Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

The Danish integration and immigration policies are among the most restrictive of their kind within Western Europe. As in other countries integration is a highly contested policy-domain and a salient policy issue. It is also one of the policy-domains most subject to change over the last two decades. This can raise the question of what has driven this development – new types of problems or the scope of the problem, political convictions or evidence-based research? It can be argued that the utilisation of external expert knowledge at national level has been minimal and the research-policy relationship can perhaps best be considered as a ‘pick-and-choose’ model where politicians and policymakers have employed the research that supports the hegemonic policy-frame of integration and definition of problems. Research thereby serves a legitimising function rather than an instrumental function (Jørgensen 2011). This characterisation at the same time stands in contrast to the importance and emphasis given to evidence-based policymaking. Subsequently, this situation may lead to disenchantment about research-policy dialogues. While the use of external research arguably has been limited there has in contrast been a proliferation of in-house research institutions within the political system itself, although often with limited budgets.

This chapter examines the role and impact of the scientific community on integration- and immigration policymaking. This analysis will be elaborated by three cases investigating, first; the ‘Workgroup on Burkas and other Similar Clothing’, which reported in 2009–2010 secondly; the development of the Action Plan against Extremism and Radicalisation of Youth from 2008/2009, and thirdly; the policy plans on social housing and segregation from 2004 onwards – in a Danish context termed ‘strategies against ghettoisation’. In all three cases the relationship between external and in-house research institutions will be discussed and taken together this will facilitate an outline of the prevailing research-policy dialogues in Denmark today. It is argued that particular research-policy infrastructures have consequences affecting not only policymaking and research-policy dialogues but also shaping perceptions of the research field itself and what it can contribute in policy processes.

The sources used for the analysis draw on a range of different materials. These include legislative texts, action plans, policy documents, newspaper articles, evaluations, political statements and speeches as well as six formal and informal interviews conducted for this piece of research.

2 Migrant Integration: An Overview

The following section gives an overview of the immigrant integration policies in Denmark over the last 40 years. The overview and following analysis draw on a policy-frame analytical approach (Benford and Snow 2000).

2.1 The Genesis of Danish Migrant Integration Policies

Until the end of the 1970s, Denmark had an extremely homogeneous population, with rather moderate immigration flows. The trajectory of Danish migration policy resembles that of many other European countries. In the late 1960s and early 1970s, substantial numbers of predominantly male migrant workers arrived in response to private employers’ demands.

Henceforth the main policy-frame was a frame building on the notions of ‘temporariness’ and ‘guest workers’. This frame was championed by both political actors and private employers who saw no need to introduce measures of integration, such as language acquisition, as the workers were there to work and not to talk. Both Danish society and many migrants themselves envisioned this stay as temporary. However, by the time of the oil crisis in 1973, many migrant workers had earned the right to stay, to state support, and to family reunification, making permanent settlement in Denmark a more attractive option than the often insecure economic conditions in countries of origin. Where previously male migrants had paid taxes but made little demands on the welfare state, recession and labour market restructuring gradually marginalised many of the original migrant workers. High levels of unemployment among unskilled workers created increased demands for public funds and services.

This created the basis for a new policy-frame as issues of integration (or rather disintegration, see Geddes 2003), became an increasingly political matter, raised particularly by the Social-Democrats in the early 1980s. This situation continued into the 1990s but was mainly set on the agenda by political actors. In 1994 a number of social-democratic mayors from suburbs of Copenhagen (Ishøj, Brøndby and Hvidovre) and Århus (all cities/areas with large concentrations of immigrants and real or perceived social problems) lost patience and brought the immigrant issue to the top of the agenda, through demands for quotas and dispersal of immigrants nationwide.

Following a commissioned study (Indenrigsministeriet 1997), the first Danish Act of Integration based on this new policy-frame was implemented in 1999 by the Social-Democratic and Social-Liberal government. The new act initiated a ‘tougher’ line with regards to integration and centralised authority over such matters, which previously had been delegated to local authorities and the private sector. The overall framing was based on a representation of immigrants as lacking the necessary human capital to achieve parity with the majority society. In this representation, self-sufficiency and autonomy were held to be valued characteristics which migrants were likely to lack. The focus on socio-economic integration inevitably led to a shift of burden towardss the immigrant, where integration first and foremost became the responsibility of the individual. The main instrument therefore was restrictions and the Minister of the Interior Thorkild Simonsen, responsible for the new Act stated: ‘I find it hard to see what the right-wing parties can implement in an orderly manner that would be more restrictive than what we are proposing here (Jastrup 1999).’

2.2 The New Era of Immigration and Integration Policies – From 2001 Onwards

The change of government to a right-wing coalition supported by the anti-immigrant Danish People’s Party in 2001 initiated a new era of migration and integration policymaking based on a policy-frame of assimilation and cultural adaptation. The new integration policy-frame was and is based on an overtly socio-economic representation of integration. The primary focus was therefore to keep people in employment in order to be able to maintain welfare services. The main claim was that ‘we must integrate those who already reside in the country’. This legitimated the restrictions in access to Denmark, especially targeted at inflow coming through family reunification and asylum seekers (Regeringen 2002). Integration policy thus was linked to general immigration policy.

Integration from then on was defined as active participation in the labour market and contribution to the welfare state. The key policy document for the new Danish integration policy was the action plan En ny chance til alle [A new chance for everybody] (Regeringen 2005) itself linked to the government strategy paper Noget for noget [Quid pro quo] published in 2004 (Regeringen 2004a). These programmes are the basis of policy today, but have been amended in different – often more restrictive – ways since, not least with the significant changes introduced in 2010 (Regeringen 2010a).

During the 2000s the political opposition was in an ambiguous position. There was a general support for the restrictive turn and ‘new realism’ within the population and, despite voicing critique of the policies, the opposition promised a continuation of existing policies. The government changed once more in 2011, and is now constituted by the Social Democrats and Social Liberal Party. The change in government also led to a shift from a predominantly centralised Ministry for Immigration and Integration towardss decentralised coordination, leading to the abolishment of the Ministry of Integration and a division of competences between different ministries. The government launched a new Integration policy in 2012 consisting of seven goals. These goals resemble the goals and strategic action plans of the past government and point to a large degree of path dependency in terms of policy goals (Regeringen 2012).

The policy trajectory initiated in 2001 evolved into a new frame that emphasises cultural adaptation to the host society and emphasises that integration should lead to a specific form of national identity. This is in line with the reframing of assimilation as the guiding principle for policy developments and provides an exclusivist mechanism that is used to legitimise especially ‘harsh’ political actions that are used to combat social phenomena that are described as ‘un-Danish’. This logic informs many of the policy initiatives relating to terrorism, radicalisation and ghettoisation, as will be discussed below.

3 Boundary Relations and Division of Labour

In the following section I will discuss and outline the boundary relations between research and policy. I will argue that the Danish boundary relations resemble an engineering model and will outline the features of the Danish model. This is followed by a concluding section on the roles of external research institutions in the Danish case.

3.1 An Engineering Model

With the realisation that migrants were settling permanently in Danish society, migrant integration emerged not only on the political agenda but also on the research agenda in the 1980s and 1990s.

Integration research initially reflected the fact that immigrants came as labour migrants and gradually turned into a perspective on immigrants as constituting social problems demanding political solutions. Research was diffuse and came mainly from papers and essays from specific educational institutions and/or as (local) reports with a focus on practical issues originating in concrete immigrant projects. During the 1980s socio-economic issues gained prominence alongside an increase in studies on marginalisation and inequality. Structural labour market explanations are prominent in describing the immigrants’ position. From the 1990s research is polarised between two main types. Firstly, studies focusing on immigrant behaviour and attitudes to immigrants; and secondly, evidence-based mapping studies seeking to generate facts about immigrants and performance in relation to language, education and work.

Especially in the 1990s there was also an increased utilisation of expert knowledge for policy-formulation. A good example is the reports by the sociologist, and later politician, Eyvind Vesselbo for the municipality of Ishøj in 1990. In the report Vesselbo shows how the population of Turkish newcomers grew from 145 people in 1969/1970 to 1,824 in 2000 (Vesselbo 2000). The report has been much cited in the Danish parliament and has been used as an example of how things can get out of control (see Jørgensen 2011). The research was not used instrumentally but was used to substantiate, i.e. lend authority, to the new emerging policy-frame coupling immigration, human capital and socio-economic integration.

However, for a large part the focus on restrictions, and the political ambition to bring down the number of immigrants arriving in the 1990s stand in stark contrast to the dominant Danish research agenda on immigration at the time. This finding lends support to the ‘two communities’ thesis, which claims that there is a gap between the worlds of researchers and policymakers, making it difficult to translate research into policy (Wehrens 2013). The central research institution in the 1990s was the Danish Centre for Migration and Ethnic Studies (DAMES) established in 1995 in Esbjerg by the University of Southern Denmark (SDU). Its main task was to coordinate and strengthen Danish immigration research. Contrary to the political agenda its main conclusion was that Western societies including Denmark indeed needed migrants and should focus on how to manage migration and avoid social marginalisation (see Jørgensen 2011). The research coming from DAMES was – unsurprisingly – not utilised in policymaking. The relation between politics and science in the period from the arrival of the migrant workers in the 1960s until the late 1990s can be characterised as a pick-and-choose model resembling the ‘engineering’ model outlined by Hoppe (2005). Knowledge was mobilised for the service of the state when needed and here it first and foremost served a symbolic function by substantiating the policy-frame being developed by politicians at the time.

Thus, research-policy relations in Denmark resembled an engineering model of boundary relations, and to some extent continue to do so today. Several features in particular resemble this engineering model. The first important feature was the establishment of the Ministry of Refugee, Immigration and Integration Affairs bringing together all tasks and competences relating to immigration and integration in one ministry. Of course there were overlaps with other ministries such the ones for education, social affairs or employment, but the main responsibility for policymaking within this field was laid down in this ministry until its abolition in 2011. The ministry contained different departments and different in-house research institutions. It also provided secretariat support and steering responsibility for the commissioned- and think-tank research that was contracted to inform policymaking in this field. The in-house bodies have become more important over the years and play a decisive role in defining policy problems and developing policy-frameworks.

The emergence of new departments was established more from political will than from the belief that there was a lack of research; at a minimum it can be said that the timing was without doubt based on political interests. The establishment of the Office for Democratic Community and Prevention of Radicalisation in 2008 definitely owed its existence to political interest in the topic at that particular time. It is a good example of a new type of in-house research institution, very often being a small unit with a department leader and a (small) number of researchers. The influence on policy-formulation of these units should not be underestimated however. They often cooperate with cross-ministerial groups in formulating policy and action plans. The question that follows is how such units interact with external research institutions in the policymaking process. Here a tendency is that the work-groups from the ministries can choose to engage external research institutions or commission specific research but are not obliged to include these results. Very often this research is indeed not utilised at all or mainly used in a legitimating way to show that the ministry has sought expert opinions or alternatives to a given policy problem and thereafter chosen to stick to the already preferred policy-frame. The various departments and sections of the Ministry of Integration developed libraries and databases containing state of the art research on the particular area of interest, e.g. radicalisation and terrorism. The commissioned studies also were listed among these. It can be difficult however, to see if these extensive knowledge bases were utilised in any way; there are few references to such information. First and foremost they had a legitimising function showing that the department in charge indeed had knowledge of research.

The second feature is that some research institutions had – and in certain cases still have – a privileged position. One of the most influential research bodies was the Think-Tank on Integration. It was established in 2000 and terminated in 2007. It illustrates a preference for experts with economic and statistical backgrounds, which is perfectly in line with the preferred socio-economic perspective on integration. The Think-Tank served as a research unit to gather data on trends in migration and integration. The Think-Tank’s reports were designed to guide legislators in developing new and coherent integration policies, as well as presenting international comparisons. The Think-Tank’s perspective was based on the same economically oriented backdrop as prevailing governmental understandings of integration (Tænketanken 2001). It also ordered research from other research institutions to guide its conclusions. The normative backdrop of the Think-Tank nevertheless was convergent with the position taken by the government and the Think-Tank’s findings were used to legitimate both previous and future policies. It was given the particular task of presenting reports on the state of integration in Denmark, but was not used to provide research on specific problems. Hence the research presented by the Think-Tank was not used instrumentally but rather symbolically, in both legitimising and substantiating ways to support the government policies already defined and implemented. The Ministry therefore had a great interest in disseminating the results from the Think-Tank and emphasising the congruence between the ‘two communities’.

The third feature of boundary relations and use of expert knowledge relates to the utilisation of external research. Here the relations definitely illustrate the primacy of politics when it comes to influence on policymaking. Generally we find a situation where external (but state-funded) research has very little impact on policymaking. The most important research initiative in the 2000s, the Academy for Immigration Studies in Denmark (AMID), has had very little influence on later policymaking if we compare the research results and suggestions coming from AMID with actual policy developments (Jørgensen 2011). AMID incorporated the economic angle, but also developed an innovative inter-disciplinary approach that included researchers from political science, anthropology, sociology and other disciplines. One of its ambitions was to facilitate dialogue between researchers, practitioners, policymakers and, to some extent, immigrants themselves. Basically AMID tried to influence the research-policy dialogue in a two-way model where both sides acknowledged the competences of the other. AMID has not been actively engaged in policymaking however, and the ambition has rather been to frame the issue of integration and to highlight gaps in research and in political instruments. In 2002 AMID released a publication providing a state-of-the-art review of research on integration between 1980 and 2002. This publication concluded by providing 40 concrete recommendations on areas where there was a profound lack of research that also had consequences for future policymaking on these particular issues (AMID 2002). But since these recommendations were made, almost nothing has happened to address them; the changes following from the 2010 revision of the Foreigners Act stand in particular contrast to these recommendations. It is still too early to tell if the change of government will lead to radical changes in the policy path.Footnote 1

There have been other research institutions besides AMID of course, and since its closure in 2007, we find a number of research initiatives emerging across the country, as well as the original consortium partners comprising AMID at the time. These external institutions also include independent self-funded research units such as the Rockwool Foundation, which since 1999 has had a special task-group working on immigration and integration (see www.rff.dk/en). A common feature across all these initiatives is that their results may or may not be used. In short the main tendency has been that studies using policy-frames which conflict with those of the government most often have been ignored, while research supporting government policies has been highlighted. This does not necessarily imply that relations between politics and research are ‘bad’. For example, the research institution Social Cohesion and Ethnic Diversity (SOCED) had a good relationship with the Ministry of Integration, which acts as a partner, pointing to a fruitful research-policy dialogue.

The discrepancy between the interest in external research and actual utilisation of this is also reflected in the attitude of a research employee in the former Ministry of Integration whom I interviewed, who stated that ‘the employees at the Ministry certainly take account of research results but at the moment there is no room for changing the path taken.’Footnote 2 The truth may also be that the research provided by universities lacks the applicability that civil servants ask for when assessing usability (cf. Boswell 2009).

3.2 The Role of External Research Institutions

External research institutions have different roles to in-house research institutions: I will point to four main roles. Firstly, in the policy-formulation process they are engaged rather late and most often asked to provide assessment of a given policy proposal alongside a long list of state agencies, NGOs and interest-holders. These consultation statements are often rather short and liable to be lost among all the others. As already stated they are given at a stage in the process when the policy-problem is already formulated. This particular form of involvement in the policy-process has more to do with the formal rules of the democratic decision-making process than utilisation of knowledge.

Secondly, external research institutions may be commissioned to provide a specific type of research, most often in an open competition. Here the frame is decided beforehand but not the outcome. The risk is that the research knowledge provided is not utilised later. This obviously leads to disenchantment about research-policy dialogues on the part of the researchers when this happens. Besides the risk of being ignored the researchers involved in commissioned research become open targets for political critique when working on politically contested issues.

Thirdly, external research institutions are often involved in evaluations of existing policies. Here the key task exactly is evaluating and not formulating policy. Such contributions may be emphasised if they support government actions, or ignored and bypassed in silence if they provide conclusions going against the preferred policy-frame or if they are undertaken in a policy-area where there is consensus, because there is not any real political interest in the outcome.

The fourth feature of the utilisation of expert knowledge has to do with the perception of research and knowledge itself, and hence decisive for research-policy dialogues. During the 2000s there was primacy for policymaking based on ‘values’, which was situated in the so-called ‘value-struggle’ initiated and championed by the right-wing government and Danish People’s Party. It legitimised firm actions bordering on the discriminatory and undemocratic, and emphasised that social problems may become so intractable that the required solutions may be invented from case to case and be based on moral conviction rather than knowledge. This understanding gained strength with a controversial New Year Address at the end of 2001, in which the Danish Prime Minister Anders Fogh Rasmussen at the time stated that:

We do not need experts and ‘arbiters of taste’ to decide on our behalf. […] There is a tendency towardss an expert tyranny that risks suppressing free public debate. […] Experts can be useful in submitting factual knowledge. But when we have to make personal choices we all are experts.

This understanding has been repeated several times since and used as the guiding principle for policymaking. Basing policymaking on ideological convictions cannot but distort research-policy dialogues, as rational arguments and research evidence may be ignored and may lead to dissatisfaction from researchers and hence strengthen the gap between the scientific and policy communities.

4 Key Issues in Research-Policy Dialogues

In the following three sections I will analyse the specific boundary relations and research-policy dialogues in the mentioned cases. The three cases represent different aspects within the integration policy domain: the first case on burkas concerns a much contested and politicised issue where methodological issues gained special prominence; the second case on radicalisation likewise represents a much politicised policy issue which here has links to issues of integration, extremism and security and involves different sets of actors; and the third one on ghettoisation relates to an issue where so far there has been more consensus between policy and research communities.

4.1 Burkas

One example of research-policy dialogues in Denmark involves the commissioned research of the ‘Workgroup on Burkas and other Similar Clothing’. The work group was appointed by the government in summer 2009 after a political debate started by the Conservative party which suggested banning the burka. The work group consisted of members from the Ministries of Internal and Social Affairs, of Employment, of Integration and of Justice. The backdrop was the government’s intention to combat the repressive ‘view of human nature and women’ that the burka illustrates (Indenrigs- og socialministeriet 2009). A very normative backdrop one can add. The problem was that nobody knew the scope of the problem, so the work group decided to commission a report on the use of the niqab and burka.

The task went to a group of researchers from the Department of Cross-cultural and Regional Studies at Copenhagen University who delivered the report in November 2009 (Rapport om brugen af niqab og burka 2009). The report was based mainly on qualitative methods and included interviews with Muslim leaders and organisations and the conclusion was that their best estimate was that there were around 100–200 women wearing the niqab in Denmark and very few wearing burka. Interviews with seven of these women moreover indicated that they did so by their own choice and did not feel forced to do so. It is difficult to say if the results were valid but they were definitely not what the government had expected. The report was leaked to the news but was subjected to a secrecy clause so Copenhagen University or the researchers could not respond to the debate that followed.

The report was criticised by almost all political parties. Naser Khader then member of the Conservative Party stated ‘I cannot believe that people on campus can bring themselves to make such a report. It’s embarrassing (Agger 2010)’, and Pia Kjærsgaard, then leader of Danish People’s Party, stated: ‘I am considering reporting the researchers to The Danish Committee on Scientific Dishonesty. This is really far out (Brix 2010a).’ Both Khader, who suggested the burka ban originally, and Kjærsgaard, who was and is against all signs of ‘Islamism’, had strategic reasons in criticising the report, but other politicians also made similar statements and criticised the methodology used in the report and thereby questioned the very foundations of that particular research. Some researchers supported this critique while most other researchers and Copenhagen University supported the validity and methodology of the report, especially because the researchers only had four weeks to finish the report and were given a very modest budget (Brix 2010b). Nevertheless, the critique was repeated in the final white paper of the government’s Legal Committee where it is claimed that the report is ‘on the “edge” of good research (Retsudvalget 2010)’, despite the support from large swathes of the scientific community. The white paper concluded with a sentence again emphasising the ‘right’ of policymakers to base policymaking on moral convictions rather than evidence.

The proposed ban did not go through, however, and the proposers had to settle with an amendment of the penal code making it an offence to force others to wear the burka and similar clothing, punishable with up to 4 years in prison. In the proposal for the new bill the methodology and quality of the expert knowledge (i.e. the Copenhagen University report) was again emphasised and based on these presumed shortcomings the work group concluded: ‘Given this background, it can in the work group’s opinion not definitely be ruled out that some women are forced to wear burka, niqab or similar clothing (Lovforslag nr. L 181 Folketinget 2009–2010).’ The example of the burka report shows that there are risks involved when the research-agenda is influenced by different interests. It leads to greater gaps between the policy and research communities and to increased dissatisfaction for both parties. The politicians could not understand that the researchers could not provide the evidence they needed to legitimise the bill banning the use of the burka; and the researchers, who were two PhD students, could not stand the harsh critique following the report and the attacks on their professionalism. This dispute also spread to other scientific venues polarising the research camps between quantitative and qualitative researchers. Finally it probably shows that there may be heavy costs when engaging in research on heavily politicised topics.

4.2 The Strategy on Radicalisation

The struggle against terrorism and various forms of extremism and radicalisation has gained much attention in many countries. Spurred by international developments as well as the Muhammed cartoons in 2005, attention for radicalisation in Denmark increased both in politics and in research. However, increased political attention and research does not tell us anything about the boundary relations as such or about how that research is used in policymaking. The main dispute here has been about the definitions supporting the policy-frame. Here we find a discrepancy between government, ministry and in-house research institutions and almost all external actors. The action plan En fælles og tryg fremtid [A Common and Safe Future] adopted in January 2009 describes the main actions and attention in this field (Regeringen 2009). The process leading to the action plan included research to different degrees and there has been a clear political intention to gain more knowledge on how and why people become ‘radicalised’.

The development of this plan has a longer story though. In 2005 the government’s action plan on combating terrorism identified a need to strengthen research on radicalisation and recruitment. The government thereby set the initial research agenda. Concretely two different projects were funded with 1.3 million euros each. The Danish Institute for International Studies (DIIS) received one of the grants and the other went to the Ministry of Defence who used the money to open an independent research centre, the Centre for Studies in Islamism and Radicalisation (CIR) based at Aarhus University’s Institute for Political Science. Both projects ran until the end of 2009 and were designed to cover gaps in knowledge and inform policymaking.

The same message was repeated after the publication of the proposal for the new action plan in 2008. The proposal was created by a cross-ministerial work group led by the aforementioned Office for Democratic Community and Prevention of Radicalisation, but did not include researchers. In the proposal outlining the action plan it is stated that: ‘Research from both projects will in future make important contributions to develop, qualify and target strategy in preventive work (Ministeriet for Flygtninge, Indvandrere og integration 2008: 61).’ However, the proposal was subject to a ‘dialogue-process’, which consisted of three meetings, one with representatives from the scientific communities and the other two with educational institutions, civil organisations and municipalities. Here the use of expert knowledge was used in a very symbolic manner, inviting researchers to participate, which legitimated the policy proposal but did not change anything in the policy proposal as such. The proposals were a combination of security and social control-based initiatives and instruments known and used from the integration policy. The end result was not so much a comprehensive policy plan, as an overly broad policy plan.

The Danish Security and Intelligence Service’s Department for Preventive Security and the Center for Terror Analysis (CTA), which both conduct in-house research as well as hands-on projects, have been central actors in the policymaking process and framing of radicalisation. Their presence definitely forefronts the security dimension, which obviously is a crucial dimension when investigating causes for terror, but it champions some policy solutions and disregards other. There is also a tendency to pick-up definitions and solutions in countries with more or less the same approach and it is no coincidence that the Dutch actions against radicalisation have been a great inspiration for the Danish policymakers. The consultation statements submitted by the scientific community more or less point to the same issues. Most acknowledged interesting and well-grounded ideas but almost all criticised the definitions of radicalisation, being too broad and encompassing too many potential radicals compared to what research results show. The policy plan pointed to the ‘many’ while research pointed to the ‘few’. If support for Hamas in the Palestinian-Israeli conflict was defined as radicalisation, as the policy proposal contended, such ‘support’ would most likely qualify the majority of Muslims in Denmark as radicalised, making the definition somewhat useless. The proposed initiatives were felt as stigmatising and generalising and in reality made all Muslims seem part of the problem rather than the solution. The very strong security aspect was mentioned as a problem for overall framing. The concept of radicalisation itself originates from the political debate and not from research and there might be risks in making such a concept the basis for future research (see hearing statements on Fakta om høringen 2008).

However, the problem was also based on the fact that most researchers were unable to provide knowledge that policymakers needed (e.g. who, when, how many, what can we do, how can we detect?). The in-house research institute responsible for the action plan had fewer problems in this and proposed a model based on different stages and phases of radicalisation, which was used in the final action plan published in 2009. Such a model, in theory, provides something to look for and indicates when actions should be taken to literally prevent the ‘bomb from exploding’. This did not stop the critique from other scientific venues but it is at the same time an example of instrumental use of research but here originating from an in-house institution. The scientific criticism is mentioned and acknowledged in the final action plan but it is difficult to spot any profound changes in the definitions and aims from the first 2008 proposal. Moreover the plan created a policy-frame prioritising the focus on radicalisation as a guiding principle for policymaking in other (integration) related fields. It is for instance coupled explicitly to the 2010 bill on housing policy and prevention of segregation. As a consequence, segregation is framed as potentially leading to radicalisation, for which certain solutions are prescribed, in contrast to framing segregation as marginalisation or poverty where other solutions are proposed. Hence, popular policy definitions are diffused to other policy areas basically because there is a political will to focus on this particular issue.

The utilisation of external expert knowledge pertaining to radicalisation can be characterised as legitimating. The CIR for instance published an investigation showing that very few were potential ‘radicals’. The researchers had problems with the definitions but did attempt to operationalise them and most importantly pointed to the action plan’s negative consequences, potentially leading to discrimination and marginalisation (Kühle and Lindekilde 2009). The investigation got a lot of media attention but the ministry and policymakers ignored the main points of critique. The authors also had problems recognising their own findings in the dominant radicalisation frame set up by both politicians and media, and gradually decided not to cooperate with the media.Footnote 3 In order to present their findings and what they understood as the main contributions they wrote a feature to a weekly newspaper and a web-portal for science (Kühle and Lindekilde 2010).

4.3 Ghettoisation

Segregation as it pertains to housing is framed as ghettoisation in the Danish context. It has been a very politicised and much contested area, like many other integration related issues we could add, but the scope and visibility of this problem has put it high on both the media and political agenda. The same can be said for the research agenda. It is one of the aspects of migration and integration policies that have generated a very developed research literature. This is due to the relations to other policy fields and research fields such as urban development, gentrification, social policies, and housing policies and so on.

The urban neighbourhoods now being defined as ghettos by the former Liberal-Conservative government were not perceived as problems when they were developed of course. On the contrary they were seen as the solution to the housing problems that immigrants faced when settling in Denmark. However high concentrations of different immigrant groups, social problems, low levels of labour market participation and general perceptions of higher levels of crime and lack of security gradually developed into a new policy-frame of ghettoisation. The first policy plan using this policy frame and addressing these questions is the Strategy against Ghettoisation in 2004. The main problem is that people with better socio-economic resources move away from these areas, which in turn attract people with poorer resources, thus laying the basis for segregation. This focus is also recognised by research in the field. The main framing is that ghettoisation sets up a serious barrier for integration and therefore the government needs to develop instruments to combat this development (Regeringen 2004b). Ghettos are thereby described as a problem pertaining to immigrants particularly. In contrast, this story is not supported by the existing research (e.g. Skifter Andersen 2010). The 2004 plan did not involve any research initiatives explicitly or refer to the literature. It did however draw up a new division of labour. It set down the so-called Programbestyrelse [Program Committee] that was to create the framework in which to implement the initiatives in the strategy plan. The committee itself does not have any members from the scientific community but consisted of members from municipalities, housing associations and with practical experiences in the issues at stake.

In the same period public funding supported a large research project titled Segregering, Lokal Integration og Beskæftigelse (SLIB) [Segregation, Local Integration and Employment] which was an advisory group in connection with a strategic research programme for welfare research lead by Professor John Andersen. The advisory group formed a governance network comprising members from housing associations, municipalities, Ministries, consultants and experts. It is perhaps an example of a dialogue-based relationship between researchers and policymakers, and follows similar previous practices at local levels.

SLIB had the original purpose of building a better knowledge base for integration in vulnerable urban and residential areas, i.e. ghettos. This is basically consistent with instrumental use of expert knowledge and deviates from the otherwise dominant engineering model. SLIB cooperated with another important research institute, The Danish Building Research Institute (SBi) and the aforementioned Skifter Andersen. SBi has a broad research agenda looking at all technical aspects of buildings as well as the surrounding environment, and therefore also the social spaces defined as ‘ghettos’. One of the members describes the dialogue-process as productive not least due to innovative civil servants with actual interest in the field.Footnote 4

In the final report on employment in segregated areas the group was asked to move from analysis to policy recommendations and invited to present the preliminary findings and suggestions at different meeting with the relevant ministries. This created some tension however, as the Ministry of Employment would not support the suggestions and the researchers on the other hand would not change them. As a result, the research-policy dialogues are characterised less by mutual dialogue and more by the ‘two communities’ thesis. The same conclusion is drawn by another participating researcher who maintains that the infrastructure for dialogue is there and there are good formal relationships between researchers and actors from the ministries and municipalities, but in the end it is not the civil servants who make the final decisions about policy goals and expert recommendations are rarely followed.Footnote 5 Thus research is decoupled from the final policymaking stage.

Without going into details about the research undertaken by either research institute it should be emphasised that SLIB and SBi as units and as individual researchers have contributed massively to the existing research on this field.Footnote 6 SLIB was originally scheduled to cease work in 2010 but it was decided by the previous government to continue the unit with the task of assessing specific policies and interventions in vulnerable areas, and evaluations of interventions. This is interesting as the researchers attached to SLIB have been openly critical of the new action plan against ghettoisation presented in 2010. While researchers and research units in other integration policy fields are sidetracked – as the discussion on the radicalisation strategy illustrated – researchers in this particular field are kept in the dialogue.

Until 2011 there was no ministry for housing or housing policies in Denmark and the policies were therefore made through cooperation between different ministries, with the Ministry of Integration holding a central role. In 2011 the new government established the Ministry of Housing, Urban and Rural Affairs, which now holds responsibility for segregated urban areas. In October 2010 the government launched a new action plan, which led both to political discussions and also to an increasing discrepancy between the policy and research communities (Regeringen 2010b).

Framing ghettos more as a cultural problem than a social problem situated the ghetto-problem in a value-based policy area as distinct from evidence-based or research-informed policymaking. The action plan provides the first actual definition of the problem based on three criteria: a high proportion of residents without labour market attachment (<40 %); a high proportion of immigrants and their descendants from non-Western countries (<50 %); a high proportion of residents with criminal records (<270 convicted out of 10,000 residents) (Regeringen 2010b). If an area was characterised by two out of three markers it was by definition a ghetto. Making ethnicity and culture the basis for ghetto problems implied that the traditional socio-political measures couldn’t be of any use here. Moreover the solutions mentioned focused mainly on control, law enforcement and demolition of parts of the ghettoes, dismantling the physical space, and less on changing the composition or social space.

John Andersen contends that two minor proposals followed the recommendations of the SLIB group; a proposal for preventative removal of families with serious social problems and a proposal on employment initiatives in the segregated areas. Both suggestions were framed differently from SLIBs suggestions however, and overall the policy instruments and solutions outlined in the action plan were not aligned with previous suggestions and recommendations. More problematic, however, is the ghetto definition itself as it is difficult to see the valid causality between unemployment, ethnic composition and crime rates. John Andersen describes the action plan as driven by ideological reasons rather than research findings.Footnote 7 It over-emphasised the explanatory power of ethnicity and thereby stood in stark contrast to SBi’s findings.

Generally most experts in the field are critical of the new strategy (Lehmann and Møller 2010) and their recommendations on how to deal with the problem of segregation are very different from the governments’ strategy (Møller 2010). The former chairman for the Program Committee Jørgen Nue Møller tried to rearticulate the old policy-frame on integration and criticised the government for being too passive and not following the recommendations of the Committee (JyllandsPosten 11 February 2010; Programstyrelsen 2008). Nonetheless, the situation was closer to the two communities’ thesis than the close relations between research and politics that characterised the policy-field during the 2000s and perhaps still so at municipal level.

The new government from 2011 initially followed the same definition and revised the ghetto-list in 2012 and insisted on maintaining the word ‘ghetto’ as well. However, in 2013 the Ministry of Housing, Urban and Rural Affairs added two additional criteria (level of education in the target area and level of income in the area). It remains to be seen how expert knowledge will be used in the new policy setting. Research-policy dialogues previously characterised by informative and instrumental knowledge use and evaluation could change, with ideological and value-based policymaking leading to a more symbolical role for expert knowledge.

5 Conclusions

The institutional arrangements between decision makers and scientific communities are complex and vary in different policy fields. While this is a very simple and also unsatisfactory conclusion, it is nevertheless difficult to identify a single model to describe these arrangements. The role of research and utilisation of expert knowledge is closely related to how a given policy issue is debated in the political and public agenda. If the government or ministries have sought to promote a specific policy-frame, research has been used in symbolic ways to support the proffered policy-frame. That is, by substantiating a predefined policy-frame or by lending authority to and thereby legitimising a given frame.

In the analysis of the Danish case as presented in this chapter I have argued that the boundary relations resembled Hoppe’s ‘engineering’ model with diffuse demarcations but primacy lying in politics where knowledge is mobilised for the service of the state. The different examples of commissioned research projects support this conclusion. However, researchers at the same time enter the policymaking process at a later stage when policy goals and definitions have already been defined or are left out in the final phases. The discussion of definitions was crucial for the research-policy dialogue on radicalisation. In Denmark, the issue of integration has been defined first and foremost by politicians, and research has been used to back up their perspectives in a bureaucratic pick-and-choose model. Hence, there may be different models at different times or indeed at the same time relating to different aspects. That finding is justified considering the three cases analysed. The value of such case studies when seeking to understand the overall boundary arrangements and uses of expert knowledge is exactly that it makes it possible to outline the complexity. Researchers working on issues of urban segregation definitely have been involved more closely in the policymaking process and provided knowledge that was used more instrumentally than research dealing with citizenship and naturalisation for instance. Consequently research-policy dialogues may be more balanced in some policy issues, but characterised by disenchantment in other aspects of the same policy-domain.

The analysis also shows that there is a difference between the roles ascribed to in-house research institutions and external research units. We have seen that commissioned studies have been ignored when their results were presented. There can be several explanations for this. The results may deviate from the policy-frame and therefore be ignored. They may be difficult to translate into policymaking, due to being too ambiguous (‘on the one hand…’ etc.). Commissioned studies may also have an alternative role, insofar as they give importance not to findings but rather to de-politicising effects, i.e. a policy controversy is to be elaborated by an external research unit that is given money and time, and when they finally present their results most politicians have lost interest. In other situations research is ordered and included first and foremost to show that specific angles have been investigated but the results will not be decisive for policymaking.

If we look at in-house research institutions their mark on policy has been easier to identify. The understanding of radicalisation was developed by an in-house research group. Again, it cannot be regarded as surprising that internal research is utilised differently to external research. Nevertheless, it shows that different lines are drawn for the boundary relations between science and policy, and signals a greater distance between internal and external scientific venues than between science and policy as such. Hence the two communities’ thesis can be confirmed in some situations and policy areas but not in others.

Research-policy dialogues in Denmark have changed over the last decades. The development of the research field and the growth in both external research institutions and in-house research institutions has changed the nature of these dialogues. The research field over the years has become more polarised and the policy field has utilised research in a pick-and-choose model. The relationships between policy and research take different forms according to the particular topic at stake. This has led to productive dialogues in some cases but disenchantment and frustration in other cases. One may conclude that the more a given topic is politicised, the less there is room for dialogue.