How stakeholders of the new housing tax make sense of change during times of demographic decline : an analysis of stakeholders in Den Helder and their stories.

How Stakeholders of the New Housing Tax Make Sense of Change

during Times of Demographic Decline.

An Analysis of Stakeholders in Den Helder and Their Stories

Ruud de Joode 10115080 University of Amsterdam

17-8-2014

Supervisor: B. Hissink Muller Second reader: A. Blessing

0.0 Abstract

Stakeholders of housing policy in half of all municipalities will be confronted with two major changes in the future: shrinkage and a new housing tax. This research looks at the specific situation in the city Den Helder, which is confronted with shrinkage without official recognition from the central

government. This research shows that different stakeholders have different perspectives of what shrinkage is and what should be done about it. Only the stakeholders directly involved in the new housing tax show clear interest in its goals and consequences. This research used stakeholder identification theory as described by Mitchell et al. (1997) to show that the stakeholder position stakeholders think they have relates to their perceptions and reactions towards policies. More explicitly, the type of reaction one gives can be explained by the perception of one’s power, legitimacy and what is perceived as urgent. This research also used sensemaking analysis as

described by Weick et al. (2005). Sensemaking explains how the thought process is linked to action. It explains that with slow change without a clear start, it's hard to recognize the cues that indicate the need for adapting to this change. Also, it shows how labels like anticipation can lead to different interpretations of what needs to be done. Without judging whether the actions currently taken are justified, this research argues that without explicit labelling unintended actions may follow. For many Dutch municipalities that will be confronted with demographic decline in the future, the question whether this open interpretation is desirable should be considered.

0.1 Acknowledgement

I would like to thank a couple of people for assisting me in writing this thesis. Firstly I would like to thank the people who were interviewed: Yvette Vleugel-Wieringa, Marion van den Broek, Jeroen Pistor and Martin van der Maas, thank you so much for allowing me to interview you, without you this research wouldn’t have been possible. I also would like to thank my supervisor, Bas Hissink Muller. Firstly for helping me during the process of getting to a research topic, secondly for your understanding when my tablet crashed (I am not thanking Microsoft here for a reason). And finally, for reminding me that the debating tricks I normally teach other people, excite your public. Without your help I think I would still be stuck in writing limbo. I also would like to thank Illy van der Putten, the person who has given me feedback during the entire process. Who was willing to comment me on my draft from A to Z and turned out to be a real APA specialist. Thank you, there are many ways to show love, this must be one of them. Finally, I would like to thank Reinier de Adelhart Toorop, for being the person who was able to phone me the Dutch definition of anticiperen according to his Van Dale. I really should get myself one of those soon.

Index

2.1 Past changes in the Dutch housing policy ... 7

2.2 Narrative analysis ... 9

2.3 Mitchell et al., stakeholder identification ... 11

2.4 Conclusion ... 14

3.0 Methods and techniques ... 16

3.1 Getting to a research question ... 16

3.2 Type of research ... 18

3.3 Case selection (selecting a research domain) ... 19

3.4 Combining the two theories ... 23

3.5 Stakeholder selection ... 24

3.6 Data gathering ... 25

3.7 Data analysis via the operationalization of variables ... 25

3.8 Conclusion ... 26

4.0 Den Helder in times of major change ... 28

4.1 Den Helder, an introduction ... 28

4.2 How the new housing tax influences the tasks housing associations can accomplish ... 29

4.3 The consequences of demographic decline ... 31

4.4 Regional cooperation ... 33

4.5 conclusion ... 35

5.0 Stories of stakeholders ... 36

5.1 Central Government ... 36

5.2 The municipality ... 42

5.3 The housing association ... 46

5.4 Inter-municipal organization ... 50

5.5 Citizens ... 53

5.6 Conclusion and overview ... 55

6.0 Explaining the stories ... 58

6.1 Stakeholder positon ... 58

6.2 Sensemaking ... 61

6.4 Conclusion ... 65

7.0 Conclusion ... 66

7.1 Answering the research question ... 66

7.2 Discussion ... 67

1.0 Introduction

Chaos is a paradox: we, as individuals or groups of individuals, are confronted every day with an ever changing world. Sometimes it's fine to ignore these changes, they don’t concern us directly or their effects are so unclear that they pass us by without directly changing our lives or actions. At other times however, we might have a direct stake in the things that change. For example, when our current set of actions were dependent on them being a certain way. At those times we are stakeholders.

This research is about actors who have a stake in housing policies. These stakeholders are being confronted with a changing policy, one that might influence them in such a way that they need to change their actions. This case is specifically about the new housing tax introduced by the central government, levied on housing associations and landlords with social housing.

This change is however not taking place in isolation, because changes happening in the world usually aren't. At the same time the government is taxing social housing, the demography of the Netherlands is changing. Although the population number rises the next few decades, increasingly more

municipalities are confronted with demographic decline. Between 2004 and 2013 the population of 37% of all municipalities declined and around 34% (140 municipalities) are confronted with structural decline (Derks, 2013). The government has recognized that shrinkage needs extra investments, they therefore introduced an exception to the housing tax for housing associations that are in the officially designated shrinkage areas. These associations get a discount on the tax when they demolish or restructure housing. However, not all municipal governments, housing associations, regions and residents that are confronted with decline are in those official shrinkage areas.

1.1 Reading guide

The theoretical framework (chapter 2) will explain how changes in housing policy in the past have changed the roles of different stakeholders. How stakeholders react to these changes is explained using sensemaking theory (Weick et al., 2005) and stakeholder analysis (Mitchell et al., 1997). These theories form a hypothesis in chapter 3, where the research question is also selected and the rigid process of stakeholder selection is explained. Why a single case study approach with the thematic analysis of interviews and documents has been chosen will also be explained.

To illustrate the changes that the stakeholders are being confronted with, the research domain is described in chapter 4. Here, the changing circumstances of shrinkage and the new housing tax are also explained. During the fieldwork the importance of cooperation became apparent, why it's also mentioned in this chapter. Chapter 5 shows the different stories of the stakeholders. It'll be illustrated how every stakeholder perceives the different changes and acts accordingly. Chapter 6 first looks at the attributes of Mitchell et al. (1997), since both theories suggest perceived

stakeholders positions influence behaviour and perceptions. Next, the Weick et al. (2005)'s sensemaking is used to explain thoughts and actions. In the end, the three main conflicts between intentions (by the central government) and the actual consequences in the case study area are shown. The final chapter, chapter 7 gives the answer to the research question, suggestions for additional research and for improvements of this research. It also tries to estimate the relevance of the conclusions for other cases. It suggest the belief in the concept of “beating the odds” can’t be unique to this one case.

2.0 Theoretical framework

As the introduction describes there is a policy shift in housing policy happening. To see how stakeholders are expected to respond to such major change this chapter will show that:

1. Housing policy in the past has had multiple shifts resulting in changing stakeholder positions 2. That sensemaking analysis shows the way stakeholders deal with changes to the status quo 3. That stakeholder identification has attributes that can show the position of stakeholders This chapter will link what appears to have happened in the past to the theory of sensemaking. Sensemaking describes aspects of the way stakeholders make sense, or the lack thereof. One of the aspects mentioned by this theory of Weick et al. (2005) is the way stakeholder make sense of their surroundings. It explains what takes place within a stakeholder when the status quo is no longer and the stakeholder needs to adapt. Weick et al. (2005) explains that this is sensemaking is based on a couple of factors, including the perception one has of its own position. It is the position of

stakeholders that is further developed in Mitchel et al. (1997) stakeholder analysis. It shows that the basis of stakeholders positions are the building blocks urgency, power and legitimacy.

2.1 Past changes in the Dutch housing policy

One of the focusses of this research is the position of stakeholders. This section will elaborate on how these positions have changed since the second wold war. It’s hard to understate the dominance of housing associations in Dutch housing policy, more than 30% of all houses are owned by

corporations (CFV, 2013). Although this is not an exclusively Dutch subject, this percentage is a lot higher than in other Western European countries (Kemeny et al., 2005). To understand the scope of these associations in the Netherlands it is important to understand their basic history.

Table 1: Overview worth and assets of housing associations (CFV, 2013).

Housing associations in the Netherlands

381 (in 2013) Amount of houses in their control 2,4 million

Total worth assets €106.7 billion

Since the second world war the central government has made many changes in housing policies. Table 2 gives an overview of policy changes before the ’90. It shows how there has been a shift in the degree to which stakeholders can influence each other and policies. The foundations of the current housing associations originated right after the second world war when there was a great housing shortage (Priemus, 2009). Policies have changed the situation over those years from a highly

coordinated system of housing associations to a system with more freedom concerning finances and policies.

Table 2: Policy changes after world war 2 and their effect on actors' positions (Tesselhof, 2007).

Year Change in policy Change in position

1958 Banking function back at the national government

creates more central role for central government

1961 At new housing projects the preference will be for housing associations instead of municipalities

decline in influence municipality

1976 Commercial and social renting are equalized meant to increase influence commercial renters, instead increased influence social renters

1976 the requirement to repay the advances of the government is dropped

Increases the financial position of the housing association

1976 Municipalities can’t add provisions anymore towards housing associations

decline in influence municipality 1988 Abolishment of the Woningwet loans and

the guarantee by the WSW for loans of the associations on the capital market

Increases the financial position of the central government

Two groups clearly profited from these changes. Social housing on the one hand and owner-occupied housing on the other grew to respectively 32% and 55% of the total housing stock. This was due to the collapses of commercial renting, which went from 60% in 1950 down to 13% in 2009 (Priemus, 2009). More autonomy for the housing associations came throughout the ’70s and ‘80s but has also led to more critique. Financial risks where too high on the side of the government and the costs kept rising (van der Schaar, 1987).

2.1.2 Brutering and its consequences

The main breakpoint for housing associations was the Brutering, also known as the grossing and balancing scheme. During the Grossing Operation, they received the current value of subsidies the government had pledged them for the coming decades. This was done to solve the problem of rising operation subsidies, created with the DCRS (Kemeny et al., 2005). That same year, housing

associations had to convert their loans from public to private. From that moment on they were responsible for the financial risks that had been the government's responsibility (Priemus, 1995). Creating a unitary rental market in which non-profit associations could compete the rental market for tenants, as long as their main focus remained social renting (Kemeny et al., 2005). These new freedoms caused two important changes in the relationships between different actors.

Firstly, the consequences were mainly for the relationship between the government and the housing associations (Helderman & Brandsen, 2011). Since the government could no longer directly

command through resources, influence now came through the Social Rental Housing Management Decree (BBSH) (Brandsen, 2006; Brandsen & Karre, 2010). Forcing them for example to reinvest their profits into social housing, creating a revolving fund. But also widening their tasks, not only the houses but also the direct surrounding of their housing became their responsibility (van der Schaar, 2003; de Kam 2013).

The second implication was for the relation between corporate supervision and the housing associations. Questions that have risen are mostly related to how much responsibility and freedom both parties have, such as: how should we measure societal improvements, how much money is needed to obtain these improvements (Helderman & Brandsen, 2011)? Supervision of housing for the people was being done by VROM, whereas financial supervision was performed by the Centraal Fonds voor Volkshuisvesting (CFV). However a parliamentarian research about concluded that during

a period of more than fifteen year a lot of discussion about oversight hadn’t let to better oversight. In the meantime the risk-profile of associations was increased, using their freedom more and more for commercial endeavours (Hoekstra et al., 2012). Due to the housing associations often creating juridical connections to other corporations, the question of what the most important responsibilities of housing associations are, remained relevant (Helderman & Brandsen, 2011). Being a hybrid organisation, housing associations started to reinvent themselves. Restating their purpose to reclaim legitimacy for their privileged position.

2.1.3 Regaining legitimacy

Having only a few (not too strict) guidelines and no direct competition, housing associations are neither market parties nor governmental organisations (Hoogendoorn et al., 2006). Instead they exist as an "in between" organisation, no longer a governmental organisation but neither a market . Because of professionalization they didn’t work with volunteers anymore, but instead with professional staff they take away a lot of local participation (Hoogendoorn et al., 2006). So their legitimacy therefore can't be gained from democracy or the workings of the free market, it must be gotten from society itself (Edderouzi, 2011). Many associations have therefore adopted the profile of societal organisation, as an organisation for society. This profile has been popularized in their year reports, seminars and study meetings (Cloudt & Geurts, 2013). Others recognize this societal role for associations as well (Brandsen et al., 2006; Minderman, 2008). Although formally this recognition might be justified due to some overlapping aspects for all associations, not all associations are alike and the concept is not well enough defined (Gruis, 2000).

2.1.4 Conclusion

Since the second world war there have been multiple changes which influenced the position of different stakeholders in housing policy. The major change in the nineties, in which associations gained more freedom, required them to reinvent their own legitimacy. When confronted with the need to change because the status-quo is no longer, sensemaking takes place. Sensemaking is a theory of how stakeholders make sense of changes and then use that new found sense for action. This will now be discussed.

2.2 Narrative analysis

Sensemaking is a type of narrative analysis. Narrative analysis gives the opportunity to portray a subjective sense of reality. Bryman (2008) states that with narrative analyses the question shifts from “what has happened” to "how do people make sense of what happened". It can help to detect stories interviewees are telling throughout the interview without making these stories explicit on their own (Riessman, 1993). This makes narrative analyses an ideal analysis tool for finding these stories about how stakeholders deal with the changes to the status quo.

To see how changing conditions influence different stakeholders Rhodes and Brown (2005) used narrative analysis as a way to review business and management literature. In this story about narrative analysis as a tool, as they call it themselves, they identify five principal research areas in which narrative analysis is used. One of those is areas is sensemaking (Rhodes & Brown, 2005). Sensemaking allows for people to make sense of organisational events.

2.2.1 Weick, Sensemaking

Sensemaking is the process in which one tries to “structure the unknown” (Waterman, 1990). The simplest definition is that sensemaking is “the making of sense” (Weick, 1995). It places stimuli into some kind of framework. This frame enables us to "comprehend, understand, explain, attribute, extrapolate, and predict” (Starbuck & Milliken, 1988). This social process happens especially when discrepant cues interrupt an individual’s activities. It helps the individual to rationalise the ongoing

activity in retrospect (Weick, 1995; Weick et al., 2005). So an individual can make sense of a situation (comprehend) when a change can be explained by linking cues to a frame to explain the change in the framework he/she used to be present in.

Since this research looks at changes in a framework (in this case a policy shift), sensemaking can help to show how individuals or organisations develop their awareness of change. Therefore it is

important to look at the different phases of sensemaking so that they can be recognized during fieldwork. These phases have been described in an abstract (Basu and Palazzo, 2008) and in a more direct case study (Weick et al., 2005). The three phases combined are shown below:

1. Cognitive

Which implies thinking about the organisation’s relationships with its stakeholders and views about the broader world. This happens in periods of chaos. Sensemaking doesn’t happen in isolation, many cues are influencing the sensemaker at the same time. It is crucial to separate the relevant from the irrelevant cues. Noticing and bracketing is crucial in this phase, words for new events need to be created, bracketing is based on mental models developed earlier. A combination of salient cues and mental models draws the attention to an event that is out of the ordinary.

2. Linguistic

Which involves ways of explaining actions and the way the sensemaker goes about sharing these explanations with others. Starbuck and Milliken (1998) note that understand precedes explaining. The sensemaker does this via labelling. Labelling helps to categorize the stream of experiences. Chia (2000) explains that labelling is a strategy of differentiation and simple-location, identification and classification, regularising and routinezation with the aim to create a more functional deployment. In this case the functional deployment is the sharing of the discovered change with others. These social factors can be the sharing or discussing new knowledge.

3. Conative

Which involves the behavioural posture the sensemaking adopts. After the question; ‘what is going on here’, the question that follows is ‘what do I do next?’ These must be actions, paired with commitment and consistency that shows its perceived relationships.

These three steps show that sensemaking is a process where comprehension is translated into words and then into action. Sensemaking is as a springboard into action, it should be not be seen in a too abstract way, it's more concrete than that (Weick et al., 2005).

2.2.2 Sensemaking in times of crisis and change

Sensemaking has mostly influenced writings on crises, however it also had an impact on literature written about organisational change (van de Ven & Poole, 1995). How a change is perceived is especially relevant because there is research indicating that threat and fear lead to a decreased consideration of external cues and an increased reliance on tried modes of operating (Staw et al., 1981). When confronted with a disruptive situation stakeholders will first look for reasons that enable them to resume the interrupted activity and stay in action (Weick et al., 2005). These reasons are based on frameworks and can be anything between institutions and traditions inherited from predecessors. These subjective estimations have the tendency of over- or underestimating risks. The biases can be either motivational or cognitive. It can be motivational because people take the

consequences of their predictions into account. It can be cognitive due to deficiencies, like the lack of information but also institutions (Gruis, 2000).

Weick finds (1988; 1990; 1993) that small failures during crises will be linked and amplified. These changed circumstances become too great to manage and interrupt well-rehearsed patterns of action which then work less effective. In organisational change the same pattern emerges. A change in circumstances interrupts well-practiced patterns, because change involves movement in an organisational entity overtime (Van de Ven & Poole, 1995).

2.2.3 On what level does sensemaking take place

Research looking at changes tends to focus on the top management levels. For example, Gioia’s research into strategic change emphasizes the sensemaking of the top management (Gioia &

Chittipeddi, 1991; Gioia et al., 1994; Gioia & Thomas, 1996) Social construction should not only take place on an individual’s level of analysis and neither by just looking at “the deciders”. This would overlook the importance of the social processes through which sense is made. As mentioned before, the labels and the new frames are created through communication. Shared meaning, which emerges through the social process, plays a significant role in both crises and change (Maitlis, 2005; Weick, 1995).

Shared meaning leads to collective mental models or frames that underlie organisational

sensemaking. They influence the way the world is perceived within the organisation (Basu & Palazzo 2008). Brickson (2007) also points out that the process of sensemaking within an organisation leads the organisation to view its relationships with stakeholders in a particular way. This, in turn,

influences its engagement with their stakeholders. So the interaction with others is directly influenced by the interaction within an organisation during sensemaking. But sensemaking is not about the truth and getting it right (Weick et al., 2005). Actually, the idea is that sensemaking is driven by plausibility rather than accuracy (Weick, 1995). The enactment approach describes organisations as acting not within a “real” environment but within a perceived environment and behaving not as “real” organisations but as self-perceived organisations (Smircich & Stubbart, 1985).

2.2.4 Conclusion

So sensemaking describes the way that a stakeholder perceives his changing reality. It helps a stakeholder to make sense and use this cognitive process to create new terms and labels in the linguistic process that follows. This new understanding of reality is what leads to action in the last phase, the conative phase. The stakeholder adapts his behaviour to accommodate the new situation. When studying this process it is important to select stakeholders that are relevant to the issue, in this case a new housing tax introduced in times of shrinkage. Not just because it makes it easier to study but also because the interaction between different stakeholders is an important part of the linguistic phase and the knowledge sharing. Identifying those who matter, requires the identification of stakeholders. Mitchell et al. (1997) has created a method for stakeholder to identify who really counts.

2.3 Mitchell et al., stakeholder identification

If you want to know whose position is influenced by changing in policy, three important questions need to be answered. These are, (A) Who holds a stake in these housing policies, (B) what was their position before the policy change and (C) what is their position after the policy change. Stakeholder analysis in general and Mitchell et al. (1997) specifically give the opportunity to look at these questions. This way the change that might have occurred can be distinguished. That doesn’t mean that there is consensus about answers to these three questions. Identifying which actors are actual stakeholders and hold a stake in something (like a policy) has widely been discussed. The question is amongst others, what is the right number of stakeholders? This question is especially hard to answer

since there has been no consensus about what Freeman calls “the principle of Who or what really counts" (Mitchell et al., 1997).

2.3.1 Definition of a stakeholder

Windsor (1992) points out there has been an opinion difference between researches that take a broad or narrow view on the stakeholder universe. Making it harder to distinguish relevant and irrelevant stakeholders is the fact that the term “stakeholder” has been used indiscriminately according to Friedman and Miles (2006). Not only in the academic world but also in commerce and politics the term stakeholder is used widely. More than 66 different concepts for the term “stakeholder” exist (Mainardes et al., 2011). So a specific definition is required in order to identify the right stakeholders for this research.

The definition of Freeman and Reed (1983) might help for this purpose. They created a wide definition of stakeholder, in which they state that a stakeholder is an individual or a group who "can affect the achievement of an organisation's objectives or who is affected by the achievement of an organisation's objectives". This alludes to an interactive relationship between the organisation and its stakeholders in which both influence the other. Stakeholder analysis however has not only been used to identify relations inside a company. It also has the possibility of looking at the interaction between multiple institutions and individuals and look at their power relationships (Mayers, 2005). Since then it has been developed further by key articles including one by Mitchell et al. (1997), which made it possible to look at more than just the internal relations within companies and organisations.

2.3.2 Urgency legitimacy and power

In the article written by Mitchell et al. (1997), a specific way to identify stakeholders has been developed. In this article it is argued that there are three relevant attributes for identification, being urgency, power and legitimacy. These are separated based on the fact that an actor can have one without the other and still influence the organisation. Legitimacy is norm based and a result of operating in a socially constructed world (Mitchell et al., 1997). It also balances power for one will receive pressure when operating outside the accepted norms (DiMaggio & Powell, 1983). Power is a variable and not a constant according to Mitchel et al. (1997), meaning that one can gain or lose it. Etzioni (1964) identifies three power bases. First there is coercive power, powers based on physical threats. Secondly there is utilitarian power, which is based on the use of material means for control purposes. Lastly there is normative power based on symbols. Like prestige of esteem. Urgency is the last attribute taken under consideration. This attribute shows the urgency of the demands of the stakeholder towards the organisation in terms of time and it being critical for the survival of the organisation.

Table 3: Overview of stakeholder types defined by power, legitimacy and urgency. According to Mitchell et al. (1997) # Name of stakeholder Attributes present Description example Latent stakeholders

are barely recognized and/or barely recognize their position themselves

1 Dormant Power Have the power but don’t have a claim, power stays unused

Those who can spend, influence or command intention if wanted with little effort required. Like a government

2 Discretionary Legitimacy With the lack of power or urgency there is no

pressure to deal with these stakeholders

A interest group of a legitimate subject, non-profits who require donations

3 Demanding Urgency Annoying but ignorable. They are able to confront their interests barely to others.

protesters with a cause that is barely recognized by the general public Expectant stakeholders 4 Dominant Power Legitimacy

Most of the time is in the possession of a formal mechanism to influence the organisation owners, significant creditors, community leaders or political leaders 5 Dangerous Power Urgency

Trying to get its way through force. Are outside of the band of legality

wildcat strikes, employee sabotage, terrorism

6 Dependent Legitimacy Urgency

Are dependent of others to put their interests at the centre, can use

stakeholders with power to accomplish goals victims of misbehaviour Other 7 Definitive Power Legitimacy Urgency An expectant stakeholder can become a definitive stakeholder by gaining the last attribute and

guarantee direct attention to their needs

This will mostly be dominant stakeholders whom have a sudden urgency that needs to be dealt with.

8

Non-stakeholder

These different attributes lead to the different stakeholder types as shown in table 3. In this table stakeholders are separated according to

their attribute

combination. Some stakeholders only have one specific attribute while others have a combination of two or even all three. The amount of attributes directly influence the way a stakeholder is treated, stakeholders with only one attribute will be mostly ignored. They themselves

also tend to barely claim any direct attention. Depending on their attribute, they could claim a second attribute. It are those stakeholders, expectant stakeholders, who have a more active claim; they do expect something. The way they go about it depends on which combination of attributes they hold. The last group of stakeholders that matter are those who have all three the attributes. As table 3 shows, these are mostly stakeholders who go from dominant to definitive, since it only takes an urgent claim to make them change.

By explaining Mitchell's theory two things become clear. Firstly his theory is really focused on the perspective of a company towards others. This will have to be adapted so the theory can be used for this research (see 2.4.2). Secondly, it is very much based on the perspective of reality that this specific company has. Mitchell et al. (1997) already notes however that there isn’t one “true” reality but that reality is a construct of multiple perceptions and changes over time. What is clear though is that Mitchell can help in assessing stakeholders position before and after policy changes and can help determine which actors are actual stakeholders and thus matter. Mitchell's notion about realities also directly relates to the second part of this research. Stakeholders have unique perspectives on the expected future because of their position and their perspective on reality. Narrative analysis can help making those different perceptions explicit.

2.4 Conclusion

So what is clear from sensemaking is that how organisations (or stakeholders) perceive a change depends on a couple of factors. It depends on (1) if they are able to recognise this change, (2) how closely their current set of patrons are usable in this change and (3) and how they label the change that leads to new action. Sensemaking gives the stakeholder the opportunity to construct meaning and use this as a springboard for action. The bridge between meaning (the cognitive phase) and action (the conative phase) is linguistic. To see if a stakeholder is able to do sensemaking, not only the top management levels should be considered. The entire organisation takes part in this social process of sensemaking. This process directly influences the way the organisation sees itself, which influences how the organisation sees others surrounding it. Stakeholder analysis on the other hand shows what the self-image of stakeholders is based upon. The attributes mentioned are the building

blocks for this self-image and guide the way stakeholders with different combinations of attributes behave.

2.4.1 Combining the different theories

This makes both theories very relevant for this study since it is looking at times of change and how actors coop and react to these changes. The first part, 2.1, already showed that these shifts have occurred in the past. It also indicated that housing association have had to reinvent themselves in at least one attribute, legitimacy, when another (power) changed as well. If can be established what the coming change is large enough this would main that their current perspective on their own urgency, power and legitimacy has to change again. Mitchells theory can help establishing this. If they are actually able to see the need for this change depends or if stakeholders are able to link their frame with these cues. Weick's theory will help explain this.

2.4.3 The need for adaption

However to be able to use these two theories some of their aspects need to be adapted.

Sensemaking is now very focussed on what has changed, instead on what is about to come. Mitchells stakeholder analysis is focussed on the interaction between stakeholders instead of stakeholders position towards policy. The next chapter will explain, among other things, how these theories will be adopted for this research.

3.0 Methods and techniques

This chapter will outline how the research will be set up. The first part of the chapter will explain how the research question and the hypothesis were shaped. The second part will then outline how this question will be researched, how the theories mentioned in chapter 2 will be customized and how the variables will be operationalized. In short this chapter will state:

1. How the hypothesis based on the theoretical framework leads to the research question. 2. What research type will be used and why it has been chosen.

3. Which case has been selected and for what reasons.

4. How the theories of Mitchell and Weick have been adapted for this research. 5. Who the selected stakeholders are and how they were chosen.

6. In what way the data will be gathered.

7. How the operationalization of the variables takes place.

3.1 Getting to a research question

The aim of this research is to identify opinions and stories about shrinkage and to see what effect the new housing tax has on these stories. Additionally, it aims to find out if these opinions are related to how a stakeholder identifies itself and whether they correlate to a stakeholder's position according to the new policy. Part 3.1 will explain how (A) this aim of the research leads to a hypothesis and a research question that can best be answered through the use of qualitative research. And, it will show (B) why this qualitative research would best be done by using a case study and why in this study the nature of the units of analysis leads to a holistic single case study.

3.1.1 Linking the theoretical framework to the hypothesis

The theoretical framework indicates that sensemaking helps stakeholders to observe reality and that it is based on the current perspective of yourself and others. Stakeholder analysis has showed that self-image is based on three aspects. The theory also suggests that self-image can have

consequences on behaviour and that it doesn’t change easily when the environment disrupts it by changing. The theoretical framework has also shown that changes in housing policy have happened in the past and that those policy changes have changed the roles of different actors. If sensemaking is based on self-image, which doesn't always change when policies change, what would be the effect of the new policy on sensemaking? The following hypothesis makes a prediction based on all of these theories.

Hypothesis

The expectation a stakeholders has of the future is directly influenced by the internal and external identification of a stakeholder. Internally the stakeholder's overall position is determined by its power, urgency and legitimacy. When the situation around the stakeholder changes and sensemaking takes place, this internal view can conflict with the new reality. Instead of changing along the stakeholder holds firm to its current internal view and the view of its surrounding, hereby creating a new conflict between perceived and actual position.

This is a very general hypothesis that is based on the theories of both Mitchell and Weick. This research adds the following assumption to the hypothesis:

The new housing policy is such a change which has effect on the position of some stakeholders and therefore on their stories about the future.

When looking at this hypothesis it becomes clear that the research question in its simplest form is: Is this hypothesis with the corresponding assumption correct? However there are multiple factors that make the hypothesis and therefore the research question more complex.

3.1.2 Conceptual model

This hypothesis leads to the abstract representation as show in figure 2. The internal variables that influence the units of analysis are shown on the right, a clear representation of the variables named in Mitchell et al. (1997) that influence a stakeholder's overall position. On the left, the external variables are shown. The

housing tax, shrinkage in the region and pressure to work together are all variables that can't be controlled by the stakeholders. The way they observe these aspects will vary. The way they see them will then influence the stories they tell about the future.

3.1.3 Research question

An important factor this research will take into account is demographic decline (also mentioned as shrinkage throughout this report). Chapter 4 will discuss the three different changes that the specific actors are being confronted with, but shrinkage is most important for the research question. What also needs to be specified for this research is who the stakeholders are that will be looked at. Since the research question will be answered using Weick and Mitchell it is important to note that the research methods ask for specific and nuanced positions. Specificity and nuance are best reached through qualitative research and therefore a case and stakeholder selection needs the be made. This will further be discussed in chapter 3.2. By taking all these different aspects in mind the final research question is as follows:

Are the stakeholders' stories about shrinkage in the future in Den Helder adapted to the new housing tax based on Mitchell's three attributes, and can these adaptations (or lack thereof) be explained by Weick's sensemaking analysis?

This is a more specific research question than the previous, however some choices might still seem random. This chapter will explain some of the specifications of the research, like the choice for the city Den Helder, for the specific actors and for using Mitchell and Weick. What is clear though is that this research question cannot be answered with a simple yes or no. Therefore sub questions will first need to be answered.

3.1.4 Sub questions

This rather complex research question can be divided into multiple sub questions. Most sub questions have their own chapter. It is indicated behind every question in which chapter this question will be answered. The sub questions are as follows:

1. What are the major changes that effect the stakeholders in Den Helder? (4.0) Figure 2 conceptual model of influences on sensemaking.

2. What stories do the stakeholders in Den Helder have about their expectations of the future relating to these major changes? (5.0)

3. Can the stories of the stakeholders be explained by Mitchells stakeholder theory? (6.0) 4. Can the stories of the stakeholders be explained by Weicks sensemaking theory? (6.0) 5. How do the positions found for the different stakeholders relate to the intended new policy?

(6.0)

Now that the research question and the sub questions are established the method of answering needs to be determined. This will establish the method of conduct for this research.

3.2 Type of research

As mentioned before sensemaking is a linguistic process, it’s about finding words for observations before this leads to action (Weick et al., 2005). This asks for a way to observe those thoughts, especially when those thought are hard to define before the research and are depended on specific stakeholder positions this tends to be complex (Bryman, 2008). To make sure the answers to the research question are being convincing the research needs to be detailed enough to get those nuanced answers, while still saying something about the situation in general (Siggelkow, 2007).

3.2.1 Case study

It is case study research that is concerned with complexity and the particular nature of a case in question (Stake, 1995). A case can't effect only a single individual. It effects a collective of individuals or smaller groups that are bound by a specific aspect. This can be family relation (Lewis, 1961), employment (Burawoy, 1979; Pollert, 1981) or community (Whyte, 1955). But a case can also be a location or a specific event (Bryman, 2008). In this case all stakeholders are part of the same area and involved in the same events, although they are very different in size and nature. The fact that their opinions are treated equally makes this a holistic case study.

However, Gerring (2007) states that a (single) case study analysis does not exist, because it's impossible to conceptualize, without any form of cross-case analysis. Only by doing analysis on a wider array of cases a specific case can be selected. This means the basis for selecting a specific case needs to be solid. Bryman (2008) suggests the same thing and explains that a case study should be considered a case study when the geographical location has meaning for the research itself. Especially because, as Siggelkow (2007) notes, case study researchers cannot use the “canonical” statement ‘results are significant at p < 0.05’. They can however choose an aim by selecting a specific type of case study and show why the selected research area helps in finding this aim (Siggelkow, 2007).

3.2.1.1 Type of case study

Therefore, there is a burden of proof to show why a specific case was selected and how it compares to other potential cases. What type of case is needed for the study is important to consider. Yin (2003) gives five types of case studies: the critical, unique, representative, revelatory and longitudinal types. A research question generally asks a combination question of yes/no and why. This means the hypothesis could be falsified or confirmed. A critical case study aims at showing if a hypothesis does or doesn’t hold (Yin, 2003). Flyvberg (2006) notes that one of the common misunderstandings of case studies is that ‘case study contains a bias toward verification”. This means that if a hypothesis states that X will by Y, the case study will proof it true. This gives case study a drawback because it doesn’t use the scientific method. Diamond (1996) means that the scientific method would curb tendencies to stamp a pre-existing interpretation on data. Some researches use criteria like the reliability, replicability and validity which are supposed to prevent that (Yin, 2003). These criteria mean the research could be repeated by someone else without the expectations of the researcher.

3.2.2 Reliability, replicability and validity

Bryman points out that there is discussion in the academic world about the appropriateness of reliability, replicability and validity as research design criteria (Bryman 2008). Like mentioned earlier Yin (2003) considers them appropriate while others barely mention them (Stake, 1995). Bryman (2008) concludes that it are mostly researchers who use case study as a qualitative research method play down these criteria. They point out that every research has the risk of running into subjectivity, for example when selecting research variables (Flyvberg, 2006). They also bring up positive sides of case studies, like the fact that research with a smaller N forces researchers to go for falsification earlier and that with depth case studies the field itself is a disciplinary factor when the hypothesis is false. This is something that quantitative research doesn’t allow for (Ragin & Becker, 1992; Geertz, 1995). Geertz (1995) experienced having to adapt the hypothesis to the actual situation during the start of his fieldwork. This created the additional value of a better research question for the rest of the fieldwork.

Whether case studies can be used to generalise is widely discussed. Generalizability is a direct aim of using a case study to falsify or confirm a general hypothesis (Flyvberg, 2006). Some argue case studies can test theories in a “hard” sense, comprising explanation and prediction (Eckstein, 1975). Flyvberg, (2006) confines theory to a softer sense, to testing a hypothesis. He notes that case studies are ideal for generalizing using ‘falsification’. In case of a falsification of the theory the general strength of case studies (Flyvberg, 2006) is used. The research is given a direct incentive to aim for falsification. When doing so, a general hypothesis is considered for a specific, critical case (Yin, 2003; Flyvberg, 2006; Bryman, 2008). The research tries to show the general hypothesis doesn't apply here. Although this research agrees that the most important criteria is generalisation, it does aim to meet the other criteria mentioned by Yin (2003) and Bryman (2008). This will be done by observing before interpreting and explaining extensively how and why certain decisions were made. This provides to possibility to replicate and the transparency to decide whether it's reliable and valid.

Hopefully the process leading to the selection of a case study research is now clear. Since the units of analysis are the stories about shrinkage in the future by stakeholders in Den Helder the research will be very linguistic, especially with the theories of Weick et al. (2005) and Mitchell et al. (1997) being used. Therefore this research aims at being in-depth by using a single case study, which will be holistic because all stakeholders share common attributes in location and upcoming events. Because using a case study approach does create some risks the selection of the case will be described very extensively below.

3.3 Case selection (selecting a research domain)

When working with a qualitative single case study one has the risk of selecting a case that is (a) not representative for its intended peers (there is a bias) and/or (b) leads to results that aren’t relevant for other cases (Seawright & Gerring, 2008). The main aim for the case selection was finding an area that would show how the policy changes are effecting the plans to combat demographic decline. To prevent selecting an irrelevant case, or making it impossible to find a similar case in the future, the following criteria have been set for selecting a case in this study:

1. A clear geographical boundary needs to be set. 2. Demographic decline needs to be present. 3. It is not an officially recognized shrinkage area. 4. The area needs to have an urban character.

The reason for using these criteria will be explained and at the same time the results of the criteria will be shown. Afterwards it will be explained how the selected case meets the criteria.

3.3.1 Assessing the criteria

A geographical boundary seems necessary to restrict the case. As Bryman (2008) mentioned a case can be bound by its geographical area. Where this exact boundary should be set depends on the topic of research. Will this study look at a municipality, a Corop-region or a province? There are 40 Corop-areas based on a core municipality (Kramer et al., 2009) Firstly it's important to keep in mind that housing associations are most directly influenced because they have to pay the tax. The area that they operate in will be effected by the amount of tax they need to pay (Koninkrijk der Nederlanden, 2013). A second matter to consider is that a high level of aggregation has

the risk that one will miss changes in local situations. Figure 3 shows the development of the amount of households between 2012 and 2025 in Corop areas in the Netherlands. According to this map there will be growth almost everywhere in the country. Another conclusion shows when looking at the municipal map for the same time period. Figure 4 shows this clearly, shrinkage is visible in more areas than the Corop level allows for. The only levels lower than the municipal level are the neighbourhood, block and that of a single house. However not only would it be hard to get numbers and stories about the future at a level that low, the consequences for general policy would also be limited. This means a case is ideally at the aggregated level of a municipality, if its housing association doesn't operate in a region bigger than the declining municipality. That way their policies have to take shrinkage into account.

2. Demographic decline present

Since this research wants to look at the stories about shrinkage in the future, shrinkage needs to be present in the case study. The first thing to keep in mind when measuring demographic

Figure 3 Development amount of households per Corop from dark blue to dark red; strong shrinkage, shrinkage, stable, growth, strong growth by CBS (2013a)

Figure 4 Development amount of households per municipality from dark blue to dark red; strong shrinkage, shrinkage, stable, growth, strong growth by CBS (2013b)

decline is that the way to measure it matters for the results, which can be illustrated when looking at figure 4 and 5. In figure 4 the change in the amount of households is shown. This figure shows that there is only a decline expected in very specific parts in the country. Figure 5 however shows demographic decline spread throughout the country. This because this figure shows the decline in population. This means that although less people will live in a house, most municipalities will not be confronted with vacant housing caused by demographic decline. The second thing to keep in mind is that the decline can’t be in the distant future. If people can’t feel and see that demographic decline is happening, they won't have stories about it yet. Thus demographic decline, based on the definition of decline in population and household, must be present in the municipality.

3. No officially recognized shrinkage area

The central government of the Netherlands has divided the country in three areas based on

demographic changes (IPO, 2012). First there is the group of municipalities that are called the “Top shrinkage regions” (top

krimpgebieden). These areas expect a decline of 16% in population until 2024. The second group are the anticipation areas. In these areas there isn’t any decline yet but there will be in the near future, although less than in top

shrinkage regions. The last group is “the rest”, for these municipalities there is no- or barely any shrinkage predicted. Figure 6 shows the location of these areas throughout the country. That this research isn’t about the last group is clear from criteria 2. However, the difference between the first two types is less clear. Both will experience decline, though the first group to a larger extent. Regardless of the similarities, the government only made exceptions for the top shrinkage regions in the new housing tax; they have to pay less tax if they implement specific policies (see

4.2). So associations in these regions are spoon-fed cash-incentive driven policies, making the Figure 5 Population development per municipality from dark blue to dark red; strong shrinkage, shrinkage, stable, growth, strong growth CBS (2014c)

Figure 6 Shrinkage and anticipation areas red: Official shrinkage areas, yellow: anticipation areas by (IPO, 2012)

local “story” very government driven. For the anticipation areas the option of paying less tax doesn't exist, making the local choices more relevant. Therefore, by choosing an anticipation area it becomes possible to look at the choices that need to be made due to shrinkage on a local level, without the central government pushing these choices in a specific direction. Though anticipation areas officially don't show decline yet, this depends on the level of aggregation. A specific municipality in an anticipation area can actually show decline, while the whole area, on average, doesn’t.

4. The area needs to have an urban character

There are several reasons why an urban municipality would better serve as a case.

Municipalities of all different sizes cooperate with each other. Larger municipalities tend to act more on their own or are a leading factor in their cooperation between municipalities (Schilder et al., 2011). Also, the discussion about shrinkage becomes more relevant because it's not just buildings that are confronted with the problem but whole neighbourhoods. Additionally, urban municipalities frequently hold region wide functions that might be lost when the city’s population growths too thin. This makes the problem more relevant, not just for the individual municipality but for a larger region.

Therefore, by choosing a city with an urban character the stakeholders will tend to act more, the shrinkage will cover whole neighbourhoods instead of specific buildings and the

consequences will be felt by the entire region.

3.3.2 The choice for Den Helder

Now that the four criteria have been set and explained, a research domain needs to be selected. Den Helder (already mentioned in the research question but selected after these criteria were checked) fits all criteria of this profile.

When checking the first criteria it is clear that Den Helder is a municipality putting it at the right aggregation level. While its Corop-region shows residential growth or stability the city shows a declining trend which is only visible when looking at the municipal level. Den Helder has one big housing association that is dominantly active, almost exclusively inside the city. These characteristics cover the first criteria. The second criteria, the presence of demographic decline, is also met. Den Helder is one of the few municipalities who has had a demographic decline both in population and in households between 2004 and 2009 (CMO, 2013). Due to the population of the city decreasing to under the important threshold of 60.000, the amount of city representatives has been lowered (gemeentewet, 1992). The stakeholders will probably not only see, but also feel what is happening. There are a few other municipalities where this is the case (CMO, 2013), however only one meets the third criteria (Strijen) of not being an official top shrinkage region and still having the first two

criteria. Since the same goes for Den Helder these two municipalities are left. Since Den Helder is a municipality with a majority of people living in an area with medium or higher density and the population, although shrinking, is still 56 461people (at the end of June 2014) it can be considered to have an urban character (CBS, 2014a).

3.3.3 Canary in a coalmine

Although there are multiple municipalities that would fit at least one of the different criteria, for the combination there is only one. So the question rises if this case is not too unique, an anomaly that doesn’t have any possibility for generalization. When looking at figure 4 and 5 one can see that not only Den Helder will be confronted demographic decline. Making it unique now certainly doesn’t make it unique in the future, according to expectations for 2040. This makes the city an ideal canary

in a coalmine. It is already experiencing what is expected to happen to other anticipation areas in the coming years, while not getting a special treatment like the three top shrinkage regions. Thus a qualitative in-depth case study of Den Helder holds potential lessons about what other municipalities will be confronted with in the future.

Now the type of research has been selected as well as the case study area, what remains is the question how this research will be conducted. As mentioned in 2.4 both theories of Weick and Mitchell need to be used together and adapted before stakeholders and data gathering methods can be selected.

3.4 Combining the two theories

To be able to select stakeholders and variables both theories from the theoretical framework will be used. For this purpose the theories need to be combined. Sensemaking gives the opportunity to see the process of making sense of a new reality (Weick et al., 2005). Being able to see the change, a.k.a. the problem, as a stakeholder, and understanding which parts of the routine can stay and which must change. Sensemaking discusses the ability to see oneself. Stakeholder analysis gives the opportunity to look at what the basis for this judgement is (Mitchell et al., 1997). It provides the attributes stakeholders use to judge themselves and the situation. Combining these two methods gives the opportunity to see how stakeholders form an image about themselves. And if this doesn’t match with the new policy, see whether this might be explained by their ability (or lack thereof) to see the framework they are acting within. However these theories need to be slightly modified for this research. These modifications will be made explicit to increase the reliability of the research.

3.4.1 Stakeholder analysis

Mitchell focusses his stakeholder analysis on the perspective of a company towards the different stakeholders it has. The company uses its perspective on reality to decide the position of the other players. This research however, looks at how stakeholders who have a stake in a certain policy see their position within this policy. This means some definitions must be modified to fit stakeholders in policies rather than companies. These definition changes are listed below in table 4.

Table 4: Modifications made to stakeholder analysis. Original definitions used in stakeholder analysis and the adapted

definitions for this research Mitchell et al. (1997).

Original definition Modified definition

stakeholder can affect the achievement of an organisation's objectives or who is affected by the achievement of an organisation's objectives."

can affect the objectives of a policy or is directly affected by the achievement of a policies objectives

Objective model

to show the way an company identifies stakeholders and decide which to give priority

to show the way a stakeholder identifies its position within a policy and if this corresponds with the objective of the policy itself

reality there isn’t one “true” reality but that reality is a construct of multiple perceptions and changes over time.

there isn’t one true reality, there are however stakeholders whom can enforce their perspective of reality, like the creators of public policy

Power power to influence the firm the ability to act (or lack thereof) Legitimacy legitimate relationship with the firm the justification to act (or lack thereof)

Urgency urgency of stakeholder's claim on the firm

the motivation for action (or lack thereof)

3.4.2 Sensemaking

Because sensemaking is a process rather than actor driven, it requires a setting of a change to be able to interpret it. The setting of a policy shift will be used for the modification of definitions as shown in table 5.

Table 5: Modifications to the theory of sensemaking. Original definitions based on sensemaking and the adapted versions

for this research (Weick et al., 2005).

Original definition Modified definition

framework such as institutional constraints, organisational premises, plans,

expectations, acceptable justifications, and traditions inherited from

predecessors.

The positions that stakeholders hold according their perceived position according to Mitchell et al. (1997).

disruption when the current state of the world is perceived to be different from the expected state of the world, or when there is no obvious way to engage the world.

The new housing tax is such a disruption, it changes that what was expected into a new situation

cues Signs of that what interrupt an individual’s activities.

The exact changes that happen due to the new housing tax

3.5 Stakeholder selection

Now that the theories are customized, a set of guidelines for selecting stakeholders must be followed. In the case of this research, stakeholders are considered as someone who:

can affect the objectives of a policy or is directly affected by the achievement of a policy's objectives. When taking a narrow view this would lead to the one who can directly influence the policy (central government) and the one directly affected (the housing association). In a more broader view

(restricted by the research area) others are effected by the policy as well. Firstly the residents of Den Helder (particularly those living in social housing), will notice if the housing association has to adapt its policies (according to the housing associations, see 4.2). The same goes for the municipal

government. As mentioned in 2.1 they have lost the power to influence housing associations directly. That doesn’t mean they are less dependent of their investments, especially considering their

broadened tasks since the BBHS (2.1.3). The last stakeholder is the region surrounding the municipality which will be influenced due to the urban character of the research case. The

neighbours with whom spatial and economic cooperation is organized will provide the perspective on what the effects are on the regional function of the case.

This leads to the following overview of stakeholders as shown in table 6. There is a direct line of influence from the central government via the housing associations to the residents of their buildings. The other line (with those who are indirectly influenced in their possibilities) are represented by the regional entity and the municipality.

Table 6: Overview of the different stakeholders that are relevant for this research. A direct and an indirect line of

influence are visible as a consequence of the housing tax.

Central government Regional entity Municipality Population and/or residents Housing associations

3.6 Data gathering

Now that stakeholders are selected the next question is how they will be researched, with what data gathering methods. As mentioned before, sensemaking is a very linguistic affair. That means

stakeholders should have the opportunity to give their own interpretation of events and their consequences for the future. However, there are specific steps to check if sensemaking is taking place and there are clear attributes provided by Mitchell et al. (1997) that indicate how stakeholders see their own position. This sets the preference for semi structured interviews. With the linguistic freedom for the respondent to give its own interpretation, while still having the structure to check for certain essential aspects.

Another complication in the field was the fact that not all stakeholders could or would be

interviewed. As a consequence this research combines semi-structured interviews with analysis of policy documents. The approach and data analysis will be discussed below per stakeholder. The central government as one person that can be interviewed doesn’t exist. The main actor of reference in this research has been the minister of housing, Stef Blok. In the end he is responsible for all new housing policies and their consequences. Housing policy itself (as chapter 4.1 will show) is rather technical. His opinion and motivation as minister are well documented in letters by the minister to parliament, which are personally signed.

It is neither preferable nor feasible to interview all fifty-seven thousand residents of Den Helder. Additionally, there were stakeholders who didn’t want to be interviewed. Interviewing governmental workers was not a problem, the Alderman of housing however refused an interview because: “due to her short tenure she doesn’t feel she can be of any added value” (J. van Dongen, personal

communication, July 18, 2014). The major housing association didn’t want to be interviewed because “we prefer written interviews” (M. van den Broek, personal communication, May 27, 2014). To still cover these stakeholders and present a complete information overview, policy documents will be used together with interviews. Especially when looking at frameworks Gioia et al. (1994) shows a strong argument for using these; ‘reading, writing, conversing and editing are crucial actions that serve as the media through which the invisible hand of institutions shapes conduct’. Instead of all the citizens a representative of the neighbourhood has been selected for interviewing, He stated that he, residents and the municipality all see him more as a representative of the people than of the

municipality.

So first there are semi structured interviews for an in-depth look at certain positions, secondly there are policy documents and official announcements mentioning shrinkage and/or the new housing tax. These include documents that actors advised during the interviews for their relevance. To go through these documents certain variables first need to be operationalized. With this operationalization the interviews and documents can be systematically analysed.

3.7 Data analysis via the operationalization of variables

The attributes of Mitchell et al. (1997) have been used as well as the phases described in

sensemaking (Weick, 1995) to operationalize the variables. Thematic analysis uses them to create themes for interview analysis (Ryan & Bernard, 2003). The independent variables urgency, power

and legitimacy are being used. Urgency relates to whether problems have been identified by different actors, legitimacy comes from the relationship to others and power from the ability to influence others. They have to be related with the two subjects established in the research question which are shrinkage and demographic decline. Documents have been selected on the basis of the presence of either the word shrinkage (krimp) new housing tax (verhuurderheffing or

verhuurdersheffing) or Den Helder. In the case of the central government the word "Blok" was also used to make sure that Stef Blok (the minister of housing) was involved in the document.

All data will be used for a thematic analysis (Bryman, 2008), where they’re used to create themes. Per theme different stakeholders opinions will be categorized into sub themes. Depending on the stakeholders themselves, themes and subthemes might form or change. In the case of the municipal government, the loaded word shrinkage is not allowed so the theme would be covered by the more neutral demographic transformation. Ritchie et al. (2003) describes how thematic analysis can be used to create a schematic overview of stakeholders’ views. It requires looking at (Ryan & Bernard, 2003):

1. Repetitions

Looking for reoccurring statements in the interviews and policy documents based on one of the variables.

2. Similarities and differences

Answering; Does a stakeholder hold the same view over multiple reiteration of the same theme in interviews and documents? Do other stakeholder also hold these views? 3. Linguistic connectors

The question: do the stakeholders address causal relations? Between for example: action and policy or action and position.

This way the research can look for the positions stakeholders think they have and for the stories they tell about the future. It will help determining if there is a correlation based on the research by Mitchell et al. (1997). The variables in this research will be the themes, not words or numbers. This means that quotes, sentences describing or fitting the themes will be written down in their entirety to cover the themes. One theme was added besides positions, shrinkage and the new tax. This theme is cooperation (relating to coping with shrinkage). During the fieldwork this subtheme turned out to be so important it will be discussed alongside shrinkage and the new housing tax in chapter 4.0 and 5.0.

3.8 Conclusion

This research will be a deductive holistic case study, aiming to show if predictions made by a general hypothesis are found in a specific case study. This means that the case selected aims at being a critical case study. In case studies generalization is a very important aspect which can be tackled mainly by selecting a case based on a set of clearly defined criteria. Den Helder is unique case in which all criteria are met. However at the same time it stays relevant because other areas in the Netherlands are expected to meet all criteria in the near future. Reliability is further enhanced by showing which adaptions have been made to sensemaking and stakeholder analysis theory to customize them to this research. Thematic analysis will be used to analyse the interviews and

documents produced by the selected stakeholders. Both the themes and selected stakeholders are in first instance based on the theories, definitions and variables. During the analysis similarities,

repetitions and linguistic connectors will be looked for. In this way the current research attempts to answer the multiple sub questions in order to answer this research question:

Are the stakeholders' stories about shrinkage in the future in Den Helder adapted to the new housing tax based on Mitchell's three attributes, and can these adaptations (or lack thereof) be explained by Weick's sensemaking analysis?

Before analysing interviews and documents however, the last three variables related to future problems will be illustrated. As mentioned before they also are themes in the analysis. Chapter 4 will provide a context by describing the setting stakeholders are currently in and will be moving towards in the future.