians (Algemene Rekenkamer, 2012). 21 When former Health Minister Elco Brinkman became a lobbyist for Philip Morris and in this capacity contacted the state secretary for finance about a European taxation matter, this was exposed in the media. Soon afterwards Lea Bouwmeester, a Labour Party parliamentarian, announced that she was preparing a bill that would make it mandatory for the government to include a "lobbyists paragraph" in each proposal for new legislation. 22 Such a paragraph would require disclosure of lobbying contacts that occurred in the drafting process of laws and give details about who, on behalf of which organisation, visited which policymaker, with what intentions; similar requirements exist in other jurisdictions, including the United States, England, and the European Union. Three years later the initiative law had still not been introduced. According to Bouwmeester, "writing such a proposal is complicated and the parliament has too little support for it" (Meeus, 2015).
In 2016 journalists revealed that one quarter of ex-politicians (ministers and parliamentarians) had become lobbyists (Huisman, Kooistra, & Korteweg, 2016). In 2017 the cabinet responded for the first time to calls from the Labour Party for more transparency, and decided that there would be a restriction on ex-ministers and state secretaries accepting lobbying positions in their field of expertise for two years after leaving office. It has also become more difficult for tobacco industry lobbyists to contact government officials directly due to a stricter adherence by the government to Article 5.3 FCTC (more on this in Chap. 6).

lobbying through Different Venues
The Netherlands is not a federal state like Canada, Australia, or the United States. Federal states offer interest groups ample opportunity to place tobacco control on the political agenda. In Australia, it has been noted that "Australian states are like dominos, and when one state takes the first step, the others will likely follow shortly after" (Bryan-Jones & Chapman, 2008). In the United States when tobacco control activists have no success at the central federal level, they simply turn their attention to local and state policymakers. The advantage of such "venue shopping" has been suggested as a reason why the United States has stronger tobacco control than Denmark, which has a single-venue system like that of the Netherlands (Albaek, Green-Pedersen, & Nielsen, 2007). While there seems some truth in the single/multiple venue explanation, it is overly simplistic, since counter-examples can be found easily. Germany has a federal system but is a tobacco control laggard, while a leader like New Zealand is a singlevenue country. The next chapter will explore the increasing importance of European and global legislation for tobacco control and lobbying opportunities for venue shopping in an international context. A special "shopping venue" is the judicial system. This has frequently been used by both sides (pro-and anti-tobacco advocacy groups) in the Netherlands, and has forced breakthroughs when policymaking was slow. A case that received much media exposure was the Nanny Nooijen case of 2000: an employee of the Dutch postal office won a court case against her employer, who had failed to protect her from exposure to second-hand smoke from her colleagues. A few more court cases were needed before the workplace smoking ban finally made the legal route redundant, and one of these was in 2003, where the court forced Isala Clinics to financially compensate an employee with asthma for health damage as a consequence of exposure to tobacco smoke. 23 In March 2002, during the debate in the senate about the Tobacco Act, senator Ruers (SP) said, "I want to point out that the only real breakthrough in the Netherlands regarding the smoke-free workplace was the court case the previous year. An employee had the courage to start a lawsuit, which she won. This caused a tremendous change in the Netherlands. (…) A civilian accomplished more for society than all [other] measures taken together." 24 Another example of the importance of the legal system involved the non-smokers' rights group Clean Air Netherlands (CAN), which won an important court case against the state in 2014. CAN successfully pleaded that small bars must not be exempted from the workplace smoking ban, referring to Article 8 of the Framework Convention on Tobacco Control (FCTC) treaty. But not only did the health network successfully use the legal venue: four years before the CAN ruling, small bar owners, with help from the industry, won court cases where local judges ruled that small bars without personnel had to be exempted. Chapter 8 gives some further examples of the tobacco manufacturers' use of the court system to frustrate tobacco control or intimidate tobacco control organisations.

the story of Deregulation
This chapter about institutional factors has to include a discussion of the process of deregulation, which started in the 1980s. Later Dutch cabinets followed a neo-liberalist agenda that aimed at less government, less bureaucracy, and more free market. This section discusses the consequences for Dutch tobacco control.
The first Lubbers Cabinet (1982)(1983)(1984)(1985)(1986), which consisted of CDA and the liberal-conservative VVD, wanted to reduce administrative burdens on businesses and initiated a regulatory "reform" programme, informed by similar deregulation initiatives in the United States and the United Kingdom. An important motive for this so-called deregulation operation was to overcome the economic crisis by improving business competitiveness, but it also had ideological motives. Deregulation was one of six elements in a substantial programme aimed at reducing high unemployment rates and central budget deficits. Other elements were reconsideration of the role of the government (making it smaller), decentralisation, reorganisation of the governmental bureaucracy, reduction of the number of civil servants, and extensive privatisation ( Van der Voet, 2005).
Proposals for new laws had to be submitted to a deregulation commission. This "Commission Van der Grinten" advised on whether regulation was necessary and, if so, how it could be simplified. Not all proposals were sent to this commission: the cabinet decided on a selection. Among these was the proposal for a Tobacco Act, and the involvement of this commission was one of the reasons why the process of drafting the initial act took four years. 25 The commission's advice was not to impose smoking bans but to leave this to industry self-regulation. 26 Cabinet ignored the advice, but adopted the commissions' wish not to ban the sale of cigarettes to minors (16-year-olds) nor to ban cigarette sales through vending machines and self-service outlets. The draft Tobacco Act was adjusted to reflect these amendments before it was sent to parliament.
Commission Van der Grinten was active until 1995, when a new form of regulatory impact assessment, called the Marktwerking, Deregulering en Wetgevingskwaliteit Operatie (market competition, deregulation, and quality of legislation) (MDW Operation), replaced it. This was the new answer to the government's continued desire for a simplified regulatory environment offering fewer hurdles to businesses. Formal top-down interventions had to be constrained, and self-regulation by industry and citizens was the default. MDW was operative until 2003, when the second Balkenende cabinet (VVD, CDA, D66), which had as a motto "more participation, more work and less regulation," introduced its own version under the heading "B4." 27 Tobacco industry lobbyists regularly referred to the cabinet's wish to minimise bureaucracy and reduce administrative burdens for business, which had been estimated at €16.4 billion per year.
Since 1998 new regulation has been subject to scrutiny from ACTAL. 28 This independent advisory board assesses whether new legislation is suitable for decision-making in the cabinet (Hoppe, Woldendorp, & Bandelow, 2015) and requires policymakers to provide specifications of the exact administrative burden it will impose on businesses. In order to do this, civil servants may contact industries for information about the expected costs of new legislation to their sector. Only after a stamp of approval from ACTAL can a proposal for new legislation be sent to the cabinet.
The Dutch experience with deregulation for businesses has been considered an example to other nations making attempts to reduce bureaucracy in the EU (see Box 5.2). The tobacco industry, with support from other multinational corporations such as the oil industry, pharmaceutical companies, and the food and alcohol industry, lobbied successfully to introduce such business-friendly adaptations of legislation making in the Brussels bureaucracy, making it more difficult to pass legislation that protects public health (Smith et al., 2010;Smith, Fooks, Gilmore, Collin, & Weishaar, 2015).

Box 5.2 Dutch inspiration for a business-friendly policy agenda in Europe
Around 1994 British American Tobacco (BAT) started to lobby for regulatory reforms in the EU similar to what was common practice in the United Kingdom and the Netherlands (Smith et al., 2010). The Netherlands was one of the strongest supporters of "better regulation" in the EU, after Germany, the United Kingdom, and Ireland. When the Dutch assumed the EU presidency in 1997, BAT recognised an opportunity as "The Dutch appear even more committed to the principles of cost-benefit analysis and risk assessment … and would support a treaty amendment to achieve this objective" (BAT, 1996). BAT, with support from other multinationals, promoted the idea that the EU would be more competitive when legislation was simplified and when proposals for new legislation, including those that protect health, were subject to rigorous costbenefit analyses and impact assessments. As a consequence of BAT's lobbying efforts, a "better regulation" agenda that included a man-

DeCentralisation
Governments choose the path of decentralisation for ideological and financial reasons (De Vries, 2000). Since the 1980s the Dutch government has decentralised a great number of tasks, including health promotion, to local governments and the private sector assuming that this will improve operational efficiency and that in this way central government can downsize and reverse its growth (De Vries, 2000). This had direct consequences for how government approached national tobacco control policy making, since responsibility was increasingly given to the local level.
Decentralisation of public health policy began in 1986 with Nota 2000, a 600-page memorandum on public health (Dekker & Saan, 1990). It outlined new ambitions in the field of public health and a political commitment to WHO's global strategy, "Health for All by the year 2000" (WVC, 1986). WHO called on governments to improve health through a strategy encompassing not only individuals but also their environment, all levels of society, and all sectors that might influence health. The new idea of an intersectoral approach and organising health promotion at the lowest level caught on. Health policy and prevention was to be restructured as part of a broader administrative reform of retreating government and liberalisation of public tasks, initiated by the Lubbers cabinets (1982( -1994( ) (WVC, 1991. The idea was that the Ministry of Health would continue to coordinate, set targets, and monitor, but that execution would devolve to the local level. The Netherlands is not unique in this: many EU countries decentralised in the 1990s and later (Marks & Hooghe, 2003).
With the Wet Collectieve Preventie Volksgezondheid (Public Health Collective Prevention Act) (WCPV) of 1990, prevention of disease became a responsibility that the national government shared with municipalities. datory impact assessment of new legislation and requirements to consult industry at an early stage of the policymaking process was implemented as part of the Treaty of Amsterdam (1997) (Smith et al., 2010). This achievement was heralded by BAT as "an important victory" for the company (BAT, n.d.). It helped the industry prevent the introduction of EU-wide public smoking restrictions and delay tobacco advertising restrictions.
According to law, every four years all municipalities (there were 393 in 2015) have to develop a public health policy document in which they outline priorities for the next four years. Local policymakers are free to choose priorities and targets based on regional epidemiological data and are expected to make use of evidence-based interventions-although this is not mandatory. Municipalities have to finance health promotion from the general municipality budget, which shrank over the years due to budget cuts by the central government, and are thus required to look for additional financial means, such as applying for competitive grants or initiating private-public partnerships with insurance companies or commercial businesses. Municipalities increasingly take the lead in public health at the local and regional level, and receive support from Gemeentelijke Gezondheidsdienst (Regional Public Health Services) (GGD) with this task. Officials of the Inspectie voor de Gezondheidszorg (Inspectorate for Health Care) (IGZ) visit the GGDs every year to assess the execution of their duties within the law.
In 2003 an amendment to the Public Health Collective Prevention Act made local governments' periodic status reports on disease prevention mandatory by law. Local municipalities were also required to play a specific role in what became known as the "prevention cycle" (VWS, 2011b). In 2008 the WCPV was integrated in a new public health act (Wet Publieke Gezondheidszorg (WPG)), another step towards strengthening the prevention cycle and handing responsibility to the local level. According to a recent national health policy document, health promotion is now a decentralised task and the joint responsibility of local government and health insurers (VWS, 2011a). Although the formal responsibility for tobacco policy remains with the cabinet, because Article 22 of the national constitution stipulates that the government has a legal obligation to protect and promote public health, municipalities are expected to coordinate tobacco control activities at the municipal level and by doing so contribute to the reduction of national smoking rates.

Tobacco Control Lost Between the National and the Local Level?
The responsibility for municipalities in the field of public health is daunting, for there are great complexities and small budgets. Many municipalities are not sufficiently equipped to develop and implement prevention strategies. On 25 March 2010, IGZ reported on the lack of effectiveness. The day of publication was chosen carefully, because on the same day the Rijksinstituut voor Volksgezondheid en Milieu (National Institute for Public Health and the Environment) (RIVM) published its new Volksgezondheid Toekomst Verkenning (Public Health Status and Foresight) (VTV) report (see Chap. 10 for more about the RIVM and VTV reports). Both reports came to the devastating conclusion that the prevention cycle was not functioning and that local health promotion policies do not contribute to reductions of obesity, smoking, alcohol abuse, or depression at the national level (Inspectie van de Gezondheidszorg, 2010). The message was that the central government must take back control. At an unusually frank press conference, Marc Sprenger, director-general of the RIVM, commented to the press that "the government must have the courage to firmly and normatively take up its role. It is one thing to distribute nice folders in schools, but that is not enough. One must also be prepared to take tough measures … that support local programmes" (NRC, 2010).
The government may promote decentralisation and citizen participation when it simply wants to cut spending. At such times Dutch tobacco policy risks getting lost between the national and the local level: the national government may delegate responsibility to lower levels and emphasise citizens' own responsibility, while municipalities evade responsibility and point back to the government. The potential advantages of decentralisation, such as tailoring policies to local circumstances, encouraging citizen participation, and being more efficient, are questionable if they are not adequately facilitated by the central government. The current attempt by the Dutch government to stimulate local initiatives in public health through the National Prevention Program (NPP) is not supported by the necessary budget. Instead, the government assumes that societal and commercial organisations will pay for the health promotion programmes themselves. This may work for some issues, but is problematic for tobacco control. Municipal councillors have many other more pressing concerns than tobacco, and tobacco prevention is not a topic that will earn them credit from the public (see Box 5.3). In addition, they have to overcome resistance from local functionaries who feel uncomfortable interfering in a private habit such as smoking (Van der Meer, Spruijt, & De Beer, 2012). These problems have been recognised by the Raad voor de Volksgezondheid en Zorg (Council for Public Health and Health Care) (RVZ), an independent advisory body. It urged the government to take a stronger lead, to set quantifiable targets, and to resume coordination and control (RVZ, 2010). In one report RVZ concluded that national public health tasks can only be effectively administered by municipalities if they are supported by a complementary central policy, and if the minister has voiced a clear normative standpoint (RVZ, 2011). The advice echoes the call for action from the director-general of the RIVM, Mark Sprenger, the year before. RVZ recognised that "there are political-ideological arguments against applying the most effective interventions (especially supply side restrictions). … These have to do with the prime responsibility that a citizen has for the choices it makes, an opinion that seems to have more supporters in the Netherlands than elsewhere." The council was critical about the government's weak prevention policy and its inertness, and recommended that the shift towards local health policy be accompanied by strong measures such as higher tobacco taxes and a larger budget for prevention.

Box 5.3 Municipalities and national tobacco control goals
A 2011 study among health policy officers of 151 municipalities revealed that 39% had not included tobacco control in their policy intentions (Huijsman, van der Meer, de Beer, van Emst, & Willemsen, 2013). Of those who did, only 41% indicated that they had a distinct tobacco reduction programme. In most cases this was part of a broader addiction or lifestyle programme. The concrete activities were related to education, mostly smoking prevention programmes in schools, and occasional organised cessation support for smokers wanting to quit. The main reason for not implementing local tobacco control policy was that it is too labour intensive (especially for smaller municipalities) and that they felt that tobacco policy could best be tackled at the national level. One officer was quoted saying: "With the restricted means and time that we have, we want to make a choice in what to do in our health policy. It is better to do a few topics well, than many badly. Because tobacco control can be done much more effectively at the national level (smoking bans, smoking cessation support through national health insurance) we choose to do other topics [than tobacco]." In 2013, the low priority given to tobacco control by local authorities was confirmed in an inventory by the national organisation of community health services. This showed that one-third of the municipalities did not tackle smoking (GGD Nederland, 2013). Alcohol prevention was given more priority. A more recent study found that tobacco control is still virtually non-existent at the local level. Local decision makers see few advantages, because they "are not familiar with the possibilities they have to control smoking. Smoking is such a small issue for them that they do not take the time to get to grips with it" (Mulder, Bommelé, Branderhorst, & Hasselt, 2016).

ConClusion
Several aspects of the Dutch policy environment work against expeditious adoption of tobacco control. For many years Christian Democratic principles of subsidiarity, coupled with the corporatist tradition of policymaking and consensus seeking, combined to make it "logical" for policymakers to invite representatives of the tobacco industry to present their views on tobacco control, with business-friendly solutions (self-regulation instead of legislation) as the outcome. Dutch policymakers tried to avoid polarisation and conflicts between groups, and informal stakeholder consultation became a popular strategy to this end-which presented ample opportunity to the tobacco industry to influence and delay tobacco control. This was facilitated by calls to reduce the administrative burden to businesses during periods of economic recession, resulting in extensive deregulation operations in the 1980s and later. Another aspect of the policy environment not conducive to tobacco control was the neo-liberalist agenda of most cabinets, which preferred a limited role by the government.
Given the multiple-party nature of Dutch politics, the drafting of new policies must follow the specifics laid out in a coalition agreement, which is the end result of intense negotiations between the co-ruling political parties. The tobacco control coalition has only been successful twice in getting tobacco control proposals into coalition agreements.
In the Dutch system, where ministers are politically appointed and rarely have a background in health or medicine, the chance that ministers will become tobacco control champions is small. The political make-up of the cabinet is crucial, but the political orientation of the majority in parliament is also important. Parliament will usually support the government, but its opposition may amend or delay government proposals and influence the political agenda through the presentation of resolutions and parliamentary questions. Parliament is an important venue for lobbyists from both sides, but the industry has been most successful, since delaying or adapting policy intentions is easier than getting new legislation on the agenda in the first place. The inside lobby for many years has been unchecked and barely accountable in the Netherlands, and less regulated than in many other democracies, although this is now improving. Dutch parliamentarians are understaffed and depend on lobbyists for information and support with drafting bills and motions. For years the industry has had an advantage as their lobbyists stayed at their posts for decades, while both the civil servants who specialised in tobacco control experienced regular turnover.
A long and gradual process of decentralisation of public health tasks away from the central government to the municipal level increased the risk that tobacco control will be lost between the national and local levels, as it made it easier to escape responsibilities. Municipalities have little means and lack the motivation to take on national tobacco control tasks. The judicial system has become increasingly important as the corporatist elements of Dutch policymaking diminish, and the court has been responsible for important breakthroughs, especially regarding smoking bans.