Ethnicising activation as a standard story in a Swedish municipal labour market programme

ABSTRACT In Sweden, Denmark and Norway, activation policies are used to speed up refugee’s entry into the labour market. Previous research on activation policy documents has shown that refugees are ethnicised in the framing, development, and design of activation, but research on activation practices are lacking on how activation practices targeting refugees are organized and conducted, and why that is. This article analyses how and why refugees and ‘general unemployed’ are activated in different interventions within a municipal labour market programme, and which implications such division may have for the refugees being activated in said labour market programme. The article draws on interviews with a manager, social workers and participants within the same municipal labour market programme. The findings show that unemployed refugees were differentiated from ‘general unemployed’ by placing refugees in a specific intervention. The theoretical analysis, based on Charles Tilly’s theorizations on social categorizations and stories, shows how the differentiation maintained an inequality between the refugee participants and ‘general unemployed participants in how resources were provided, connected to the organization emulating and adapting to surrounding society. The inequal organization was legitimized through the stories the interviewees told regarding the labour market programme which in the article is concluded as a standard story of ethnicising activation. The ethnicising activation is analysed as an exploitation of the unemployed refugee participants, who in turn hoards the opportunity of participating in the labour market programme, in hope to find paths towards the labour market.


Introduction
In Sweden, as in all Nordic countries, there has been a significant increase of individuals migrating from outside of the European Union as refugees in the last decades (Djuve and Kavli 2019).Refugees living in Sweden are specifically vulnerable on the labour market due to their experiences with trauma, educational-and labour market backgrounds, and encounters with racism (Bursell 2012;Larsson 2015;Vernby and Dancygier 2019).Establishment in the Swedish labour market is often slow for refugees, and even after finding a job, refugees still have lower incomes than other groups in the labour market and are overrepresented in receiving social assistance in comparison to the general public (Forslund et al.,2017).
Sweden, Denmark, and Norway have implemented a range of activation policies that are aimed at speeding up refugees' entry into the labour market.Activation is part of active labour market policy and targets individuals claiming income support -who often fall within the lowest tier of public welfare -with interventions and policies that aim to assist the target group to enter or reenter the labour market (Hvinden 1999).Previous research on activation policy documents has shown that constructions of ethnicity and race matter in the framing, development, and design of activation, such as participating in labour market programmes (Fernandes Guilherme 2013;Gubrium and Fernandes Guilherme 2014;Hungler and Kende 2021;Vesterberg 2016).Based on ethnicised assumptions, refugees are framed by policy makers as in need of rigid activation interventions to become worthy recipients of welfare by reproducing an image of refugees as a weak, disadvantaged category that passively depends on welfare benefits (Dahlstedt and Neergaard 2016;Vesterberg 2016;van Riemsdijk et al. 2021).For instance, this have been the case in Swedish, Danish and Norwegian activation policies over the last few decades (Fernandes Guilherme 2013;van Riemsdijk et al. 2021).
Although research findings have shown that ethnicising assumptions regarding refugees are reproduced in activation policies, research is limited regarding activation interventions in labour market programmes responsible for activating refugees.Therefore, this article seeks to complement and give nuance to previous research by focusing on the organization of activation interventions and how activation is conducted.This article is framed by a qualitative study of a municipal labour market programme in Sweden that activates refugees and 'general unemployed' in two separate interventions within the programme.By comparing the two interventions, the aim is to analyse how and why refugees and 'general unemployed' are activated in different interventions within a municipal labour market programme, and which implications such division may have for the refugees being activated in said labour market programme.The empirical material was analysed using Tilly's (1999;Tilly 2006) theorizations on social categorizations and stories.Through interviews with the manager, social workers employed within the programme, and participants of both interventions within the programme, a standard story emerged about why refugees were activated in a particular intervention, but also how this categorization reproduced an ethnicisation of the individuals categorized as refugees that generated an unequal allocation of resources for the two categories of unemployed individuals, which is concluded as an ethnicising activation.
The concept of ethnicisation is used in the article to capture structural social processes of inequality which are ideologically constructed based on biological attributes and cultural essentialisation which have materialized consequences; for instance, ethnic discrimination in the labour market (cf.Neergaard 2004).By using the concept of ethnicisation, the article contribute to the research tradition of critically studying institutional practices which are dependent on political, financial, and cultural claims and power relations that lead to the systemic exclusion of ethnic minorities; i.e. materialized consequences (Schierup 2006;Ålund and Schierup 1991).The article can provide an empirical understanding on the institutional demand of ethnicising activation as a legitimate way to organize activation interventions and activate unemployed individuals, and how it interplays with the provision of resources between socially exposed groups of unemployed individuals who often participate in labour market programmes in hope to find paths towards the labour market.

Empirical background and setting
Sweden is a unitary state where legislative power is held at the national level alongside constitutionally guaranteed local self-governance in 290 municipalities, with the government controlling the scope of municipal authorities (Jacobsson et al., 2017).Labour market policy in Sweden is a centralized field, and policies are drafted by the national government and implemented by Public Employment Services (PES).PES is also the organization in charge of integration policy, through the Establishment Programme, since 2010.The Establishment Programme is optional for refugees with residency permits in Sweden, and participation runs for a maximum of two years (Larsson 2015).Municipalities are the main welfare providers, like social services, which includes social assistance (Jacobsson et al., 2017).Related to social assistance and the activation demands placed up social assistance claimants are municipalities, who are also providers of labour market policies through municipal labour market programmes, often targeting social assistance claimants and/or individuals unestablished in the labour market such as refugees (Forslund et al. 2019).There is thus a dual activation field, as unemployed individuals can be activated through PES or their municipality of residency (Jacobsson et al., 2017).This can also occur in collaboration between PES and the municipality, as exemplified in this article.
The municipal labour market programme 1 that was studied for this article is located geographically in a municipality with over 100,000 registered residents.In the municipality there were a range of activation alternatives, which all were governed by an administrative board.There were no official goals for the municipality relating to the integration of refugees, this was instead discussed in the policy document of the administrational board as part of the goals of increasing selfsufficiency and decreasing social assistance claims.Refugees were, alongside women with short education backgrounds and youths, pinpointed as a specific target group for the labour market administration in 2018.So, even though PES had the official responsibility to activate refugees, the municipality also activated refugees, which coincides with what Forslund et al. (2019) describe in their mapping of Swedish municipal labour market programmes.
The programme in the article originated in 2010 and the study was conducted in 2018.The programme enrolled about 300 participants each year, with the official aim of supporting unemployed youths 16-29 years of age in entering the labour market.The programme was influenced by the international phenomenon of 'supported employment' (see Drake et al.,2012).Thus, participants had subsidized employment during a designated time with a fixed salary, while receiving support and assistance in becoming more equipped for the regular labour market, such as coaching, CV-writing, and job-search support.Subsidized employments are an activation intervention that, in general, are subsidized by PES in part or in full (Forslund et al. 2019).Municipal labour market programmes can act as an intermediary between PES and the employer.Municipalities can also finance subsidized employment themselves, often targeting social assistance recipients.In the labour market programme, there were examples of both municipal-and government-funded subsidized employment.The two forms of subsidized employment mentioned in the empirical material are municipally funded subsidized employment and 'Extra-jobs' (Extratjänst).Extra-jobs, subsidized and granted by PES, target refugees and long-term unemployed individuals and are only eligible for public and non-profit employers (See Mulinari 2018).
In 2018, the programme started a collaboration with PES to target refugees in the designated organizational aim of the labour market programme.The result was the intervention that in this article is named 'Intervention 1'.Participants, between 16-29 years old, from the Establishment Programme at PES were directed to participate in Intervention 1 as part of their programme period within the Establishment Programme.This meant that a sorting process was created in the programme, where unemployed individuals were sorted into different interventions depending on whether they were categorized as refugees.I refer to these two categories as refugees and general unemployed.The term general refers to the general sampling of the programme's participants Intervention 2, meaning youths who have been unemployed for a long period or are assessed to have high thresholds in entering the regular labour market.Thus, 'general unemployed' can include unemployed individuals from an ethnic minority.The two categories I am referring to are not understood as natural, solid categories, but rather reproduced social categories that are used by the programme to sort and handle unemployed individuals.

Previous research
Activation is a large research area that spans several focuses and perspectives (See Clasen and Mascaro 2022).Activation within labour market and social policy systems is often a mix of enabling and demanding interventions (Caswell et al. 2017).A tailor-made intervention design consists of a focus on the unemployed individual's needs and choices, while a pre-packaged intervention is more rigid, with little room for individual needs or choices (cf.Jacobsson et al., 2017).In relation to measured effects on activation, which are scarce and ambiguous, findings suggest that tailor-made interventions render the most success in supporting unemployed individuals towards finding and securing employment (Hasluck and Green 2007;SBU 2022).
In previous research, activation practices are often mainly understood as 'doing' the labour market policies the organization and 'street-level bureaucrats' are set to implement (Brodkin 2011;Caswell et al. 2017;Larsson 2015).The strong research focus on activation in relation to employment and national policy descriptions also means that little attention has been paid to what happens in activation practices besides attempts to support unemployed individuals in entering the labour market and, if so, why that is and how it can be understood, leaving much of the practical activation work with unemployed individuals 'a black box' (Caswell et al. 2017, 1).Consequently, more nuanced aspects of activation are required; for instance, examining participants' experiences of being activated or by studying activation practices from another perspective than simply an arena for implementing and negotiating labour market policies.
Previous research on ethnicisation within activation has mainly focused on discourse analysis on policy texts (Fernandes Guilherme 2013; Gubrium and Fernandes Guilherme 2014;Hungler and Kende 2021;Vesterberg 2016).This research provides important insights on ethnicising as institutionalized in labour market policies in Europe, but lacks understanding as to how and why ethnicising activation is organized and conducted within the organizations that activate unemployed individuals.Using the related concept of racialization, Larsson (2015) shows, through the perspective of street-level bureaucracy, how public employment officers implement racialized unemployment policies.In similar vein, Neergaard (2004) analysed the PES role in the racialized labour market through interviews with public employment officers.More recently, Söderqvist Forkby (2022) analysed anti-racist social work within a labour market project focusing on integration of unemployed migrants.Söderqvist Forkby discusses how the project -even though there are attempts to do anti-racist work -contributes to racial oppression related to the project by targeting stereotypical and ethnicised migrant categories.Together, the aforementioned research highlights the need for further research on activation of unemployed migrants or ethnic minorities, which this article contributes from an organizational perspective.
Theoretical framework Tilly's (1999) defines inequality as an unequal distribution of resources within organizations.Unequal categorical pairs (men/women, citizen/non-citizen, black/white etc.) become institutionalized in organizations, thus constituting inequality (Carlsson Stylianides et al.,2021;Tilly 1999).Categorisations are also relational.One category's hierarchical position is in relation to other categories, which will change over time and by location (Tilly 1999(Tilly , 2002)).Categories are more easily incorporated in an organization if they already exist outside the organization.When such external categories are combined with an organization's internal categories, society's alreadyestablished assumptions, practices and relationships facilitate inequality (Carlsson Stylianides et al.,2021;Tilly 1999).Tilly's (1999) describes that there are four mechanisms that uphold inequality in relation to allocating resources.Exploitation is a mechanism where resources are distributed unequally among those categories of individuals who partake in the gathering of the resources.Exploitation is supplemented by the mechanism of opportunity hoarding, where an unequal category of individuals gains access to resources that can lead to a monopolization of the resources.Emulation is when organizations borrow or copy organizational structures that have unequal categorizations.Adaption holds the categorical inequality in check, where actors within the organization use stories to legitimize and make sense of the categorization made within the organization, which in the end grants more power to the unequal structures backing categorization.To analytically capture how resources are provided between the categorical pair of refugees and general unemployed and the inequality it may uphold, I use stories as an analytical tool.
In the article, the analysis focuses on the descriptions, accounts, and explanations given by the interviewees regarding activation in the labour market programme, which are conceptualized as stories (cf.Tilly 2002).Stories are here understood to contribute to upholding social boundaries by providing conceptions regarding differences and similarities that separates and joins together groups of individuals (Montesino 2022;Tilly 1999).Thus, stories confirm existing relationships (Tilly 2006).Collective stories, like the interviews in this article, legitimize how categories are constructed and reproduced within organizations (Montesino 2022).Stories as an analytical tool do not capture the accuracy of 'reality' but rather how the story is used and why (cf.Tilly 2006).For instance, how stories told by organizational members contain information, such as the organization of the programme, how resources are allocated, and why categorizing unemployed individuals to activate them is suitable in supporting employment.As such, the stories connect to the social processes surrounding the storytellers to fit a standard story in society (Tilly 1999(Tilly , 2006)).Storytellers: i.e. the interviewees, use established concepts within the organizational context such as myths, symbols, and 'common sense' to uphold the standard story.By analysing how and why refugees and 'general unemployed' are activated in different interventions within the programme these stories can be encapsulated as a collective standard story that has formed in response to the social setting of the storytellers.

Research design and methodology
The material presented in this article are part of a bigger study which is part of a research programme funded by Forte (2016-0723).The study has been ethically approved by the regional ethics review board in Uppsala (Reg.no.2016/173).
The study had an abductive strategy, meaning it employed a combination of inductive and deductive reasoning in an attempt to go beyond both strategies by altering theories and being reflexive in how the material was analysed (cf.Alvesson and Sköoldberg 2008).The study started inductively, as there was no predetermined theory that governed the collection of empirical material.Rather, the study was guided by gaining an enhanced empirical understanding of activation practices.This means that theoretical concepts such as ethnicising activation were applied to make sense of the empirical material, rather than the other way around.In relation to the abductive approach, this means that different research foci and theoretical choices have been considered and tested throughout the research process to interpret and reinterpret the material in a reflexive manner (Alvesson and Sköldberg 2008).
The article is based on semi-structured individual and focus group interviews with the first-line manager of the programme (individual), ten social workers employed within the programme (one focus group), and six unemployed participants (one focus group).The social workers were recruited with the help of the manager mentioned above, and the participants were recruited with help from the social workers.The interview guides were designed to guide the conversations but left opportunities for the interviewees to expand upon their own perspectives and angles on the themes (cf.Brinkmann 2014).The themes could, dependent on interviewees, include the following: 'describe an ordinary day within the programme', 'what is the goal of the programme' or 'what can social workers and/or participants decide in the programme'.
Focus group interviews were suitable for study's exploratory point of departure, as the ambition of focus groups is to begin a discussion between interviewees.This can be a fruitful way of gathering new information (Lambert and Loiselle 2008).For instance, an important discussion between the participants arose regarding unemployment benefits.The focus group interviews were also a reason as to why stories were later used as a theoretical tool, as the focus group interviews provided an opportunity to analyse interviewee's collective stories.All ten social workers were interviewed in one focus group, and all six unemployed participants were interviewed in one focus group.The size of the focus groups, especially that of the social workers, was not ideal, as such large groups can easily lead to some individuals dominating the conversation (cf.Wilkinson 1998).Thus, I used different strategies to include all interviewees.For instance, by being attentive to non-verbal communication between social workers and asking specific respondents to answer or to elaborate on topics.
The interviewed social workers all had diverse educational backgrounds within the social sciences, and all had previous work experience with practical social work, which is why they are referred to as 'social workers'.The social workers were all employed by the programme, working directly with the participants to activate them towards the labour market.
The interviewed participants represented both Intervention 1 (Younes, Amira, Hussein and Mohammad) and Intervention 2 (Moa and Caroline).All participants were between 16 and 29 years old and openly unemployed prior to participating in the programme, with no to little labour market experience in Sweden.None of the interviewed participants had finished upper secondary school in Sweden.The interviewees from Intervention 1 had all completed corresponding upper secondary school before coming to Sweden.Participants had been in the programme between two and eight months.As participance in the programme required Swedish language skills on a level in which participants could make themselves understood at their workplace, there was no perceived language barrier in the group interview with the participants.All interviewees (manager, social workers, and participants) have been provided with fictive names by the author, all interviews were transcribed verbatim, and quotes have been translated to English by the author.
The analysis of the material is based on thematic analysis.The analytical approach was influenced by Bazely's (2009) three analytical steps: (i) describe, (ii) compare, (iii) relate.First, I described the themes contextually, by noting the background of the analytical category; for example, the programme organization, as described in the interviews.After that, themes were compared, depending on the respondent.Last, themes were related to previous research and theoretical concepts, where I noted whether the content of the theme aligned with previous results and how it could be understood from the standpoint of the theoretical concepts selected a priori.Throughout, a comparison between the two interventions has been guiding the analysis.In the final step of the analysis, the differences and similarities of the two interventions within the programme were connected to Charles Tilly's theorization of durable inequality (1999) and stories ( 2006), which created sub-themes where quotes from the interviews were connected to theoretical concepts.

Resources in daily activation
In the first theme, the analysis focuses on the allocation of resources in daily activation within the labour market programme, from the point of view of how refugees and 'general unemployed' are activated in different interventions, and which implications such divisions may have for the refugees activated by the labour market programme.Four resources have been analytically noted and are presented in Table 1 below.

Resource one: flexibility in design
The first resource concerns the flexibility of the intervention.Having access to a flexible activation intervention is a resource that better can suit the needs of the unemployed participant and lead to faster employment (Hasluck and Green 2007;SBU 2022).Intervention 2, for general unemployed, was tailor-made to the individual participant and their needs.
We work with [general unemployed] who for some reason need a job, you could say.And what we do is that we meet the person and then we find a place based on their wishes [. ..] we try to meet the person's wishes you know.[. ..] if it works well, it becomes an employment for 6 months, and we are there all the way and support both the employer and the participant.
(Johan, social worker) Johan was explaining how Intervention 2 was designed.In his explanation, he placed emphasis on how the intervention works for those who need a job and on the importance of following their wishes in employment.The needs and choices made by the participant thus governed Intervention 2, as Ahmed said: 'There is no successful practice if we control them and do not listen to their wishes.They are the ones in control of their wishes, so it is a balance to get in [. ..]', (Ahmed, social worker).
The 'balance' in the quote relates to designing the intervention so the participant is motivated to participate, hence creating a long-term plan for staying employed.Together, the quotes display how the social workers talked about Intervention 2, reproducing the perceived benefits of participating in such an intervention (cf.Tilly 2002).However, within Intervention 1, for refugees, participants had a pre-packaged intervention, with a specific type of job offered to them, within specific municipal organizations: 'And then we have . .

. another, another concept, what they should work as [Intervention 1]
, so therefore there is no freedom in choosing . . .' (Isak, Social worker).When Isak was describing Intervention 1, he described a low flexibility, with little possibility for participants to influence what they want to do, but without mentioning it being, or not being, beneficial for participants to participate in such an intervention.The resource of flexibility is understood to be connected to how participant needs are perceived.Isak described how Intervention 1 was a possibility for refugees, ' [. . .]to prepare them to work in the Swedish labour market'.Based on the quote, my understanding is that in Intervention 1, participants' needs were described to primarily involve preparation, and a more rigid design would ensure.The contrast between the two interventions aligns with previous research on activation of refugees, where refugees are constructed as an unemployable and excluded category before being activated and are thus in need of preparation before entering the labour market (cf.Vesterberg 2016).The idea that the intervention will prepare refugees to work in the Swedish labour market, as Isak described it, can be understood as an adaption to the categorization of refugees, where social workers in the interviews tell the myth of how refugees should be activated within the Swedish labour market, without considering the diversity of individuals within the categorization (cf.Tilly 1999Tilly , 2006)).

Resource two: time
In Intervention 1, refugee participants had subsidized employments at 50%, with no possibility of working more if they wanted to.They were simultaneously participating in the Establishment Programme at 50%.'[. ..] it's half-time, and most of them combine it with language studies in the morning' (Isak, social worker).As I interpret the quote, Isak gives a straightforward description; unemployed refugees only can participate at 50%.In contrast, Lisa explained regarding Intervention 2: 'Then there can be participants who do not think they can handle participating full time [. ..] "My goal is that I should be able to participate 75%".Okay, then that's the goal' (Lisa, manager).Lisa described in the quote that participants in Intervention 2 could choose to participate full-time or part-time, depending on their needs.Thus, in Intervention 2, time became a resource that was provided to the participants, based on the needs of each individual participant.How time differs between the two interventions can, like flexibility, be understood as a myth of how refugees need In the interviews, participants in Intervention 1 and 2 had different explanations regarding how much they were paid and why, that related to if they were activated as refugees.Above, Caroline said that they all were being paid the same, since they all had the same amount of income at the end of the month.This explanation legitimizes differences in how time and resources are provided, since the total amount in the end was the same (cf.Tilly 2006).Furthermore, income for refugees in Intervention 1 comprises welfare benefits that will not be counted as income in the social security system, meaning that if participants in Intervention 1 and 2 become sick or go on parental leave, they will also receive different benefits from the social security system, and it also affects future pension.Thus, income distributed in the programme had direct implications for the participants, which disadvantaged the refugee participants' future income in relation to participants in Intervention 2.

Resource four: access to unemployment benefits
The last resource concerns participant's access to unemployment benefits after employment, a resource that became evident in the group interview with the participants and had significant value for them.Participants in Intervention 1 were provided with the subsidized Extra-job and participants in Intervention 2 were provided with subsidized municipal employment.As it turns out, being employed with an Extra-job did not grant access to unemployment benefits when the employment was terminated.
For example, for us who work [through Intervention 1] we had no idea we had no right to unemployment benefits, so it was like 'oh my god' after like two three months.[. ..] my boss told me 'you have no right to unemployment benefits' (Amira, Intervention 1) In the quote, Amira explained that the participants in Intervention 1, who all had Extra-jobs, would not have access to unemployment benefits after their employment.Not having access to unemployment benefits is controversial in Sweden, since unemployment benefits are often described as a cornerstone of the rights-based universal welfare state (Lindellee 2018).The participants in Intervention 1 thought this would include them, as well.Again, Amira described; 'and it was like okay . . .we think everyone who works in Sweden has the right to unemployment benefits, even if it is Extra-job or not. ..it was strange to us.Really'.The quotes highlight a clear unequal distribution of resources between unemployed refugees and general unemployed.Participants in Intervention 2 who had municipally subsidized employment had the right to collect unemployment benefits; 'as I understood it, then you have it [unemployment benefit] as a [participant within Intervention 2] but then as an Extra-job you do not have it, but I do not know why [. . .].' (Moa, Intervention 2).Hence, this implies an unequal distribution of resources that affects the participants even after leaving programme (cf.Tilly 1999).Participants in Intervention 1, if unemployed after leaving the intervention, are referred to means-tested social assistance and the stigma that comes with it (Angelin 2009).Living on social assistance also implies that one must use up possible savings before being eligible and live in uncertainty regarding whether social assistance will be granted and for which living expenses it will be allocated.In contrast, unemployment benefits are a right with no restrictions on savings or means-testing (cf.Lindellee 2018).Thus, the fourth allocation of resource provides insight on how the activation of the refugee participants may have implications for them after leaving the programme.

Organizing activation
In the second theme, two sub-themes emerged in the analysis, that together highlight the organizational aspects of why refugees and 'general unemployed' are activated in different interventions within a municipal labour market programme.

The categorization process
Categorization of the unemployed individuals was, as described previously, connected to collaborating organizations surrounding the programme.In fact, the programme accepts unemployed refugees from PES in Intervention 1, and general unemployed from municipal social services in Intervention 2, as Lisa the manager describes.In the quote, Lisa provided insight on the categorization process' connections to the collaborating organizations: PES and municipal social services.Based on the quote, the categorization seems like a common-sense practice for the organization (cf.Tilly 1999).Emulation occurs when an organization borrows or copies structures that include unequal categorization, that thus becomes commonsense (Tilly 1999).The Establishment Programme within PES is such an example.'I talk to my case manager [PES] and then he says, "okay there is a job [within Intervention 1], you can try and maybe work with them".So, I said okay'.(Hussein, participant).As described by Hussein, the programme recruited participants through PES regarding refugee participants in Intervention 1.As described by Gubrium and Fernandes Guilherme (2014), there is no need for the individual social workers within the programme to categorize participants into Intervention 1 or 2, as they are already embedded in the organization as categorical pairs, emulated from external collaborating organizations (cf.Tilly 1999).However, it also means that organizing activation within the programme is built upon an unequal categorical pair (cf.Carlsson Stylianides et al.,2021).
The categorization processes can also be seen in light of how subsidized employments were used in the programme.Based on the interviews, participants in Intervention 1 were only provided with the subsidized Extra-jobs funded in full by PES.Participants in Intervention 2 were provided with municipally funded subsidized employment.In the interview with the manager Lisa, a point was made of how Extra-jobs resulted in economic revenue for the programme.
But now, for the first time, we now have more revenue than we had in previous years from PES because of . . .Extra-jobs.Ehm. ..because we hire more and more refugees, and they have Extra-jobs and so.It has only been refugees who have had Extra-jobs.So, we have record revenue now, because of the Extra-jobs.(Lisa, manager) In the quote, Lisa described how the programme has a record revenue because of their use of Extrajobs for unemployed refugees in Intervention 1, since Extra-jobs are fully funded by PES.The economic benefit of Extra-job also spills over for the sectors of the municipality that hire refugees through the programme, as they receive employees at a fraction of the actual cost (see Frödin and Kjellberg 2020).The way the programme has chosen to use the Extra-jobs can be analysed as an exploitation of the refugees participating in Intervention 1, as the programme makes economic revenue while providing participants with an unequal distribution of resources in contrast to Intervention 2. Thus, resources were being distributed unequally among those categories of individuals who partake in the gathering of the resources (Tilly 1999).All in all, this provided a story of the benefits of activating unemployed refugees for the municipality that at the same time upheld inequalities between participants in Interventions 1 and 2.

Participation in relation to other activation options
Even though the organization of the programme maintains an inequality between the categorized pair of unemployed individuals through the two interventions in the programme, the opportunity to be granted subsidized employment can indeed be a resource for unemployed individuals who previously had low or no income.This was also recognized by the manager, Lisa.'We usually talk about how [the programme] is a bit like sitting on the idol jury with the golden ticket.You know, we have jobs to hand out'.The quote from Lisa emphasizes a story of participation in the programme is an opportunity that many won't have, by referring to participation as a golden ticket.In relation to other types of activation, where social assistance or unemployment benefits are the only income available during participation, participance in the programme becomes an opportunity for unemployed individuals in the municipality, as it both provides a salary and the status associated with working.In the group interview with the participants, this was mainly reflected by participants in Intervention 1.In the quote, Mohammed and Amira described how they told their friends and family members that they were working, but also how they tipped others who could be interested.Caroline and Moa seemed more hesitant to recommend participation to their friends, and it is also unknown whether they told their friends and family that they participated in the intervention.Thus, based on the quote above, the opportunity of participating in the programme seems to have greater relevance for the refugee participants in Intervention 1.Hence, for the individual unemployed refugee, participating in a municipal labour market programme that provides subsidized employment may seem like an opportunity to enter the labour market, have a salary, and gain some social security; a 'golden ticket' as Lisa said in the quote.In that sense, opportunity hoarding as an inequality-generating mechanism emerges, where an unequal category of individuals gains access to resources that can lead to a monopolization of the resources (Tilly 1999).Here, the story of the programme becomes important.Being described as an opportunity for a category of unemployed individuals that have a disadvantaged position in the labour market also legitimizes the organizational inequality, where resources were allocated based on the division of the participants as refugees or general unemployed.

Concluding discussion: ethnicising activation as a standard story
In line with previous research (Fernandes Guilherme 2013; Gubrium and Fernandes Guilherme 2014; Hungler and Kende 2021; Vesterberg 2016), I argue that constructions of ethnicity and race matter in the framing, development, and design of welfare services, such as activation of unemployed individuals through labour market programmes.The aim of this article was to analyse how and why refugees and 'general unemployed' are activated in different interventions within a municipal labour market programme, and what implications such a division may have for the refugees being activated in the labour market programme.The analysis in the article contributes to previous research by providing insight on how the ethnicisation of refugees within the labour market programme has taken materialized forms within the organization and how it maintains an inequality that disfavours the refugee participants.In the programme the resources flexibility, time, income and access to unemployment benefits were allocated unequally between the categorical pair of unemployed refugees and 'general unemployed'.Further, the inequality was connected to how the categorized participants were offered different subsidized employments which were highly connected to the resources available for the participants.I conclude this as an ethnicising activation.
Ethnicised activation, based on this article, can be understood in relation to the categorization of refugees and general unemployed and how that categorization relates to how refugee participants receive fewer resources in the programme.However, the study highlights that it is through the stories from the manager, the social workers, and the participants that the legitimization of ethnicised activation is made visible (cf.Tilly 2006).Not least because the intervention offered to the refugee participants is highlighted as an opportunity for both employees and refugee participants in relation to other activation alternatives that are available.Thus, ethnicised activation becomes a standard story of success for the actors who participate.In turn, a success story becomes all the more difficult to contradict.That unemployed refugees should be offered a specific intervention that involves fewer resources than other unemployed individuals therefore gain legitimacy, and the myth of how refugees 'should' be activated is maintained and inequality along with it.
Participation in the labour market is an important benchmark for politicians in assessing refugees' worthiness to claim welfare, as is the case overall in the global North and specifically in the Nordic countries (Djuve and Kavli 2019;Vesterberg 2016).After the Swedish election of 2022, with a right-winged and nationalistic government, there are no indicators that this will change but rather become more dichotomized in relation to the perceived needs of the majority society.Further research is thus needed.For instance, a deeper narrative analysis could provide additional insights.There is also a need to take gender, class, age, sexuality and more categorizations into consideration to develop an understanding of how and why inequality becomes manifested in activation practices.As ethnicising activation only has been concluded based on one labour market programme, with a specific set of interventions and resources, further empirical studies in Swedish and international contexts are needed to test the analytical strength of the concept.While this study has limitations, it illustrates the institutional barriers for unemployed refugees who are striving to enter the Swedish labour market, when opportunities for refugees in activation interventions still mean less opportunities than they do for general unemployed.For practitioners within labour market programmes the study highlights the need to look beyond success stories in activation interventions and reflect upon what the interventions contain, who the target group is and how the interventions may help specifically vulnerable individuals to find paths towards the labour market.

Notes
1. Further referred to as 'the programme'.2. 1000 Swedish crowns correspond to approximately 91.7 Euros.
work 75% and paid approximately 10,500 after taxes.Moa: I work. ..well yeah, I work 50% but get salary for 100% because I get paid to study, so I get 19 [thousand] before taxes. . .(Hussein & Amira, intervention 1, Moa & Caroline, intervention 2) Interviewer: Have you told family and friends that you work and what you do?Mohammed: Absolutely Amira: Yes, my friends and I have like two whowant to work [in Intervention 1], I said go to the case officer at PES [. ..]Interviewer: do you tip your friends [Caroline]?Caroline: hm. . .well, I probably don't.[. ..]I think it's great if, for example, you want to try a job or want to learn better Swedish and such, but. ..they look for jobs on their own.Interviewer: do you tell your friends or tell friends that you work here [Moa]?Moa: Yeah. ..I guess I would tip, but they kind of have jobs so. . .[. ..] (Mohammed & Amira, intervention 1, Moa & Caroline, intervention 2)

Table 1 .
Distribution of resources between the two interventions.
PES can direct them[refugees]to this with [Intervention 1] because they are within the Establishment and there is none [within the municipality].So, it is PES that directs you there.But whether PES assigns [them] or the municipality assigns [them], then we don't start thinking about whether they would actually have something else, as this is our mission [. ..].