The construction of ‘socially marginalised Greenlanders’ as a target group in Danish welfare policy and practice

This paper examines the emergence of ‘socially marginalized Greenlanders’ as a distinct target category in Danish welfare policy and practice. It builds on analysis of policies targeting Greenlandic minorities in Denmark and interviews with welfare professionals in charge of implementing these. The paper shows how Greenlandic minorities are represented as characterized by markers of difference viewed to set them apart from other socially marginalized citizens. These relate to 1) structural differences that impact on the ability to receive and benefit from welfare services, 2) to the perceived cultural origins of the problems that socially marginalised Greenlanders face, and, finally, 3) to the excessive social problems associated in policies and by professionals with an upbringing in Greenland. The paper shows how policies and welfare professionals both reject and continuously resort to the notion of the target group as distinct from other socially marginalized citizens. In continuation of this, the analysis further shows how ambivalences and contradictions are not so much found 1 Research for this paper was funded by a research grant from Offerfonden [Eng.: The Victim’s Foundation]. Den grønlandske Fond [Eng: The Greenlandic Foundation] further supported preliminary research.


Introduction
It looks as if socially marginalised Greenlanders have a harder time profiting from social interventions. … There is a much larger proportion, when you think about all Greenlanders in Denmark, there is a much larger proportion of these who are homeless. So, something indicates that they have very complex problems, and that they profit quite poorly from social interventions. (Welfare professional in charge of coordinating strategies targeting marginalised Greenlanders) In Danish welfare practice and policy reports on social marginalisation, citizens with a Greenlandic background are regularly pointed to as a group that stands out as especially vulnerable to social exclusion in Denmark (Baviskar, 2015;Rambøll 2008;Socialstyrelsen, 2003Socialstyrelsen, , 2013. Among other sources, this is documented in statistical findings that show a significant overrepresentation of Greenlandic minorities among the homeless, social benefit recipients, citizens enrolled in drug and alcohol treatment, victims of violence and other forms of crime, and those with little or no higher education (Baviskar, 2015(Baviskar, , 2016. The understanding of Greenlandic minorities as extraordinarily marginalised compared to other disadvantaged citizens is echoed among welfare professionals who further describe ordinary services as inadequate for securing the access of Greenlandic minorities to appropriate welfare assistance. In response to this problem, a range of policy and practice initiatives have been crafted over the past decade and a half to improve services to Greenlanders in the Danish welfare system. Taking our empirical point of departure from policy documents and strategies as well as interviews with professionals within the field, we examine policies and professionals' perspectives on Greenlandic minorities in the Danish welfare system. Analytically, we draw on studies on the complex relations between policy and practice and on social categorisations in policy-making in order to explore the ways in which policies and professionals define 'socially marginalised Greenlanders' as a distinct target group that requires unique forms of intervention. Across the material, three themes are identified which demonstrate how marginalised Greenlanders are identified as a specific target group with assumed shared characteristics: structural barriers, culture, and excessive social problems. As indicated in the quotation above, Greenlanders in the Danish welfare system are typically represented as a group that is especially vulnerable. Both policies and professionals thus describe a target group characterised by 'markers of difference' (Keskinen, 2017, p. 154) that set them apart from the majority population and other marginalised citizens, as well as non-marginalised Greenlanders in Denmark. On the one hand, highlighting these perceived differences allows for a contextualisation of social problems rarely seen in other services to citizens with complex problems . On the other hand, we will show how this contextualisation sometimes privileges a culturally focused framing of the challenges associated with the target group that risks concealing other explanations for the overrepresentation of Greenlandic minorities among socially marginalised citizens. Finally, the article demonstrates the problematic nature of singling out 'socially marginalised Greenlanders' as a group in need of intervention. We go to the heart of this tension by discussing how the process of defining target groups is never clear-cut. Policies and professionals' statements about Greenlandic target groups alternate between emphasising their distinctiveness, at times resorting to stereotypes, and soon afterwards rejecting such claims of difference.

Background -Greenlandic minorities in Denmark
Some 16,000 Greenlanders currently live in Denmark. Greenlanders who move to Denmark do so for a variety of reasons, for instance, to accompany a spouse, partner, or parent(s) or to pursue education or employment opportunities, while others move to Denmark without a clearly formulated plan (cf. Flora, 2017). The majority of Greenlanders who move to or live in Denmark mostly get by on the same terms as the majority population.
However, a minority of Greenlanders is highly overrepresented among socially marginalised citizens. Among welfare professionals working with socially marginalised citizens, Greenlandic clients are often described as a group that stands out as especially difficult to reach through ordinary welfare services . A 2015 statistical mapping of Greenlanders in Denmark estimated that the proportion of homeless among citizens with a Greenlandic background was nearly 50 times higher than among the Danish majority population, amounting to 6% of Greenlanders living in Denmark (Baviskar, 2015, pp. 11, 47). Moreover, five times as many received social benefits and 10 times as many received alcohol treatment (Baviskar, 2015, pp. 11, 47). Furthermore, socially marginalised Greenlanders are exposed to crime three times as much as the majority population, especially as victims of crime (Baviskar, 2015).
Greenlandic encounters with the Danish welfare system have a long history. It is beyond the scope of this article to cover such encounters in Greenland.
Instead, we focus exclusively on recent welfare attempts to support marginalised Greenlanders living in Denmark. In 1878, a boarding house for Greenlanders was established in a Copenhagen residential area that functioned as a pension house for Greenlanders arriving to study in Denmark (Rud, 2009;Togeby, 2003). Today, Greenlandic Houses exist in the four largest cities. They function as cultural venues which host events relating to Greenlandic society and help Greenlanders moving to Denmark. Moreover, they form part of the broader institutional landscape through which services targeting socially marginalised Greenlanders in Denmark are currently implemented in collaboration with private and municipal actors. Another central institution that offers targeted services to Greenlanders is Kofoeds Skole, an institution that offers training and rehabilitation to people with various social problems, ranging from unemployment to homelessness and problematic substance use. In the 1970s, the school began to single out Greenlandic students as constituting a distinct target group (Meldgaard, 2011).
This was due in part to the proportion of Greenlanders at Kofoeds Skole, which in the 1970s included some 16% of the students and towards the 1980s comprised 6% of Greenlanders living in Denmark (Meldgaard, 2011, 168).
Moreover, the identification of Greenlanders as a particular target group was due to a sense that the services required to assist them were different from those needed by other students. Around the turn of the millennium, the focus on Greenlandic students intensified (Meldgaard, 2011), much in line with the development of some of the first broader strategies aimed at establishing a more concerted focus on what is now termed 'socially marginalised

Greenlanders.'
Two defining moments can be identified as central in framing current policy directions, which have had an impact on the formation of services within the field. The first was the 2003 formulation of the 'White Paper for Socially Marginalised Greenlanders' (1) 2 . The second occurred a decade later with the formulation of the 'Greenlander Strategy 2013-2016' (10). Both were nationallevel policy responses to the growing focus on the 'special needs' of Greenlandic minorities, both at the political level and among professionals who requested better tools to respond to these problems (7). Both the 'White Paper' and the 'Greenlander Strategy' were accompanied by funds from the special pool for the social area (Satspuljemidler) to support new initiatives specifically targeting this citizen group (7). In addition to these broader strategies, a series of studies of the target group was conducted at the municipal (3, 5, 8, 14, 16) or regional (4) level, and the conditions of Greenlanders in Denmark were documented through statistical mapping and qualitative findings (2, 12, 13). Perspectives of socially marginalised Greenlanders themselves were relayed in qualitative reports (11,15). Finally, guidelines with suggestions for how to approach Greenlandic clients were produced (5) as well as evaluations of concrete interventions (6, 17, 18).
One characteristic of the development of the strategies and services such as the 'Greenlander Strategy' and the subsequent services resulting from it is that many professionals in charge of implementing strategies have been directly or indirectly involved in developing policies. In other words, there is an ongoing dialogue and significant overlap between professionals and policy makers within the field of marginalised Greenlanders. An array of different private, voluntary, and municipal services are currently in charge of implementing these strategies. These range from counselling to drop-in centres and alcohol treatment, as well as municipal services, such as social workers with Greenlandic language skills, and cross-sectoral networks that aim at supporting collaboration between the different sectors working with socially marginalised Greenlanders.

Analytical framework -categorisations and stereotypes in policies and practice
An often-repeated criticism within policy analysis is that decision-making and policy planning take place far from the 'crowded offices' of the professionals working with a given problem or group of citizens in their daily practices (Scott, 1998;Lipsky, 2010;Mosse, 2005). Therefore, much policy research has focused on the mismatches and inconsistencies between problem representations at the levels of policy and practice. This is often referred to as a 'vale of implementation' (deLeon & deLeon, 2002, p. 469). Such studies have demonstrated how policies have 'gone sour' and risk never having an impact in practice (Kettle, 1993;Meyer & Rowan, 1977) showing how 'actors … devote their energies to maintaining coherent representations regardless of events' (Mosse, 2005, p. 2). That is, people working within policy implementation spend a lot of time supporting policy representations that are not necessarily relevant to practice.
In contrast to such analyses and critiques, policies and strategies directed towards marginalised Greenlanders have, as shown above, been developed in close collaboration between policy makers and professionals in the field. In fact, in some instances, welfare professionals are themselves policy makers or at least contribute significantly to the development of policies. Further, as will be shown in the analysis, professionals' representations often mirror those of policies when they engage in discussions of marginalised Greenlanders.
Therefore, we draw analytically on research that includes both perspectives on policy production and implementation as a dialectic process (Tate, 2020, p. 84;Moore, 1978). That is, we examine the emergence of 'marginalised Greenlanders' as a distinct target population and discuss what such representations entail with a view to both policy documents and professionals in charge of implementing them (see also Mosse, 2005). Hence, we regard the policies in focus not as finished end products but as 'a continuous process of contestation' (Wright & Reinhold, 2011, p. 86). Here, the contestation is not so much between differently positioned interest groups, but more present as an inherent part of ongoing policy and practice conversations about how to understand and best assist the target group.
Scrutinising the specific content of the categorisations of marginalised Greenlanders, we also take inspiration from the growing body of studies of the social construction and categorisation of target groups (Bacchi, 2009;Bjerge et al., 2020;Schneider & Ingram, 1993, 2015Harrits & Møller, 2011;Møller & Harrits, 2013). Using generalisations and grouping citizens into different categories by emphasising collective features and bypassing individual variations are inevitable parts of defining a target group. Generalisations about certain problems can thus be viewed as necessary in order to formulate societal interventions and to enable professionals to 'manage complexity, by foregrounding certain kinds of knowledge and information and selectively erasing others' (Yarrow, 2008, p. 440). However, as has also been documented, such practices risk reducing the complexity of a group to a specific 'kind of …', hierarchically organised in relation to other social groups (Bacchi, 2009;Dovidio et al., 2010. That is, the categorisation of a target group as distinct from others inevitably involves making claims about what sets it apart from others (Harrits & Møller, 2011). Schneider and Ingram describe this process as referring to '1) the recognition of the shared characteristics that distinguish a target population as socially meaningful, and 2) the attribution of specific, valence-oriented values, symbols, and images to the characteristics' (1993, p. 335). stereotypes that can be both positively and negatively constructed (1993,335). In policy categorisations, stereotypes are typically constructed through the use of binaries, although often emphasising the more negative characteristics associated with the perceived 'problem group' (Bacchi, 2009;Fairclough, 2006). One side of such binary oppositions is represented as common-sensical and superior (Bacchi, 2009), e.g. 'socially excluded' versus 'socially included', or 'passive' versus 'active'. Often, however, one side of these binaries is not visible in policies. Instead, it works as a silent or implicit opposition, positioned in contrast to taken-for-granted ideas about what constitutes 'normal' or 'good' ways of living; terms like 'socially marginalised', 'unemployed', or 'victim' are cases in point (cf. Nielsen, 2012;Møller & Harrits, 2013). It should become clear in the analysis that professionals also draw on such oppositions in the process of singling out marginalised Greenlanders as a distinct target category. What Schneider and Ingram (1993) refer to as stereotypes, Harrits and Møller describe as 'social categories', which, they argue, work to inform political categories and legitimise policy interventions (2013, p. 157). They thereby distinguish between a 'political category', which 'describes the subject in a target population and is identifiable in the texts", and a social category, which 'consists of references to culturally shared knowledge about social categories or even as references to stereotypical behaviour' (2013, p.

Data and methodology
The empirical data for the article consist of two types of data that are analytically separated out for our description of the methodology. First, we examined 18 documents (see Figure 1) produced between 2003 and 2019, a period in which national, municipal, and private organisations began to specify 'socially marginalised Greenlanders' as a distinct policy category in need of intervention. We applied an explorative, descriptive analysis of key documents consisting of white papers, guidelines, strategies, and reports. Data were identified through an extended research process, which included the identification of inter-contextual links between references to documents focusing on socially marginalised Greenlanders (cf. Fairclough, 2006). For example, if one strategy referred to another strategy or report or repeated specific categories similar to those in other documents, these documents were subsequently included in the analysis. Further, we contacted well-established professionals and experts within the field and asked them to indicate key documents. We did not conduct a systematic search but used different sources and expert knowledge to gain an overview of the documents. The two kinds of data were coded separately. However, we applied a similar coding strategy to the data: rather than using preconceived categories (Hsieh & Shannon, 2005, p. 1279), we first conducted an open coding of documents and interviews, in which we identified themes and keywords. We then discussed the first impressions of the data and their representations of marginalised Greenlanders and the problems and solutions related to these representations (Bacchi, 2009). Subsequently, we conducted a more detailed coding of documents in NVivo. This approach allowed us to identify themes, patterns, and keywords to support an overall content analysis of the documents and the interviews (cf. Bjerge et al., 2019;Hsieh & Shannon, 2005;Shore & Wright, 1997). Finally, we compared the codes from the documents and the interviews to find patterns in how marginalised Greenlanders as a target group were represented in policies and by professionals. Three overall themes emerged in the coding: 1) structural differences that were seen to impact the ability to receive and benefit from welfare services, 2) descriptions of the perceived 'cultural' origins of the problems that socially marginalised Greenlanders face, and, finally, 3) the excessive social problems or traumas associated by professionals with an upbringing in Greenland. Below, we explore the content of each of these themes.

Structural barriers
A main concern in policy and among professionals working with Greenlandic minorities is the barriers they encounter when trying to access services in the Danish welfare system. As Danish citizens, Greenlanders have the same rights to social services as the majority population when living in Denmark.
However, reports have pointed to barriers that might impede their ability to fully benefit from these rights (11). These include a lack of knowledge about the Danish welfare system and linguistic difficulties in encounters with the welfare system. It is further noted that due to their Danish citizenship, Greenlandic minorities 'aren't offered an integration intervention upon arriving in Denmark' (8; see also 1, 11). It has further been suggested that the lack of so-called integration services might hinder citizens' opportunities to receive Danish language training (9). Some reports go as far as to suggest that because of this, their 'Danish citizenship might be a hindrance rather than of help' in terms of integration (1; see also 8,9,16).
Attention to such structural barriers stretches back to some of the first studies of the status of Greenlanders in Denmark. In a power and democracy study,

Lise Togeby described Greenlandic minorities as a 'hidden' population within
Danish society: Greenland is a part of the Danish Kingdom, which gives Denmark a special responsibility for ensuring that Greenlanders are included as citizens on equal terms as the rest of the citizens in Danish society. Against this background, the lack of knowledge of the conditions of Greenlanders in Denmark is surprising. In the holy name of equality, we have made the Greenlanders in Denmark invisible (2).
According to this argument, the group is, paradoxically, not recognised on its own terms -as Greenlanders -precisely because Danish society disregards differences between the Danish majority population and Greenlandic minorities, with the broader aim of ensuring their inclusion. More recent policies have documented the effects of this invisibility for citizens in need of welfare assistance, describing how citizens may go for long periods without receiving any formal help -ranging from social benefits to housing -after moving from Greenland to Denmark (15). This is echoed among welfare professionals, as seen in the following interview excerpt: I've also met some who disappeared over a period of years. We found out that this person was missing. I mean disappeared, not missing, but disappeared. Like, hasn't received any public assistance for three years. And it was like, 'What have you lived off for all those years?!' Well, she had lived on the street. Received whatever help friends can offer, and then you think … 'But what has she gone through?' Right, to survive? (social worker employed in intervention specifically targeting Greenlandic minorities) As has been thoroughly documented in reports, due to their Danish citizenship, Greenlanders are not registered as 'Greenlandic' upon moving to Denmark (16) and thereby do not register as anything other than Danish citizens. Linguistic barriers and a lack of knowledge about the Danish welfare system further add to this 'invisibility' (8, 9, 10), as does some welfare professionals' lack of knowledge about rights that could support more appropriate help. For instance, policy reports have documented how, in practice, Greenlandic minorities' right to free interpretation services in the Danish welfare system often comes to rely on social workers' or health staff members' discretion or knowledge of this possibility (1, 3, 8, 11). This is echoed in interview findings with welfare professionals: Well, recently I accompanied a client to the hospital, where she was told that if she wanted an interpreter she had to know that she would get billed afterwards. And not … we had just left there when I find out that legally, Faroese and Greenlandic citizens are exempted from that. So that's annoying, right? (Social worker employed in intervention targeting Greenlandic minorities) Thus, the right to interpretation services, for instance, may be lost somewhere between citizens' or professionals' knowledge of the right or willingness to either request or offer the service, as suggested here by another social worker: So I think [lacking the ability to demand an interpreter] … has been an obstacle for many citizens' integration process, that they don't dare say it, because they expose themselves to a degree, and because they have so much faith in authorities that they don't feel they can say it. That it's up to others to decide. … And it has also been interesting to see how often the system tries to get around it [offering translation services in Greenlanders' encounters with the welfare system when a citizen speaks no or little Danish]. (social worker employed in an intervention targeting Greenlandic minorities) Above, we see how the need for targeted services to Greenlandic minorities is to a degree legitimised with reference to the setup of the welfare system itself.
This relates not so much to any perceived characteristics of the target group itself, but to the way in which Greenlandic minorities' status as neither foreign nor fully Danish (11), with all the rights that that entails, obstructs their ability to access and profit from services.

The promotion of culturally framed interventions
The sense that Greenlandic minorities too often fall through the cracks of the Danish welfare system is further attributed to perceived cultural characteristics viewed as setting them apart from other socially marginalised citizens, as seen in the following interview excerpt: It's really difficult for some of them to come down here [to Denmark] and meet the kind of treatment system that we have. And it's not because I'm saying there is anything wrong with the treatment system. I'm just saying that it isn't certain to be geared to work with people who have a different cultural background. … Up there you have completely different cultural values, cultural ways of being together, which are much more grounded in nature and community norms and strengths. (employee in job training intervention to socially marginalised citizens) Here, then, it is perceived cultural difference rather than structural gaps that obstruct citizens' ability to benefit from services. In some policy reports, perceived cultural differences are highlighted as a main barrier preventing citizens from profiting fully from welfare support. Thus, one report states that 'they are marked by the same multi-social problems as socially marginalised Danes, but are at the same time, cultural strangers … Greenlanders are culturally, historically, demographically, and geographically very far from Danes (6).' In a similar manner, the White Paper frames differences between opportunities of Greenlandic minorities and those of other socially marginalised citizens as cultural: 'the problem is … that Greenlanders based on their cultural and social background have a series of needs which differ from those of other Danes' (1). It further links social marginalisation directly to these cultural differences, stating that it is 'up for debate whether enough is known about what constitutes a unique "Greenlandic identity", and what this means for the social breakdown' (1).
While the issue of culture is highlighted in most policy documents and all interviews conducted with professionals, it is also treated with ambivalence.
Thus, even while insisting on differences, the White Paper argues that 'the typical Greenlander in Denmark does not exist. Many Greenlanders are well integrated into Danish society, with a spouse, children, and work (1).' Another report engages critically with dominant descriptions of 'the Greenlandic folk character' as being opposed to 'the typically Danish' and rejects stereotypical narratives about Greenlanders as 'intuitive, relaxed, empathetic and impulsive', but, as noted above, it also describes Greenlandic minorities as 'cultural strangers' (6). In this way, documents move back and forth between emphasising and soon afterwards rejecting culturally framed explanations for the social problems experienced by Greenlandic minorities. Likewise, how such perceived cultural differences come to expression is fraught with contradiction, as seen, for instance, in the way Greenlandic minorities are simultaneously described as reserved and as overly social.
These documents reveal a strong focus on the social environments, networks, and practices of Greenlanders, pointing among others to a strong need for socialising as characteristic of the target group. For instance, a 'tradition for sharing one's apartment with friends and family' is problematised because it is viewed as a 'barrier towards housing-related development' of the citizen (3, p. 18), and risks contributing to extending the time in which individuals may go without housing and income (11) and to a worsening of alcohol or drug use (15). Paradoxically, however, this description operates alongside a characterisation of Greenlandic minorities as quiet or reserved. For instance, Greenlanders are described as 'often more withdrawn' in comparison with other socially marginalised citizens in encounters with welfare professionals (11, p. 23). Similarly, reports problematise Greenlandic clients' lack of participation and unwillingness to make demands or thoroughly explain social or health problems to social workers (4, 16). The silence of Greenlandic clients, one report cautions, should not be understood as 'acceptance and understanding of what has been said', but 'could just be a Greenlandic politeness in the face of an overwhelming Danish flow of talk, which they might not even really understand (11, p. 23).' The White Paper does not only attribute this 'silence' to linguistic barriers. Instead: The silence reflects a lack of faith in one's own opportunities, a lacking trust in the understanding of Danish authorities, and a lack of will to verbalise problems that it might not be possible to solve anyway. It is therefore possible that, to a certain degree, the contentment expressed [by the Greenlandic citizen] could be interpreted as a result of a silence that conceals a degree of powerlessness (1).
In some reports, this silence or reservation is in turn viewed as a result of the colonial relationship between Denmark and Greenland and related difficulties in regard to speaking up against a colonial ruler (2, p. 74). This perceived silence is viewed as contributing to the inability of socially marginalised Greenlanders to exercise agency over their own situation.
In practice, these struggles over how to make sense of Greenlandic clients' problems and needs result in markedly different proposals for what appropriate interventions might look like. Group-based interventions are often promoted in response to an understanding of Greenlandic minorities as particularly loyal towards other Greenlanders (3,5,7,16). Thus, it is suggested that facilities and drop-in centres should offer opportunities for interventions that target Greenlanders collectively, instead of at the level of the individual (7). One report cautions, in relation to this, that 'it can be discussed if the gathering of Greenlanders is problematic in an integrational perspective', but continues by arguing that 'it has been assessed that the need for being together … is absolutely central, especially regarding the most marginalised part of the group' (3, p. 7). However, the opposite is also the case when reports or professionals emphasise the 'silence' or reservation of the target group. Thus, a report notes that Greenlanders sometimes 'need air ' (3, p. 72) and that these dual needs must be attended to in services. Likewise, a professional employed in a health intervention cautioned against group-based interventions with Greenlandic minorities in relation to drug or alcohol use: I think it's because if you are going into treatment for alcohol or drug use …, then it's often group based; you often sit a number of people together and talk about personal stuff like feelings and so on, and many of them, they really don't like that. … Generally, a lot of people don't like that. But I think especially Greenlanders don't like it. It's not because you necessarily have to have a special offer for Greenlanders, but on the other hand, linguistically, if there was something, if there was an offer in Greenlandic that took into consideration what is … what kind of Greenlandic mentality is it that you need to take into consideration. What it is we need to get a hold of. What kinds of emotions you have to talk about. (employee in health offer to socially marginalised citizens) As this illustrates, policy proposals and professionals do not agree on how to understand the target group defined in such interventions. Likewise, although attention to the cultural background of clients is repeatedly pointed to as intrinsic to the development of appropriate interventions, explanations about what perceived cultural differences mean in practice are continuously unsettled. Thus, one social worker, when prompted to reflect on such culturally framed explanations, responded: We just haven't succeeded in describing it properly yet. I mean, it's definitely an aspect of what it means to be socially marginalised. That's difficult for everyone. … So yes, it is most definitely really, really, really complex and … but what to call it ... I don't know. (social worker employed in job training to intervention for socially marginalised citizens) These somewhat contradictory positions -often within a single report or interview -show how the issue of culture is continuously negotiated and contested at the level of both policy and practice.

Excessive social problems and the limits of targeted services
Finally, the social problems associated with Greenlandic clients are characterised as different in character -and as excessive -in comparison with those of other socially marginalised citizens, in particular with a view to clients' experiences prior to moving to Denmark. According to the White Paper, 'Abuse and violence also form part of a Greenlandic everyday life in Greenland' (1). Another report suggests that 'A majority has, prior to coming to Denmark, experienced a childhood marked by problems such as alcohol use, incest, violence, and suicide' (6, p. 6), further suggesting that 'They bring that history with them' (6, p. 8). This is further viewed to impact welfare encounters between Danish professionals and Greenlandic clients and to influence the former's ability to properly respond to clients' problems and needs, as suggested in the White Paper: 'The Greenlandic client may have had an upbringing in Greenland that is so traumatising that it can be difficult for a Danish social worker to relate to it' (1). Among welfare professionals, there is growing attention to perceived differences between the social problems experienced by Greenlandic minorities and how they are seen as distinct in relation to other socially marginalised citizens: If you look at Greenlanders in relation to social marginalisation …, maybe there's an additional step [of marginalisation] that our social interventions simply haven't been capable of … we simply don't have social interventions that can capture that or have the concepts to talk about. It's not something about Greenlanders that's particular, but something about our interventions that we haven't figured out yet. (social worker in charge of implementing strategies for Greenlandic minorities).

The idea of additional layers of complexity is echoed in professionals'
characterisations of Greenlandic minorities as distinct from other marginalised citizens, which positions them differently in terms of accessing and benefitting from services: They have a hard time taking care of themselves and … they have enormous debts. They drink up their money if it isn't under administration. They can't find out how to go to the doctor by themselves and they can't … see that they need to see a doctor. So like, they are marginalised in all kinds of ways. (social worker employed in a Greenlandic association in Denmark) While 'socially marginalised Greenlanders' are thus continuously singled out as a distinctive target group, understandings about appropriate interventions in the group are less clear-cut. Both policies and social workers, however, repeat the need for professionals with specialised knowledge about the target group: 'Our advantage is', said one social worker, 'that we have a history with this target group. … You need to have a particular understanding in order to work with this group'. In this instance, 'a particular understanding' or history with the target group facilitates such specialised knowledge. Many policies further emphasise how the 'specific needs' of Greenlandic clients should be met by welfare professionals with specialised knowledge about Greenland and Greenlandic culture and language. This idea has informed documents since the first formulation of 'Greenlander strategies', as seen in the White Paper, which recommended 'that municipalities and counties where there are many Greenlandic clients receive more and qualified consultancy support from a person with deeper knowledge about Greenlandic affairs than what is present today' (1).
As already noted, professionals supported the idea of targeted services to Greenlandic minorities. However, the notion of Greenlandic clients as distinct from other socially marginalised citizens also caused uncertainty and sometimes outright irritation, for instance, in questions about what types of demands social workers could put on clients. This was seen in considerations about the degree to which the target group should be treated differently from, for instance, other socially marginalised citizens: But this is what provokes me sometimes when you work with Greenlandic minorities; the leash they get is so long, I mean. And we put so few demands, I mean now I'm saying exactly how I feel about it. And other communities, especially Muslim, they have to do this and they have to do that. … And we don't put those demands at all to Greenlanders. Because … and it's not only the citizenship, it is everything else under the surface, right? And we've talked about … this guilty conscience that I sometimes hear social workers and treatment providers say, like 'but we feel bad for them'. We just feel a bit more pity towards the Greenlanders. (social worker employed in intervention targeting Greenlanders) As described above, strategies for improving the access of Greenlandic minorities to welfare services have led to a range of different interventions and service offers targeting them. However, even while insisting on the necessity of such interventions, both policies and professionals suggest that there are limitations to their effectiveness if clients are exempted from the demands normally accompanying, for instance, social benefits. The idea of Greenlandic minorities having a 'longer leash' than other clients was echoed by other social workers, suggesting that at times, targeted interventions are viewed with some resistance if they are assessed as unfair in relation to demands and requirements placed on other disadvantaged citizens.

Discussion
Above, we have examined how 'socially marginalised Greenlanders' have been singled out as a distinct target group in Danish welfare policy and by professionals. This is done with reference to structural, cultural, and social differences viewed in such a way as to set clients apart from other socially marginalised citizens in Denmark. As demonstrated above, the definition of 'socially marginalised Greenlanders' as a distinct target group goes beyond more objectively observed differences, such as structural barriers, towards the attribution of specific characteristics that evoke either cultural or social differences. In this way, strategies targeting Greenlanders, more often than ordinary strategies targeting citizens with complex problems (Bjerge et al., 2020), approach the individual client as situated in -and caught up in -wider cultural, historical, and social contexts. These contexts are regarded as contributing factors in relation to many of the problems and challenges experienced by Greenlandic clients, but also as key in the formulation of welfare responses to them (Bacchi, 2009;Schneider & Ingram, 1993;Bjerge et al., 2020;Møller & Harrits, 2013). In other words, 'culture and cultural difference are thus characterized at once as both source of dysfunction and therapeutic panacea' (Santiago-Irizarry, 1996, p. 9; see also Johansen, 2011).
It is worth considering whether aspects of individual clients' behaviour, problems, or needs are perhaps too easily categorised under the label of 'Greenlandic culture'. In an analysis of ethnic minority patients in the Danish psychiatric system, Katrine Schepelern Johansen (2011) suggests that clients' problems and needs were often understood in the context of 'culture', especially when they appeared unfamiliar or difficult to categorise for staff.
Hence, 'for some staff members, categorizing some aspects of patients' behaviour or problems as cultural or ethnic is a way of saying that this is beyond the limits of what they can or want to deal with professionally' (Johansen, 2011, p. 178). Moreover, she suggests, 'categorizing a patient as non-Danish was often based on a diffuse feeling of not being able to establish a proper relationship with the patient' (Johansen, 2011, p. 178). In a similar manner, we have seen above how the nature of Greenlandic clients' problems is sometimes described as beyond the capacity of the individual social worker to relate to -as unintelligible -and as being of a different nature from those of other socially marginalised citizens. Likewise, attempts to make sense of the 'problem' of socially marginalised Greenlanders with reference to cultural differences draw on what Møller and Harrits (2013) referred to as social categories, often building on assumptions or common-sense ideas about 'Greenlandic culture ' (cf. Møller & Merrild, 2019;Nielsen, 2012). It is worth noting, of course, that 'culture' is only viewed as being of relevance in regard to welfare interventions if clients are of other ethnic backgrounds (cf. Johansen, 2011;Nielsen, 2012). This means that differences between Greenlandic minorities and other socially marginalised citizens may sometimes be exaggerated, sometimes with the latter as a 'silent opposition' (Bacchi, 2009;Fairclough, 2006;Nielsen, 2012) as described above, and sometimes explicitly, as demonstrated in the example below: Yes, and then you could say that many Danish, ethnically Danish, substance users or socially marginalised citizens are somehow raised in the institution. The Danish system, right? So they know where to go. It's really easy, because maybe they have a harder time doing all the house-related stuff, but they know how to say if they need help, where Greenlanders maybe live more in their own local communities of assistance (hjaelpesamfund), so they don't get around to asking for help. (social worker employed in intervention targeting homeless citizens) Here, Greenlandic clients' difficulties in navigating the Danish welfare system are contrasted with ethnically Danish clients, who are in turn described as clients who 'know where to go' or 'know how to say if they need help' and whose knowledge about and access to services is therefore straightforward.
This, of course, overlooks the difficulties other socially marginalised citizens might experience in relation to expressing demands for the right type of assistance or 'knowing where to go' .
The relation between policy and practice examined in this article differs from the top-down policies often examined in policy analyses (Scott, 1998;Mosse, 2005;Lipsky, 2010;Moore, 1978). This means that some professionals hold much more influence over the content of policies -as when network meetings among professionals working with Greenlandic minorities are used to discuss what future policy initiatives might look like -than other areas of policy making regarding socially marginalised citizens (cf. Tate, 2020;Bjerge et al., 2020).
As suggested by a social worker employed in an advisory capacity, this tendency has been reinforced by fewer demands for legitimisation and documentation than seen in other interventions for socially marginalised citizens: I mean it has been a tendency that when you had to make homeless strategies or substance use strategies or social strategies, you had to make long background stories and long passages about change theory and justifying the spending of money and what it all has to led to, where, in the area of marginalised Greenlanders, it has been like, now the old strategy has expired, and then we earmark some money for the next one. And then we figure out within the network what needs to be prioritised. It has been more like, this area needs to be prioritised, and then the people who work with it probably know best what the money should be spent on. … We haven't really seen any demands for documentation until the most recent strategy in 2017. … That is the first time we have seen any documentation demands from politicians in relation to the Greenlander area. And that's really unusual when you think about how much we have talked about 'effect' since 2007, 2008, that demand just hasn't been there in the area of Greenlander strategies.
While this means that policy understandings are close to those of professionals, it also means that social categories informing them tend to flow more freely between policy and practice. In this sense, the policies examined in this paper and their representations of 'socially marginalised Greenlanders' are not fixed or settled end-products that give shape to a clearly defined target group, nor are policy representations coherent (cf. Mosse, 2005). Instead, uncertainty, contradiction, and ambivalences are laid bare and remain unresolved within policies, as well as in interviews with welfare professionals.

Concluding remarks
This paper has offered an examination of limitations as well as potential benefits surrounding the singling out of 'socially marginalised Greenlanders' as a separate target group (Bacchi & Goodwin 2016). On the one hand, turning a particular group of people of concern into a concrete policy 'problem' enables the formulation of concrete practical interventions, or, in other words, something that can be acted upon (cf. Schneider & Ingram, 1993;Bacchi, 2009;Bacchi & Goodwin, 2016;Møller & Harrits, 2013). At the same time, insisting too rigidly on the particularities of the group risks alienating Greenlandic clients further and risks overshadowing more diverse representations that foreground the experiences of Greenlandic minorities. To be sure, more recent policies and reports increasingly highlight such experiences, but as demonstrated in this paper, there remains a tendency to evoke culturally framed explanations in attempts to understand processes of social marginalisation relating to Greenlandic minorities. As the paper has also shown, both policies and welfare professionals relate ambivalently to such claims of difference, oftentimes rejecting them but soon after resorting to the notion of 'culture' as that which constitutes Greenlandic clients as different from other socially marginalised citizens with similar problems.