Deliberating sex work : exploring the challenges and opportunities pertaining to setting up a collaborative governance process for Amsterdam's sex work policy

(1)

Deliberating Sex Work:

Exploring the Challenges and Opportunities Pertaining to Setting Up a Collaborative

Governance Process for Amsterdam’s Sex Work Policy

Nina Tom (10971955)

MSc. Conflict Resolution and Governance University of Amsterdam

Supervisor: Dr. Floris Vermeulen Second reader: Maria Kranendonk

(2)

(3)

Introduction

In the summer of 2018, Femke Halsema was appointed mayor of Amsterdam. Sex work in the Red Light 1 District has been one of the central issues she seeks to address during her term. Based on the premise that “the showcasing of vulnerable women is not acceptable” (Couzy, 2019), Halsema has embarked on a mission to improve the position of sex workers in order to enhance their strength and independence, thereby reducing their vulnerability to trafficking and other forms of exploitation (Couzy, 2019). The whole country is watching as Halsema is about to present different scenarios for the future of sex work in the city to the city council — the so-called ‘ _{Wallen scenarios}’ (Koops, 2018). She is not the first to attempt to ‘improve’ sex work policy and/or the Red Light District. Its strong moral character and high visibility make sex work an attractive policy field for ambitious local politicians. However, despite the apparent effort that goes into the creation and implementation of sex work policy, these policies tend to fail consistently (Wagenaar, 2017). This is not only the case in Amsterdam, but in the whole of the Netherlands.

This elicits the question; why is sex work such a challenging policy field? In the following section, I describe what I call the vicious cycle of ineffective policy. I set out three factors that constrain policy, describe how these factors generate repressive policies, and how such this feeds into the marginalisation of sex workers.

1.1 The Vicious Cycle of Ineffective Policy

1.1.1 Challenges of designing sex work policy. Sex work is a complex policy field. It is characterised by low predictability, uncertainty, and disagreement on problem definitions and policy goals. However, it is not unique in this sense; it shares these characteristics with many other policy fields in modern society. Wagenaar (2017) discusses numerous domain-specific challenges of sex work policy that contribute to its complexity and constrain the possibilities for public policy. I briefly discuss three of

1_{In this paper, I use the term ‘sex work’ instead of ‘prostitution’, unless referring to prostitution as a phenomenon}

rather than the act or occupation. Furthermore, I use the term ‘sex worker’ instead of ‘prostitute’. It should be noted that the term ‘sex work’ is sometimes broader than the term ‘prostitution’. However, I will use the term ‘sex work’ to refer to prostitution in the classical sense, namely where one party provides sexual services to another party in return for money or goods. The reason why I use the term sex work instead of prostitution is because the latter term is considered derogatory by some, as it refers to someone’s body (passively) being used for sex. The term sex work,

(6)

these. First, sex work policy is the archetypical example of _{morality policy} sex work policy generally revolves around ideology in the first place; practical or evidence-based considerations tend to be disregarded (Mooney, 1999). The moral character of sex work policy furthermore results in its high salience to the general public, and abnormally high levels of citizen participation (Mooney, 1999). The deep conflict among the public about how sex work should be approached — in which individuals’ place on the political spectrum does not necessarily predict their stance on sex work — constrains policy options. In light of sex work policy’s moral character, pragmatism becomes an ideological stance of itself (Wagenaar, 2017). Second, there is a lack of reliable data relating to the sex industry resulting from the invisibility of large segments of the sex worker population. It is nearly impossible to establish the most basic of statistics such as the number of sex workers active in a given city at a given point in time (Wijers, 2018). There is also disagreement about the definition of concepts such as ‘forced prostitution’ and ‘human trafficking’. As a result interpretations, estimates regarding human trafficking in the sex industry vary from three to ninety percent (SOURCE). The moral nature of sex work policy makes the lack of reliable data especially problematic, as parties to the debate can employ whichever numbers or facts suit their position. Finally, the sex industry is burdened with a widespread and severe degree of stigma. It emerges in two ways. The first type of stigma directly emanates from morality. It portrays prostitution as a vice, and the sex worker as a deviant. Clients and third parties such as brothel operators are accordingly seen as criminals. The second form of stigma stems from the conception of sex work as inherently exploitative. Sex workers are perceived as agenciless victims of male-instigated violence. More specifically, sex work is increasingly associated with human trafficking. . In other words, debate about of these diverging

1.1.2 Repressive policy._{These three factors severely constrain options for concerted action with} regard to sex work in the shape of public policy. Conceiving of sex work as a problematic phenomenon — either in terms of morality or due to perceived criminality and exploitation in the sector — results in tireless attempts of public officials to minimise the sex industry through the adoption of repressive measures, and a tendency for over-regulation. Yet, repressive policy and high degrees of regulation do not appear to reduce the prevalence of sex work; it merely drives large portions of the sex worker population underground (Vanwesenbeeck, 2017). Working illegally negatively impacts sex workers’ societal position and their access to health care, and increases the risk of exposure to violence (Wijers, 2018). Sex workers consequently also become more vulnerable to exploitation — precisely the opposite effect from that envisioned (Wijers, 2018). In theory, sex work is considered work in the Netherlands and sex workers are considered citizens with agency and rights. In practice, however, repressive regulations often enhance the

(7)

marginalisation of sex workers (Rottier, 2018). The figure below depicts a simplified version of the above-described vicious cycle.

Figure 1. The Virtuous Cycle of Repressive Policy From the above, it is safe to conclude that repressive policy and increased regulation should not be the cornerstone of the _{Wallen scenarios if Halsema indeed seeks to improve the position of sex workers and} enhance their strength and independence. One may ask, then, what sort of policy should be implemented instead? It is not the purpose of this paper to explore this question. In fact, the line of argument presented throughout this paper illustrates why I am not in a position to make conclusive statements about this.

1.1.3 Breaking the cycle. _{I propose that instead of continuously attempting to improve policy,} public officials should rethink the policy process itself. Classical modes of governance have proven to be ineffective in this field. I propose that in order to break the vicious cycle of ineffective (repressive) policy, sex workers need to be structurally included throughout the policy process. At the moment, the three aforementioned factors — particularly the dismissal of sex workers’ agency resulting from their victimisation — prevent collaboration with sex workers. Where sex workers (and other stakeholders) are invited to the table, they are presented with the opportunity to contest prevalent misconceptions about the industry and to counter stigmatising perceptions that result in ineffective, restrictive policy. Hence, their inclusion could mitigate the effects of stigma, morality, and a lack of data that currently dominate the

(8)

1.2 Research Questions

The research presented in this paper stems from the premise that in order to devise ‘effective’ sex work policy, sex workers (and other stakeholders) should be involved throughout the policy process. It explores one particular mode of governance that allows for such far-reaching inclusion: collaborative governance. This research is guided by the following research question:

1. How do sex workers in Amsterdam experience their participation in the policy process?

2. What _{are the challenges and opportunities pertaining to setting up a collaborative governance} process regarding sex work policy in Amsterdam? 2

In light of this former question, I have phrased the following sub-question:

a. How has the establishment of PROUD affected sex workers’ (perceived) participation in 3 the policy process?

The first research question serves to illustrate the current situation, upon which the second (main) research question builds. The following sub-questions belong to the second research question:

b. Why might collaborative governance be useful in this context?

c. Could PROUD fulfill the role of representing sex workers within a collaborative governance process?

d. How_{can potential challenges for collaborative governance process be overcome?}

In line with the research questions formulated above, the aim of this research is twofold. Firstly, I aim to shed light on how sex workers currently perceive their participation in the policy process within the city of Amsterdam. The principal aim of this research, however, is to highlight factors that need to be taken into account when setting up a collaborative governance initiative for Amsterdam’s sex work policy.

2_{The exact indicators for the potential effectiveness of a collaborative governance process will be discussed in detail}

in the theoretical framework.

(9)

1.3 Relevance

1.3.1 Academic relevance. _{This research intends to add to existing literature on sex work} (policy). Existing academic literature on sex work tends to dismiss sex workers’ agency. It often reaffirms the dominant perception of sex workers as vulnerable individuals incapable of making informed decisions about their occupation or lives in general. Sex workers’ perspectives, on the other hand, are rarely at the core of analysis. Research that is directly informed by sex workers’ experiences tends to explore their life stories and is aimed at identifying who sex workers are and what made them pursue this line of work, or serves to illustrate the exploitation or abuse they (have) experience(d). Van der Meulen (2011) observes that “conventional research on sex work does little to help illuminate questions such as how sex workers engage in community change and in struggles for human rights” (p. 372). Literature that takes a more liberal approach towards sex work tends to discuss the topic in terms of law, discourse, or sociology (Wagenaar, 2017). Sex work policy, on the other hand, is a relatively under-researched topic. Even where sex work policy is discussed, the question of how sex workers and other stakeholders perceive policy is usually neglected. Instead, most academic literature on sex work policy revolves around questions such as how political-economic forces affect policy or how discourses around sex work affect policy, and how policy relates to human rights (abuses) (Wagenaar, 2017).

In their book _{Designing Prostitution Policy}, Wagenaar, Amesberger and Altink (2017) discuss sex work policy in the Netherlands and Austria. This research is largely inspired by their findings. The authors suggest the initiation of collaborative governance in the sex work policy context as a remedy for ineffective policies within this field. This suggestion has gained support among sex work academics since (see e.g. Wijers, 2018). However, I noticed that practical questions relating to the establishment of such an initiative remain largely unaddressed. Hence, this paper seeks to contribute to the sex work (policy) literature by exploring such practical questions.

1.3.2 Societal relevance. _{The societal relevance of this research is twofold. Firstly, before a} collaborative governance process or any similar initiative is established in practice, it is necessary to explore where opportunities and challenges lie. As will be illustrated throughout the rest of this paper, the municipality of Amsterdam is increasingly making an effort to deliberate with sex workers and other stakeholders. Where a collaborative policy process would be initiated without sufficient consideration for challenges, opportunities, and boundaries, the process may not only turn out to be ineffective, but could be counterproductive. Increased conflict in the sex industry, further marginalisation of sex workers, and policy deadlock are examples of negative consequences that could be generated by a collaborative attempt that is not thought through, and for which the time may not be right. Accordingly, one of the purposes of

(10)

this research is to shed light on what needs to happen in order to create a ‘ripeness’ for collaborative governance in the context of sex work policy in Amsterdam.

Secondly, this research is largely informed by the views of sex workers. I believe that the inclusion of sex workers’ voices has great societal relevance. The sex worker community is severely marginalised, even in the Netherlands — a state that is widely viewed as one of the world’s most progressive places with regard to sex work. As discussed above, the stigma inherent to the occupation results in the victimisation, infantilisation, and subsequent silencing of sex workers. One of the purposes of this research is to counter the prevalent image of sex workers as victims. The sex workers I have been in touch with, and the insights I have gained during this study have only served to contradict this image. Throughout this paper I hope to convey to others the message that sex workers are an incredibly diverse occupational group comprised of strong individuals whose agency needs to be recognised and whose voices need to be heard.

1.4 Outline of the Paper

The following chapter provides an overview of the methodology of this study. It offers an account of why I selected the city of Amsterdam as the subject of my case study, describes the process of data collection, data analysis, and ethical considerations and challenges pertaining to sex work research. I also discuss how the results of this study should be treated in light of the limitations. Chapter 3 is dedicated to an extensive discussion of the theories that this research is grounded in. I first describe how ‘pluralised ungovernability’ — manifested as increasing distrust in and disaffection from the democratic system resulting from a multitude of interrelated challenges inherent to modern society — results in the inadequacy of traditional modes of governance. I propose ‘collaborative governance’ in response to this inadequacy, and set out a model for collaborative governance in detail. Finally, I briefly theorise representation within the collaborative governance process, specifically in relation to marginalised groups such as sex workers. Chapter 4 subsequently provides the reader with background information on legislative approaches to sex work. It further illustrates how sex work policy is shaped in the Netherlands, and explains how a strong human trafficking frame on the sector is central to the attempted repression of the sector described above.

The findings of this study are presented and analysed in Chapter 5. I first discuss how sex workers experience their participation in the policy process, which seems to be increasing (in part due to PROUD), yet insufficient. I then discuss the value of collaborative governance in the sex work context, and how it reacts to certain of the aforementioned challenges inherent to sex work policy. The bulk of the analysis is dedicated to discussing the challenges for collaborative governance based on the model set out

(11)

in the theoretical framework. The three main challenges I identify relate to (1) the difficulty of proportionally representing the sex worker population and subsequent issues relating to legitimacy of the process and its outcomes, (2) the balancing of the municipality’s incentives to initiate a collaborative governance process with stakeholders’ incentives to participate, and (3) the high levels of conflict and low levels of trust within the sector, in particular vis-a-vis the municipality. Chapter 6 summarises the paper and discusses the implications of these challenges and how they should be dealt with.

(12)

Chapter 2.

Methodology

In this chapter I justify my choice for a case study design and explain why Amsterdam is an interesting case to study. In section 2.2, I explain how I collected my data and the context-specific challenges that I ran into during my fieldwork. Section 2.3 explains how I analysed the interviews I conducted. I then briefly discuss how I have dealt with certain ethical considerations that are of particular importance when researching a sensitive context such as sex work (policy). Finally, I discuss the limitations of this case study in terms of objectivity and generalisability, and the implications thereof for the conclusions presented in this paper.

2.1 Research Design

Through this research, I have sought to obtain insights into sex workers’ views on the policy process and shed light on the challenges and opportunities pertaining to setting up a collaborative governance process in Amsterdam. I have chosen a case study design because it allows me to “elucidate the unique features of the case” (Bryman, 2012, p. 69), and thus to appreciate its complexity. Amsterdam is a particularly compelling case to study for numerous reasons. The city is world-(in)famous for its Red Light District and its liberal approach towards sex work. It is a crucial case to study, as Amsterdam has the potential to influence other municipalities — and perhaps even the national government — in their governance of sex work. Furthermore, the relatively progressive coalition in the city council and the broader liberal political climate in the capital suggest room for experimentation in terms of policy and governance. The council’s relatively pro-sex work stance has resulted in the adoption of BIJ1’s motion to ‘structurally’ include PROUD in the policy process less than a year ago (“Amsterdamse sekswerkers moeten betrokken worden bij sekswerkbeleid”, 2018). It is interesting to explore where the boundaries of the city’s apparent progressivity lie. This is not to say that Amsterdam is the likeliest place for collaborative governance to be implemented. This setting is simultaneously a challenging one precisely because of its high visibility and media attention. The inseparability of sex work in the Red Light District and the mass tourism the city is burdened with, and the municipality’s vested (financial) interest in the Red Light District further complicate the case.

(13)

2.2 Data Collection

Before I started field work, I familiarised myself with sex work (policy) literature and some (collaborative) governance literature. Attending a conference on the decriminalisation of sex work signified the commencement of my fieldwork phase. Many of the stakeholders that this research focuses on were present, including PROUD and other sex workers; government officials; police; SWexpertise and academics specialising in sex work; health professionals from _{Prostitutie en Gezondheidscentrum}4

(‘Prostitution Health Centre’, P&G) and Soa Aids, and many others. Attending the conference allowed me to hear different stakeholders’ positions, as well as to observe their interaction. I subsequently met informally with a prominent Dutch sex work expert. Both the conference and the meeting provided me with valuable insights into the industry, governance processes, and relationships between different actors that I could explore further during my fieldwork.

In search of answers to my research questions I have conducted in-depth qualitative interviews with different stakeholders whom I believe should be incorporated in a hypothetical collaborative governance process. This research is primarily based on the opinions and views of stakeholders because these factors are better indicators for the potential success of a collaborative governance initiative than any other subject of analysis. For this purpose, qualitative interviews are most conducive because they allow for the perceptions of respondents to guide the interview (Bryman, 2012). The interviews were of a semi-structured form, thus allowing for the prior formulation of a few questions or broader themes to be addressed in the interviews, while allowing me to remain flexible and be guided by respondents throughout the interview (Bryman, 2012).

I have conducted seven individual, face-to-face, semi-structured, in-depth interviews, and one semi-structured telephone interview. The interviews lasted approximately half an hour up to an hour. I also briefly spoke to a brothel operator on the phone who initially indicated that he was not interested in the interview, but nevertheless provided his opinion on the matter. Finally, I held a focus group with multiple sex workers and a few health professionals . I recorded the interviews (except the telephone 5 conversation with the brothel operator) and the focus group with permission of the respondents. The interviews were held in Dutch; the focus group was held in English. Excerpts from Dutch interviews used in this paper have been translated into English.

2.2.1 Challenges inherent to sex work research. Sex work is not only a notoriously difficult field in terms of policy, but also in terms of research. Shaver (2005) discusses numerous methodological

4_{P&G is an organisation funded by the municipality of Amsterdam that provides STI checks, vaccinations, legal}

(14)

challenges inherent to sex work research. These challenges largely emanate from the same factors constraining sex work policy set out in the introduction. For example, the lack of data on the sex industry makes it difficult to obtain a representative sample of sex workers (Shaver, 2005). In fact, it is not even clear what a representative sample looks like in the first place. The severe stigma resting on the sex industry further complicates data collection, as it often results in a lack of willingness to cooperate, among other things (Shaver, 2005).

Indeed, I experienced difficulty reaching the sex worker population, even in a country where sex work is legalised. This is not only a consequence of the community’s historical marginalisation and stigmatisation, but also due to the fact that sex work is a rather over-researched topic. Most research is rarely of any value to the sex worker population, and sometimes even contributes to the marginalisation and victimisation of sex workers. Therefore, many sex workers tend to be reluctant about participating in studies. In order to overcome this bottleneck, I had to convince certain ‘gatekeepers’ to the sex worker community of the value of my research. I first interviewed Velvet December, advocacy coordinator of PROUD, after briefly meeting her at the aforementioned conference. She recognised the potential value of the research and subsequently put me in touch with the Amsterdam-based Prostitution Information Centre (PIC). Even though the PIC usually does help students contact sex workers, I was able to arrange a few interviews through the PIC due to December’s positive referral. Similarly, the respondent from P&G referred me to P&G’s confidant for male sex workers, who helped to arrange the focus group.

2.3 Data Analysis

After each interview, I transcribed the recording. I subsequently coded the transcripts according to prevalent themes. More specifically, I engaged in what Bryman (2012) calls open-coding. He defines this type of coding as “the process of breaking down, examining, comparing, conceptuali[s]ing and categori[s]ing data” (Bryman, 2012, p. 569). For the first research question, I predominantly coded in accordance with sub-questions, for example ‘views on the policy process’ and ‘views on PROUD’. For the analysis of the second research question, I largely coded along the lines of the sub-questions as well, for example coding for ‘value of local knowledge’. I further used many indicators set out in Ansell and Gash’s model for collaborative governance, such as ‘trust’, ‘stigma’, ‘conflict’. Once I had sufficiently familiarised myself with the interviews and I had a document with collected quotes for all the different themes and sub-questions I wanted to address, I reflected on how these themes were related and how the data could be conceptualised in order to create a coherent story. I then analysed all factors chronologically. Throughout the writing process, I went over the transcripts numerous times in order to make sure that no relevant data was disregarded, and I continuously engaged with the theory.

(15)

2.4 Context-Specific Ethical Considerations

Shaver (2005) identifies other implications of stigma besides difficulty to find sex worker participants. Namely, he warns that

because membership in hidden populations often involves stigmati[s]sed or illegal behavio[u]r, concerns regarding privacy and confidentiality are paramount. This, it is often argued, may lead individuals to refuse to cooperate or give unreliable answers to protect their privacy. Alternatively, participants may say what they think you want to hear. (p. 297)

Indeed, protecting the privacy of the respondents — particularly that of sex workers — has been a paramount concern throughout this research. I have sought to limit participants’ concerns regarding this matter by guaranteeing sex worker respondents anonymity from the outset . Other respondents were6

provided with the option to have the interview processed in an anonymised manner. In line with the principles of ethical research, the informed consent form respondents were required to sign before each interview further included information about the research (aims), requested participants’ permission to record the interview, and pointed out that respondents could at any point withdraw their participation (“Principles of Research Ethics”, 2012). I furthermore spoke to sex workers in a ‘safe space’, either at the PIC/PROUD office or at P&G.

For the purpose of respecting respondents’ privacy, I did not ask the respondents any demographic questions, nor did I inquire into their personal background or the background of their occupation as a sex worker. Nevertheless, I have come to understand that despite the interviews revolving around policy-related issues, this did not (at all) preclude the interviews from being emotionally laden. Even for non-sex worker respondents, frustration with the current state of affairs was often apparent. Because sex work policy is in itself a sensitive topic, I have tried to show (genuinely felt) empathy towards all respondents, and clarified that they did not have to discuss matters with me that they did not want to speak about. Yet, overall respondents seemed quite keen to share their stories and experiences, perhaps because they appreciated someone listening to their perspective.

(16)

2.5 Limitations

Social science research is usually evaluated based on whether it meets criteria such as reliability, replicability, and representativeness (Bryman, 2012). These factors are often difficult to comply with in qualitative research. Qualitative research is often criticised — if not dismissed — due to a lack of proper measures for assessing its quality (Leung, 2015). Case studies are sometimes regarded as a particularly problematic design within qualitative research. Case studies, many argue, lack validity because (1) the design relies too much on the researcher’s personal interpretations, resulting in a lack of objectivity, and (2) a single case is insufficient basis for generalisability (Flyvbjerg, 2004).

2.5.1 Objectivity. _{Payne and Payne (2004) explain that}

Objectivity in social research is the principle drawn from positivism that, as far as is possible, researchers should remain distanced from what they study so findings depend on the nature of what was studied rather than on the personality, beliefs and values of the researcher. (p. 153)

The research presented in this paper is not objective according to this definition, as few qualitative, interview-based research are. Respondents are likely to have been influenced in their responses by how they perceive of me and what they believe is a desirable message to convey, perhaps not disclosing the full truth. Furthermore, the codes I used are subjective. A different researcher would likely have chosen different codes, interpreted excerpts in a different way, and consequently possibly reached different conclusions than I have. Objectivity is an elusive goal in any research, whether qualitative or quantitative. In the study of human affairs, there will always be context effects (Burawoy, 1998), and the knowledge yielded is always (to some extent) context-dependent (Flyvbjerg, 2004). Burawoy (1998) argues that rather than attempting to limit the researcher’s involvement in the world she studies, the extended case study method asks for the acknowledgement of a researcher’s place within the world she studies — it asks for ‘methodological self-consciousness’. Indeed, it is not the purpose of this research to present the ultimate, objective truth; there is none in this context, and even if there were, the limited scope of this research would not have allowed me to discover this full truth. Rather, I have merely sought to gain an understanding of (perspectives on) the current sex work policy process in Amsterdam, and to explore where potential challenges for collaborative governance may lie. In order to cope with a lack of objectivity, claims made in the analysis are substantiated with quotes from interviews where possible so as to allow the reader to personally assess — and perhaps disagree with — the inferences I draw.

(17)

2.5.2 Generalisability_. Generalisability can be understood in two ways here: generalisability of this case study to other contexts, and generalisability of the sample population to the broader population. In relation to the former understanding, it should be noted that it is not the purpose of this study to generalise the results beyond this specific case. I have not chosen to do a case study of Amsterdam’s sex work policy based on the delusion that the conclusions derived can be applied to other contexts. Instead, the purpose of this case study is merely to evaluate the case at hand in order to derive insights that may have practical value within this context.

The second form of generalisability largely relates to representativeness of the sample. As explained above, I experienced difficulty recruiting respondents for this research, as is usually the case for sex work research. Consequently, my sample is an opportunistic one rather than a representative one. Although I spoke to a somewhat diverse group of sex workers (in terms of gender, age, nationality, type of sex work, degree of political engagement, etc.), I do not operate based on the pretense that this sample is representative of Amsterdam’s sex worker population. As for other respondents, in most cases I only spoke to one person belonging to each stakeholder group. It is impossible to say whether these individuals are representative of their stakeholder groups. Hence, representativeness (and consequently generalisability) of the sample is certainly a limitation of this study, as I was aware it would be from the outset. However, while the research outcomes presented in this paper may not reveal all possible challenges and opportunities, it does provide an account of some of the potential challenges and opportunities that should be taken into account when setting up a collaborative governance process for sex work policy in Amsterdam.

(18)

Chapter 3.

Theoretical Framework

This chapter lays out the theoretical foundation from which this research stems. Section 3.1 offers a theoretical conceptualisation of the background against which the policy process ensues. More specifically, I use Warren’s (2009) concept of _{pluralised ungovernability in order to make sense of} Amsterdam’s sex industry. Many of the factors that Warren identifies as contributing to this pluralised ungovernability manifest themselves in the sex work context, resulting in the inadequacy of electoral democracy and other classical tools for policy making. Collaborative governance can be conceived of as a form of governance-driven democratisation — Warren’s envisioned solution to pluralised ungovernability. In section 3.2, I discuss why collaborative governance might be a fruitful approach for complex policy issues such as the one at hand before extensively discussing Ansell and Gash’s (2007) model for collaborative governance. Finally, I briefly touch upon representation literature, and the importance of _{descriptive representation} within this context.

3.1 Theorising Complex Policy Problems

3.1.1 Pluralised ungovernability. _{Warren (2009) observes that contemporary democratic} systems increasingly deal with democratic deficits; electoral democracy has become an inadequate tool for policy making. There are high levels of distrust towards and disaffection from the political system. There are numerous reasons for this. Firstly, globalisation, among other factors, has resulted in an increasing discrepancy between territorial constituencies and issue constituencies. In other words, the territorial boundaries that determine which issues a citizen has the right to vote on do not adequately overlap with the issues that a citizen is concerned with or affected by. Thus, the ‘all affected’ principle — “the basic norm of democracy that holds that those potentially affected by a collective decision should have some influence over the decision” (Warren, 2009, p. 9) — is not always adhered to. Secondly, electoral systems fail to accurately reflect issue-specific citizen preferences. Voters tend to vote for representatives based on their stance on a few specific issues that they find particularly important. Hence, their views on certain ‘secondary’ matters — such as those on sex work — are not effectively reflected by their vote. Thirdly, society is becoming increasingly pluralised, resulting in dense and increasingly competent civil societies. Electoral institutions’ capacities are challenged by civil societies’ capacities for organisation and advocacy (Warren, 2009). Fourthly, citizens are becoming less deferential to authority while their democratic instincts are increasing, as illustrated by citizen’s personal involvement in

(19)

(political) causes (Warren, 2009). _{Finally, Warren (2009) explains, organised groups nowadays often} possess expert capacities, making them capable of contesting government expertise. These societal phenomena are severely undermining the legitimacy of our democratic institutions. While one cannot speak of a general system crisis (yet), these challenges emerge “issue by issue and policy by policy” (Warren, 2009, p. 7). Warren (2009) refers to this state as one of ‘pluralised ungovernability’.

3.1.2 Governance-driven democratisation._{Government officials are generally the first ones to} experience the consequences of pluralised ungovernability (Warren, 2009). Innes and Booher (2010) explain that

When government policies fail to solve problems, the typical reaction is to try to fix the policy or to tinker at the edges of the system. Very seldom do leaders of the public question the institutions that have failed, nor do they often ask whether different kinds of practices and structures could be more effective, much less look for ways to transform the existing model of government. (p. 8)

A paradigm shift that would transform the existing model of government so fundamentally is not imaginable without a radical, transformative and internalised movement challenging the status quo. Rather, _{Warren (2009) observes that government officials are increasingly responding to pluralised} ungovernability by establishing processes of public input or community representation during the policy process. He refers to this rebirth of democratic ideals in the field of policy and administration as ‘governance-driven democratisation’. It can be conceived of as a supplement to electoral democracy where the latter experiences functional weaknesses.

One of the main differences between governance-driven democratisation and electoral democracy is that the latter relies on purely territorial constituencies, whereas the policy-focused nature of the former allows it to generate dynamic, fluid, issue-focused constituencies. This reveals one of the main opportunities of governance-driven democratisation: its capacity to bring democratic processes closer to the aforementioned all-affected principle (Warren, 2009). Governance-driven democratisation offers a number of other opportunities besides contributing to the realisation of the all-affected principle: it has the capacity to (1) empower participants specifically and directly, (2) design representation in a way that includes marginalised people and unorganised interests, (3) provide specific kinds of deliberative opportunities that exist neither in the electoral nor in the public sphere, and (4) connect state power and civil society (Warren, 2009).

(20)

3.1.3 Pluralised ungovernability in Amsterdam’s sex work context. _{The situation surrounding} sex work in Amsterdam can be conceived of as one of pluralised ungovernability. There is conflict between (as well as within) different groups of stakeholders, a lack of coordination between different agencies, distrust between actors (especially towards the government), anger experienced by both sex workers and brothel operators resulting from past grievances and perceived injustices and misunderstanding between different actors. The media is furthermore constantly seeking to capitalise on instances of exploitation or other sensational stories from within the sector, and the public is highly involved and tends to have strong (though often uninformed) ideological opinions about sex work, as tends to be the case with morality policy (Wagenaar et al., 2017).

Some of the opportunities provided by governance-driven democratisation may be particularly valuable in the context of Amsterdam’s sex work policy. For example, the all-affected principle is currently far from adhered to in this context, mainly because the large proportion of immigrant sex workers in the city are able to voice their opinions through standard democratic processes to a very limited extent only. Additionally, the fact that governance-driven democratisation provides the opportunity to include marginalised people and unorganised interests is important in this context, as the main groups affected by these policies are marginalised groups — first and foremostly sex workers, but also brothel operators. Finally, the civil society surrounding sex work is highly developed, with numerous non-governmental organisations (NGOs) fighting for sex worker rights and doing research into the sex industry. These civil society groups deserve a seat at the table in the policy process, and may be able to provide valuable insights into the reality of the sex industry, having worked closely in collaboration with sex workers and being aware of the community’s competences, needs, and struggles.

3.2 Collaborative Governance in Response to Complex Policy Problems

3.2.1 The emergence of collaborative approaches to policy making. _{In Planning with}

Complexity, Innes and Booher (2010) analyse collaborative practices that are emerging in the context of planning and policy making. Like Warren, they observe that traditional, linear modes of policy making are making way for “nonlinear socially constructed processes engaging both experts and stakeholders” (p. 5). In such processes, diverse groups of actors, including government officials, stakeholders, experts, and the public, collectively attempt to address policy issues. Such processes can be understood as expressions of governance-driven democratisation. The background against which these processes occur is not only one of pluralised ungovernability, but also one in which notions about what constitutes relevant knowledge within policy making are being redefined; scientific knowledge is increasingly supplemented by local knowledge. Innes and Booher (2010) explain that

(21)

lay and local knowledge is at least as critical as science in the making of sound planning and policy decisions. Local knowledge is more pervasive and more persuasive than scientific knowledge because it offers the detailed, situated knowledge that is crucial to effective action (p. 21)

3.2.2 Defining collaborative governance._{Collaborative governance is a specific mode of collaborative} policy making. Ansell and Gash (2007) define collaborative governance as

a governing arrangement where one or more public agencies directly engage non-state stakeholders in a collective decision-making process that is formal, consensus-oriented, and deliberative and that aims to make or implement public policy or manage public program[me]s or assets. (p. 544)

This definition reveals six characteristics of collaborative governance. First, the process is typically initiated by public agencies or institutions. Second, non-state actors participate in the process, either as individuals or as organised groups. Third, the process is never merely consultative; stakeholders must be directly engaged in decision making. This does not preclude public agencies from having ultimate authority, but merely that the decision-making outcomes closely resemble, or absorb, stakeholder’s advice. Fourth, the process is collective in the sense that stakeholders formally meet together in a deliberative process, and formal in the sense that there must be an explicit public strategy of organising stakeholders’ influence in a structured manner. Fifth, deliberation is aimed at consensus. This does not imply that consensus will actually be reached, but merely stipulates an aspiration of the process. Finally, the collaborative governance process focuses on the governance of public policies and affairs which distinguishes such processes from other consensus-based decision making procedures, like alternative dispute resolution (Ansell & Gash, 2007).

Ansell and Gash (2007) explain that collaborative governance is fundamentally distinct from adversarial and managerial modes of policy making. Whereas adversarial decision making leads to a winner-take-all situation, the potentially competitive relationships between stakeholders are meant to transform into more cooperative ones through a collaborative governance process. In managerial modes of policy making, “public agencies make decisions unilaterally or through closed decision processes, typically relying on agency experts to make decisions” (Ansell & Gash, 2007, p. 547). On the other hand, stakeholders are directly included in decision-making in a collaborative governance process.

(22)

3.2.3 When to initiate collaborative governance. _{Collaborative governance is a mode of} governance that is often employed where traditional approaches to the policy process have failed persistently (Ansell & Gash, 2007). Innes and Booher (2010) explain that setting up a collaborative governance initiative “is not needed when there is already agreement about ends and means, when cause and effect relationships are well understood, and where there is relative certainty about how the decision will play out in the system” (p. 7). It is best applied to wicked policy problems where no single solution can be shown to be optimal and it is impossible to definitively establish causality (Innes & Booher, 2010). Thus, it fits perfectly with cases that exhibit signs of pluralised ungovernability. In such cases, it may be fruitful to have diverse stakeholders engage in dialogue in order to be able to build on a wide variety of (local) knowledge, and hence a wide range of potential approaches to explore in order to improve the current situation.

3.2.4 The value of collaborative governance. _{One of the main values of collaborative} governance is its capacity to produce desirable policy outcomes. Its incorporation of local knowledge — something that tends to be neglected by traditional modes of governance — is central to this. Local knowledge is particularly valuable when deliberating complex policy problems because it “fills gaps, provides information about context, and offers pragmatic, experience-based insights from these who know a situation firsthand” (Innes & Booher 2010, p. 170). This is imperative for identifying realistic approaches that have the potential to produce desirable outcomes, as well as filtering out options that are unrealistic or may have problematic consequences unforeseen by those who lack firsthand knowledge (Innes & Booher, 2010). Thus, local knowledge of stakeholders functions as a feedback loop in the policy process. Simultaneously, the inclusion of “lay voices, especially of marginalised people who seldom have an impact on the decisions that affect their lives, is a matter of justice and authentic democracy” (Innes & Booher, 2010, p. 170).

The value of collaborative governance transcends mere policy implications; the process itself may also benefit participants. First, stakeholders may come to recognise the reciprocal nature of their interests. Once they conceive of their own interests in terms of mutual gains rather than zero-sum games, they may find that their own interests cannot be achieved without the interests of other parties being achieved. Subsequently, they can start to explore approaches that satisfy multiple interests and result in mutual gains. Second, stakeholders may come to reevaluate and reframe their interests, allowing them to discover means to achieve their interests that would previously have been unimaginable. Hence, it stimulates more creative solutions. Third, jointly engaging in a collaborative process allows the stakeholders to develop relationships that often continue after the process has ended. Where stakeholders relationships may have been defined by conflict prior to the process, engaging in what Innes and Booher (2010) call ‘authentic

(23)

dialogue’ may change the nature of these relationships. As parties are exposed to each others’ perspectives and experiences, they may develop more empathy and understanding towards one another.

Through these developments, collaborative processes may contribute to the adaptivity and resilience of the community in question resulting from individual and collective learning. More specifically, such processes can

build capacity for self-management in communities, improve policy knowledge, and creative innovative strategies tailor-made to the unique conditions of particular situations. They can transform intractable problems into tractable ones as participants come to see in new ways and develop new values and goals and a sense of common purpose. They build social, political and intellectual capital and can be used not only for the immediate problem, but for much more over time. (Innes & Booher, 2010, pp. 7-8)

Finally, for participants in a collaborative process — especially typically marginalised or invisible groups — having a seat at the table can be empowering. They tend to develop new skills and build new networks that may be useful in the future, perhaps allowing them to do things that they never considered before. In fact, Innes and Booher (2010) have found that “many of the most valued outcomes of collaborations were of the systemic variety — that is, they not only produced specific outcomes, but they also affected the system by changing attitudes, relationships, and capacities of players” (p. 35).

3.2.5 A model for collaborative governance. _{Based on a meta-analysis of existing collaborative} governance literature, Ansell and Gash (2007) set out a general model of collaborative governance. The model distinguishes between four categories of variables: starting conditions, facilitative leadership, institutional design, and collaborative process. The model is set out below.

(24)

Figure 2. A Model of Collaborative Governance (Ansell & Gash, 2007, p. 550)

3.2.5.1 Starting conditions. _{Ansell and Gash (2007) identify three critical variables present at the outset} of the process that influence cooperation: resource and/or power imbalances, underlying incentives, and a history of cooperation of conflict. First, potential resource and/or power imbalances between stakeholders may stand in the way of wise decision making. It is worth noting here that power is an elusive concept that holds different meanings to different people and across different contexts; it is an intersubjective social construct (Innes & Booher, 2010). Nevertheless, (perceived) power imbalances may generate distrust or a lack of commitment. Some actors may lack the capacity, organisation, status, or resources that other actors do possess, leaving the stronger actors in a position from which they are able to manipulate the process towards their own means. Especially where stakeholders are not organised, and hence not (well) represented in the collaborative governance process, this is a risk. Within a collaborative governance process, power should not be perceived as a zero-sum game where one party’s gain results in another’s loss. In fact, a collaborative process may create what Innes and Booher (2010) refer to as ‘network power’ — the mutual empowerment of participants resulting from understanding each others’ perspectives and the development common heuristics and shared purposes (Innes & Booher, 2010).

(25)

Second, the underlying incentives of stakeholders to participate in the collaborative governance process must be understood. These incentives largely depend on cost-benefit analyses of time and energy invested versus expected outcomes. Stakeholders’ incentives to participate increase where they perceive a direct relationship between their participation and policy outcomes that they deem desirable, and decrease where they understand their position as merely advisory or symbolic. Incentives further depend on perceived interdependence between stakeholders. High (perceived) interdependence increases the likelihood of a successful collaborative governance process (Ansell & Gash, 2007).

Finally, a history of conflict or cooperation influences collaboration. Intuitively, a history of cooperation is promising for collaboration. However, high levels of conflict may also stimulate collaboration as long as levels of interdependence are deemed to be high as well. Still, a prehistory of conflict may result in low levels of trust among participants. Where this is the case, steps must be taken towards reconciliation, possibly in the form of mediation (Ansell & Gash, 2007).

3.2.5.2 Facilitative leadership. Leadership is considered to be a crucial component in the collaborative governance process. It may take the form of facilitation, mediation, or nonbinding arbitration. Leadership entails setting ground rules, building trust, facilitating dialogue, and exploring solutions. Good leadership has the potential to remedy unfavourable starting conditions, for example by using empowerment strategies to create a more level playing field, increasing participants’ incentives to participate by reframing the situation and stressing interdependence, and by reconciling conflict by means of mediation (Ansell & Gash, 2007). Finally, the facilitator furthermore serves the function of guarding information (Innes & Booher, 2010).

3.2.5.3 Institutional design. Institutional design is understood as “the basic protocols and ground rules for collaboration, which are critical for the procedural legitimacy of the collaborative process” (Ansell & Gash, 2007, p. 555). One of the most fundamental questions with regard to institutional design is which stakeholders should be included in the process. A basic condition for successful collaboration is that it must rely on the aforementioned ‘all-affected’ principle — namely that all stakeholders who are affected by the outcomes of the process must be included. The spectrum of stakeholders must be sufficiently broad to represent the actual context. This is crucial for the legitimacy of the process and its outcomes. Innes and Booher (2010) further explain that

contrarian and disadvantaged stakeholders are necessary to help achieve robust agreements that break open the unacceptable status quo that brought people to the table in the first place. These individuals and interests see the world from a different perspective than others and ask the

(26)

In fact, the authors argue that conflict and tension between stakeholders is essential in the context of collaborative governance, as agonism enables them to move beyond the status quo, which is necessary in order to reach creative solutions (Innes & Booher, 2010). Other important design features include clear ground rules and transparency of the process, which can be accomplished through the formalisation of governance structures.

3.2.5.4 The collaborative process. Successful collaboration appears to depend on a number of interconnected factors. Rather than a linear process, Ansell and Gash (2007) visualise the collaborative governance process as a virtuous cycle; all factors are inter-connected. One of the factors within this virtuous cycle is communication. Effective communication requires stakeholders to engage in face-to-face dialogue (Ansell & Gash, 2007). Effective dialogue requires that all participants are fully and equally informed, and that everyone has the opportunity to express their views and be listened to. All major interests and knowledge must come forward during deliberations. Actors must engage in such a way that the accuracy and legitimacy of claims is assured and that these are comprehensible to all participants. Accuracy can be tested by expert knowledge; comprehensibility is something that needs to be assured by the facilitator (Innes & Booher, 2010). Besides providing stakeholders with the possibility to identify potential solutions, face-to-face dialogue is crucial for breaking down barriers to communication such as negative stereotypes and antagonism, and for establishing trust, mutual respect, and shared understanding (Ansell & Gash, 2007). It differs from negotiation in that dialogue is aimed at mutual understanding. Rather than debate or argument, dialogue is a conversation in which ideas are exchanged.

Building trust is a second critical component of the process. Especially where levels of trust are low from the outset, this may be a time-consuming process (Ansell & Gash, 2007). Stakeholders’ commitment to the process is a third central variable for explaining success or failure of collaborative governance processes. It requires willingness of the stakeholders to accept the results of the process, no matter the outcome. This, in turn, requires trust. Commitment is further enhanced by perceived ownership of the process, as well as perceived interdependence (Ansell & Gash, 2007). Fourth, the collaborative process needs to encompass a learning process in which stakeholders develop shared knowledge, as well as a shared understanding of what they can collectively achieve (Ansell & Gash, 2010). All participants bring their own experiences and theories to the table. Learning from experience and engaging in activities such as storytelling are fundamental for the construction of (relevant) knowledge. Together, participants subsequently seek and contextualise facts in a manner that is aimed at problem solving (Innes & Booher, 2010). Innes and Booher (2010) explain that “[o]nly by uncovering what is hidden under socially constructed understandings can there be any hope of seeing past the disempowering language and expectations of society and getting out of impasses” (p. 93). Dialogue thus helps participants escape the

(27)

trap constituted by their own thoughts, helping them to reflect upon the frames that they hold and recognise that their perspective is merely that — _{their perspective (Innes & Booher, 2010). Finally,} intermediate outcomes influence collaboration; where small gains are possible, this encourages the productive collaboration cycle. This is especially important where prior antagonism is high and commitment to long-term trust building is necessary.

3.2.6 The need for collaborative governance Amsterdam’s sex work context._{Chapter 1 set} out three aspects of sex work policy that inhibit the policy process: it being the epitome of morality policy, the lack of reliable data about the sector, and stigmatising perceptions on sex work and sex workers. Initiating a collaborative governance process in this context may answer to these challenges to some degree. First, incorporating sex workers’ and other stakeholders’ local knowledge contributes towards a more accurate and complete picture of the sex industry, thereby partly mitigating the lack of data. Second, engaging with sex workers (and brothel operators) may alter stigmatising perceptions held be certain other participants. This leaves the issue of sex work policy revolving around morality. Perhaps a collaborative process may open the eyes of those involved to the need for pragmatism instead of morality. However, as mentioned before, pragmatism in this context is an ideological stance of itself.

3.3 Representation within the Collaborative Process

Above, I explain that the potential success of collaborative governance is in part dependent on the organisation of interests. Mansbridge (2000) explains that for marginalised groups, it is especially important to be represented by people belonging to that group (‘descriptive representation’) for three reasons. First, she explains that

members of groups embedded in a tradition of domination and subordination often experience faulty communication: the dominant group has not learned to listen, and the subordinate group has learned to distrust. Members of subordinate groups may thus often find that effective substantive representation depends on their being represented [...] by members of their own group with whom they can communicate easily and by whom they can reasonably expect to be better understood. (Mansbridge, 2000, p. 99)

Second, marginalised groups have to put new, ‘uncrystallised’ issues on the political agenda. Representatives who have lived the experience of the marginalised group and consequently have understanding of a deliberative issue through personal experience will be better able to contribute to

(28)

salience of the issue (Mansbridge, 2000). Third, the inclusion of members of the marginalised group in the deliberative process have the ability to change the social meaning and implications of membership to the group in question (Mansbridge, 2000). In the context of sex work, this means that sex workers’ presence in collaborative decision-making processes could reduce the stigma surrounding the occupation, and the prevalent image of sex workers as agenciless victims.

3.3.1 Sex worker interest organisations. Across the globe, sex workers organise themselves through labour unions or interest organisations. In the Netherlands, sex workers are organised under PROUD since 2015, a union with approximately four hundred members. As a result of the stigma resting on the occupation and the marginalisation of sex workers, sex worker interest organisations and unions differ from those of other occupational groups. The latter usually merely focus on protecting workers’ rights and campaigning for safe working conditions and other matters that contribute to safeguarding workers’ well being. For sex worker organisations to be able to improve the working conditions and wellbeing of sex workers, on the other hand, they need to address the deeply rooted and widespread societal stigma towards sex work and sex workers. Consequently, sex worker unions tend to take the _de facto shape of advocacy organisations; social protest and pressure group tactics being the main tools of choice (Wagenaar et al., 2017).

Mere unionisation, Gall (2014) argues, is insufficient in this context for three reasons. First, improvement of the working environment in this context is not as much dependent on the employer (in this case, the brothel operator) as it is dependent on national and local policies and regulations. Second, most sex workers are self-employed. This fact, in interaction with the fact that a significant portion of the sex worker population is ‘hidden’, makes unionisation much more difficult. Third, many sex workers are recent immigrants who often do not master the language of the country they work in, and are not deeply embedded in local (sex worker) communities (Wagenaar et al., 2017).

3.4 Summary

In this chapter I have laid out the theoretical background of this research. I have explained the main theory to be applied in this research — collaborative governance — as a form of governance-driven democratisation. Governance-driven democratisation entails the rebirth of democratic ideals in the form of processes of public input or community representation in response to the pluralised ungovernability that results in the inadequacy of electoral democracy as a tool for policy making. Collaborative governance is a form of governance in which public agencies directly engage non-state stakeholders in public policy processes. In sum, Ansell and Gash’s model of collaborative governance stipulates that the potential for an effective collaborative governance process depends upon its institutional design, starting

(29)

conditions, facilitative leadership, and the collaborative process. Institutional design revolves around questions such as which stakeholders to include. Here, the all-affected principle should be adhered to. In accordance with Mansbridge’s concept of descriptive representation, it is especially important for marginalised groups to be represented by someone belonging to that group within the collaborative process. Starting conditions include potential power/resource imbalances or a prehistory of conflict or cooperation, both of which influence parties’ incentives to participate. Finally, the collaborative process includes the following interrelated factors: face-to-face dialogue, trust building, commitment to the process, shared knowledge, and intermediate outcomes. These five factors all influence one another; the collaborative process variables should be understood as iterative or cyclical rather than linear. Not all above-mentioned factors need to be fulfilled; where this is not the case, effective facilitative leadership could still ensure a fruitful collaborative governance process.

(30)

Chapter 4.

Sex Work Policies

This chapter provides the reader with background information on different sex work policies. Section 4.1 discusses three different legislative approaches towards sex work worldwide and the underlying rationales and possible outcomes of these approaches. Section 4.2 provides an overview of sex work policy in the Netherlands, and describes how sex work policy came to be repressive in the Netherlands resulting from a human trafficking frame that the industry is perceived through.

4.1 Three Legislative Approaches Towards Sex Work

Broadly speaking, there are three legislative approaches to sex work: criminalisation, legalisation, and decriminalisation. However, Östergren (2017) points out that it is not entirely accurate to speak of policy regimes as such. Sex work is generally regulated at a regional rather than a national level. Consequently, laws and regulations concerning the sex industry can (and often do) vary within a country to a contradictory extent. The way states deal with sex work is furthermore not consistent over time, either (Östergren, 2017). Therefore, Östergren (2017) suggests that rather than in legislative terms, sex work policy should be categorised in terms of its purpose: to repress, restrict, or integrate. Below, I do discuss legislative approaches, but relate those to Östergren’s categories.

4.1.1 Criminalisation. _{In most states, the commercial sex industry is criminalised, although to} differing degrees. Criminalisation can either take the form of prohibition or neo-abolition. First, under a prohibitionist model, all aspects of sex work are criminalised; both the sex workers and the client are considered criminals. The United States, Russia are among the states that adhere to such a model, as well as many Asian and African states. Second, the neo-abolitionist model — also known as the Swedish or Nordic model — criminalises the purchasing of sex, but not selling of sex. Thus, the client of a sex worker commits a criminal offense when buying sexual services, but the sex worker providing the services does not. Prohibition and neo-abolition can be conceived of as repressive models in Östergren’s terms in the sense that both aim to eradicate the commercial sex industry as a whole. The perceived need for the eradication of prostitution emanates from the stigma. The prohibitionist model appears to stem from morally-grounded stigma, whereas the neo-abolitionist model seems to flow from the conception of sex work as inherently exploitative. Criminalisation policies fail to acknowledge sex workers’ agency, as well as the social reality that economically, sex work is simply the most attractive alternative for many people — in particular women — worldwide (Vanwesenbeeck, 2017).

Deliberating sex work : exploring the challenges and opportunities pertaining to setting up a collaborative governance process for Amsterdam's sex work policy