Underemploying highly skilled migrants: An organizational logic protecting corporate ‘normality’

Why do highly skilled migrants encounter difficulties getting a skilled job? In this study, instead of searching for an answer in migrants’ characteristics, we turn to organizations and ask: why do organizations underemploy highly skilled migrants? With an in-depth qualitative study of a programme for highly skilled migrants’ labour integration in Sweden, we show that highly skilled migrants are perceived as a potential threat to organizational norms and practices. Using the relational theory of risk – approaching risk as socially constructed – the study provides a novel explanation for highly skilled migrants’ underemployment. It shows an organization logic protecting corporate practices seen as ‘normal’ from a perceived disruption that employing highly skilled migrants could possibly cause. Theoretical contributions to the understanding of highly skilled migrants’ employability are threefold: (1) the field assumption that organizations are in favour of hiring migrants is challenged; (2) highly skilled migrants’ underemployment is explained through a protective organizational logic; and (3) we stress the necessity to problematize an implicit reference to organizational normality when recruiting.


Introduction
Despite Europe's shortage of labour market skills (EU Skills Panorama, 2014), tertiaryeducated migrants and those with extensive professional experience (Al Ariss and Syed, 2011) usually find themselves in one or both of two situations: finding a job in their host countries is difficult or they are not able to utilize their full potential once employed (OECD/European Union, 2015). Consequently, the highly skilled migrants are more likely to be 'working in a job that is below the employee's full working capacity' (McKee-Ryan and Harvey, 2011: 963) than workers born in the host country. Multiple explanations for this situation have been proposed, citing macro-level situational barriers (e.g. Aten et al., 2016) or, more commonly, providing micro-level explanations based on individual characteristics, such as, available capital (Al Ariss and Syed, 2011), linguistic disadvantage (Ogbonna and Harris, 2006) or the migrant's career choice (O'Connor and Crowley-Henry, 2019).
Our study contributes to this conversation about migrants' underemployment by considering the neglected mezzo-level of organizations (Hajro et al., 2019;Tharenou and Kulik, 2020). We shift the focus from the current micro-level emphasis on the individual (why are highly skilled migrants underemployed?) to a mezzo-level, in which we ask: why do organizations underemploy highly skilled migrants?
To answer this question, we use the case study of a Swedish (civil society) organization, anonymized as 'Diversity for Profit'. As part of its activities, Diversity for Profit runs a mentorship programme aimed at strengthening the employability of highly skilled migrants in the Swedish labour market. Sweden faces challenges of skills shortage (Ahlberg et al., 2019) as well as the underemployment of highly skilled migrants (Eurostat, 2019;Stirling, 2015), both of which are experienced in Europe more widely. The unemployment rate for this group of migrants is four times that of those born in Sweden with similar education and approximately 45% of highly skilled migrants are underemployed compared to 20% of highly skilled workers born in Sweden (Statistics Sweden, 2019).
Our in-depth qualitative study shows that a novel explanation for migrants' underemployment can be found in the narratives of mentors, mentees and Diversity for Profit employees. Current research from an organizational perspective offers two broad explanations. The first is that migrants are subject to discrimination in society, which is unreflectively reproduced in organizations (see Ghumman and Ryan, 2013;Turchick Hakak and Al Ariss, 2013). The second is that intrinsic organizational limitations lead to involuntary discrimination in recruitment (e.g. Almeida et al., 2019;Derous et al., 2017). In other words, research rests on the assumption that organizations are willing to hire migrants and if underemployment is taking place, this is due to other factors. We challenge this taken for granted field assumption (Alvesson and Sandberg, 2011) by considering the possibility that organizations may lack the incentive to recruit migrants or that they even have reasons for underemploying them. An organizational basis for the underemployment of migrants is precisely what is found in the narratives of the interviewees, pointing to a believed organizational harm. Using the theoretical framework of the relational theory of risk (Boholm and Corvellec, 2011;Corvellec, 2010), we show that the narratives implicitly portray migrants as a threat to local organizational practices, to what is seen as 'normal', and therefore, underemploying migrants can be a way for the organization to protect itself from this perceived risk.
This study contributes to the literature in at least three ways. First, it provides an understanding of the underemployment of highly skilled migrants by offering a mezzoorganizational argument: that the phenomenon can be explained as a response to the organizational logic of protection from a risk. The relational theory of risk (Boholm and Corvellec, 2011) therefore enriches our understanding of employability by questioning the taken for granted assumptions that organizations want to hire migrants. It shows that when migrants are perceived as a threat, organizations have grounds to underemploy them. Second, it provides a theoretical contribution to the nascent literature on the ways in which migrant employability is shaped by organizations (Hajro et al., 2019;Tharenou and Kulik, 2020). We argue that employability -the ability to gain, maintain and develop in employment -is also dependent upon organizational aspects and we propose the addition of a new organizational factor: organizational practices that extend beyond organizational size, the nature of the organization's clientele or features of the recruiters. We posit that organizational practices shaping everyday normality at work influence how migrants are narratively constructed as a potential threat to this normality. It is not merely the individual characteristics of migrants that shape their employability (see Forrier et al., 2018;O'Connor and Crowley-Henry, 2019); it is also the organizational practices of the employing organization. Finally, the study further provides the theoretical contribution of problematizing (Alvesson and Sandberg, 2011) the role of organizational normality in a recruitment process. When current research points to involuntary bias and organizational limitations in the recruitment of skilled migrants (e.g. Almeida et al., 2019;Crowley-Henry and Al Ariss, 2018;Ghumman and Ryan, 2013;Ogbonna and Harris, 2006), we show that this normality is much more than a simple state of affairs to which the candidate is expected to fit. Rather, it is a valued norm that leads to the reproduction of structural discrimination.

Migrant employability: Normality as an implicit aspect
Research on employability is conducted from one of three broad perspectives (Thijssen et al., 2008): societal, organizational and individual. Researchers adopting the societal perspective are concerned with full employment and study migrants' access to jobs in view of labour market structures and policies (e.g. Kogan, 2016). The organizational perspective primarily addresses the effective management of human resources (HR) and its researchers study the support for a better organizational integration of migrants (see Zikic, 2015). The individual perspective is by far the most prominent in today's literature on employability (see Thijssen et al., 2008;Van Harten et al., 2017). Its proponents are concerned with individuals' characteristics that link them to their jobs; for instance, the study of migrants' identity work, when they are confronted with the experience of misidentification (see Fernando et al., 2020).
In recent decades, an employability discourse has developed in society and academia alike, shifting the responsibility for employment from the state (e.g. education) and employers (e.g. professional training) to individuals (Koyama, 2013;Mancinelli et al., 2010;Roets et al., 2012). Individuals do not evaluate their own employability, however.
Employability is ultimately assessed in relation to an organization: the organization willing to recruit and promote the individual. Yet, few studies examine the relational and contextual nature of employability (see Forrier et al., 2018).

The role of organizations in (co-)constructing migrant employability
Organizations play two key roles in the construction of employability for migrant workers. They support the migrants shaping their employability and they assess and sanction relevant criteria for employability. With regard to support, civil society organizations or local employment offices can help migrants present their skills and competences in forms that employers can easily recognize. Garsten and Jacobsson (2013) underline the notion that individual employability is based on ideas of normality, such as the conventions guiding the content of a CV or cover letter. Activities to develop employability are aimed at shaping individuals or their presentation of skills and competences as close as possible to this standard. These activities may include degree validation, mentorship programmes or internships. Thus, migrants' relationships to what is perceived as normal is part of an organization's assessment of their employability (Garsten and Jacobsson, 2013). Diedrich et al. (2011) investigate how Swedish municipalities and public and private organizations assess the competencies or skills that the migrant has acquired abroad, with the objective of building a qualification portfolio demonstrating employability to local employers. They stress that there are unintended consequences of the validation process because it has the potential to perpetuate existing lines of discrimination, leading to underemployment. Perceptions of the individual's foreignness, for example, can result in evaluators disregarding experiences that are not aligned with their perception of -in this case -Swedish standards and characteristics. Diedrich et al. (2011) also found that local language competence becomes increasingly decisive for employment, regardless of the type of job and tasks involved (see also Ogbonna and Harris, 2006;Turchick Hakak and Al Ariss, 2013). They show how the process of validating foreign skills is a process of classification, 'in which "Swedishness" is implicitly valued and "non-Nordicness" is marked as deviancy' (Diedrich et al., 2011: 288). Similarly, as Dahlstedt and Vesterberg (2017) highlight, in programmes aimed at the labour inclusion of individuals with foreign backgrounds, ability and willingness to adapt to the norms of Swedishness is deemed essential for employability. In particular, racialized migrants are obliged to become employable by embracing certain values -values of the hardworking citizen and the free, independent individual -considered key characteristics of the ideal citizen (see also Vesterberg, 2013). These expectations of adaptation to the local culture -to what is seen as normal -are also found in mentorship and internship programmes for migrant labour integration (see, for example, literature reviews in Bergström and Omanović, 2017;Reeves, 2017).
Organizations play a second role in the employability of migrant workers: they assess and sanction what are considered relevant criteria of employability. Migrants are often subject to different evaluation from that of non-migrants in their recruitment, selection and career advancement (Guo and Al Ariss, 2015;Hajro et al., 2019;Lee et al., 2020). Studies such as that of Holgersson et al. (2016) stress that job applicants with a migrant background are rarely considered potential candidates for high-status positions. Many studies suggest that the human capital of skilled migrants (their education, training and accumulated professional experience) is perceived as different and poorer compared to that of local candidates (see Crowley-Henry and Al Ariss, 2018) and is therefore devalued (Almeida et al., 2019;Zikic, 2015). When migrants are recognized as potential candidates, then, they may not face an equal chance of success. Wearing religious attire or having an Arab-sounding name or darker skin tone than other candidates decrease the chance of a call-back during recruitment interviews in some countries (e.g. Blommaert et al., 2014;Derous et al., 2017;Ghumman and Ryan, 2013). Once employed, migrants tend to face fewer career prospects within the organization (see Behtoui et al., 2017;Ogbonna and Harris, 2006;Romani et al., 2019) because of human or linguistic capitals that differ from those of their colleagues. In summary, studies of the roles played by organizations in defining employability suggest that a migrant's deviation from what is considered normal appears to constitute a key aspect of perceived employability.

Organizational factors influencing the perceived employability of migrants
Research on organizational factors affecting migrants' employability is currently in its infancy (Hajro et al., 2019;Tharenou and Kulik, 2020) and researchers have primarily concerned themselves with HR support to migrants once they are employed. Research tends to focus on the aspects of employability that touch on retention and promotion (see Hajro et al., 2019;Lee et al., 2020;Shirmohammadi et al., 2018) or investigate ways that organizations can support migrants to adjust to the new work and living environment (e.g. Hajro et al., 2019;Howe-Walsh and Schyns, 2010). So although many migrants often are not even able to find employment, organizational-level studies on the employability of migrants mainly have addressed HR actions for employees. We argue that an organizational role in shaping migrants' employability is also present prior to employment, at the recruitment stage.
Recruitment studies involving migrants point to three factors influencing the organizational perception of their employability: recruiter characteristics; organizational norms and size; and the international orientation of the organization. Recruiters' tendencies to hire persons similar to themselves are observable in the recruitment of migrants. Åslund et al. (2014) demonstrate across sectors and across company size that recruiters' backgrounds have an impact on their hiring practices, such that the manager's origin is highly correlated with the origin of the recruit, even when hiring occurs from the same local labour market (see also Fernando et al., 2016). Similarly, recruiters who have been exposed to international business and cultures are more likely to hire migrants than are recruiters with little international exposure (Almeida et al., 2015(Almeida et al., , 2019. In short, migrants' deviance from the characteristics that recruiters are used to experiencing influences the perception of their employability. A second organizational aspect in assessing migrants' employability is informed by literature on person-organization fit, indicating the organizational norms against which a candidate's suitability is assessed. Horverak et al. (2013) show that only when migrants are perceived to fit the local societal norm -when they meet the condition of societal acculturation -are their (limited) differences valued by the organization. Similarly, Almeida et al. (2015) stress that migrants' cultural attributes (e.g. their accent, attire, last name) that differ from perceived normality play a stronger role than their professional qualifications in employers' evaluation of their organizational fit. Farashah and Blomquist (2020) point out that if migrants' skills make them suitable for a job, it is their perceived cultural adaptation to the host country that renders them acceptable as future employees. In other words, recruiters assess a perceived fit according to what is seen as normal in their country and organization. Employers' negative attitudes towards migrants are reduced, however, when the organization is used to employing migrants (Lundborg and Skedinger, 2016). Recruiters in organizations with an ethnically diverse clientele are also less preoccupied with the person-organization fit of migrant candidates (Fernando et al., 2016).
A third organizational aspect playing a role in assessing migrants' employability is the size and international orientation of the employer. Large organizations have more resources for screening candidates (see Almeida et al., 2015;Lundborg and Skedinger, 2016) and seem to be less concerned about migrants fitting the existing pool of employees or whether they will meet clients' expectations (Fernando et al., 2016;Lundborg and Skedinger, 2016). The international orientation of organizations also appears to matter in recognizing the benefits of hiring skilled migrants (Makkonen, 2017;Shchegolev et al., 2016).
In sum, although current studies on employability often centre on individuals (Forrier et al., 2018;Van Harten et al., 2017), we see that organizations play a significant role in co-constructing migrants' employability, in relation to a preferred implicit normality. Yet, many studies take this preference for granted and few provide theoretical support for it. In studies of recruiters, personality traits or social identity theory -with a preference for in-group -are considered (e.g. Almeida et al., 2015). At the organizational level, most research centres on corporate recognition of migrants' forms of knowledge or human capitals (Almeida et al., 2019;Cameron et al., 2019;Crowley-Henry and Al Ariss, 2018), based on the argument that, in the case of migrants, this recognition is hampered simply because these capitals differ from the organization's perception of the normal. In other words, it is assumed that organizations are willing to hire migrants, but it is the psychology of recruiters or corporate limitations in assessing candidates' potential that stop them from doing so. We argue that the expected match of migrants to local, societal and corporate normality can also be explained as an organizational logic of protection from perceived risk.

Research site
Diversity for Profit is a not-for-profit organization based in one of the most multicultural cities in Sweden. Founded in 2003, with a vision to promote stronger social integration of ethnic groups in this urban area, it has from the start been a collaboration among multiple local actors (municipality, companies, county council). The original goal of Diversity for Profit was to change public opinion and the discourse on diversity and inclusion of migrants in the labour market. At the time of our fieldwork, its goal was strongly oriented towards facilitating the labour market integration of highly skilled migrants by supporting the development of inclusive work practices through corporate education and mentorship programmes. These activities are inspired by a norm-critical perspective. Despite the relatively small size of Diversity for Profit (two full-time and two part-time staff members when we performed the study), this network organization is well known in the region of South Sweden.
Diversity for Profit is a membership-based organization with a mix of members that include private companies (accountancy or IT consultants, banks, law firms, multinational retail companies and recruitment companies) and public organizations such as municipalities, county councils and universities. Some member organizations are major employers in the region and the active membership of some of these companies is evident from Diversity for Profit's board of directors. Multiple organizational voices and needs regarding the employment of migrants are thus well represented in Diversity for Profit.
At the time of our fieldwork, Diversity for Profit offered activities such as workshops, networking events and seminars, focusing its efforts on key regional employers. Some of the workshops are included in the annual membership fee, but most of the products and services are sold to member and non-member organizations. Diversity for Profit also grants a well-publicized national award for Diversity Employer of the Year. These activities indicate its clear corporate orientation, which provided a unique opportunity to study how the employability of highly skilled migrants is understood and crafted for organizations active in multiple sectors and heavily involved in regional business activities.
Diversity for Profit's flagship activity for labour integration of highly skilled migrants is its mentorship programme, bringing together migrant mentees and mentors well established on the Swedish labour market. In its marketing materials, Diversity for Profit presents this programme as competence development for the mentors and individual development for the mentees. The 10-month programme sets up the mentor-mentee relationship and includes training sessions with 20-40 participants per programme. The programme followed in this study started with 22 participants (11 mentors and 11 mentees). The mentee sessions are focused upon skills such as ways of orienting oneself in the Swedish job market and how to apply for jobs using social media and career sites. Mentor sessions introduce the participants to their mentorship role and train them in leadership skills and the benefit of diversity at work. All mentors receive a diploma of mentorship at the end of the programme.

Mentees and mentors
Diversity for Profit selects mentees through an application process that requires a CV and a cover letter outlining the potential mentee's motivation. Selection is based partially on the participant's potential to be matched to a mentor. Most of the mentees have tertiary-level education from their home country and represent many countries -Syria, Cuba and Ukraine, to name a few. As with migrants to Sweden in general, Eastern Europe and the Middle East are strongly represented among the mentees. The mentees interviewed and observed in this study were refugees, economic migrants and migrants following or rejoining a partner. They were from a range of age groups, but most were at their early career stages or had been when they left their home country, with work experience rarely exceeding 10 years. The mentees' education was primarily in business administration (two were IT specialists), which coincides with the general Swedish trends regarding underemployment of highly skilled migrants and those with an education in social sciences (of which business administration is a part) are overrepresented (Statistics Sweden, 2019). The gender and social origin of the mentees varied, but they primarily belonged to the middle class or upper-middle class in their home countries. The mentees' Swedish proficiency ranged from beginners to those who were fluent in Swedish.
The mentors were generally employed by Diversity for Profit's member organizations. They had often been encouraged to join the programme by a previous mentor, their direct supervisor or someone above them in their organizational hierarchy. Mentors were usually at the middle-management level of their organization, although a few were toplevel managers, with work experience usually exceeding 10 years. The mentors were predominantly women over 40 years of age who grew up in Sweden and have Swedish as their mother tongue. Three of the male mentors were themselves immigrants and three mentors had been expatriates, but all were working locally at the time of the study. In the analysis, we explored whether mentors with experience of working abroad presented a different pattern in their views on employing migrants. They did not, but they all expressed a motivation to participate in the programme based on their own personal experience, expressing a willingness 'to help'. Several of the mentors were either recruiters (CEOs of small and medium-sized enterprises (SMEs)), HR professionals or middle managers involved in recruitment. In the mentor interviews, some of their responses noticeably adopted an organizational perspective about the difficulties of hiring highly skilled migrants. We believe that the identified patterns from our interviews can be linked to an organizational perspective in general rather than a specific industry or organizational culture, because the mentors were from various types of organizations. These patterns conveyed general experiences of mentors from organizations considering a migrant hire. Likewise, when mentees shared their experience of recruitment processes, they conveyed their experience of organizations in general. Thus, the case of Diversity for Profit provided three advantages for gaining a corporate perspective on employability: the variety of mentors' organizations, the various experiences of mentees and the fact that Diversity for Profit works in close collaboration with multiple corporate actors, knows their needs and has developed its programme accordingly.

Method
This case study is positioned in the tradition of interpretive research. Because our goal was to gain an understanding of actors' experiences through their own frames of interpretation, we followed the criteria for the interpretivist research process and analytical rigour (Gehman et al., 2018;Gioia et al., 2013;Symon et al., 2018).

Collection of empirical material
We collected material through over 80 hours of observations and 27 interviews with four members of Diversity for Profit, 12 mentors and seven mentees, to investigate their experience of the mentorship programme (see Table 1 for an overview of field material used in this study). All informants have been given pseudonyms to protect their anonymity. In addition to multiple participant observations at Diversity for Profit's premises conducted by the first author, organizational archival data (Board Meeting minutes 2012-2016), promotional material and information collected from the website of the organization captured the context of the programme and how it is marketed. The fieldwork began in the spring of 2015 and the main collection of empirical material (observations and interviews of participants) continued until autumn 2016. Following interpretivist criteria for analytical credibility (see Symon et al., 2018), we undertook additional collection of empirical material until the spring of 2019: interviews with the managing director of Diversity for Profit, recording of promotional activities and attendance at major Diversity for Profit activities to follow the development of the mentorship programme.
The fieldwork focuses primarily on two cohorts of the mentorship programme. The first author conducted participant observations of the second cohort and both the first and second author interviewed participants from both cohorts. All interviews with mentors were conducted in Swedish. Four of the seven interviews with mentees were conducted in Swedish and three in English. 1 The events and meetings observed were all in Swedish. All field notes were transcribed after the observations and all interviews were transcribed verbatim. Complementary information on the mentees was collected through 23 LinkedIn posts to determine who had subsequently gained employment in a skilled job.

Analysis
The analysis adopted an abductive process (Dubois and Gadde, 2002) typical of a qualitative case study, with the first inductive analytical step followed by a deductive one. We started with a thematic inductive analysis (see Gehman et al., 2018;Gioia et al., 2013) to progressively discern emerging patterns in our material. Using NVivo software, we coded and analysed the field material stepwise. Interviews and observation notes were coded inductively (443 in vivo codes) and then considered in light of archival material and website content. The focus of this first coding round was to determine how the various participants experienced the programme and to map the programme activities.
For the next step, we sorted the codes into emerging first-order concepts: the participants' experience of the programme, the mentees' situations as job seekers, the mentormentee relationship and the content of the programme, which included topics such as 'How to Write CVs' and 'The Importance of Networks'. This step led to the development of second-order themes linked to mentors' and mentees' experiences and to components of the programme (e.g. 'employability-raising activities'). This coding showed how both the mentorship programme activities and the stories about the successful migrants touched on normalizing workplace behaviour or normalizing job-seeking activities. We use the term 'normalization' in a broad sense, referring to a process of 'becoming normal'. During this step, it became clear that codes linked to migrants were associated with problems and with such notions as risk taking, harm or threat. A systematic exploration of all the problems mentioned by the interviewees led to the development of second-order constructs showing the multiple uncertainties perceived in connection with hiring migrants and how these uncertainties could affect the normal course of operations in an organization. In sum, inductive coding of the empirical material shows two strong patterns: the hiring of migrants is associated with problems and uncertainties linked to normal organizational activities and the mentorship programme activities centre on normalizing the migrants (see Figure 1).
These emerging patterns prompted us to consider the literature on perceptions of uncertainty, harm and risk, and we started the second, deductive phase of the analysis by engaging with the relational theory of risk. We considered this theory as relevant because, in contrast to positivist research traditions concerned with the evaluation or measurement of a given risk, the relational theory of risk centres on an elucidation of 'why and how something is considered a risk' (Boholm and Corvellec, 2011: 176, emphasis added). It is inscribed in an interpretivist approach to the study of risk (e.g. Hardy and Maguire, 2015;Hilgartner, 1992), consistent with our positioning. A basic tenet posits that an entity is never a risk in and of itself, but is socially constructed as such (it is considered a risk) in relation to something else -something valued. This means that in the interpretive study of risk, equal attention is placed on understanding what is perceived as potential harm along with what is seen as valuable (see, for example, Brown and Van den Broek, 2017). In our case, we wondered why and how migrants were associated with a Normalization: Migrants display local norms Normal job seeking activities

Learn how to apply for job
Mentor provides advice on how to write CV and application letters, how to follow up on an application, and how to act in a job interview. The migrants learn the importance of online platforms such as LinkedIn.

Build and use local network
Migrants are encouraged to build local networks and use them to get a skilled job.
Job references with Swedish names can help to get a skilled job.
Enacting normal workplace behavior

Learnt Swedishness through Swedish institutions
The hiring company could verify the university degree because the applicant had a Master's degree from Sweden. The migrants studied Swedish intensively for two years to perfect it.

Learn workplace behaviour
The purpose of an internship is to gain experience in a Swedish workplace. Trainee job makes it possible for migrants to prove ability to do the job. After two months of internship, the CEO (and mentor) offered the migrant a job.

Potential performance problem
The hiring company may get the 'wrong' competence.
The migrant may do a low-quality job, due to low-quality education. Foreigners' competences are often narrower.

Potential language problem
Poor Swedish may cause misunderstandings, and mistakes may be made. The customer did not want a 'foreigner'.

Potential problem to current employees
One does not want to change culture; it costs too much. One feels uncomfortable with those who are different.

Potential problem with fitting in
Everybody should feel comfortable and feel that they belong to the group.
It is important that one fits with the rest of the group. threat and in relation to what valued object. In addition, the relational theory of risk highlights the importance of a relationship (Boholm et al., 2015;Cadet and Carroll, 2019;Corvellec, 2010). This is indeed a central aspect: the necessity of having a relationship between the entity that is valued and the entity perceived as a putative threat. It is not enough that an entity is socially constructed as posing a threat; in order for the entity to become a risk, it must also be linked to or put in the presence of the valued object. Without this relationship, there can be no risk (Boholm, 2003;Boholm and Corvellec, 2011;Corvellec, 2010). We therefore deductively analysed the empirical material in view of the relational theory of risk. In this analytical step, a clear relationship appeared between the strong patterns previously identified (migrants linked to uncertainties regarding organizational activities and the programme's normalizing activities for migrants) and a novel explanation for the underemployment of migrants emerged.

Organizational perspectives on hiring highly skilled migrants
Our inductive analysis shows two main themes linked to the hiring of migrants: potential lower performance and potential disturbance to organizational normality.

Organizational performance perceived to be at risk
Interviewees outlined the organizations' problems in evaluating a migrant's degrees and competences, which made it difficult for them to know how these recruits would contribute to organizational performance. As one mentor representing top management in an IT firm said: I believe [different education standards] are one of the greatest barriers (. . .) We have noticed it with programmers from India too. Now I am talking only about the IT industry. Unfortunately, I believe their educations are not as good as ours. They are not really where we believe they are.
Participants also pointed to language proficiency as a potential problem. Less than perfect language skills were perceived as hindering job efficiency. A mentor, Lenny, related a conversation with HR management in his company. HR tells us that when they suggest a foreign candidate the hiring managers say, 'Well, no, I prefer someone with the name Andersson who can start to do the job from day one, with whom I can feel sure.' Even though it says in the CV that the applicant speaks Swedish well, the managers think more of their own needs and it is coloured by what they are used to.
There is consequently a perceived risk of lowered efficiency if the newly recruited person does not speak Swedish well and an assumption exists that communication between manager and migrant employee will be negatively impacted, as will the person's ability to do the job well. Spending time explaining and clarifying is believed to be inefficient.
Other participants noted that while they were open to hiring migrants, customers and clients sometimes expressed reluctance. A manager (mentor) in a consultancy firm outlined how he once assigned a consultant with a foreign sounding name to a job, but the customer refused to work with him and he had to be replaced with a consultant with a Swedish sounding name. Other mentors, but also a mentee, mentioned it would be more costly to hire migrants in general, because, they believed, the migrants' perceived lack of certain skills has to be addressed before they can fully perform their role effectively. The hiring of migrants is clearly associated with potential difficulties in maintaining optimal organizational performance.

Organizational normality perceived to be at risk
When talking about highly skilled migrants' difficulties in finding a job, mentors tended to provide explanations from the perspective of organizations. One mentor, Lova, talking about recruitment reflected upon the unknown: One is indeed a bit curious about foreign competencies. But how can I know that it's going to work? Maybe not; and does one dare to take that chance? And [if] you have another one who has [an application with no foreign or unfamiliar components]. No, I don't know; there are a lot of companies that unfortunately choose not to take the chance. (. . .) I believe one is a bit afraid of the unknown. (. . .) Because, for many, the first step in recruiting is to hire someone like yourself. That's the way it is usually.
Lova points to the idea of organizational fit and the value given to the known -to what we call organizational normality.
Anna-Karin, a mentor and project manager, emphasizes how a person's potential fit into an existing group is part of her reasoning when hiring: [When I recruit], I start with thinking: 'Well, this person would probably fit with the others in the group' (. . .) The four of us have worked together for at least a year -then you know each other pretty well. You know who knows what, who's good at what, and then you want the new person to fit into the group as well.
She concludes by stressing that conforming to a certain norm is preferable for all parties: 'In the end, you want a group of people to be able to work together without anyone feeling uncomfortable or outside.' Masha, a mentee, talks about the organizational perspective privileging the known when describing how hiring a migrant may impede organizational convenience: But if it is [a] local company (. . .) why would they hire someone, even with higher education, but who is not a Swede? Why? I mean it's safer, more comfortable, and, em, more convenient for them to take Swedes.
These quotes highlight the implicit assumption that risk is automatically associated with hiring people who do not fit organizational norms, group culture, values and, broadly, organizational practices, thereby reinforcing the perceived obstacles to the employment of migrants.

Conditional hiring
Many of the migrants who achieved a qualified job went through an internship or trainee programme or started at a lower, less skilled position before achieving a full-time job commensurate with their skills. The participants claimed that highly skilled migrants could not expect to start at the same hierarchical level or the same level of competencies and skills as they had enjoyed before their migration. As Khakan, a mentor said, 'There is a general attitude that one must start low.' His statement was reinforced by another mentor, Jacob, who said, 'Well, if you arrive with somewhat wrong education, maybe you have to accept an easier job.' Mentees appeared to confirm the mentors' views, albeit less keenly. As Angela, a mentee with a Swedish Master's degree stated: 'I'm like starting from zero. Regardless of my education, regardless of anything that I've had in the past. This is where [I] start.' Zuzanna expanded on that position by saying, 'My first mentor told me that if you cannot find a job as an accountant, maybe you can start as a care assistant and then work your way up.' Shamun, a mentee who found a position consistent with his qualifications, also mentioned his difficulty in getting established: 'The first job I got, yeah, it wasn't, let's say it wasn't an improvement from my previous career (. . .) [I] continued from what I was doing. Maybe less than what I was doing.' When highly skilled migrants are associated with employment, it seems to be generally accepted that they should enter the organization at a lower position than might have been expected in their home country.
The overall picture that emerges from the text of the interviews is that, when taking an organizational perspective, both mentors and mentees tell us that the employment of migrants is clearly associated in the minds of recruiters with putative harm to organizational performance and disruption of organizational normality. It is also associated with conditional entry into the labour market -entry at a lower level seems to be expected by the mentors and reluctantly accepted by the mentees.

Narratives on employability
In our field material, the participants linked the development of highly skilled migrants' employability to the necessity of addressing limitations associated with their foreignness. A large part of sessions in the mentorship programme focused on activities aimed at overcoming these perceived deficiencies in order to move the mentees closer to an implicit Swedish local standard. These activities are understood to be normalizing.

Adopting local job-seeking practices
The field material shows that many participants saw the highly skilled migrants as needing to replace their previous home-country job-searching skills with Swedish skills. The programme provided several sessions focusing precisely on those activities: networking, recruitment processes and writing a strong CV that meets Swedish standards.
During one of the mentee workshops, local features of the Swedish job market were presented by a representative of Diversity for Profit: '75 to 80% of all jobs are mediated through networks and contacts' (Observation notes, Mentee Workshop No. 1).
The mentees assimilated this notion and referred to it to explain why they had difficulty finding a job and what they needed to do to address this problem. One of the mentees, Masha, told us that it was valuable to make contacts on LinkedIn: 'Especially in Sweden as a country. It's not often one finds jobs by advertisement. Very often, it is through connections.' Mentees must learn local ways of presenting themselves through letters, interviews and interview follow-ups. As Alicja explained about her job-hunting experience in Sweden: 'You have to be very flexible [in your cover letter] and you need to answer [each aspect of the job ad].' Zuzanna believed that she had to sell herself in Sweden, unlike her experience in her home country, where 'you don't have to sell yourself. It's not the same culture.' When Angela told her mentor that she had not heard back from the recruiter since her job interview, she was advised to contact the HR person directly: And then I went home that night, I went into LinkedIn (. . .) and I opened the profile of the HR that I met and I wrote to him that I've had an interview and I would just like to (. . .) Just to say hello. And just to remind him that there's someone waiting for a response. The following day they called me [to offer the position]. If I had not spoken to my mentor and she'd told me [to make contact via] LinkedIn, I would never have thought about that.

Developing local workplace behaviour
One of the mentees, Hadi, found a job commensurate with his qualifications after an internship and assured us that it was the internship that made the difference, because it enabled him to access local workplace behaviour: I believe I got the job because I am here at [internship company]. I have four months of experience from a bank and I am already in the labour market. Now I know the rules and how to do things in Sweden (. . .) Earlier, I applied for 50, 60 jobs. I got nothing, not even an interview. But now, with this internship on my CV, I applied for three jobs and got one of them. So the difference is the internship.
The internship also offered Hadi an opportunity to gain knowledge of local work practices and expectations. During her internship, Maria Alejandra realized that she needed IT skills other than those she had learned in her country of origin and turned to IT vocational training: 'I searched for a training programme and found one and studied a while to update myself a little bit.' Annelise was a mentor who pointed to the differences between workplace behaviours in Sweden, workplace behaviours in her mentee's country of origin and the importance of mentees understanding local Swedish workplace behaviour: It could be little things, small nuances, or how one understands hierarchy in a workplace, or how one thinks about who one can talk to or how to contact people. The differences are like night and day.
Mentees were well aware of the importance of speaking Swedish as an aspect of conforming to local workplace behaviour. Many worked diligently to improve their proficiency in Swedish, undertaking multiple language courses at various levels. One of the mentees outlined how he volunteered for the Red Cross in the afternoons to practise Swedish and develop a better understanding of Swedish workplace behaviour (Observation notes, Introduction Workshop for Mentees and Mentors). Swedish language and culture are often linked in the mentees' accounts, as shown by Angela's story of how she was accepted for a trainee position: They wanted someone who spoke Swedish. I worked on my language skills and got comfortable in the Swedish culture and stuff like that. So I thought this was the job for me. If I get it. Yeah, I got it . . ..
In sum, both the programme activities and the narratives of interviewees referred to a process of learning local workplace behaviour -local knowledge, Swedish languagefor the migrants to be seen as employable. The employability of migrants, therefore, is not linked only to recognizable job-search behaviour and CV contents, but also to adherence to a local Swedish norm.

The organizational logic of keeping migrants away
In the narratives about migrants, two major patterns were identifiable: (1) the strong connection between hiring these highly skilled migrants and perceived potential harm to the organization; and (2) the need for highly skilled migrants to normalize -to work on their fit with a local Swedish norm. Using the relational theory of risk (Boholm and Corvellec, 2011) and its three major concepts of object at risk, risk object and risk relationship, we see how these two major themes from the narratives come together to display an organizational logic for the underemployment of migrants. We use the term 'organizational logic' in a broad sense, referring to what is perceived as rational, coherent and which makes sense in the context of the organization.
An object at risk is a valued entity that one wants to protect from threat (Boholm, 2003;Boholm and Corvellec, 2011). What is valued and considered at risk is contextual and variable as people understand and judge risks in terms of emic, locally defined values and concerns (Beumer, 2018;Boholm, 2003). In our case study, the narratives present the hiring of migrants as potentially harmful to organizational performance and everyday work, the object perceived to be at risk in this case being organizational normality. An organization's everyday work and current normality is not simply a norm or a state of affairs, then; it is a condition valued by interviewees who describe the recruitment of migrants as a disruption of this sense of normality.
A risk object is a possible threat, which can take multiple forms -a phenomenon, a process or a thing considered potentially harmful to a valued object: the object at risk (Boholm and Corvellec, 2011). In this case, highly skilled migrants are implicitly depicted as the risk object by means of what they represent: the level of their unknown skills and competences, their local language proficiency, their cultural differences and their general unfamiliarity.
Risk occurs when those who value the object at risk believe there is a possibility of a harm. There is no certainty of harm; that would constitute a hazard. Rather, one could say, uncertainty is a fundamental dimension of risk assessment (Boholm, 2003). It appears from the interviews that it is the unknown and the unfamiliar associated with migrants that create this uncertainty and perceived risk. The migrants are implicitly seen as risk objects. One way of reducing this uncertainty is to shape the migrants closer to the perceived organizational normality -to limit the uncertainty associated with the risk object. This explains the current practices of the mentorship programme: shaping the migrants' job search and their work behaviours in line with local Swedish norms, thereby reducing their unfamiliarity (or, put more bluntly, their foreignness) and the uncertainty attached to the unfamiliar. Many of the employability-shaping activities depicted in the interviews (learning local practices of candidate profiling, the local language and everyday work behaviour) are clearly about normalizing these highly skilled migrants.
The risk relationship is that which is established between the object at risk (what is valued) and a potential threat (the risk object). This relationship can simply take the form of a person who is in proximity to the object one values (Cadet and Carroll, 2019). The risk relationship reflects the observer's knowledge and understanding of what is of value and what is a threat (Boholm and Corvellec, 2011), which means that the relationship need not actually endanger the object at risk; it is enough that the observer perceives it to be a danger. In the interviews, the employment of highly skilled migrants establishes a relationship between the object at risk (organizational normality) and the risk object (the unfamiliar migrant). If highly skilled migrants are seen as a threat and employment is seen as placing them in a relationship with the object at risk (organizational normality), then it makes sense from an organizational perspective to limit this relationship. The obvious way to minimize the relationship is not to recruit them, not to hire them, to place them in internship positions or to promote them at a lower level than they enjoyed in their home country. By underemploying highly skilled migrants, organizations can support the local norms and the local normality, thereby indicating that the foreign characteristics associated with the migrants are of lesser value than familiar characteristics: the organizational normality. From this organizational perspective, underemployment of migrants makes sense.
In sum, using the relational theory of risk, it becomes clear that if people talk about risk and uncertainties in relation to organization normality, it is because they value it. In the narratives, this threat (risk object) is embodied in the unfamiliar migrants and it becomes understandable that one wishes to normalize them through the mentorship programme's activities, to reduce the uncertainty they represent. The employment of migrants creates a relationship between this putative threat and organizational normality and, from an organizational perspective, it is sensible to limit or avoid the situation. The underemployment of highly skilled migrants can thus be explained as an organizational logic to control the risk relationship between the object at risk (organizational normality) and the perceived risk object (the highly skilled migrant). This is illustrated in Figure 2.

Discussion
While organizational willingness to hire migrants has largely been taken for granted in employability research, viewing employability through the lens of the relational theory of risk (Boholm and Corvellec, 2011) challenges this field assumption (Alvesson and Sandberg, 2011). When migrants are seen as unfamiliar, organizations do not want to engage with them. Indeed, our analysis highlights that if the unfamiliarity of migrants is perceived as a threat, it makes sense from an organizational point of view to want to normalize it and to limit it by either not starting, or making conditional, the employment relationship. The relational theory of risk, by allowing us to see a perceived threat to normality, highlights how important normality appears to be in the construction of migrant employability. Consistent with other programmes working on the development of migrants' employability (see Bergström and Omanović, 2017), a goal of Diversity for Profit's mentorship programme is to shape migrants' CVs so that they are consistent with local norms. Likewise, non-governmental organizations or state organizations validate migrants' previous experience according to local Swedish standards (see Diedrich et al., 2011) by normalizing practices in the construction of migrant employability. Yet, we see that the migrants who secured a job are those who could demonstrate the enactment of this normality and their acculturation to the local norms (see Horverak et al., 2013). It is not enough for the migrants to appear to conform to local norms. Employers want evidence that those norms are enacted through activities such as internships or volunteering. They want this evidence before they recruit migrants to their organizations. In light of the high value placed on those norms, it makes sense that organizations require this additional evidence. This also provides an explanation for the higher barriers that migrants must jump and why they must work harder to demonstrate their employability by accepting positions lower in the organizational hierarchy than for which their education and experience qualifies them.
The relational theory of risk also leads to two additional theoretical contributions in the understanding of employability from an organizational perspective. It shows how organizational practices socially construct employability of migrants and second, it problematizes organizational normality. Because risk is a social construction, it explains why some organizations perceive highly skilled migrants as a risk object while research has shown that other organizations perceive them as a resource (e.g. Ortlieb and Sieben, 2013). Migrants' characteristics such as available linguistic or social capital (Al Ariss and Syed, 2011;Ogbonna and Harris, 2006) may not be core to their perceived employability. Rather, it is the recruiters' narrative construction that is central -a construction made in relation to organizational aspects. Our contribution to employability theory is  Figure 2. Using the relational theory of risk to explain the underemployment of highly skilled migrants as an organizational logic. thus positioned at the organizational level (Thijssen et al., 2008), in contrast to the currently dominant view that approaches employability as individually shaped (e.g. Forrier et al., 2018;O'Connor and Crowley-Henry, 2019). Previous research investigating the organization's shaping of a migrant's employability tends to consider HR support practices for employees (e.g. Hajro et al., 2019;Shirmohammadi et al., 2018). We contribute to broadening this view by showing how the employability of migrants is also constructed by organizations prior to employment. The literature on recruitment points to organizational characteristics that shape the evaluation of the employability of foreigners: for example, the characteristics of the recruitment team or the clientele (e.g. Almeida et al., 2015;Carlsson and Rooth, 2007;Fernando et al., 2016), or the international orientation of the organization (e.g. Makkonen, 2017;Shchegolev et al., 2016). In these studies, similar to our case, local organizations associate migrants with factors such as a lower local human capital (Crowley-Henry and Al Ariss, 2018), thereby allowing migrants' perceived lack of knowledge about local practices and language to be linked to inefficient work habits (Makkonen, 2017). What these studies have in common is their approach to organizational factors as given variables that will affect recruitment and they do not consider the socially constructed aspect of the appreciation of employability. In contrast, the interpretivist approach taken here has allowed us to study how things can receive meaning in relation to each other and indicates how the perception of employability occurs in relation to something else: normality. We show how organizational practices, rather than given features, shape the perception of the risk (Boholm and Corvellec, 2016;Corvellec, 2010) and the narrative construction of the migrants as a risk object. It is because employees are not so used to communicating with non-native Swedish speakers that they discursively present the lack of Swedish language proficiency as a threat. We point to the ways in which organizational practices -rather than features such as the organization's size -and the resulting perceived normality are core to the shaping of employability. Our contribution to the organizational stream on migrants' employability is to show that employability is also shaped by organizations prior to recruitment and employment and by a normality perceived by the organization and based on everyday practices.
The relational theory of risk leads to a second theoretical contribution to the understanding of employability from an organizational perspective, as it enables us to explicate why normality or local cultural norms are an implicit (and sometimes explicit) point of reference in the literature on the employability of migrants. Subsequently, this helps us to problematize this taken for granted normality. Indeed, our study suggests that normality is much more than the way we are used to working. If this were the case, migrants would be depicted not as a threat, but in more neutral terms -simply as change agents, for example. When viewed through the lens of the relational theory of risk, we learn that normality is valued and that employers want to protect it. Normality is thus not simply what people are used to and what unfortunately hampers them from seeing the merits of migrants in recruitment and promotion (e.g. Cameron et al., 2019;Crowley-Henry and Al Ariss, 2018). Normality is valued more highly than the diversity that migrants can bring to the organization (see Horverak et al., 2013), indicating the existing power inequality present in society and organizations alike. Indeed, the mezzo-level of organizations is not independent of the broader societal environment in which political debates on migration, legislation or economic trends -to name a few -influence views on migrants (see Guo et al., 2021;Krings, 2020;Syed, 2008). In Sweden, at the time of the study, migrants are seen as occupying a low(er) ethnic and thus social status and more so if coming from outside Europe. Workers born outside Sweden have in general lower salaries and lower status at work (Manhica et al., 2015;OECD/European Union, 2015). They also have a lower employment rate and lower matching grade between their education or skills and jobs and lower rates of full-time employment compared to those born in Sweden (Statistics Sweden, 2019). This status is directly connected to their underemployment (see Gorodzeisky and Semyonov, 2017;Tomaskovic-Devey et al., 2015).
The presentation of migrants as a threat to organizational norms thus reflects the status position of migrants in Swedish society. Indeed, when Nygren et al. (2017) study the relationship between norms and the perception of risk, they show how both are inextricably intertwined with status and power. They illustrate the ways some individuals are ascribed certain risks or discursively constructed as a risk or risky, whereas others are not (often those seen as the norm). They stress how the performativity of risk occurs along lines of differences that (re)produce power relationships along with established societal norms (Nygren et al., 2017: 420-421). In our study, migrants are constructed as risky in view of their differences. Although a migrant's lack of Swedish proficiency is presented as a problem, the non-migrant's lack of intercultural communication skills is never problematized. This shows the prominence of (societal and corporate) norms; a 'Swedish' way of working is posed as a norm that must be followed. Migrants are thus constructed as a risk in their deviation from this norm and this perceived deviation leads to ethnic structural discrimination at work between migrants and non-migrants (Behtoui et al., 2017;Romani et al., 2019;Tomaskovic-Devey et al., 2015). Our second contribution to studies on organizational approaches to employability, therefore, is the elucidation that normality is not a given: it is a socially constructed norm that reproduces ethnic structural inequality in the workplace. This means that the central reference to normality or organizational 'fit' in our understanding of employability needs to be problematized rather than considered a simple state of affairs.

Conclusion
This study investigates the case of an organization that is actively working to shape the employability of highly skilled migrants so they can meet regional employers' expectations and be integrated into the labour market. By studying how the employability of migrants is perceived, particularly in the central activity of a mentorship programme and through its participants' narratives, we contribute to the understanding of highly skilled migrants' underemployment.
When approached from an organizational perspective, underemployment of migrants has to date been studied with a focus on migrants' 'damaged' human capital or a 'biased' assessment of applicants' skills and characteristics (e.g. Crowley-Henry and Al Ariss, 2018;Ghumman and Ryan, 2013;Ogbonna and Harris, 2006). With this study, we bring to the fore an alternative focus: an organizational logic for the underemployment of migrants. We hear that migrants are implicitly depicted as representing a threat to the normal way in which organizations function, a threat that could materialize into a risk to organizational performance and everyday work if migrants were to be employed. By applying the relational theory of risk (Boholm and Corvellec, 2011) the study contributes a novel explanation for the underemployment of highly skilled migrants: organizations limit the recruitment of migrants in order to control for the putative threat they represent to organization normality.
The relational theory of risk thus challenges and enriches our understanding of employability in three major ways. First, it challenges the taken for granted assumptions that organizations are willing to hire migrants: when the latter is seen as a threat, there are organizational grounds to underemploy them. Second, showing that normality is presented as an object at risk, the theory stresses the centrality of organizational normality in the shaping of employability. It thus adds to our current understanding of organizational aspects that contribute to migrants' employability. Finally, the relational theory of risk stresses that existing power positions are key elements in the shaping of what is seen as a risk. This enables us to problematize (Alvesson and Sandberg, 2011) the focus on existing organizational normality and its expected fit, as it creates an implicit norm and structural discrimination at work, to the disadvantage of migrants. In view of these findings, we invite further studies to consider the centrality of organizational normality in other aspects of migrant employability, such as retention and promotion.