Jump to ContentJump to Main Navigation
Designing Prostitution PolicyIntention and Reality in Regulating the Sex Trade$

Hendrik Wagenaar, Helga Amesberger, and Sietske Altink

Print publication date: 2017

Print ISBN-13: 9781447324249

Published to Policy Press Scholarship Online: September 2017

DOI: 10.1332/policypress/9781447324249.001.0001

Show Summary Details

(p.265) Appendix: research design and methodology

(p.265) Appendix: research design and methodology

Source:
Designing Prostitution Policy
Author(s):

Hendrik Wagenaar

Helga Amesberger

Sietske Altink

Publisher:
Policy Press

Goals of the project

The research that forms the basis of this book was a collaboration between Platform 31 (a knowledge and network organisation in the area of urban policy), the Dutch cities of Rotterdam, Utrecht and The Hague, the Austrian city of Vienna, and the University of Leiden in the Netherlands. The Dutch partners contributed one third of the budget each; Vienna contributed a smaller amount. The research was designed to compare the implementation, outcomes and effects of prostitution policy in two countries, Austria and the Netherlands, that have a discernible and more or less coherent approach to regulating prostitution. The focus on policy implementation issues from the fact that in prostitution policy, as in most policies, a considerable gap exists between policy formulation and its actual implementation in the field, and that it is the actual delivery of public policy that determines the impact on the target group and on society at large. The main goal of this study is therefore to provide a detailed, in-depth analysis of the implementation of prostitution policy in the countries of study. A second goal of the study is to bring policy theory to bear on our data about prostitution policy, for the purpose of obtaining a better understanding of how such policies are designed and implemented.

Research questions

The project focused on the following research questions:

  1. 1. What is the size and state of the prostitution market in the Netherlands and Austria?

  2. 2. How is prostitution regulated in both countries at the implementation level? Which policy instruments are used? Which rationale is given for the use of different policy instruments? Are the policy instruments used in an effective and efficient manner? Which conditions have to be fulfilled (organisationally, financially, and from the perspective of political-administrative relationships) for the effective and efficient use of various instruments? Is the use of the mix of instruments consistent, effective and efficient?

  3. (p.266) 3. What are the effects of the mix of instruments in the prostitution field? Effects are based on the stated goals of the policy. We would also look for general effects such as public order, trafficking and illegal prostitution, labour rights, work conditions and human rights of sex workers, and unintended consequences such as transfer of prostitution to different areas (national or international) or impacts on the social position of sex workers.

  4. 4. What unintended consequences does the policy have? Here we think of increases or decreases in the number of sex workers or sex facilities, transfer effects, the disappearance of prostitution into an invisible underworld, or an increased risk for sex workers.

We want to emphasise that the participating cities were interested in the possibilities of effectively regulating prostitution. That implied that they were less interested – at the time at least (see Chapter Seven) – in a study of trafficking.151 Although trafficking was not the subject of the project, if it emerged in the course of the research as a topic that respondents were dealing with (for example, local policies that were designed to recognise and fight trafficking), we of course analysed it and integrated it in our conclusions.

Design of the study

The research was not designed as a formal outcome study with a controlled pre-test and post-test design. The complexity of the prostitution domain, in combination with the rapid developments in this field, prohibited, in our opinion, a formal, quasi-experimental design (Shadish et al, 2001). Such a design would provide us with a mere semblance of scientific control while the rigour of the design would result in missing out on most of the dynamic complexity that characterises this field. Moreover, authoritative knowledge about actual approaches to the regulation of prostitution is so scarce, that a mixed-methods, process-based research design seemed to be indicated. For this reason, we opted for a comparative case study design (Landman, 2000). A mixed-methods, comparative case study allowed us to map regulatory approaches, the societal and administrative context in which regulation operates, and the impact of regulation on the prostitution field. Policies are put into effect to remedy or alleviate a particular (p.267) problematic situation. By studying the prostitution field in depth we aimed to obtain both an understanding of the societal origins and validity of the original problem formulation, the administrative context in which policy and regulation is designed and implemented, and the impact, or lack thereof, of policy measures that have been designed to address the problem. Policy effects are the intended and unintended consequences that follow from the implementation of a range of policy instruments with the goal of solving problems and/or attaining aspirations in the field of prostitution.

The main comparison was between the Netherlands and Austria. The participating Dutch cities (see below) expressed the desire for a most similar systems design (MSSD; Landman, 2000). The scientific rationale of a MSSD is that, when most independent variables are held constant, differences in the dependent variable are explained by those independent variables that do differ (Landman, 2000, p 71). An MSSD, as well as its counterpart, a most different systems design (MDSD), are strong designs that yield important comparative knowledge. One must distinguish between large n and small n comparative studies. Clearly this is a small n study: a comparison of two countries. Such a comparison does not allow the researcher to hold the independent variables constant, simply because there are too many of them; this is the problem of ‘too few cases/too many variables’ problem that dogs so much social science research (Goggin, 1986). The value of small n comparative studies must be sought in the rich contextual comparison. Or as Landman puts it: ‘The comparison to the researcher’s own country is either implicit or explicit, and the goal of contextual description is either more knowledge about the nation studied, about one’s own political system, or both’ (Landman, 2000, p 5).152

The cities’ argument for an MSSD was that a comparison with a most different policy regime (in particular, one of client criminalisation, as in Sweden) would generate important information, but from a practical point of view it would be more useful to see what the policy implementation and effects are of a country with a similar regime. The cities thought it unlikely that the Netherlands would move towards a (p.268) client criminalisation policy any time soon. On the other hand, we know so little of the effects of regulation (and decriminalisation) that it is expected that an in-depth comparison of such policy regimes would yield more interesting and relevant knowledge. Austria was selected for various reasons. Similar to the Netherlands prostitution is legal in Austria and in most provinces brothels are a more or less accepted part of society. Austria regulates prostitution. It has also experimented with decriminalisation-type policy measures, such as improving the working conditions of sex workers, sex workers’ taxation and their eligibility for social security. Interesting to the Netherlands is that Austria has a long experience with the registration of sex workers. Similar to the Netherlands, Austria has a large number of migrant sex workers, mostly from eastern Europe. In an administrative sense the countries also show similarities. Each can be called a decentralised unitary state, although the Netherlands does not have a federalised structure such as Austria, and the Dutch cities are less autonomous than the Austrian Länder. Prostitution policy therefore shows more regional variation in Austria than in the Netherlands. Politically, both countries are governed by coalition governments in which, until recently at least, Christian Democrats and Social Democrats are the dominant forces.153 Both are relatively small and affluent; they do not have the extremes of poverty and degradation that are among the drivers of prostitution in other countries. Finally, there were various pragmatic reasons behind the choice of countries, such as language, available contacts and accessibility.

Partnered research

A major funding condition of Platform 31 was that the research team and the participating cities formed a consortium. As a result the research has been developed and executed in close collaboration with the three Dutch partner cities Rotterdam, The Hague and Utrecht; Platform 31 participated in an observing role. Soon after the start of the project the city of Vienna also joined the consortium. The three Dutch partner cities indicated that they would like to obtain policy-relevant information early on in the research and not at the end of the study. The study was therefore practice-oriented and designed and executed in close interaction between the research team and the partner cities. What did this co-production of research entail? As there are no blueprints or handbooks available that instruct researchers and partners (p.269) how to engage in co-producing usable knowledge, the interactive design was developed ‘in vitro’ in the course of the research. In 2009 the research questions and research design were formulated in a series of meetings with the Dutch cities. At the meetings the participants jointly formulated questions that were practically relevant (that is, emerging from and speaking to practical concerns of policy makers) and scientifically feasible (meaning they can be researched in a scientifically adequate manner). In hindsight this resulted in a process of mutual learning and trust building. With regard to learning, the researchers obtained unique, ‘hands-on’ insight into the challenges and particulars of designing and implementing measures for regulating prostitution at the local level (shaped by national legislation and extraneous developments). For the representatives of the cities it meant a growing awareness of the possibilities and limitations of scientific research. The cities, aware of the lack of knowledge about the regulation of prostitution, were above all interested in more precise and reliable data about the size and nature of the sex industry, in learning about policy theory, and in general in obtaining a knowledge base for effective administrative action in regulating prostitution.

From the start of the project the research team and the Dutch partner cities met every three months. At the meetings a programme manager from Platform 31 was always present. A scientific advisory board was created to guarantee the scientific quality of the research. The board agreed to review all reports that were to be produced by the research team. The board consisted of Professor Joyce Outshoorn, emeritus professor of women’s studies at Leiden University, Professor Ronald Weitzer, professor of sociology at George Washington University, and Dr Anton van Wijk, senior researcher at Beke, a research agency that has done extensive, policy-oriented, research on prostitution in Amsterdam.

Data collection

A key feature of the project is the way that data were collected. Data collection in prostitution research faces numerous threats to validity and is affected by numerous practical obstacles. A large proportion of sex workers are very mobile. Many migrant sex workers do not speak English. Because of the stigma attached to prostitution, sex workers protect their anonymity and are reluctant to be interviewed. Sex workers distrust government officials, as they expect nothing good to come from them. They also distrust researchers, as they believe that nothing is done with the results anyway. Given these challenges, data (p.270) collection in prostitution research will always be a compromise between ambition and convenience, and results and conclusions of any study will be surrounded by uncertainty and problems of interpretation. In fact, only a series of carefully designed and executed studies comprising different policy contexts will gradually converge in more robust results and conclusions.

In this project the emphasis has thus been as much as possible on original data collection by field workers who are familiar with the local situation. Our insiders were outreach workers from local NGOs in Austria (some of them former sex workers), one of the authors (Sietske Altink, former worker of the Rode Draad) and an active sex worker in the Netherlands. The interviewers received training in qualitative interviewing. The advantages of using insiders over academic field workers are obvious. Insiders find it easier to gain trust and are in a better position to interpret the answers. In addition, insiders are in a unique position to obtain the serendipitous insights that come with moving about in the field. A risk of using insiders is that they share a context with their respondents and/or identify with them. In both cases this can compromise the quality of interview and observation data (Weiss, 1995). In general local researchers conducted the research in the two countries: Helga Amesberger in Austria and Hendrik Wagenaar and Sietske Altink in the Netherlands. Not only will local researchers have better access to national and local data sources, but they are also in a better position to situate what the respondents say in its proper political-cultural-historical context.154

Sampling turned out to be strongly dependent on the specific situation in the respective countries. Since 2000 the Dutch prostitution market has been the subject of a considerable number of research studies (Daalder, 2002, 2007 and Van Wijk et al, 2010 are the most well known, but numerous smaller studies have been conducted by local commercial research agencies). National and international researchers and journalists regularly approach Dutch field organisations for cooperation in gaining access to sex workers. Also, licensed sex facilities are subject to frequent inspections from various kinds of officials. In general the Dutch prostitution field is much more interconnected with (p.271) official organisations than in most countries. Outreach work is largely done by municipal health and social work organisations that have long-established, stable trust relationships with the prostitution field in their city. Professionals often feel that they put this trust relationship at risk by mediating for research access. In a relatively small market where many actors know each other personally, this has resulted in a certain research ‘fatigue’, a reluctance with field workers to mediate among sex workers to collaborate with researchers. In those cases where we did get access to sex workers we discovered that the interviews were not without risk to the sex worker. In at least one case the interviewee lost her job in the brothel where she worked after being interviewed by us. Although we could not establish a clear causal link, it nevertheless proved to be demoralising to our insider interviewer who felt guilty about this. The Austrian prostitution market on the other hand has been much less researched. In addition, outreach work is largely done by independent NGOs who have to go out and contact sex workers for purposes of providing information. Officials have a more arm’s-length relationship with the field than in the Netherlands. Sex workers and outreach workers suffer less from research ‘fatigue’; in fact we had the impression that our fieldwork was one of the first instances that Austrian sex work had been researched ‘on the ground’. We found two major outreach organisations (LEFÖ and maiz) willing to conduct interviews with sex workers. For most interviews specific appointments were made; a few interviews were done in the course of outreach work.

The different situations in Austria and the Netherlands have led us to use different sampling and data-collection strategies, each with their own strengths and weaknesses. In the Netherlands the emphasis was on the use of insiders for recruiting respondents. In addition to interviewing, the two Dutch field workers engaged in extensive observations of management practices in sex facilities and of management–sex worker interactions. The two field workers hung out in clubs where possible and had numerous conversations – although short of a formal interview – with hostesses, bar personnel and clients. From this participant observation we learned a lot from actors’ off-hand comments. The Dutch sex worker insider also kept a diary.

In the Netherlands we interviewed 44 sex workers, and six to eight administrators, professionals and police officials for every city, covering administration, public health and the police. Three of the sex workers had participated in the so-called B9 arrangement for victims of trafficking (a national referral system) and were therefore officially designated ‘victims of trafficking’. Interviews were held in Dutch or English. Two of the authors (Sietske Altink and Hendrik Wagenaar) (p.272) interviewed officials, administrators and police officers. In an earlier study (Wagenaar, 2006) Hendrik Wagenaar interviewed around 25 facility owners and over 50 municipal officials; we have drawn on those data in this study. In Austria we interviewed 85 sex workers (59 in Vienna, 26 in Oberösterreich), three facility owners, and 29 administrators, professionals (NGOs), politicians and police officers. Outreach workers from two NGOs, LEFÖ in Vienna and maiz in Oberösterreich, conducted the interviews with sex workers; our Austrian researcher (Helga Amesberger) did four additional interviews with Austrian sex workers. The interviews covered topics ranging from the pathway into prostitution to earnings and working hours. The Austrian interviewers were fluent in the languages the sex workers are speaking (Bulgarian, Romanian, Hungarian, Czech/Slovakian, Polish, Spanish, Russian, English and German). In both countries we paid the sex worker a fee of €50 for the interview. Table 1 gives an overview of the characteristics of the sex workers interviewed in both countries.

Table 1: Austrian and Dutch sex worker interviewees: main characteristics

Austria (n=82)

The Netherlands (n=44)

Gender

Female

80

38

Male

0

4

Transgender

2

2

Nationality

Austria

8

0

Belarus

1

0

Bulgaria

15

3

Czech Republic

5

0

Hungary

11

1

Italy

1

0

Israel

0

1

Latvia

2

0

Moldavia

0

1

Nigeria

4

0

Poland

1

0

Portugal

0

1

Romania

15

3

Russia

1

0

Slovakia

6

0

Spain

1

0

The Netherlands

1

31

Ukraine

3

1

Meso-and South

7

2

America (Venezuela,

Cuba, Dominican

Republic, Chile, Jamaica)

(p.273) Work venue

Brothel, (night)club

29

5

Studio, private club

11

11

Escort

2

1

Streetwalking

9

1

Window, Laufhaus

22

4

Massage parlour, sauna

4

10

From their own home

0

5

Mixture (more than one work venue)

6

0

Others (hotel prostitution, illegal brothel/bar)

2

7

Age at the time of interview

18-20

0

2

21-29

39

16

30-39 40-49

25 14

12 11

50+

4

3

Years in sex work

<1 year

8

5

1-5 years

34

9

6-10 years

27

10

11-15 years

7

4

16-20 years

3

4

>20 years

1

7

Unknown

2

5

Age at the time of commencing sex work

15-17

4

0

18-19

10

7

20-24

27

12

25-29

12

11

30-39

18

4

40-49

3

0

>50

3

0

Unknown

5

10

Interviewers in both countries were equipped with the same interview guidelines. The interviews with sex workers took place at the workplace of the interviewee, in cafés or at the offices of the NGOs. One guiding principle for conducting the interviews was not to disturb the sex worker’s business. That is, if the sex worker solicited a client during the interview, the interview was terminated. This happened in a few cases. The interviewers tried to speak alone with the sex worker, but sometimes colleagues and friends were present who also (p.274) provided information.155 The interviewers took care that the manager or facility owner was not present during the interview – not an easy task as in one case the interviewers were requested to lower their voices because otherwise the owner could listen to the interview via the surveillance cameras.

Because of the growing importance of the internet as a platform for recruiting clients in the Netherlands, the cities asked us to do an analysis of internet prostitution. A major problem with monitoring internet prostitution is double-counting. From client forums we learned that sex workers only advertise when they are actually available; this offered us an opportunity to estimate the number of sex workers who are active on the internet per day.156 Sex workers provide the prospective client with one or more mobile numbers and ask him to call using the number recognition facility. Sex workers indicate that they will not react to text messages or emails. This led us to try to google mobile numbers. This turned out to be a successful method for tracing sex workers. As a mobile number is linked to the same person (or persons), it offers a good indication of the number of unique sex workers who are actively recruiting on a given day. We learned that the same number appears on the same site at successive times of the day. Sometimes the same number advertises on the same site with different ads (although the site managers discourage this). Less than half of the mobile numbers appeared on only one site. The conclusion is that the term ‘double-counting’ is an understatement in internet prostitution. It was no exception to see numbers appear 200 to 300 times in one day (2011) when entered as a Google search, with some generating 2,000 hits or more. On 10 different days we counted the number of unique mobile numbers on a single sex site in the Netherlands. By double-checking two major sites on later days, we learned that the figure for unique numbers is quite stable. Figures can be an underestimate, as we cannot exclude the scenario where sex workers share a phone number.

We collected and analysed policy and administrative documents, and newspaper and other media reports on prostitution. As one of the main goals of the study consisted of conceptualising prostitution policy, we consistently used our empirical findings to confront the policy literature. Often the literature suggested insights and explanations of (p.275) the data, but almost just as often the data suggested important revisions of the standard theory. Finally, Hendrik Wagenaar interviewed New Zealand politicians, officials, professionals, scholars and members of the New Zealand Prostitutes’ Collective (NZPC) on two occasions (in 2013 and 2015) about the New Zealand Prostitution Act and the role of the NZPC in it.

Generalisability

While the data on which this book is based are collected in two relatively small European countries, we believe that their importance exceeds these two cases. Despite obvious differences in political culture and structure, both the prostitution markets and the policy dynamics surrounding prostitution in Austria and the Netherlands have a lot in common with those in other north-western European countries. Many countries in this area experience high rates of immigration, predominantly from eastern Europe, but also from Africa, southeast Asia and Latin America. Some of these immigrants end up in prostitution. In almost all countries in the north-western corner of Europe, prostitution occupies a prominent position on the political agenda. The Swedish approach of criminalising clients acts as strong beacon to pundits and politicians in many countries (including those in the EU), and leads to venomous public debates over the moral nature of prostitution and the direction of prostitution policy. Both Austria and the Netherlands have made concerted efforts to regulate prostitution with the stated purpose of reducing harm to sex workers and containing public nuisance that is caused by prostitution. To this end local policy makers in both countries deployed a range of policy instruments. These policies contain lessons for every nation or locality that tries to regulate prostitution in one way or another. The upshot is that Austria and the Netherlands are representative of the efforts of many affluent, highly developed countries to give prostitution and its direct stakeholders, sex workers and proprietors, a place in society – morally, spatially and institutionally.

Notes:

(151) Like all EU countries, the Netherlands and Austria have a National Rapporteur in Trafficking in Human Beings. This agency collects data on trafficking. The reports of the National Rapporteur have been used in our work.

(152) Landman speaks of a trade-off in international comparative research between a design with a large n, few variables, a high level of abstraction and rather robust conclusions versus a design with a small n, many dynamically interrelated variables, and detailed data and conclusions. Differently put, there is a trade-off between hard knowledge on a few, abstract variables versus less certain but more contextual, meaning-oriented and practical knowledge about the dynamics of everyday societal systems. Given the paucity of knowledge about prostitution policy, there are strong arguments to choose an in-depth comparison of a few countries.

(153) In both countries’ politics the grand postwar equilibrium has become destabilised by the increasing popularity of right-wing populist parties.

(154) One of the major problems of qualitative international comparative research that is conducted by one researcher is the problem of understanding the history and, often tacit, cultural understandings of a research site. In our opinion it is simply impossible to obtain a valid and solid grasp of the meaning of an issue by spending a limited time in a country, no matter how careful one is in selecting spokespeople. This problem is enhanced when researchers do not speak the language of the country where they are undertaking their research.

(155) When more than one sex worker participated in the interview, it was counted nevertheless as one interview.

(156) This would apply in Austria more or less only to webcam sex. Internet ads are paid on a weekly or monthly basis, so they are on the website for the requisite amount of time. This does not necessarily mean the sex workers work each day. (p.276)