“We Kiss Everyone’s Hands to Get a Permanent Job, but Where Is It?”: The Failure of the Social Inclusion Narrative for Refugees

Humanitarian migrants, while required to prove their vulnerability to gain entry to a country of settlement, rapidly become subject to an integration narrative where self‐sufficiency is the primary aim. In the integration narrative, language learning is conceptualised as an individual endeavour that will inevitably lead to employment, while linguistic fluency and social inclusion tend to be presented as the inevitable outcomes of engagement in the labour market. Lack of success is attrib‐ uted to individual failures and is typically addressed through policies designed to incentivise the individual to try harder. Drawing on a qualitative study involving refugees, language teachers and settlement brokers in New Zealand and Sweden, this article critiques the integration narrative by contrasting it with the voices of those who have sought to conform to the ideal narrative yet failed to reach the idealised outcomes. Using M. M. Bakhtin’s notions of monologue and epic dis‐ courses, it challenges the view of language learning and integration as “a test of virtuosity” (Sullivan, 2012, p. 49) which the deserving are guaranteed to pass. Instead, it argues that a range of exclusions prevents successful language acquisition, labour market entry, and social engagement and that incentives, while potentially increasing the individual’s desire for success, are insufficient unless structural inequalities are addressed.


Introduction
The process of receiving or resettling humanitarian migrants is subject to contradictory expectations of verifiable deservingness. On the one hand, to prove their need for asylum or third-country resettlement, individuals are expected to demonstrate considerable vulnerability and are authored as victims and as powerless; on the other hand, once granted residence, they rapidly become subject to an integration narrative where selfsufficiency is the primary aim and where failure becomes a marker of unworthiness (Darrow, 2018). Social inclusion for humanitarian migrants is all too often framed in terms of a duty to integrate rather than as a right to access markets, services and spaces (World Bank Group, 2014). This is particularly salient in terms of access to the labour market and to language education, where the duty to learn the language and gain employment is often emphasised above the human right to work and to education (UN General Assembly, 1948, Arts. 23 and 26 respectively). Social exclusion is thus seen as a deficiency on the part of the individual that needs to be remedied through a range of incentives and disincentives (Lindberg & Sandwall, 2017;Morrice et al., 2021).
This article is based on research in two contexts: a post-industrial/rural municipality in Sweden (Nyfält) and a designated resettlement area for quota refugees in New Zealand (Jonestown). Sweden and New Zealand were selected as they are both refugee-receiving firstworld countries with clear integration programmes but with significantly different histories of migration and refugee reception. Sweden is historically a country of emigration, conceived of as culturally homogenous, and with a traditionally generous asylum policy that has become increasingly restrictive since 2015 (Migrationsinfo, 2023;Stern, 2019). New Zealand, on the other hand, is a settler country, officially bicultural, and with a historically very restrictive refugee reception, which has only recently started increasing (Immigration New Zealand, 2023a). While the contexts have significant differences, there is also a range of similarities in terms of how refugees are perceived, the barriers they face to social inclusion, and how the journey to self-sufficiency is conceptualised. Both contexts have certain narratives of what "integration" looks like and presume a straightforward process for newcomers who possess sufficient motivation and desire to be included in their new social context. Both contexts also have a range of barriers that complicate this process and an apparently limited understanding of these among policymakers.

Refugee Reception in New Zealand and Sweden
While New Zealand has significant numbers of immigrant arrivals, both permanent and temporary, refugees constitute a very small part that is nevertheless significant in terms of settlement needs (Mortensen, 2011). New Zealand receives primarily UNHCR-approved quota refugees, with a yearly quota of 1,500 (Immigration New Zealand, 2023a) and an additional 600 under the family support category (Immigration New Zealand, 2023b). A small number of asylum seekers-generally fewer than 150-are also accepted yearly (Immigration New Zealand, 2023c). New Zealand's quota is small partly to be able to incorporate categories of refugees who may require additional support during settlement, such as women at risk, medical and disabled cases, and cases requiring special protection or additional support (Beaglehole, 2013;Marlowe, 2018). The controlled nature of refugee resettlement in New Zealand means that individuals and families are allocated to one of thirteen resettlement locations (Immigration New Zealand, 2023a) and are often settled with members of the same linguistic and ethnic background in order to facilitate support.
In contrast to New Zealand, Sweden has traditionally been a country with very generous asylum policies, but significant changes have taken place in the years since 2015 and policies are becoming increasingly restrictive. Generous asylum policies and conditions for settlement meant that large numbers entered Sweden over the years (Hagelund, 2020), culminating in 2015 with a total of about 160,000 asylum seekers (SCB, 2023). The large numbers led to revised migration laws (Stern, 2019) and in 2022 the numbers had decreased to about 9,000 quota refugees and asylum seekers with just under 3,000 family reunification cases (Migrationsinfo, 2023). The current government is reducing numbers further, with the quota decreasing from 5,000 to 900 and fur-ther restrictions on asylum seekers to be implemented (Tidöavtalet: Överenskommelse för Sverige, 2022).
Both Sweden and New Zealand place a significant emphasis on labour market participation for refugees (Lindberg & Sandwall, 2017;Marlowe et al., 2014) although Sweden views education as a necessary precursor to employment (Hernes et al., 2022). A high threshold to employment in Sweden means that labour market entry can be delayed (Hernes et al., 2022), but outcomes improve steadily the longer an individual resides in the country. While employment rates are only at 45% for refugees who have been in the country for 0-9 years, these increase to 65% after 10-19 years and 80% after 20 years or more (SCB, 2021). Outcomes are impacted by a range of factors, including gender, age, area of origin, and earlier access to education (SCB, 2021). In the New Zealand context, an expectation of rapid labour market integration is reflected in the frequent monitoring of results, which occur one year, two years, three years, and five years after settlement. However, time is still an important factor and only 10-18% of working-age refugees are in paid employment after one year in the country, while after three years the figures increase to 30-40% and 40-50% at the five-year mark (MBIE, 2021). Outcomes are impacted by genderwith women being less likely to be engaged in the labour market-and country of origin, with refugees from Iraq, Myanmar, Syria, and Afghanistan finding it most difficult to gain employment (MBIE, 2021).

Understanding Social Inclusion, Language, and Employment in Refugee Settlement
Social inclusion for refugees, in the form of access to the labour market and language education, has received increasing academic attention in recent years but remains comparatively under-researched, and undertheorised (Garnier et al., 2018;Lee et al., 2020;Morrice et al., 2021). While learning the host language has been demonstrated to be essential for self-sufficiency, well-being, and integration (Blake et al., 2017;Morrice et al., 2021;Tip et al., 2019) there is a need for a greater understanding of language learning for adult refugees and how various countries' policies may impact learning opportunities (Morrice et al., 2021). Similarly, while a range of studies has been undertaken on refugees and workforce integration (see Lee et al., 2020), significant gaps remain. Lee et al. (2020) suggest that there is a need for further investigations that take into consideration multiple levels, cross-country contexts, and a range of stakeholder perspectives. Further, while there is significant evidence of refugees being subject to barriers to employment and advancement, many studies focus exclusively on individual agency and the improvement of human capital rather than providing any theorisation on structural barriers (Lee et al., 2020).
Investigating the structural barriers and the underlying ideologies that motivate them involves a critique of the idealised integration narrative which presupposes that individual motivation is the only requirement for success and that lack of success is best addressed through incentivisation. In the context of refugee language learning, incentivisation has been investigated by Kosyakova et al. (2022). The results of the quantitative study demonstrate that in the refugee context incentives have very little impact on language learning in comparison to other factors such as exposure, although the authors found that language learning was considerably improved for those being granted residence. The guarantee of continuous residence then appears to have increased the learners' investment in the language, and perhaps also their ability to invest when no longer in a state of precarity. Incentivisation in relation to labour market entry has been investigated in a quantitative study by Hernes et al. (2022) providing comparisons between Sweden, Norway, and Denmark. The authors suggest that the incentivisation model, which at the time the data was captured was only used in Denmark, leads to comparatively high employment outcomes in the short term, but decreases in outcomes over time. On the other hand, the focus on education prior to employment in Norway and Sweden delayed access to employment but appeared to lead to more stable long-term outcomes. While these studies are significant, they are quantitative studies, and the findings would benefit from being supplemented by a qualitative perspective, as also suggested by Hernes et al. (2022). Further, recent changes in Swedish policies have meant that an incentivisation model has recently been adopted in Sweden (Emilsson, 2020), and the impacts of this on individuals is an important area for exploration.
To address some of the research gaps outlined above, I have chosen to investigate social inclusion, language learning and employment in refugee settlement from multiple stakeholder perspectives, engaging with two significantly different contexts, and exploring ideological, institutional, and individual dimensions. To do this, a dialogical perspective will be adopted as it provides a framework to investigate relational as well as ideological dimensions of social inclusion.

A Dialogical Perspective
From a Bakhtinian dialogical perspective, identity is intersubjective, that is, constructed in the multiple interactions between self and other so that our sense of who we are is dependent on how others perceive and understand our identities (Holquist, 1990;Sullivan, 2012). In our interactions we not only author our identities or "create ourselves" as we address and respond to others (Holquist, 1990, pp. 28-29); we also author others and, in turn, are authored by others (Sullivan & McCarthy, 2004). In the act of being authored, subjects may be authored as valued individuals, enriched and unique, or as "impoverished" or reduced to a type without individual uniqueness and value (Sullivan & McCarthy, 2004, p. 298).
I will argue that in the context of refugee reception or resettlement, there is a tendency to author refugees as a type designed to fit into a monological, epic narrative of integration. In a monological narrative, the subject "acts, experiences, thinks, and is conscious…within the limits of his image" as defined by the author (Bakhtin, 1984, p. 52) and is not seen as an individual with their own uniqueness. The subject is not entitled to their own truth but is subjected to the singular "truth" imposed by the author, and is also subject to "a reification" to fit into a market-oriented understanding of reality (Bakhtin, 1984, p. 62). In the epic narrative, the objectified subject also becomes the hero of a predetermined plot. Their character will be tested through challenges where success is always guaranteed if they have the character of a hero and "pass a test of virtue" (Sullivan, 2012, p. 46). Therefore, an epic discourse does not contain any space for debate as to the validity of the quest or the fairness of the obstacles (Sullivan, 2012).

The Epic Language Learner
An epic construction of the "language-learning other" provides a narrative that precludes any recognition of structural constraints or economic inequalities and, instead, attributes success or failure solely to the effort and motivation of the self-actualising individual. This reductive view of the language learner fails to recognise how formal language learning is impacted by unique individual circumstances including age and gender (Morrice et al., 2021), physical or emotional health (Field & Kearney, 2021), and earlier educational disadvantage (Field & Kearney, 2021;Morrice et al., 2021). It also fails to take into consideration the sociocultural aspects of language learning (Rydell, 2018) and the less-thanfavourable conditions for informal language learning that refugees encounter in society and the workplace (Piller, 2016). Viewing language learning as a personal responsibility and attributing slow progress to individual motivational factors, leads to the implementation of solutions that do not address affordances for language learning, but individual responsibility and motivation. Campaigns aimed at newcomers assume that they do not understand the value of acquiring the dominant language (Piller, 2016), and measures to promote language learning often include financial incentives or punishments even though research suggests that refugees rate language learning as a high priority (Lindberg & Sandwall, 2017;Morrice et al., 2021).

The Epic Job Seeker
While entry into the labour market is a high priority for refugees (O'Donovan & Sheikh, 2014), it also tends to be used as the exclusive measure of social inclusion by policymakers and politicians (Lindberg & Sandwall, 2017;O'Donovan & Sheikh, 2014). The epic job seeker narrative is evident in both contexts of the study through a policy-level focus on rapid labour market integration as well as practices designed to incentivise entry into work, even at the expense of language learning (Cooke, 2006;Lindberg & Sandwall, 2017;Morrice et al., 2021).
In New Zealand, which has a long history of immigration to meet labour market demands (Peace & Spoonley, 2019), the initial resettlement of refugees was explicitly designed to cater to labour market needs (Beaglehole, 2013). While humanitarian criteria are currently predominant in the selection process, the New Zealand Refugee Resettlement Strategy (Immigration New Zealand, 2023d) focuses strongly on rapid entry into (any) employment and on refugees rapidly decreasing their needs for government support (Marlowe et al., 2014). In addition, refugees are also subject to welfare initiatives that favour a work-focused approach, with high expectations of work readiness and sanctions for welfare beneficiaries not meeting the requirements placed on job seekers (O'Donovan & Sheikh, 2014).
In the Swedish context, integration has increasingly come to be measured in terms of employment rates (Lindberg & Sandwall, 2017) and engaging in employment is framed as a duty for the reluctant job seeker to realise. Thus, Swedish for Immigrants (Sfi) courses have increasingly become "labour market tools" (Lindberg & Sandwall, 2017, p. 124), and participants have been required to divide their time between language learning and work experience placements. Unremunerated engagement in the workplace is expected to yield significant benefits for the individual, though empirical studies in the Swedish (Lindberg & Sandwall, 2017) and Danish (Arendt & Bolvig, 2020) contexts have demonstrated that students have extremely limited interactional opportunities in these settings. In addition to these measures, permanent residence, as well as family reunification, is now contingent on sustainable financial self-sufficiency, meaning that there is considerable pressure to gain long-term employment in order to secure long-term residence and to bring one's spouse and children (Emilsson, 2020).

Methodology
The data that forms the basis for this article was collected through qualitative, semi-structured interviews in New Zealand (2021) and Sweden (2022). Interviews were conceived of as dialogical interactions between interviewer and interviewee, embedded in the broader sociocultural context (Tanggaard, 2009), and included open-ended questions that allowed the interviewer to follow up on matters of interest, but also allowed participants to redirect the conversation to topics of interest to them. Because of the potentially emotional nature of the interviews, the interviewer and interpreters prioritised the well-being of participants by retreating from topics they were reluctant to speak about (e.g., the refugee journey), and redirected the conversation when required. Interviews were conducted in either Swedish or English, with interpreters as required or requested by participants. Interpreters in New Zealand were selected based on recommendations from the settlement support agency, and in Sweden by contacting interpreters through an open database until a suitable interpreter was found. Criteria for selection included availability for face-to-face interviews, experience, and ability to deal with a range of dialects (specifically the Arabic interpreter). Aside from interpreting, the interpreters also assisted with cultural brokering before, during and after the interviews, and additional cultural advice has been sought from members of ethnic and religious communities as required. Data was transcribed by the researcher, focusing on lexical and syntactic features. Analysis was done in English and Swedish, and quotes included in this article from the Swedish data set are translated by the researcher. A full ethics application was submitted prior to data collection and was approved by the Massey University Human Ethics committee. In addition, ethics approval was sought and obtained from the Migration Research Working Group at Red Cross in New Zealand and by the Swedish Ethical Review Authority.
Data are based on interviews with a total of 85 participants who were approached through educational institutions, ethnic organisations and agencies. Participants included former refugees ( = 56), language teachers and tutors ( = 14), and individuals working in roles supporting refugee settlement ( = 15). Refugee background participants in New Zealand included Afghan, Rohingya, and Palestinian refugees who were either quota refugees or had arrived under the family reunification category. In Sweden, they were primarily Syrian and Eritrean refugees but also refugees of other Arabic-speaking and African backgrounds and included quota refugees, family reunification cases, and asylum seekers who had been granted asylum and now were Swedish residents or citizens. Educational backgrounds ranged from no formal education to university degrees, and participants' reported proficiency at the time of the interviews ranged from pre-A1 to C2 on the Common European Framework of Reference (CEFR) in English or Swedish respectively. There were no exclusion criteria, but as the main vehicles for recruiting participants were language schools and settlement organisations, many were in their early stages of language learning and/or settlement. No participant had been settled for more than 10 years. All names of participants and places are pseudonyms and ethnic and gender identifiers have been used only where relevant to understanding the data.

Dialogical Data Analysis
To understand data, I draw on a dialogical analysis that is both semantic and structural. Semantically, I have chosen to adopt a method that looks for the dialogical processes expressed in interview data, as well as the generalised or larger-scale dialogical interactions taking place outside the interview context. Using the NVivo software, semantic analysis was undertaken through a coding methodology that is based on the grounded theory method (Bryant & Charmaz, 2007) but deviates from this in that it adopts a developed theoretical perspective as a guiding framework for analysis. Like GTM, coding was done inductively, looking for processes and actions and connecting these together into larger processes and concepts (Charmaz, 2006). Once higher-level categories were created, these processes were then analysed dialogically. The semantic analysis was supplemented by dialogical narrative inquiry, paying attention to the way utterances are always in response to the positions and utterances of others, both present (e.g., the interviewer) and absent (Holquist, 1990) thereby discovering the discourses that have influenced the responses of the participants.

Findings
Findings from the interviews indicated that refugee background participants were familiar with dominant epic narratives, and resented these, instead describing themselves as highly motivated both as language learners and as job seekers, but facing a range of obstacles and difficult choices. The epic construction of integration was not only viewed as a misrepresentation but it also became clear that incentivisation measures did not foster rapid social inclusion but often had the opposite effect. Although many of the themes and processes were common to both the Swedish and New Zealand contexts, interviews in the Swedish context offered particular insights into a pathway to sustainable employment that was lengthy, inflexible, and disrupted, while the New Zealand interviews provided insights into rapid labour market entry and its consequences.

Desire and Motivation
In both contexts, refugee background participants were aware that they were authored as reluctant language learners and job seekers, and contested these representations. Language was described as "key" and language learning as common sense: Since we came to this country, learning the language of this country is important. (NZ) Do you think there is any person who doesn't want to speak the language? (SW) Language was seen as a means of social inclusion and as a requirement for a viable dialogical identity in the new context. Even when language support was available from professional interpreters or the first language community, individuals desired the independence arising from competence: I have to speak from my own tongue. It's better, it's reliable, it's convenient. (NZ) I want to, without interpreters, speak myself. (SW) However, the desire to speak did not necessarily lead to proficiency. Participants quoted a range of barriers, including age, family responsibilities, trauma, transnational responsibilities, health, and lack of prior education and/or literacy.
The desire for social inclusion and a viable dialogical identity extended to a desire for employment. At the most basic level, engaging in employment was a way of authoring a contributing identity, as opposed to the assigned identity as deficient: We're refugees but we're just the same humans as you guys, like, you helped us, we can help you guys. (NZ) I want to become someone. I don't want to be a problem. I want to show that we, we came here to be like ordinary people, to help, to integrate in society. (SW) Individuals who valued their identity as a hard worker, struggled particularly with not being able to realise this identity in a new context: I like working. When [I] have work, I am happy. When not have work I not happy….I understand working.
The desire to work was clearly also conditioned by financial need, as financial resources were seen as indispensable for social inclusion, and particularly crucial in assisting their children's inclusion: They play sport and they like to be, they're trying to be like [New Zealanders], like others. And they need more label clothes. Of course it's optional, but they don't think it's optional. (NZ) Importantly, establishing employment as a desire, refugee background participants often framed work not as a duty but rather as a right that they were denied: And they say that refugees do not want to work, but this is not correct. They come here to work and provide for their families….They come to find security. To find a good life. But they don't find security, not a good life. They couldn't find a job. And it's-all humans have the right to work. (SW)

New Zealand: Doing the Jobs That Nobody Wants
Data from the New Zealand interviews exemplified clearly the ideology of rapid entry to the workplace and that refugees should assist in addressing the labour shortages in New Zealand rather than seek careers leading to social mobility. This ideology was so embedded that it was used to promote a positive narrative of refugee reception by language teachers and support workers: Yeah, I'd say they are happy to do any menial task that we think is beyond ourselves and won't do them.
[People say] they are taking Kiwi's jobs and that's when I can at least say, well, that's absolutely rubbish.

Because Kiwis aren't taking their own [entry-level] jobs
It was also clear to refugees, who contested it: So what I am seeing, about the refugee people didn't have perfect education. [Some organisations] like to use these people to work on, like, part time jobs, season jobs, and [entry-level] jobs.
For those intending to enter skilled or semi-skilled work, there was a range of barriers, even for those with qualifications and professional experience. In Jonestown existing networks were a strong factor in obtaining employment, and participants also reported that aspects that made you appear culturally different in your application and interview were likely to impact your success, including your name, your overseas work experience (depending on the country), cultural differences in interactions, and whether you were wearing a hijab. Improving career prospects through undertaking further study was primarily complicated by the financial aspect, and choosing between the obligation to provide for family members (on location or abroad) and personal ambitions: That's the moral conflict that I went through….Do I go for my labour job, like being a waitress forever for the rest of my life? Do I go for a social work degree where I have to compromise many days of the week just sitting in school and I wouldn't be able to work?
This was aggravated by a rigid student allowance system, limited to 120-200 weeks, depending on age (Ministry of Social Development, 2023). If some of this allowance was required for language study, this could limit the time available for mainstream tertiary study: For language learners, the student allowance could only be utilised for courses that led to nationally recognised qualifications and therefore had higher requirements on literacy levels and academic ability. For other English courses, students remained on the Jobseeker Allowance, subject to obligations designed to promote rapid financial self-sufficiency. This, combined with a poor understanding of how long language acquisition for adults may take, meant that language lessons could be curtailed by social welfare employees: And they say, oh, you already two years, you already three years, you no need to learn anymore. You have to work. If you [do] not work we'll stop your benefit.
With labour shortages and employers willing to take on workers with very low English proficiency (CEFR pre-A1), Jobseeker beneficiaries with limited literacy and language could be offered employment, which they were then obligated to accept. Once in entry-level employment beginner users of English found it difficult to improve their work situation as there were few pathways available and incentivisation measures prevented workers from re-engaging with language learning. A work broker explained: We get people and they're pushing us to get them into work, so we get them into you know that entry-level work, be it horticultural or other sectors, labouring type [of] roles. But there's no avenue later down the track when they come off the benefit to be re-engaging in stair-casing these people into other roles.
Progression within one workplace was complicated by the fact that there were few in-work opportunities for literacy and language development and most other inwork training opportunities relied on written materials in English.
To discourage workers from abandoning or losing their employment, the social welfare agency had a 13-week stand-down period before the Jobseeker benefit could be recommenced (Work and Income, 2022). There was therefore little possibility for leaving employment to re-enter English study unless you were able to undertake the more demanding courses leading to national qualifications. With limited understanding of systems and processes, and limited access to information, not all individuals understood their work obligations, and particularly the 13-week stand-down period. This excessive penalty then affected them and their families: If you stop their benefit thirteen weeks, how the family want to eat and survive? If you say two weeks, three weeks, we can understand, oh, this is my wrong, my fault….If you stop the family thirteen weeks, better you give the family poison.
For those who entered the workforce with lower levels of language and limited literacy, there was also little opportunity for informal language learning in the workplace. To manage the language challenges employers would hire groups of people from the same linguistic background and appoint the most proficient speaker as the communicator for the group, limiting workplace exposure to English: They are constantly using their own language at work and at home and they're not getting any opportunity to develop their language or understanding in a social-you know, work and social environments. (Work broker) With a limited command of English, they also struggled to access their rights as employees: One said, oh, my arm's a bit sore. Just, you know, that sort of thing….But for them to actually say, look, this is RSI, you know, understanding what their rights are, I don't know. (Language teacher) For this group, engagement in the labour market facilitated financial inclusion, in that workers were better able to provide for themselves and their families and become taxpayers, but not social inclusion in the broader sense. Rapid entry to the workforce instead contributed to assigning to them a particular social location with limited scope for social mobility.

Sweden: The Long Road to Employment
Despite the policy goals of rapid labour market entry, many of the Swedish participants found that there was a long road to employment, where compliance was expected but did not necessarily lead to the desired outcomes.
Those with higher levels of education and professional backgrounds, while able to gain employment, were often barred from working in their original field. Qualifications gained abroad were not accredited as equivalent to Swedish (or EU) qualifications, and work experience abroad was not ranked very highly: I was a computer engineer and I have earlier…really strong experience. With many international firms, companies, like Dell, HP, eh Cisco, yes, I have strong experience, I have a very strong CV….
[But] if you have nothing, education from Sweden, you are zero.
There appeared to be a general expectation that professionals should be willing to re-evaluate their options. Examples included teachers becoming teacher aides, university lecturers becoming childcare workers, journalists becoming bilingual tutors, and IT experts working in administration. Even then, retraining to fit into the labour market could be time-consuming due to the rigorous education pathways, as was the case for this university-educated computer scientist: And she studies from the beginning. She studies supplementary high school courses, studies childcare, so she loses three years through studying. And now finally she's working as a child carer.
Those who had arrived in Sweden with little or no previous education, and often no or limited literacy, had to deal with even greater obstacles. They faced a labour market that was tightly regulated and highly technical, with a lack of "simple jobs" and high expectations on language and education even for entry-level work: We have incredibly high levels of education in Sweden. And then if you are illiterate, as many are, which we don't have in Sweden, the politicians don't get how long it takes to learn this language in order to take a course, in order to get a job….If you are going to be a cleaner you still need to understand pretty good Swedish because you then need a course. (Language teacher) Policymakers and bureaucrats lacked an understanding of the needs of this group, as exemplified in the centralisation and digitalisation of the state employment agency: They have transferred to make it digital services. And that you apply for work digitally. But a person who is illiterate from Somalia doesn't handle digital services, so right there, integration has failed. (Integration facilitator) Individuals who genuinely wanted assistance in obtaining work were frustrated with only receiving directives and no actual support: Not working. Employment agency: You must work. Which work?…I must myself [find] work….I don't know work. (Refugee) For these individuals, a significant period of language learning could also be necessary before they were work ready, which was difficult for policymakers to appreciate. In addition to the difficulties associated with age and lack of earlier education, many also lived segregated from fluent Swedish speakers: Actually, do our students study Swedish as a second language, or are they that segregated from the Swedish that they rather study Swedish as a foreign language? (Language teacher) Even for those with Swedish neighbours, interactions were difficult, as these generally politely declined opportunities for interaction: When I invite my neighbour, please, drink coffee, yes, again, again, please, I want to. "No, no, thank you very much" (Refugee)

Incentivisation
While maintaining the high threshold to labour market entry, recent policies have sought to promote rapid entry into sustainable employment through the implementation of a number of incentivisation measures and projects. However, findings suggest that rather than facilitate labour market inclusion, these measures served to discourage individuals and often prolonged the road to sustainable employment.
Permanent residence and family reunification were both contingent on individuals becoming sustainably self-sufficient-in practical terms, gaining permanent employment. Seasonal work, subsidised employment, and fixed-term contracts did not count in this context, which meant individuals could be reluctant to take these positions. Additionally, the precarity experienced by those on temporary residence permits impacted their language learning motivation and focus: Often [motivation] is connected to a goal or a belief in the future…and if then you only have one or two years why on earth would you learn this complicated little language that is only useful for you in this tiny spot of the earth? (Language teacher) The impact was perhaps the greatest for those who were still trying to qualify for family reunification, and whose spouses and children therefore lived in a state of long-term precarity elsewhere: We have participants who haven't seen their family for seven [to] eight years. Wife and children, only kind of Facetimed sometimes. And some have financial support yes send money to the family and are about to starve to death themselves. Of course it is a challenge to study and move forward. (Work broker) Much thinking, what shall I do, every day call to me children. "Daddy, daddy," they say. So you think about it. So you can't sleep. What shall I do? You think a lot. (Father with wife and children in displacement in Sudan) The focus on rapid self-sufficiency also meant that a range of initiatives was carried out to create exposure to the labour market through work experience, seasonal work, various work schemes, and subsidised employment, in which the individuals often had limited choice. While there were examples in the data where these had been successful and lead to permanent employment, they were more often seen as disruptions to the language learning journey: Then they place them in some work experience or some job or something and then they come back…and then they have lost a lot and so we start over again. So yes, it's a bit of a never-ending story with them. (Language teacher) Many who worked in workplaces with others who spoke the same first language struggled to retain what they had already learnt, and even when working in a Swedish-speaking environment, linguistic development was often limited: They've maybe learnt some of those phrases that make them understood, because you pick that up…"I can communicate what is needed to cope with the work here." Yes, but maybe you can't cope outside of your workplace. (Language teacher) Unpaid work experience, which again individuals were obliged to accept, was particularly contentious, and interpreted as exploitation: There are many who are exploited. They go to work experience and are promised to get a job…and when time is up, out they throw them, and say, no, we cannot afford it, we cannot hire. So you are exploited.
As the temporary financial support is contingent on compliance with prescribed activities, compliance is high. However, compliance did not guarantee sustainable employment: And now I'm on activity support from early in the morning until the time I arrived [4 PM] every day and it's work without pay….We kiss everyone's hands to get a permanent job, but where is it?

Acknowledging the Uniqueness of the Individual
While the intended solutions in the shape of incentives and interventions did not produce the desired results, findings suggested that results may be obtained through a different approach-one that acknowledged the uniqueness of the individual and focused on interpersonal contact.
Nyfält municipality, after becoming discouraged with the state employment agency, set up their own employment unit, where they work with refugees, employers and trainers to ensure each individual finds a pathway that is achievable: The biggest difference is that we work very close to the individual In Jonestown, the Pathways to Employment organisation filled the same function, brokering between refugees, employers, and training providers. However, they also felt that further work was needed for individuals to continue to develop: It would be nice to know that we could at some stage…take them out and use some funding to train them into better long-term employment The need to understand and work with the individual was also echoed by refugees and language teachers who spoke of the need for tailored approaches to enable professionals to remain in their fields or to find realistic pathways for those with lower levels of education and literacy.

Discussion and Conclusion
The current study suggests that refugees are subject to epic, monological narratives that portray the journey towards social inclusion as straightforward and contingent on the compliance and virtuosity of the individual, but that these narratives, and the ensuing processes, are contestable and, indeed, contested. The refugee background participants authored themselves as motivated language learners who desired to access the labour market and develop an identity as fully participating subjects. However, their efforts were thwarted by a range of barriers, as well as by the policies brought in to incentivise and fast-track their entry into the labour market. In the New Zealand context, the expectation and incentivisation leading to rapid labour market entry could confine refugees to a particular social location with limited possibility of social mobility and limited access to language learning opportunities. In the Swedish context, the failure to understand diverse needs could lead to a lengthy pathway to sustainable employment, which was often disrupted by projects and schemes that did not ultimately benefit the individual. Additionally, incentivisation relating to residence and family reunification caused considerable distress for individuals and impacted negatively on their progress.
The findings support earlier studies that have indicated that language learning provision has increasingly become a tool for labour market integration (Cooke, 2006;Lindberg & Sandwall, 2017;Morrice et al., 2021) and that rapid entry to employment may negatively impact language acquisition and social inclusion (Piller, 2016). They also further exemplify the high levels of motivation among refugees (Lindberg & Sandwall, 2017;Morrice et al., 2021) as well as the lack of understanding among policymakers of the diverse needs of language learners (Field & Kearney, 2021;Morrice et al., 2021). Further, they confirm that incentives that increase precarity lead to a decrease, rather than an increase, in motivation and focus (Kosyakova et al., 2022;Scarpa & Schierup, 2018). Finally, they extend O'Donovan and Sheikh's (2014) argument that the solutions appear to lie in one-on-one support and interaction with individuals, rather than in high-level measures of incentivisation.
The dialogical perspective adopted in this article serves to further illuminate the ideology that underlies social inclusion policies for refugees but also hints at possible solutions. Social inclusion is unlikely to be successful as long as one group is treated as distant, "otherised" subjects, without individual form and uniqueness and without voice in their own destiny. In the words of Bakhtin (1984, p. 58): A living human being cannot be turned into the voiceless object of some secondhand, finalizing cognitive process. In a human being there is always something that only he himself can reveal, in a free act of self-consciousness and discourse, something that does not submit to an externalizing secondhand definition.
Thus, engaging dialogically with unique individuals, working one-on-one and creating viable pathways, is likely to produce greater social inclusion and employment outcomes than financial and social incentives and disincentives. This would also require an acknowledgement of language learning and employment as human rights rather than as duties that the individual is obliged to comply with.