A Stakeholder Analysis of the Dutch National Prevention Agreement

(1)

A Stakeholder Analysis of

the Dutch National

Prevention Agreement

Master Thesis for Public Administration:

Economics and Governance

Hanneke Bouts (s1737252) Supervisor: G.E. Breeman Second reader: A. Timmermans

(2)

(3)

1 –Introduction

1.1 The consequences of unhealthy behavior: a growing global problem

1.1.1 Overweight and obesity

In 2018 a much discussed news story in the Netherlands was that once again the number of people who are overweight and the number of people who are considered to be obese has again increased. Statistics Netherlands (CBS) put out a survey and released the new numbers. It turned out that 50% of Dutch people above 20 years of age are overweight, and that 14% is considered to be obese. Since the CBS started this yearly survey in 1981, the number of people with obesity has increased by 150% (CBS, 2018). Another survey showed that 75% of the Dutch population does not have a healthy eating pattern according to government recommendations (Rijksinstituut voor Volksgezondheid en Milieu [RIVM], 2015).

Overweight and obesity is not a problem limited to the Netherlands. In European Union countries 30-70% of men and women are overweight, and 10-30% op people are affected by obesity (World Health Organization [WHO], 2018c). According to the Organization for Economic Co-operation and Development (OECD) the United Kingdom is the country in western Europe where overweight and obesity is most prevalent. The OECD’s yearly Health at a Glance report shows that in the UK 63% is overweight, of which almost half, namely 26.9%, is obese (Organization for Economic Co-operation and Development [OECD], 2017).

According to the World Health Organization (WHO) global estimates, 1.9 billion people (39% of all adults) were overweight in 2016. 650 million of these adults were obese, which compares to 13% of the world’s adult population (WHO, 2018c). The problem of overweight and obesity is often associated with developed high income countries. However, estimates on overweight and obesity have increased for almost every country since 1980 (Hruby & Hu, 2015). Thus, overweight and obesity is a problem that also affects low- and middle income countries, especially in urban areas. Low- and middle income countries even face a so-called “double burden”. This means that these countries while facing the rising problem of overweight and obesity, are still dealing with the problems of undernutrition and diseases such as malaria and aids/HIV. Only in sub-Saharan Africa and some parts of Asia underweight is still more prevalent than obesity is (WHO, 2018c).

Moreover, overweight and obesity is increasingly a problem associated with children and adolescents. This is the age group from 5 to 18. The prevalence in this group has risen from merely 4% in 1975 to 18% in 2016. This number compares to 124 million children and adolescents. The number of overweight children under the age of 5 was estimated to be 41 million in 2016. Obese children experience many health related difficulties, such as breathing difficulties. Moreover,

childhood obesity increases the risk of obesity, premature sickness and death at an adults age (WHO, 2018c).

The WHO defines obesity and overweight as “abnormal or excessive fat accumulation that may impair health” (WHO, 2018c). People are classified as overweight or obese based on their body mass index. This is a person’s body mass in kilograms divided by the square his or her height in meters. It is considered to be an appropriate indicator to measure if a person’s weight is healthy. It is easy to calculate and therefore a commonly used indicator to measure the risk on health problems due to overweight. A person with a BMI higher than 25 is considered to be overweight. A person with a BMI higher than 30 is considered to be obese, and a person with a BMI higher than 35 is considered

(6)

2

morbidly obese. A distinction is made between overweight, obesity, and morbid obesity as each weight range is associated with additional health risks (WHO, 2018c).

According to the WHO (2018c) the rise of obesity and overweight in the last decades is a consequence of environmental and societal changes associated with economic development, such as an increased intake of foods high in calories, a decrease of physical activity due to a replacement of physical labor by more sedentary desk jobs, motorized transport, and increased urbanization.

According to the WHO (2018c), obesity is one of today’s most neglected public health issues as it goes hand in hand with a variety of serious health disorders. The organization even speaks of a global obesity epidemic. The problem is spreading over many parts of the worlds, including, as mentioned before, parts of the world where the condition paradoxically coexists with

undernourishment. Practically it affects people of all ages, and all socioeconomic groups. Overweight, but especially obesity, increases the risk of serious diet-related non-communicable diseases, such as cardiovascular diseases and musculoskeletal disorders, type 2 diabetes mellitus, high blood pressure, stroke, and some types of cancers. The consequences range from early death to chronic health issues that diminish quality of life. According to the WHO (2018c), serious action should be taken to prevent millions of people from suffering from a variety of serious health disorders. Overweight and obesity, as well as the diseases that are caused by it, are largely preventable (2018c).

1.1.2 Tobacco and alcohol

Tobacco and alcohol consumption are two other major threats to public health. In most parts of the world alcohol consumption is very customary used, especially at social events. Worldwide average alcohol consumption per capita aged 15 years and older was 6.4 liters of pure alcohol in 2016 (WHO, 2018a). According to the WHO (2018a) each year thee million people die as a consequence of harmful use of alcohol. This corresponds to one in every twenty deaths. The WHO (n.d.-b) stresses that alcohol consumption comes with health risks, even when used in moderation. The WHO and other health organizations even state that there is no risk-free amount of alcohol that one can drink (Gezondheidsraad, 2015; WHO, n.d.-b). According to the WHO (2018a), alcohol consumption is the highest in the European region, namely 9.8 liters of pure alcohol per capita aged 15 and upwards in 2016. There it is responsible for a major share of health problems and premature deaths. Alcohol causes non-communicable diseases such as cancer and cardiovascular diseases. Next to health related problems, alcohol also causes behavioral disorders such as mental problems, (domestic) violence, and injuries as a consequence of violence and accidents (WHO, 2018a).

In the Netherlands the amount of alcohol consumed in liters per capita has increased a lot after the increase of wealth since the Second World War. However, since the beginning of this century the amount of alcohol consumed per capita has steadily declined (Ministerie van

Volksgezondheid, Welzijn en Sport [VWS], 2018). Between 2010 and 2016 it dropped from 10.4 liters to 8.7 liters (WHO, 2014). This decline came to a halt after 2016. The Dutch government is, however, not content with the current situation as only 40.1% of the population follow recommendations for alcohol consumption (VWS, 2018), i.e. not drinking at all or at most one glass per day. Also a large part of the population drinks excessively (that is over 14 units a week for women, and over 21 for men) or heavily (that is over 4 units for women and 6 for men during one occasion at least once a week): i.e. 9.2% and 9% in 2017. Moreover, despite changing the legal drinking age from 16 to 18 in 2014, 44.8% of children aged 12 to 16 have consumed alcohol at some point in their lives (RIVM, 2017).

(7)

3

The WHO (2018d) describes tobacco as ‘one of the biggest public health threats the world has ever faced’ and even speaks of a tobacco epidemic as its usage is so common around the world. Direct tobacco usage is responsible for the deaths of over 6 million people a year since smokers run a higher risk of non-communicable diseases such as cancer, cardiovascular diseases, and different types of lung diseases. About one million more people die of second-hand smoking. The WHO states smoking is the greatest preventable cause of death in the world. Globally around 21% of people smoke. Tobacco users mainly live in low- and middle-income countries, namely around 80% of them (WHO, 2018d). However, in Europe the average prevalence of tobacco smoking among adults is highest (28%) (WHO, n.d.-a).

In the Netherland 23% of adults smoke tobacco, with annually 20.000 tobacco related deaths as a consequence. Also, every week hundreds of children start smoking cigarettes. This is a major problem according to Trimbos Institute, a Dutch institute focused on mental health and addiction, as two-thirds of smokers start smoking before their eighteenth birthday. 50% of smokers deaths are caused by a tobacco related disease (Trimbos-instituut, 2017; VWS, 2018).

1.1.3 Healthcare systems increasingly under pressure

So overweight and obesity as a consequence of unhealthy diets, alcohol consumption, and tobacco usage pose an increased risk for different health problems. These health problems have economic consequences. There are direct costs in the form of extra healthcare costs which in turn impose a burden on healthcare systems, but there are also indirect costs such as lost work days, less

productive work days, early retirement as a result of disablement, and premature deaths (Candari, Cylus, & Nolte, 2017).

The WHO states that especially in the European region, where healthcare systems have an important social security function, healthcare systems are more and more under pressure due to increasing costs (Candari et al., 2017; Thomson, Foubister, & Mossialos, 2009). Securing these systems’ fiscal sustainability in the future while also meeting the common values defined by the Council of the European Union, which are universal coverage, solidarity, equity and quality, is already a challenge and subject of debate in many EU states as healthcare is a key head of expenditure at the expense of other public services (Baarsma & de Boeck, 2017; Thomson et al., 2009). One of the main reasons for increasing healthcare costs in Europe is the spread of unhealthy lifestyles, next to an ageing population, expensive modern treatments, and educated demanding patients (Economist Intelligence Unit, 2011).

In the Netherlands the increasing healthcare costs have been a point of discussion for years. The past 40 years the growth of expenditure on healthcare was bigger than economic growth. As a consequence people have had to pay an increasing share of their income on healthcare: healthcare premiums and the amount of money people have to pay themselves before health insurance applies have both increased over the years. It is predicted that in the Netherlands healthcare costs will double to 174 billion euros in 2040, compared to the expenditure in 2015 when 85.8 billion euros was spent on healthcare. The share of GDP spent on healthcare will rise from 12.7% in 2015 to 16.4% in 2040. Per capita this increase corresponds with 5100 euros in 2015 to 9600s euros in 2040 (RIVM, 2018). If nothing is done to stop this trend it can lead to unsustainable public finances and more public discontent (Baarsma & de Boeck, 2017; Economist Intelligence Unit, 2011). It is feared that this will lead to a part of the population no longer willing to continue to subsidize a healthcare system that is inclusive (Thomson et al., 2009).

(8)

4

Europe is the spread of unhealthy lifestyles. Alcohol, tobacco, overweight and obesity increase the risk of many non-communicable diseases, better known as chronic diseases. These types of health problems tend to be permanent or of long duration. The main types are cardiovascular diseases, cancer, different types of lung diseases, and type 2 diabetes mellitus (WHO, 2018b). I have

mentioned examples of these above. With an increasingly elderly population this increase of chronic diseases are a burden on the healthcare budget, simply because these diseases do not kill quickly (Economist Intelligence Unit, 2011). Treatment costs for these types of diseases are of longer duration with an increase in life expectancy, which contributes to healthcare spending. Moreover, a chronic disease like type 2 diabetes mellitus is also increasingly seen at a young age (Candari et al., 2017; RIVM, 2008).

It is expected that in the Netherlands the number of people with at least one chronic disease will increase to 9.8 million people in 2040, which equals to 54% of the population (RIVM, 2018). The appearance of type 2 diabetes mellitus will increase the most, and cancer and cardiovascular disease will cause the greatest burden of disease, which is a combination of loss in healthy life expectancy, and overall life expectancy. It is estimated that each year about 9 billion euros of the Dutch

healthcare budget is spent on the consequences of alcohol and tobacco use, overweight and obesity. At this moment this number equals about 10% of the total budget as calculated by the RIVM (2018). 1.1.4 Prevention is key

The WHO as well as the OECD urge governments to focus more on health promotion and disease prevention to reduce the increasing pressure on national healthcare budgets as a consequence of lifestyle related health problems (OECD 2017, Candari et al., 2017). There is growing evidence that certain prevention policies are effective to keep costs down, but that today a minimal percentage of national health budgets is spent on prevention (Candari et al., 2017).

In recent years during the discussion about increasing healthcare expenditure, the discussion about prevention has become more prominent in the Netherlands. Critics argue that healthcare is too focused on treating people who are already sick, and that the growth of healthcare costs can be stopped by preventive measures, such as health promotion, government regulations, and so-called lifestyle medicine (Baarsma & de Boeck, 2017, “Leefstijlgeneeskunde is nodig”, 2018). In 2017 the RIVM already said that the government had to come up with a comprehensive approach to improve the health of our diets (RIVM, 2017). The Netherlands Scientific Council for Government Policy (WRR), an independent advisory organization for public policy, also advised the Dutch government to give disease prevention strategies a new impulse (The Netherlands Scientific Council for Government Policy [WRR], 2017).

1.1.4 National Prevention Agreement

On 23 November 2018 the government of the Netherlands launched the National Prevention Agreement. This is a document which contains a set of agreements between over 70 Dutch organizations such as businesses, industry associations, municipalities, health organizations, and educational institutions. The National Prevention Agreement focuses on three subjects, namely overweight and obesity, tobacco, and alcohol. The document states that these three subjects were chosen because these factors cause by far the greatest loss in healthy life expectancy. The

government’s goals are to achieve a tobacco-free generation, to reduce the number of people that are overweight from 50% to 38%, and the number of people that are obese from 14.5% to 7.1%, and to reduce the number of people who drink too much alcohol from 8.8% to 5% according to

(9)

5

1.2 Goal of this research

The National Prevention Agreement is the product of the current cabinet of the Netherlands formed in October 2017. The aim of this agreement is to improve the overall health of the Dutch population by diminishing overweight and obesity, smoking, and alcohol consumption. The agreement consists of various policy measures (VWS, 2018). Every stakeholder that contributed to the agreement has had a chance to express their policy preferences to tackle the discussed health problems. Each stakeholder represents different interests, and have therefore different preferences. For example, representatives of the industry will often put their economic interests first, and may therefore lobby against taxes and legislation which might hurt them. Health organizations on the other hand may put public health as an interest first, and might be willing to use taxes as a mean to achieve

improvement.

The politicians and policymakers involved with the National Prevention Agreement assume it will have an effect on the health of the Dutch population. However, immediately after the

publication of the National Prevention Agreements critics argued the measures in the agreement do not go far enough and are too voluntary.

This chain of events led me to be interested in in the outcome of this process: namely which type of policy instruments are part of an agreement to which so many stakeholders contributed? Have all the different interests led to a consistent agreement, or is the wide variety of interests clearly reflected in the final agreement? In other words: is it the policy part of the agreement consistent? This led me to the following research question and four sub-questions:

Which policy instruments have been adopted in the Netherlands to reduce overweight and tobacco and alcohol consumption by means of the National Prevention Agreement, and are these policies consistent?

1. What type of policy instruments are part of the current coalition government’s National Prevention Agreement?

2. Which policy instruments were preferred or ill-favored by different stakeholders? 3. What will be the effect of the National Prevention Agreement according to various

stakeholders’ policy theories?

4. Are the policy instruments which are part of the National Prevention Agreement consistent? In the second chapter I discuss the literature which I have used to answer my research question. In the third chapter I explain how I have used the theories that I will discuss in the next chapter in order to conduct my research, by operationalizing the needed theoretical concepts and explaining my research methods. In the fourth chapter I make an analysis for every sub-question in order to formulate an answer to them. In the final chapter I will conclude my research by answering the research question.

(10)

(11)

7

2 –Theory

In this chapter I discuss the literature which I have used to answer my research question. In the first section of this chapter I describe the policy cycle. By doing this I aim to show where in the process of political decision-making stakeholders are able to exert their influence. Consequently, I briefly explain the history of stakeholders’ influence on the political decision-making process in the Netherlands. In the second section of this chapter I discuss Christopher Hood’s (2007) toolkit perspective which can be used to list and categorize different so-called policy instruments. In the third section of this chapter I discuss five relationships between the development and effects of policy which can be studied according to a framework by van Schendelen (1990). In the final section of this chapter I discuss how to reconstruct a policy theory and how to determine if policy theories are consistent according to van de Graaf and Hoppe (1989).

2.1 Stakeholders in political decision-making

2.1.1 The policy cycle

Public policy is the total of principles a government uses to try and steer to certain outcomes. The choice for a certain policy is partially responsible for what a society looks like and partially explains why societies are different from each other. Public policy is the product of a political process. It is a continuous flow of big and small decisions on diverse topics, with potentially many different actors involved (Deschouwer & Hooghe, 2011).

In order to understand the complex process of the development of public policy one often uses the policy cycle. There are multiple variants of the policy cycle. The one represented in Figure 2.1 below is considered the most basic form:

Figure 2.1. As based on Figure 9.1 in Deschouwer&Hooghe (2011, 206).

In this model the development of policy is divided into five steps, the final step of which culminates into the first step. However, a model is a simple representation of reality. So it must be noted that reality is often more complex. In reality the process can be in multiple stages at the same time, the process can be stopped at a certain point in time, or steps backwards may be taken. The model makes it possible to study the different steps in the process. The cycle starts at agenda setting and

(12)

8 ends with evaluation (Deschouwer & Hooghe, 2011).

During the first stage, the agenda setting stage, a problem is identified. This means that a particular undesirable situation which needs to be addressed is identified. The issue will appear on the so-called public agenda: this means the issue is debated in society. During this stage the idea grows that this problem might be solved by government intervention. Political ideologies structure the public debate on the issue. A consequence of this is that problems which easily fit contradicting ideologies will appear on the public agenda more easily. Accordingly, framing an issue as a problem which fits certain opposite political positions is a useful tactic to get an issue on the public agenda. Topics can appear on the public agenda from top-down and bottom-up. Top-down refers to a topic being placed on the public agenda by actors who function within political institutions, like for example politicians and members of parliament. Topics are brought to the attention bottom-up when these are brought to the attention by actors that function outside of political institutions, such as advocacy groups and academics. The agenda setting stage is very important because during this stage it is determined in what way the issue is discussed. The perception of an issue is very

important, and strongly determines what the next stages of the policy cycle will look like. During this stage of the policy cycle the most actors are involved (Deschouwer & Hooghe, 2011).

In the second stage, the stage of policy formulation, it is considered what all the conceivable and realistic approaches to a problem are and who should be involved in the process. This is a competitive stage as different interests and beliefs collide. Many groups and organizations take part in this stage. However, fewer than in the first stage. The government itself is present in this stage, but in many policy areas it is customary to involve so-called stakeholders. Stakeholders are those directly involved or affected by a particular policy. The many stakeholders involved come up with different possibilities for new policy, and different interests are weighed against each other (Deschouwer & Hooghe, 2011).

In the next stage, the policy adoption stage, the necessary decisions are made. It is decided which policy or set of policy measures will be adopted. At this stage the actors from outside

government who were involved during former stages are no longer involved. This, however, does not mean the final decisions are made in isolation, cause of during former stages stakeholders from outside political institutions brought different solutions to the table. It is possible that the policy cycle comes to an end at this stage when policymakers cannot agree on a solution (Deschouwer & Hooghe, 2011).

The fourth stage is the policy implementation stage. During this stage the policy is further specified and defined, and eventually implemented. In the policy implementation stage the battle between competitive political views and alternative solutions continues. This is the case when civil servants do not feel the need to prioritize the enforcement of a particular policy or refuse to enforce a particular policy (Deschouwer & Hooghe, 2011). The latter was the case with same-sex marriage in the Netherlands as for a long time a group of civil registrators refused to marry same-sex couples. This was however determined to be illegal in 2014.

During the final stage the implemented policy is evaluated. This is actually a continuous process carried out by every actor involved in the policy cycle. This stage permanently influences potentially new policy cycles (Deschouwer & Hooghe, 2011).

2.1.2 Neo-corporatist systems and the influence of interest groups

In western democracies there is a clear dichotomy between two models aiming to explain the

(13)

9

and neo-corporatism. In pluralist systems there are many independent interest groups present. According to this model there should be no concentration of power in such systems: power is spread out over many different organizations, and these organizations function autonomously. Moreover, the power these organizations have is domain specific. This means that e.g. an umbrella organization for health insurers can only exert power over issues related to this topic (Deschouwer & Hooghe, 2011).

The opposite situation is the so-called neo-corporatist system. In a neo-corporate system the number of participants during the decision-making process is restricted: the government only negotiates with a limited number of important and big interest groups. These so-called peak associations are often able to extend their power since they are often the ones responsible to enforce the agreed policy measures. In systems like these it is relatively difficult for smaller organizations to influence policy (Deschouwer & Hooghe, 2011).

In a neo-corporate system the involvement of stakeholders takes place in a very structured manner. This means that interest groups from different involved organizations have the opportunity to be involved with the development of policy, and if necessary also with its implementation. Critics argue that this way of policymaking does not lead to the best outcome possible as each stakeholder puts the interest of its own sector first instead of other principles such as the appropriateness, logic and financial feasibility of governance (Deschouwer & Hooghe, 2011).

At first sight the neo-corporatist system looks less appealing since this systems leads to a closed circuit of decision-making with a limited number of interest groups involved. As it is difficult for new interest groups to become involved in the process in a corporatist system, innovations in policy-making can be slowed down in such systems. On the other hand, other scholars defend the corporatist system by arguing this system leads to better economic conditions, such as low

unemployment and price stability (Deschouwer & Hooghe, 2011). Deschouwer and Hooghe (2011) argue that both systems have its advantages and disadvantages.

However, in recent decades the influence of civil society on the decision-making process has changed a lot in Belgium and the Netherlands, two countries with a neo-corporate tradition.

Deschouwer and Hooghe (2011) define civil society as the total of organizations which fulfill an intermediary role between citizens and the political system. Deschouwer and Hooghe (2011) mention two reasons for this change. First, there has been a trend of so-called ontzuiling. This is a Dutch word which describes the break-down of traditional religious and sociopolitical barriers. Before this change most people in these countries were member of a collective group based on religion or a political ideology. Ontzuiling led to the de-ideologization of many established interest groups. As a consequence many of these eventually merged. For example, in the Netherlands the catholic trade union merged with the socialist trade union into the still existing Federation of Dutch Trade Unions (FNV) in 1976. The second development Deschouwer and Hooghe (2011) mention is that in recent decades established interest groups were challenged by many new interest groups. These new interest groups are concerned with postmodern and post material values, such as the environment, international solidarity, and LGBT rights (Deschouwer & Hooghe, 2011).

As a consequence of the trends mentioned above the Netherlands has a neo-corporatist tradition in which civil society, which exists of many different types of organizations, has a relatively big influence on the decision-making process. This derivative of neo-corporatism is known as the so-called polder model. In this model decision-making is ideally done on the basis of consensus, whereby every involved societal group is allowed to contribute to and decide on new policy. These vary from interest groups which merely guard its supporters’ material interests, to very ideologically motivated

(14)

10

organizations which e.g. defend human rights. These organizations vary in size and level of professionalism (Deschouwer & Hooghe, 2011).

2.2 Categorizing policy instruments

The completion of the policy cycle ideally leads to the formulation of (a number of) policy instrument(s). A widespread model categorizing different policy instruments is the toolkit

perspective by Christopher Hood (2007). According to Hood (2007), answering the question of what governments do can be answered by describing the tools that governments use, rather than by looking at what these tools are used for, or how these tools are decided on. According to Hood (2007) “what government does to us is to try to shape our lives by applying a set of administrative tools, in many combinations and contexts, to suit a variety of purposes” (p. 2). Hood’s (2007) toolkit perspective scheme is about what government does in this sense. It is about the tools that

government uses and which the public directly come into contact with. Hood (2007) focuses on the workings and on what governments try to achieve, rather than on the course of events within government.

It is a difficult task to make sense of the many complex policy instruments used by governments. Hood (2007 tries to makes sense out of a big collection of policy tools used by

governments by following two steps. First, Hood makes a distinction between government’s tools for ‘detection’ and its tools for ‘effecting’. Detectors are instruments used to collect information.

Effectors are the tools used by government to make people do something. The terms detector and effector come from cybernetics, the science of general control systems, and are both significant abilities that literally any system of control needs (Hood, 2007). A control system needs to collect information about what is going on, in order to take measures based on that information to influence the world outside to which the information relates. Therefore, according to Hood (2007), these terms are inherently connected to government, as government aims to control a society.

The second step Hood (2007) takes answers the question on what foundation government forms its detectors and effectors. This yields the NATO scheme, an acronym that sums up four basic abilities governments inherently possess. The acronym stands for nodality, authority, treasure, and organization. These four resources governments can rely on for detecting and effecting. Thus, combining detecting and effecting with the four types of recourses provides us with eight types of tools that can be used (Hood, 2007).

Furthermore, Hood (2007) categorizes different types of detectors and effectors. Starting off with detectors, Hood (2007) presents how government’s four types of resources can be used to collect information. When it comes to nodality government draws in information simply by being nodal. This comes down to being in the middle of an information network. The detectors that draw in the information are called nodal receivers. Authority can be used as resource by simply demanding information. Hood (2007) calls this officially requested information requisitions. Treasure can be used as a resource to collect information by simply paying for it. Government thus uses rewards to collect information. Finally, organization can be used as a resource to collect information by way of physical effort of doing research. These are what Hood (2007) calls ergonomic detectors. Subsequently, Hood makes a further distinction by placing detectors on a scale from passive to more active, based on the amount of physical effort and the difficulty of collecting the concerning information (Hood, 2007)

When it comes to types of effectors, Hood also offers four basic types of effecting tools of which each one relates to one of the four resources of the NATO scheme. Nodality is used for

(15)

11

payments are used for effecting by using the resource of treasure. Finally, organization is used for effecting by using direct actions, which Hood (2007) in this context has named treatments. Accordingly, Hood (2007) makes a further distinction at which level the effecting tool is applied, namely particular, group, and general. Particular effecting tools are meant to influence e.g. a certain person or company. General effecting tools are meant to influence everyone or everything to which the measure applies. An example of such a measure is a speed limit. An effecting tool meant to influence a group lies between particular and general effecting tools. The NATO distinction, accompanied by the distinction on these three levels, offers twelve types of effectors.

Each of the four resources that provide governments with different options to steer society, can be depleted in a different way, and all have different limitations. Nodality enables government a strategic position from which to spread or gain information in the role of a ‘figurehead’. The biggest constraint on nodality is credibility. Authority enables governments to determine a legal framework, and is limited by the extent of legal standing, i.e., the legal right to initiate a lawsuit. Treasure enables governments to exchange money or something with money-like property. This resource is thus limited by something not possessing this quality, which as a consequence cannot be exchanged. Finally, organization allows governments to directly intervene instead of indirectly, and its limitation is a government’s capacity (Hood, 2007).

Hood (2007) points out that distinguishing between the four resources gives one a rough distinction on the types of tools government can use, and that the four are by no means completely separate. The model offers a means to distinguish between different types of policy tools (Hood, 2007). In the following sections I will separately discuss each of government’s four basic resources, and how these are used for detecting and effecting more elaborately.

2.2.1 Nodality

Nodality depicts the quality of being in the center of an information network or a social network. The meaning of ‘node’ is a point or area where multiple lines or paths intersect or part. Typically,

governments are nodal to some degree (Hood, 2007).

Detecting

So, how does government use its nodal position to collect information? Government’s trait of nodality comes down to government having a central position, and being visible and interconnected. In this network the so-called nodal receivers detect information. This can be in a passive manner, by simply just being there. Hood (2007) names this type of detection unsolicited tenders: information which is picked up without actively searching for it

Information acquired like this can have many different forms. For example, the business community that lobbies for particular legislation, or researchers alerting governments about a problem like global warming. No matter what, government does not pay for this form of information. The pay-off for the one offering the information only might exist in the form of influence (Hood, 2007).

Free information received by government can also be a byproduct of its nodal quality. An example of this is how modern public transportation systems offer the byproduct of big data on its trajectories, which then can be used to decide on certain investments (Hood, 2007).

Government can pick up information based on its nodal quality in an active manner as well. Governments can simply ask parties for information, and accordingly encourage them to deliver this information by making itself more accessible. Hood (2007) calls this way of collecting information ear trumpet, and accordingly makes a distinction between channels where people can and cannot offer

(16)

12

the information anonymously. For example, in many countries a phone line exists where people can anonymously report certain crimes. When anonymity cannot be guaranteed, this is often because the information offered concerns a citizen’s experience with that government. This is the case with a government division’s complaints department (Hood, 2007).

A more direct means than the ear trumpet is scrutiny of free media which means that government employees devote time to the observation of media sources. This mean has become more relevant with the arrival of electronic media (Hood, 2007).

Direct inquiry is a more active type of nodal receiver. Direct inquiry is when government goes outside to collect information from people at first hand. Examples of this can be something simple as a survey, or something more controversial like an undercover operation carried out by

law-enforcement. However, nowadays the carrying out of surveys and research has grown into a

substantial part of government. If this is the case, they are no longer purely ‘nodal’ receivers as they rely on government’s capabilities to fund them. Such surveys and studies are rather active

propositions, which rely on treasure as a resource. These will be discussed in the section about treasure (Hood, 2007).

Effecting

Government’s quality of being nodal can be used for effecting as well. Information has been used as an effecting tool since way back, and continues to play an important role (Hood, 2007). Hood (2007) first mentions the suppression of information as an effecting tool. Government can suppress information by keeping silent, presenting misleading information, or by disinformation.

So called bespoke messages are a very distinctive type of information distributed by

government, and they have the most particular form. The most active form is direct notification. This is when citizens do not actively have to look for the information distributed. An example of this is a message with which government informs the public about something that is going to expire. This can be by letter, or by an electronic governmental service. Two other forms of bespoke message are unprompted and prompted query response. Here the public is triggered to get access to certain information. Unprompted query response is when government provides information in such a form that the question asked cannot be predicted, or when people looking for certain information

probably do not want their identity to be revealed. It differs from prompted query response because in that case the questions can be predicted. Here, government notifies the public that certain information is available (Hood, 2007).

The opposite of a bespoke message is a broadcast message. This is a standard message distributed to whomever it may concern. Hood makes a distinction between three types of broadcast messages, namely privished messages, the packaged self-serve message, and propaganda. Privished messages stay in a relatively small circle, because the information is only picked up by certain people. An example of this is an ad put in a paper only read by a certain audience. The packaged self-service message is meant for a less exclusive audience. This is standardized information available to almost everyone looking for it. Many government websites offer this kind of information, e.g. in the form of ‘frequently asked questions’. Propaganda, the third type of broadcast message and least exclusive one, is often associated with autocratic regimes (Hood, 2007). Hood (2007) however denotes a more neutral meaning of the word, and talks about “the kind of general information for which government does not wait to be asked, but rather seeks to press upon the world at large” (p. 37). An example of this can be warnings about tobacco consumption (Hood, 2007).

(17)

13

regarding who it is directed at. Group-targeted messages are directed at a group instead of at someone or everyone. Group-targeted messages come in the same form as bespoke messages: government may use replies in order to spread certain information, or government may use the direct notification principle aimed at certain groups (Hood, 2007).

2.2.2 Authority

Authority refers to government’s legal or official power. That is the power “to demand, forbid, guarantee, and adjudicate” (Hood, 2007, p. 5). It is often viewed as one of the most characterizing qualities of government (Hood, 2007).

Detecting

Hood (2007) names detectors used by government that are based on authority requisitions. Here government uses the weight of sanctions to compel citizens to offer certain information (Hood, 2007). Again this tool’s methods range from passive to active. Here I will present them from the most passive to the most active.

Two more passive methods to request information are named obligation to display and obligation to notify. Obligation to display is when a particular individual, group or organization is required to display symbols as a result of which something can be determined in a relatively easy manner. Personal identification and the listing of food ingredients are a commonly used example of this. Obligation to notify is when government insists that certain information should be brought to their attention without the display of symbols. Government may e.g. request people of a particular age to register for military draft purposes, or request parents to register their newborns (Hood, 2007).

A return is a more active way for government to request information based on authority. Here the role of government is more active as it often sends out standard forms to be returned by the person it is sent to. The distinction between return and other forms of requested information by government lies in its frequency. It often concerns notifications that citizens have to report monthly, quarterly, or annually. The income-tax return is a classic example of a return (Hood, 2007).

The most active forms of information-gathering are the ones in which government not only indicates certain information is requested, but also goes out to obtain the information wanted. This means that in this case government and citizens meet. Two types of this kind of information-gathering can be distinguished: interrogation and inspection. These two vary in terms of where government and citizens meet. In the case of interrogation, the citizen’s attendance at government is required in a specific place. An example of this is the questioning of returning nationals at the airport. The second type, inspection, requires an even more active role for government as this type entails government to go out. Examples of this are the inspection of factories, restaurants and other businesses (Hood, 2007).

Effecting

Authority, like the other basic resources of government, can also be used as an effecting tool. Using authority as an effecting tool is often done by a request that prohibits or demands certain behavior. When it comes to using authority as an effecting tool Hood (2007) makes two distinctions to tell apart different kinds. The first distinction Hood (2007) makes is between particular and general uses. The particular applications of authority are named directed tokens, and are aimed at specific people, organizations or places. In those cases authority is used in general to adjust behavior, the

(18)

14

directed and blanketed tokens, there is a variety, which is aimed at particular groups. These are called group-targeted tokens. Hood (2007) makes a second distinction related to the degree of constraint that the demands or prohibitions put on the behavior of subjects. Low-constraint requests are often merely suggestions or approving of something. High-constraint measures are the opposite, as these are obligatory and restricting (Hood, 2007).

Examples of low-constraint directed tokens are different types of certificates. These are an “authoritative declaration about the properties of an individual or object, often in terms of fitness or unfitness” (Hood, 2007, p. 57). A familiar type of certificate is one that provides certain information about an individual, document or other object. These are not only used by government, but also by private actors. Examples of these are employee I.D. cards and certain stamps used for official documents. Less frequent but well-known examples of certificates are about less straightforward qualities of individuals or organizations. An example is a certificate for an academic or technical achievement (Hood, 2007).

According to Hood (2007) certificates are the go-to tool in government’s toolkit to adjust a subject’s behavior in three types of situations. The first situation is when there is a need for officially approved standardized information. An example is the energy efficiency labels used for electrical household appliances. By obligating companies to display this information in a standardized manner consumers are able to make an informed decision, whatever decision this may be. The second situation is closely related. This is when government utilizes a so-called ‘seal of approval’ to designate certain behavior as desirable. This is the case for farms that are labelled as organic when certain rules are adhered to. The third and final situation in which certification is often used is when a certain government needs something from another government. One often needs particular certificates to sell meat across the border. The examples mentioned show that it is possible for government to merely use certification to accomplish certain behavior (Hood, 2007).

Directed tokens of medium-constraint are conditional tokens and enablements (Hood 2007). A conditional token is when government takes action when something has happened. This action can be positive as well as negative for the subject. Examples of these are bonds, guarantees and

contracts. Conditional tokens do not explicitly forbid unlicensed activity. Enablements, however, allows one to undertake an activity that would have been forbidden without it, or it may exclude one from certain rules that would otherwise apply. An everyday example is a parking permit.

Enablements are a very common tool to regulate businesses. Government may demand you to be licensed for certain activities, which may be valid for only a certain period. Government may use these enablements to change the activities or the way in which a business operates. A license is an example of an enablement, but enablements can go under many different labels. Examples of these are permits, warrants, and quotas. The relevant difference can be found in the way government hands out these enablements. Government hands out certain enablements in unlimited quantities. An example of this is a driving license: everyone who passes a driving test gets one. But government may also limit the distribution of some enablements. Reasons for this can be that government wants to maintain an overview or limit certain activities. Limiting certain enablements also makes them more valuable and this may inspire good conduct from businesses. Casino licenses are an example. Another difference in the way government hands out enablements can be found in whether they are qualitative or quantitative. The examples mentioned above, a driving license and casino licenses are both qualitative. Qualitative means that you either get them or you do not. Quantitative

enablements however appoint a quota or a quantity. Well known examples are production quotas used to control the amount of dairy produced or the catching of fish. Businesses are allowed to

(19)

15

produce a certain amount and no more. Conditional tokens and enablements can be considered as medium-constraint varieties of tokens of authority because in the case of conditional tokens the outcome depends on the subject’s behavior, and in the case of enablements certain activities are allowed but not obliged to be carried out (Hood, 2007).

Finally, constraints are the most constraining form of a directed token. They demand or forbid certain activities, can be negative or positive, and sometimes both. Constraints in most societies are commonly used to maintain law and order. Constraints with this goal are often negative. Examples of this are constraints that prohibit violence, and drug dealing. Positive

constraints require positive action. For example, most countries government’s demand some form of compulsory education. Constraints that are positive and negative at the same time demand and prohibit certain behavior. For example, national broadcasting networks may be bound to certain educational goals, and are limited in the amount of content they are allowed to broadcast that is merely entertainment (Hood, 2007).

As mentioned earlier group-targeted tokens are in between directed tokens and blanketed tokens. In this case government e.g. utilizes group organization to distribute certain tokens or only allows groups that possess a certain certificate to perform particular tasks. In the case of group-targeted tokens government may also apply constraints to certain groups. In the early 2000s the Australian government e.g. excluded sex offenders from travelling abroad as the problem of sex tourism in Southeast Asia became apparent (Hood, 2007).

Each type of directed token discussed here can be used as a blanketed token as well, and it may also range from a low to a high degree of constraint. Certification in a blanketed form can be named standard approval. Certificates of this type are different in two ways. First, these types of certificates are transferrable. This is the case with roadworthiness certificates for cars. The second difference is that government sets out unbinding targets and norms, such as accepted driving behavior (Hood, 2007).

An open compact is a more general form of a conditional token. Here government offers a deal to whomever it may concern. Amnesty is a common example. In this case, government promises reduction of sentence to anyone who acknowledges past crimes (Hood, 2007).

The general form of an enablement can be named an open permit. The same as with certificates, enablements can be considered general if they are transferable. There are many examples of tradable permits. Production quotas are often organized by distributing permits which allow parties to produce a particular amount of something. In case a producer wants to end his business he is allowed to sell the permit (Hood, 2007). An example is the introduction of phosphate rights in order to regulate dairy farming in the Netherlands.

Standard constraint is the term for the general form of a constraint. Standard constraints demand or prohibit certain behavior from anyone. Standard constraints are the same since they can be both positive and negative. Negative examples are easy to find: there may be a ban on certain types of fishing methods. An example of a standard constraint which demands positive action is the requirement to wear a helmet on a construction site. Another use of a standard constraint is related to nodality: government may require businesses to transmit a message to the public. An example is the requirement for tobacco companies to put warnings regarding their product on the packaging. The use of standard constraints has become very common since the nineteenth century (Hood, 2007).

(20)

16 2.2.3 Treasure

Treasure is the third basic resource of government and describes the possession of money or anything that can be exchanged for money. In the case of government the term indicates its stock of anything that can be freely exchanged. Thus, treasure refers not only to money in the everyday meaning of the word, but to anything that can function as money. Hood (2007) uses the term to “denote whatever positive incentives or inducements government can use to secure information or change behavior” (p. 76). In other words: treasure can be used for detecting as well as for effecting. Also, the use of treasure is a common feature of modern government (Hood, 2007).

Detecting

Treasure can be used as a detecting tool. Here government uses its treasure as a positive incentive to receive information. However, the reward for this information does not necessarily come in the form of money (Hood, 2007).

Unsolicited propositions are the most passive form for a government to use treasure for detecting. This is because government in this case waits to be offered certain information in

exchange for a certain price. Political asylum in exchange for confidential information is an example (Hood, 2007).

Advertised rewards and applications are somewhat less passive type of transactions then unsolicited propositions. In these forms government offers some kind of reward and waits to be approached. The term advertised reward is easily understood: a ‘wanted’ poster to inform the public about an alleged criminal is a well-known example. When it comes to applications the rewards are less straightforward. Here, the information gathered is not the main objective and the reward given in exchange for the information is not explicitly presented as such. There are many examples of applications as it occupies a central position in government: forms you have to fill out when you want to enter certain countries, questions you have to answer when applying for welfare benefits, if you want to get a patent you have to hand over documentation on your invention (Hood, 2007).

Information exchange is again a somewhat more active subtype for government to use treasure for detecting. According to Hood (2007) this one falls right between passive and active. Here, the only reward for the handing over of information is the access to the pool of data which the information is added to. A dating website is a known example in the private sector. In government employment counseling is an example (Hood, 2007).

An active proposition, as the name implies, is again a more active way for government to use rewards as a detecting tool. Contrary to unsolicited propositions government here addresses subjects and offers them a reward in exchange for something else. A present-day example is government paying private companies to do research on what is in people’s minds (Hood, 2007).

Effecting

One can make a distinction between different types of effecting with treasure on different levels. First, according to Hood (2007) treasure can be used by government to change behavior in two ways. Government can use treasure to make an exchange of goods or services, or government can ‘hand over’ treasure without expecting a compensation from those they want to influence. It is difficult to distinguish the effecting tools discussed below on this level. Also, one can have both of the traits at the same time. Second, the same as with government effectors originating from government’s other basic resources, treasure can be used from a very particular level to a very general level. This can be about a payment made to a subject with a specific identity or about a transaction where identity plays no part, as well as about an exceptional transaction (Hood, 2007).

(21)

17

Customized payments are a very particular form of using rewards as an effecting tool. The term explains itself: it concerns highly particularized payments. Contracts and transfers are two types of customized payments. Contracts are particularized payments because they are directed at a particular person or organization. Also, contracts are characterized by a quid pro quo: it always concerns an exchange for something in return. Something in return may include not to do something or to bring something to an end (Hood, 2007). According to Hood (2007) contracts are widely used by government to alter behavior. Grants and loans also fall within contracts as certain conditions have to be met in order to qualify for these. Accordingly, Hood (2007) makes a distinction between

contracts with substantive and procedural quid pro quo. Grants and loans are an example of contract with procedural quid pro quo. A substantive quid pro quo is a more tangible reciprocation, for example a supply of something. Transfers are different from contracts since these do not require something in return. Hence there is no question of quid pro quo. In this sense transfers resemble gifts. Examples are promotional gifts exchanged by heads of state, but also sickness benefits. Moreover, transfers are not only unconditional. They are also particular, as transfers are decided upon on a case-by-case basis and the subjects’ identities are of importance. Hood (2007) argues that especially today transfers are “one of the most common instruments of government” (p. 88).

Conduits are less particularized effecting tools than the different types of customized payments mentioned above. Conduits use a middleman to relocate treasure to a subject. Moreover, conduits can be used as a quid pro quo as well as a transaction that does not require a favor in return. The principle is used a lot in taxation, where government uses companies and organizations as conduits to collect these. Government does this to prevent to be seen as too much involved in who gets what. Conduits are used for other types of government payments as well. One example is compensation payments. These are made via e.g. farmers’ associations to farmers in case of failed harvest. Aid is another example of a type of transaction which is often paid out via another organization (Hood, 2007).

Open payments refer to government handing over treasure without worrying about who is receiving the money, or in this case more often about who is receiving the goods. Hood (2007) mentions food drops in inaccessible war-torn areas as an example. Government almost never distributes treasure like this. What it does, however, is making payments to subjects that these are not specifically customized for. Hood (2007) makes a distinction between bounties and ‘bearer-directed’ payments. Bounties are made in exchange for a quid pro quo. This quid pro quo is

undetermined in that the provider of this quid pro quo is not specified. The bounty goes to anyone who fulfils the requirements. There are many examples of government choosing to pay out bounties which have many different goals. Examples of these are to stimulate certain activities such as fishing and growing crops. In practice however, bounties often lead to subjects behaving in a way which was not intended. This was the case in France after the First World War when the French government attempted to support its merchant navy by introducing a subsidy which was positively correlated to the distance travelled. The outcome was that many non-French merchant sailors switched over to the French flag, which made the subsidy unsustainable (Hood, 2007). ‘Bearer-directed’ payments are the more generalized payments which do not require a quid pro quo. The term refers treasure in some kind of form which makes it not bound to be used by the subject it is given to. In other words, it is easily transferrable and able to be used by someone else. ‘Bearer-directed’ payments to some extent stem from some kind of entitlement. The advantage of these kinds of payments is that from the side of government there is almost no favoritism involved. Examples are the food stamps widely used in the US welfare system (Hood, 2007).

(22)

18 2.2.4 Organization

The final resource of government Hood (2007) considers is organization. Organization depicts

government’s ownership of or access to a group of people with a specific set of skills, whatever these skills may be, with access to a building, and the right material. The term refers to capacity and capability (Hood, 2007). Hood (2007) mentions countries having an army as an example of organization as a government resource.

Hood (2007) considers organization as a separate resource of government. However, organization is often linked to the other three resources in the ‘NATO’ scheme: it is often seen as a subdivision of those three instead of as a fourth element of government’s basic resources. However, Hood’s (2007) reason to treat organization as a separate element is that “it is perfectly possible to derive nodality, authority and treasure from organization rather than the other way round” (p. 102).

Detecting

Just like nodality, authority and treasure, organization can be used to collect information ranging from passive to active modes. Ergonomic detectors is the term for detecting tools which stem from organization, in particular physical devices such as equipment and trained staff. The information here is collected without the help of an informant. This is of course always the case when the interest is not human (Hood, 2007).

Turnstiles and fixed scanners are the most passive detecting tools from information-gathering through organization. A turnstile is a detector which is located in a place that is almost impossible to avoid. These can be related to modern technology. An example is passport control at the airport when you want to enter another country. A fixed scanner is a device which overlooks a particular area of potential activity. A fixed scanner is thus not impossible to avoid, and slightly less passive as these to some degree require an active search process. An example is a watchtower because these watch over a certain area and require an observer (Hood, 2007).

Mobile scanners and hidden scanners are somewhat less passive. Mobile scanners require government to physically move its people in order to conduct surveillance. Police cars patrolling the streets are an example of a mobile scanner. In comparison with mobile scanners, hidden scanners are not visible to the public. These range from espionage to hidden cameras to telephone taps. Hidden scanners are more active since these rely on an informants work activities. Hidden scanners can be distinguished from inspections, propositions, and directs request as these assume a voluntary transfer of information (Hood, 2007).

Effecting

Effecting with organization allows government to move directly on individuals, organizations, their assets and their surroundings. These kinds of activities are inherent to government. In order to effect with organization as a resource, government uses it in the shape of direct action or treatment. Hood (2007) mentions four subtypes of treatment, namely: marking, storage and custody, transportation and distribution, and processing. Hood (2007) accordingly discusses these four to offer an insight into the range of these applications.

Marking literally refers to marking something that is of interest to a certain cause. Examples are marking a sick tree, marking certain dog breeds as dangerous and seal tobacco that went through customs. Applications of marking are often related to authority (Hood, 2007).

Storage and custody is when government provides custodial and storage facilities for optional to more obligatory uses. For example, many countries have or used to have government-owned banks. Storage and custody facilities offered by government can be physical in nature, as is

(23)

19

the case with strategic oil reserves in the United States. A more compulsory application of storage and custody is involuntary detention such as children’s homes. The same as with marking, storage and custody is often related to authority (Hood, 2007).

Transportation and distribution refers to government being engaged with activities that involve the relocating and distributing of goods, and sometimes even people. This includes garbage collection and distributing ballot tickets. An example which involves the movement of people is an evacuation (Hood, 2007).

The term processing refers to government changing an item’s status. A well-known application of processing is national vaccination programs for children. Other examples are pest control and the destruction of livestock in case of a threat of a rapidly spreading disease (Hood, 2007).

As with the effectors described before, these direct actions or treatments also range from particular to group to general applications. Individual treatment implies government taking actions against an individual or a particular organization. Group treatment can be found in between

treatments aimed at particular individuals and treatments directed towards anyone. Crowd control is an example of a group treatment. At-large treatment is the term for general applications of effectors stemming from organization, such as water fluoridation and biological weapons. But the

improvement of welfare by providing public goods such as parks and roads is an example as well. Many commonly used applications of at-large treatments are not very exciting, such as clearing snow and sewage maintenance. Another traditional example is when government monopolizes the

production of a particular good, such as alcohol. However, it is easy to confuse a measure like this with the distribution of a license (Hood, 2007).

2.2.5 Comparing Hood’s toolkit perspective

Hood (2007) toolkit perspective was first developed 25 years ago. Other scholars also tried to illustrate the tools used by government. A good question to ask is how Hood’s (2007) approach compares with these alternative analyses of tools of government. Hood (2007) offers three advantages that come with studying government from the toolkit perspective.

First, according to Hood (2007) this perspective allows one to make sense out of what seems an endless possibility of policy tools. Hood’s (2007) toolkit perspective simplifies an overwhelming variety of policy tools by categorizing its content, which for example allows for comparison.

The second advantage mentioned by Hood (2007) is that the toolkit perspective not only allows for a better understanding of government, but also provides a framework which can be used to solve policy problems. If a government e.g. wants to positively influence a public health issue, one ought to use tools that work. Hood’s (2007) toolkit perspective makes it possible to compare policy tools that have been used over time in order to tackle the same issue, to compare individual cases, and to study if a certain tool might do the job.

Another advantage of the toolkit perspective according to Hood (2007) is that this perspective is relevant especially today. In the past two to three decades there has been a major increase in people having access to digital technologies. This development has changed many things, and how government interacts with citizens and companies is one of the major changes. Digital technologies lead to the appearance of old policy problems in a new form. An example of this mentioned by Hood (2007) is online bullying. Studying how government makes use of its toolkit to counter problems like this allows us to track change in which instruments are used or in how they are redeveloped to address these new types of issues (Hood, 2007)

(24)

20

2.3 Studying policies and their effects

The change a policy measure causes is what we would call an effect. These effects can be categorized into different types of effects. Intended effects are the effects of a policy it was meant to achieve. Unintended effects on the other hand are the effects of a policy which policymakers did not intend to achieve (van Schendelen, 1990). An example is a government implementing a new wage policy increasing the minimum wage, which has led to a deterioration of the relationship with employers’ associations as a consequence. It is likely that this was not intended and probably also undesirable. It is also possible that the implementation of a new policy has the opposite effect as a consequence, namely an unintended effect which is desirable afterwards. A decrease in stolen bicycles after the implementation of a new policy which requires cyclists to wear a helmet is an example. The

observation that policies often have unintended effects led to an increase in research on this subject. According to van Schendelen (1990) this body of research can be classified into five categories, each of which each focusses on a different relationship between the development and effects of policy measures. Figure 2.2 below shows a schematic representation of these five relationships.

Figure 2.2. As based on Figure 3 in van Schendelen (1990, 163).

The field of research focusing on the relationship between agreed policy measures and intended effects (a) studies policy theory. One can presume policymakers and politician ought to achieve something. They have a policy theory about particular policies, which is a logic that explains how they think these policies will achieve a particular goal. However, when studying the policy theories of these politicians and policymakers it turns out that the basic assumptions they make are often contradicting and changing. Also, the inferences they make are often ill thought-out (van Schendelen, 1990).

Implementation studies is the field of research focusing on the relationship between agreed policy measures and the actually delivered policy measures (b). These studies show that there is often a significant gap between agreed policy measures and the policy measures that were actually implemented. There are various reasons for policy measures to be often ineffective. Politicians may (unintentionally) give a symbolic meaning to their promises, and do not really consider that words and deeds must coincide at some point. Policy measures may be lost sight of or change form due to bureaucratic processes. Another reason is that policymakers may turn out not to be able to realize agreed policy measures because of limited resources or changing circumstances (van Schendelen, 1990).

A Stakeholder Analysis of the Dutch National Prevention Agreement