The multilevel intelligent career framework: an exploration and application to skilled migrants

Purpose – Skilled migrants ’ careers are heterogeneous, with existing theories capturing only some of their diversity and dynamic development over time and circumstance. This paper aims to draw out the multilevel (macro,mesoandmicrolevels)influencesimpactingskilledmigrants ’ careersbyusingthelensoftheintelligent career framework. Furthermore, structuration theory captures the agency of skilled migrants facing different social structures at and across levels and explains the idiosyncratic nature of skilled migrants ’ careers. Design/methodology/approach – Following an abductive approach, this paper examines the career influences for a sample of 41 skilled migrants in three different host countries. Individual career stories were collected through qualitative interviews. Important career influences from these narratives are categorised across the intelligentcareer competencies (knowing why, how and whom) at the macro, meso and micro levels. Findings – Findings illustrate the lived reality for skilled migrants of these interrelated multilevel career influences and go some way in elucidating the heterogeneity of skilled migrants ’ careers and outcomes. The interplay of individual agency in responding to both facilitating and challenging social structures across the multilevelsfurtherexplainstheidiosyncraticnatureofskilledmigrants ’ careersandhow/whethertheyachieve satisfying career outcomes. Some potential policy implications and options arising from these findings are suggested. Originality/value – By considering multilevel themes that influence skilled migrants ’ career capital, the authorswereabletobetterexplainthecomplex,relationalandidiosyncraticshapingoftheirindividualcareers. Assuch,theframeworkinformsandguidesindividuals,practitionersandorganisationsseekingtofacilitateskilledmigrants ’ careers.


Introduction
Skilled migrants are tertiary educated or possess experience (Damas de Matos and Liebig, 2014;Hajro et al., 2019;Iredale, 1999) and are legally entitled to live and work permanently in a host country (Zikic, 2015). Studies on skilled migrants' careers have proliferated over the past decade with increasing interest in global talent management (Al Ariss, 2014;Collings et al., 2019;Guo and Al Ariss, 2015; as well as the distinctive career challenges faced by skilled migrants (e.g. Al Ariss and Syed, 2011;Crowley-Henry et al., 2021). There have been various studies on skilled migrants working in different professions and countries (e.g. Grigoleit-Richter, 2017;Winterheller and Hirt, 2017). Typologies of skilled migrants have been put forward (e.g. Delli Paoli and Maddaloni, 2021;Ramboarison-Lalao et al., 2012;Zikic et al., 2010). Different populations have been explored (e.g. Lebanese in France, Al Ariss and Syed, 2011;Eastern Europeans in Ireland, O'Connor and Crowley-Henry, 2020), and different career lenses have been used (e.g. the kaleidoscope career model in O'Connor and Crowley-Henry, 2020; protean careers in Gunasekara et al., 2021; bounded boundaryless in Richardson and McKenna, 2003). However, a framework that more generally captures the myriad factors impacting migrants' careers remains elusive (Crowley-Henry et al., 2018, p. 375). This makes it difficult to gain a full picture of the social dynamics involved in migrants' careers or to inform policy and practice for effectively supporting skilled migrants to negotiate careers in the host country.
Past research has identified individual-level factors influencing the careers of skilled international migrants (Richardson and McKenna, 2003;Gunasekara et al., 2021), such as bicultural identity, a value-driven approach, desire for travel and adventure, personal resilience, etc. Other factors relating to political and economic contexts and individual and family factors have been identified (O'Connor and Crowley-Henry, 2020). Zikic et al. (2010) suggested six key themes as objective barriers (and resources) to the careers of self-initiated expatriates (SIEs). These were viewed as operating at the individual level, such as motivation, identity, local know-how and social networks. While they discussed the "interdependence" between "objective" and "subjective" factors, between individuals and "institutions, occupations and gatekeepers" (p 0.3), no overarching or integrative framework was proposed. Rather, three subjective career orientations of individuals and Bourdieu's concept of capitals were used to explain how objective and subjective factors are negotiated by individuals. A more integrative model covering individual, institutional and economic/political factors was proposed by Al Ariss et al. (2012a) who identified four levels of influencing social factorshistory, macro, meso and micro. However, no empirical data were presented. Nor did the model detail how the relational interplay between individuals and meso/macro processes occurs. Similarly, Crowley-Henry et al. (2019) distinguished between influencing factors at the macro, meso and micro levels but went no further than providing researchable propositions. In Al Ariss et al. (2012b), Bourdieu's social and cultural capital concepts are employed at micro and macro levels, but their study largely overlooked the organisational level (Al Ariss et al., 2012b).
In our proposed framework, we seek to integrate the three multilevels of social influences and, within our case stories, the temporal dimension. We also seek a fuller understanding of how these influences operate within individual career stories by recourse to intelligent career theory. In our conceptualisation of the individual skilled migrant, we follow Zikic et al. (2010) and Al Ariss and Syed (2011), who emphasise their individual agency in managing their career. The literature on career self-management (CSM) positions a proactive career orientation and individual agency as central to the dynamic process of CSM (Hirschi and Koen, 2021). CSM includes the different cognitive and behavioural activities that influence one's career over time (Hirschi and Koen, 2021). We extend the notion of CSM by applying the intelligent career theory at the multilevels. This approach is in keeping with Hirschi and Koen's (2021) recommendation that the environmental factors (at the macro, meso and micro levels) influencing CSM or career agency need to be considered from a multilevel perspective. CDI Our use of structuration theory in the framework is also in line with Garbe and Duberley (2019), who found that there is an "inherently recursive nature of individual careers and broader social structures" (Richardson et al., 2022, p. 101). More specifically, our multilevel framework builds on Al Ariss et al. (2012a) by "filling in" the theory about how the interplay between structure (e.g. institutions) and individual agency operates through the "knowing why/how/whom" concepts of intelligent career theory viewed together with structuration theory. This approach answers calls to consider and incorporate the dynamic interplay of context and structures more closely in international career studies (Andresen et al., 2020;O'Connor and Crowley-Henry, 2020;Hirschi and Koen, 2021).
Our study focuses on culturally diverse skilled migrants in English-speaking host countries. This is based on literature which suggests such migrants encounter structural barriers in their career progression in the host country, including precarity of employment (Al Ariss et al., 2012), discrimination and under-employment (Almeida and Fernando, 2017;Kostenko et al., 2012) and lack of recognition of their talents by the powerful influencers (in organisations and at national policy level) in the host countries (Almeida and Fernando, 2017;Crowley-Henry et al., 2019). Such barriers often lead to significant social and economic losses for skilled migrants and their host countries (e.g. Almeida and Fernando, 2017;CEDA, 2021).
As such, the purpose of our paper is two-fold: 1. to develop a multilevel framework for understanding the broad range of factors which influence migrants' careers and 2. to inform skilled migrants, practitioners and policy makers who work with them on ways they may achieve successful careers. Our first research question asks: How can the lived career experiences of heterogeneous skilled migrants in different host countries be more holistically conceptualised in career theory? A second and related question is: How can we better understand the ways that multilevel factors are negotiated by skilled migrants to advance and shape their careers?
To address these questions, we followed an abductive approach (Saetre and Ven, 2021), a "form of generative reasoning" based on evaluating hunches raised by data and subsequently using deduction and induction to learn more (p.1). Our study involved oscillating between our three-country interview data and the extant literature on skilled migrants' careers and career theories. In this process, we found two theories best explained the data: intelligent career theory (Arthur et al., 1995) and structuration theory (Giddens, 1986). These theories illuminated the strategies individuals used to negotiate their careers as they encountered objective and subjective barriers. Our analysis considered the interplay of these theories at and across different levels of social/influences: macro/country/state level, the meso-/ organisational/profession level and the micro/individual level (see Crowley-Henry et al., 2019, p. 6). In this way, we were able to capture in one framework the multi-level structures and processes that appeared to influence these individual skilled migrants' career trajectories, together with the agentic responses made by them.
The following section introduces the intelligent career framework and structuration theory. We then share the research approach of our three-country empirical study. We highlight the relational, integrated interplay of structure and agency across levels by presenting tabulated data from our participants' career stories and highlighting one representative in-depth case using our multilevel framework. We then discuss the theoretical contribution and limitations of our framework, future research opportunities and practicerelated/policy implications and recommendations.

Intelligent career framework and structuration theory
The intelligent career framework proposes that one's knowing why, how and whom competencies constitute an individual's career capital across organisational contexts (Parker et al., 2009). First introduced by DeFillippi and Arthur (1994) and subsequently developed by Arthur et al. (1995), this framework shifted attention to the individual unit of analysis from the The multilevel intelligent career framework organisational, during a period in the early 1990s when scholarship had been focusing on firm competencies and the intelligent enterprise (Quinn, 1992). Specifically, three types of knowing are posited (Arthur et al., 1995;Parker et al., 2009;Zikic, 2015), originally termed an individual's "know-why, know-how and know-whom competencies" (DeFillippi and Arthur, 1994). These were subsequently articulated as verbs, "knowing why", "knowing how" and "knowing whom" (Parker et al., 2009). The first, "knowing why" focuses on an individual's motivations and ambitions, which spur them into their careers. The second, "knowing how" equates closest to human capital theory in that it represents the knowledge, skills and abilities an individual possesses to complete their work tasks. Finally, the third, "knowing whom", considers an individual's social capital, comprising both strong and weak ties (Granovetter, 1973).
Notably, intelligent career theory encompasses some but not all of Bourdieu's capital (human capital, social capital), together with individual motivations and aspirations. Cultural capital, mentioned earlier in this paper, appears to be omitted. However, we would argue that both knowing how and whom rely in part on the exercise or manifestation of cultural capital. This is because it is difficult to exercise either technical know-how or forge relationships at work without the involvement of relevant cultural capital. Conceptualising intelligent career theory in this way answers the calls by Al Ariss and Syed (2011) and Zikic et al. (2010) to acknowledge the role that various capitals play in negotiating career barriers and opportunities (Zikic et al., 2010, p. 3).
In our framework, we conceive the macro level operating at either national, state or provincial levelswhere government policies and other institutions relating to immigration, multiculturalism, education, employment and settlement are generated and subsequently impact skilled migrants. Further, in our study, the meso level encompasses either the organisational or professional level; for example, an employer/business or a professional association. Finally, the micro level refers to individual persons and groups operating either within an organisation, a professional association, a family context or more broadly in society. The following paragraphs describe each of the intelligent career "knowings" at the macro, meso and micro levels, respectively, drawing on extant literature in the domain of skilled migrants and careers.
In relation to knowing why career competency at the macro level, we focus on the attraction policies that countries offer to encourage skilled migrants to enter and remain in the host country. Froese (2012) explains the initial motivation to move internationally under three categories: the job market, family and regional interest. When the job market in the country/region is buoyant, or if living standards and quality of life (for family as well as the individual) are better than the home country, this may promote the motivation to relocate to another country. Global indicators of cost of living, educational standards and happiness indices, for example, may enable skilled migrants to make a more informed decision with regard to why they want to move to a particular host country.
We consider the knowing how factors at the macro level to include policies for recruiting and developing skilled migrants with the skills required in the respective country. Immigration policy that enables skilled migrants to both study and work at a high skill level in the host country before settling permanently is an important factor (Chiou, 2017; Green and Hogarth, 2017). However, the attraction of skilled migrants alone is not always sufficient. Once onshore, these migrants may need assistance negotiating an unfamiliar labour market. Agencies such as AMES (Adult Multicultural Education Services) in Australia provide bridging programmes and mentorship opportunities for newly arrived skilled migrants. Industry and professional associations (such as Engineers Australia) have also been known to provide valuable research and skilled migrant development programmes (Engineers Australia, 2018). In this way, skilled migrants' host country-specific knowing how and whom skills are developed.

CDI
Remaining at the macro level, the knowing whom career competency focuses on the social networks that migrants have within the host society. Typically, skilled migrants tend to have more bonding capital via networks with other migrants from their own country, but lack bridging capital which is the strong social ties with host country employers (Briggs, 2003), thus hindering their integration into "mainstream" society (Fokkema and de Haas, 2015). However, national and regional (i.e. macro level) policies may help to address this problem. For instance, in Canada, TRIEC (Toronto Regional Immigrant Employment Council, 2017) runs the Connector programme, which helps skilled migrants build their professional networks, by connecting them with a local, occupation-specific mentor and matching them to employers who are looking to hire (combination of macro and meso initiatives).
Next, moving to the meso level, we consider the knowing why career competency focusing on the organisational or professional motivations for skilled migrants. Organisations, through their policies and activities, may inspire and motivate skilled migrants to develop their careers internally within the organisation or alternatively may create disincentives and blocked career paths which may lead to migrants moving to self-employment and establishing their own businesses (Miller and Le Breton-Miller, 2017). Knowing why relates to the skilled migrants' awareness of the rationale for pursuing particular careers within organisations. This could require organisational policies that inform skilled migrants about the organisation's expectations and opportunities (Stoermer et al., 2018)..
Turning to knowing how at the meso level, when hiring skilled migrants, organisations gain access to resources critical to a firm's success (Hillman and Dalziel, 2003). These may be technical/professional skills or valuable cross-national language skills and cultural intelligence which they can apply to international business or local niche markets (Zikic, 2015). In practice, such know-how can be an asset or a limitation depending on the organisation's policies towards and recognition of skilled migrants' competencies.
We describe the meso level knowing whom career competency as the social ties at the organisation/profession level in a host country. An expatriate study conducted in China (Wang and Kanungo, 2004) indicates that building networks with peer expatriates and local partners comprising people from different backgrounds facilitates faster learning and adaptation to the local environment. Many multinational organisations have formal mentoring programmes where they assign a "buddy" for new recruits to ensure a greater level of acculturation and social engagement (Almeida and Fernando, 2017). In addition, professional societies that facilitate the knowing how and knowing whom may include networks that skilled migrants develop through their academic and alumna affiliations. These alumni networks can provide spaces for skilled migrants to acquire the know-how, particularly, local skills and knowledge and also knowing whom that can be useful when seeking career opportunities (Sandoz, 2020). The micro level encompasses the level associated with intelligent career theory as originally postulated (Zikic, 2015). Here, knowing why relates to the individual motivations a skilled migrant has to move internationally. The work opportunities available in the host country, a better quality of life and a higher standard of living are pull factors that attract skilled migrants, and indeed all immigrants, to move internationally (Harrison et al., 2019). Push factors, on the other hand, are reasons for skilled migrants to leave their home countries, which may be due to comparatively disadvantageous economic or employment prospects (Zikic, 2015). However, Yoko Brannen in Horak et al. (2019, p. 5) cautions against stereotyping based on cultural generalisations. Some skilled migrants may work for purely financial reasons (Parutis, 2014), some may strive for position attainment to satisfy their home country families' expectations or their own expectations (Fouad et al., 2015) and some may seek a better quality of life, while others may be motivated by the "pure challenge" (Schein, 1990) offered by the need to cope with novel environments and cultural demands in a host country.
The multilevel intelligent career framework Knowing how at the micro level concerns the human capital skills of skilled migrants. According to human capital theory (Schultz, 1961), skilled migrants should gain access to positions within the labour market which are commensurate with their knowledge, experience, skills and qualifications (Zikic, 2015). However, much of the individual knowing how that enables skilled migrants' career progression seems to emphasise communication skills, soft skills and cultural skills (or cultural capital) rather than the academic qualifications of the individuals. Ellis (2013) found that skilled migrants with superior linguistic skills (i.e. speaking several languages) benefit from that cultural capital more than educational qualifications or work experience, at least initially in finding work in the host country.
Finally, the micro level knowing whom career competency reflects the social ties that skilled migrants have and build in the host country, which positively impact their career capital. Caligiuri and Bonache (2016) suggest that the longer a migrant is amongst another cultural group (e.g. host country nationals in a host country), the better their social network and capital with that cultural group will develop and become.
Having presented how the three knowings of the intelligent career framework may be aligned to career-influencing factors at the macro, meso and micro levels, we draw on Anthony Giddens' (1986) structuration theory to further explain the heterogeneity of skilled migrants' careers and how they play out. Here, structure refers to the already-existing rules and resources which are recursively integrated and negotiated with human agents (Giddens, 1979), underscoring the duality and interdependence between human agency (the actors) and the underlying social structure shaping society (structure) in the accepted ways of doing things. "Society only has form, and that form only has effects on people, in so far as structure is produced and reproduced in what people do" (Giddens and Pierson, 1998, p. 77). Structure, therefore, both constrains and enables individual action. Due to reflexivity and evolving understandings, the predictability that agents will engage in certain actions is not absolute, nor is the result of actions undertaken predictable given the dialectic nature of structure and agency influencing each other.
A limitation of contemporary career theories is the lack of attention to contextual factors that constrain or facilitate agency (Mayrhofer et al., 2020). Hirschi and Koen's (2021) review found that few studies included the socio-organisational context as an influential factor in career agency or CSM. They see this as a gap in the literature and argue that career progress results from the dynamic interplay between the individual careerist and their environment (Hirschi and Dauwalder, 2015). Rather than suggest that everyone reacts to structures in precisely the same way, Giddens' (1986) structuration theory allows us to position skilled migrants as actors (with different degrees of agency depending on their particular circumstances) within different social systems (family/close relationships, organisational systems, country societal system). Skilled migrants, influenced by and within these different contexts, make sense of and/or make decisions regarding the various social practices (Giddens, 1986). Figure 1 depicts our conceptual framework incorporating intelligent career theory and structuration theory across the three levels. It shows the recursive relationship between structural forces and agency across macro, meso and micro dimensions, with each focusing on themes and concepts related to the different types of knowing as proposed by intelligent career theory. It attempts to provide a clear and integrated conceptual overview of the interactions between the different elements of the framework.

Methods
While careers literature remains dominated by quantitative and positivist designs, qualitative research has made a significant contribution (Richardson et al., 2022). To answer the main research question in this study -"How can the lived career experiences of CDI heterogeneous skilled migrants in different host countries be more holistically conceptualised in career theory?"a qualitative research method was appropriate (Russell and Gregory, 2003). Typically, qualitative approaches are more suitable for exploring complex, behavioural "how?" questions (Yin, 2013). While engaging with careers literature, we simultaneously drew on the career narratives of a sample of skilled migrants across three host countries. Career narratives are the stories that participants tell regarding how they navigated their careers using their past and present experiences and future aspirational self-concepts (Ashforth et al., 2008;Ibarra and Barbulescu, 2010). The process of narrating is influenced by the research participants' values, motivations and experiences and is underpinned by the relevant social and cultural factors (LaPointe, 2010). We used an interview-based approach to gather stories from our participants ensuring that we followed a process that focused on participants' lived experiences.

Sample
The countriesthe USA, Australia and Irelandwere selected because each is a developed, white majority, English-speaking country, and we were interested in how culturally diverse skilled migrants from different countries negotiated their careers in WEIRD (Western, Educated, Industrialised, Rich, Democracies (Muthukrishna et al., 2020)) countries. We also chose these specific three host countries for practical reasons (the authors lived and had access to the skilled migrants living in these three countries, and we could conduct most interviews face-to-face within the resource constraints of the project). A total of 41 participants were recruited: 11 in Ireland, 17 in the USA and 13 in Australia.
Following Zikic et al. (2010), we include a brief overview of the three host countries that our interviewees were sourced from to foster generalisability of our findings. The three countries are on three different continents (North America, Europe and Australia), have English as their main language and were at some time members of the British Empire. Australia and the USA are "traditional" immigrant-receiving countries, whereas Ireland is a "new" immigrantreceiving country (Zikic et al., 2010, p. 672). Both the USA and Australia are "nations of immigrants" (Higley and Nieuwenhuysen, 2009;Zikic et al., 2010, p. 672), having large-scale immigration programmes extending from colonial times to the present. Australia has seen mass immigration diversify and expand its population since the Second World War. Both the USA and Australia are multicultural countries favouring skill-based immigration through systematic immigration selection criteria and programmes, both temporary and permanent. Ireland, in contrast, has traditionally been a country of emigration, though since the 2000s  The multilevel intelligent career framework strong technological and economic growth, together with membership in the European Union, have attracted significant inward migration (Mac Einri and White, 2008).
All three countries have strong, diversified labour markets, high per capita incomes and, as developed economies, a continuing need to meet labour skills such as doctors, nurses, engineers and IT professionals (Chand and Tung, 2019). These skill gaps are increasingly being filled in Australia and USA by immigrants from the developing countries of South Asia, South East Asia and China. Ireland, in contrast, relies more on culturally diverse immigration from Europe.
Hence, our three-country study represents host countries that have both similarities and differences at the macro migration level. While sharing similarities as WEIRD countries, they each have their own socio-cultural, political and economic landscapes and are obviously not representative of all countries globally. Our interest in this qualitative study was in how the skilled migrants in these different host countries experienced career support at the macro, meso and micro levels, in order to highlight common themes and sub-themes that were articulated across our sample. Future studies may extend our analyses to other host countries and skilled migrants.
Our sample size was determined through the information redundancy achieved during data collection and theoretical saturation (Hagaman and Wutich, 2017;Sim et al., 2018). Theoretical saturation of insights occurred when the later interviews in each country only reconfirmed the previously disclosed factors that shaped skilled migrants' careers (Glaser and Strauss, 1967). The participants in Ireland originated from the USA (white American), Brazil, South Africa, Scotland (2), Italy (2) Argentina (2), Spain and Lithuania. All were white Caucasians except for the South African participant. The participants in the USA were from India (7), Pakistan (2), Bulgaria, Ireland (6) and Switzerland, representing a more equal mix of white (8) and non-white (9) participants. The Australian participants were from South Korea, Vietnam, Malaysia, China, Sri Lanka (5), India, Singapore, Pakistan and Colombia, all non-white. 55% of the sample were males and 45% female, with ages ranging between 30 and 50 for most (70%), 20% were over the age of 50 and the rest were under 30. This age distribution is quite typical of skilled migrants, who tend to be concentrated in the prime working-age group.
Participants were employed in a wide range of industries. The sample from Ireland worked in the finance and banking, public sector, manufacturing, information technology, research and tourism industries. The USA sample was employed in the health, construction, information technology, manufacturing, finance, banking and public sector. In Australia, the participants worked in finance and banking, information technology, manufacturing, transport and the public sector. Over 74% of the participants had a postgraduate qualification (Master's degree or PhD), while all had, at a minimum, an undergraduate degree relevant to their occupation. Overall, therefore, the participants in this research were heterogeneous, with unique career narratives. Our focus was on the common themes and sub-themes mentioned across the diverse sample, detailing the career-influencing factors at the macro, meso and micro levels in relation to our high-level conceptual framework ( Figure 1).

Data collection
The researchers used purposeful sampling (Creswell, 2007) and personal contacts across the three countries to source the initial interviewees for the study. Snowball sampling then increased the range of participants in the study (Browne, 2005), where the original participants suggested other potential interviewees for the study. Only skilled migrants from culturally diverse backgrounds who had worked for more than 10 years in the host country were included due to the study's focus on careers over a significant time period.

CDI
We used in-depth semi-structured face-to-face or online video interviews using WhatsApp, WebEx or Skype platforms. The first author conducted the first three pilot interviews in the USA and fine-tuned some questions based on the pilot interview experiences. The first author then conducted all the interviews in Ireland and the USA, while the other authors conducted the interviews in Australia. Over 90% of the interviews were conducted face-to-face. These interviews were conducted at the participant's organisation, a local caf e or a public space. Each interview was approximately 1 h-1.5 h long. All interviews were audio-recorded, professionally transcribed and reviewed for accuracy by the authors. The authors also kept notes about our personal observations and experiences during the interviews. We used the same topic guide to assess the convergence and divergence of career narratives across national boundaries. Our topic guide included questions on the background (family, motivation to migrate) and career (current position, how they progressed, career aims, factors that influenced their career progression). As such, the questions focused on why they decided to migrate to the host country, how their careers unfolded, and what their migration and career experiences were like (Rubin and Rubin, 2012). A scoping literature review was also undertaken on the careers of skilled migrants to ensure that our questions were aligned with the purpose and goals of the study, namely, how we can understand the individual career stories, particularly, the macro-, meso-and micro-level social influences on skilled migrants' careers in the host countries (Brinkman and Kvale, 2015).

Data analysis process
Although qualitative data analysis is generally described as a nonlinear iterative process, we broadly followed the phases illustrated by Lester et al. (2020) to ensure the transparency of our process. The first phase was to prepare and organise the data for analysis which involved gathering all of the transcripts of the audio-recorded interview files into one location using a structured naming protocol for each file, as well as creating an Excel master sheet to catalogue the data source, interviewer/creator, including all demographic details of the participants. While all authors were involved in the interview data collection process, it is important to note that we lived in two different countries and three cities and worked at four different institutions. Therefore, we needed to ensure that a common approach to naming and recording the participant demographic details was followed while also de-identifying (anonymising) relevant information for participants we interviewed individually.
During the second data analysis phase, the authors became familiar with the data. All four authors read all the transcripts individually and reflected as a team about our personal observations and experiences of conducting interviews. This process took several months as we read over 40 transcripts during our teaching semesters. The authors originated from western and non-western cultural backgrounds and built a strong relationship based on mutual respect and trust. We practised good listening skills, which helped us to authentically reflect and question each other about our observations and experiences during the interviews, our biases and how these biases may have influenced our interpretations of the participants' non-verbal behaviour during the interviews. These critically reflective discussions focused on why and how questions and helped us understand the issues we faced concerning potential social desirability biases. For example, we discussed how cultural values, norms and gender-based behaviours may have influenced how we conducted interviews.
The third phase of the data analysis was to code the data. The initial aim was to compare the data during open coding to determine the extent to which the host country's national country context facilitated or impeded the respective skilled migrants' careers in those particular host countries. Firstly, three researchers read and manually coded one transcript from the USA to identify patterns in the data. Secondly, all four academics coded at least one The multilevel intelligent career framework to two interviews they conducted manually using open coding. Thirdly, two Australian researchers read at least seven to eight transcripts from USA and Ireland and coded based on the Intelligent Career framework (why, how and whom) across macro, meso and micro levels. We placed material that we considered relevant to careers, such as gender, which did not fit exclusively into the intelligent career theory into a separate folder for separate consideration and reflection (Richard and Hemphill, 2018).
Specifically, we moved from first-order codes (sub-themes) to second-order codes (metathemes) by collapsing the codes across our findings into macro-, meso-and micro-level themes. On analysis of our codes and themes, we found that intelligent career theory made the best sense of our findings when extended to incorporate the multilevels. We also drew on structuration theory (Giddens, 1986) to highlight the idiosyncratic nature of individual migrants' careers. As such, the researchers attempted in the most authentic way possible to navigate between a conceptual analysis of the stories collected and the empirical evidence through the co-construction of experience and meaning.
The fourth phase was to move from codes to categories which required us to move away from isolated cases to broader interpretations. All four authors had frequent online meetings to compare the skilled migrants' career narratives and used a reflective dialogue to compare and contrast the meaning-making of the emerging themes. To avoid subjective positionalities, we questioned our assumptions (two authors were white while two of the authors were women of colour) and how our lived experiences may influence our interpretations and analysis (Stuart, 2017). Making sense of data from diverse perspectives can increase the richness of the analysis. We used an iterative approach to data analysis, which encourages researchers to visit and revisit the data, connect them to emerging insights and progressively refine the understandings (Srivastava and Hopwood, 2009;Tracy, 2013).
According to Lincoln and Guba (1985), the trustworthiness of a research study is central to maintaining rigour. Collaborative data analysis is one way to enhance trustworthiness (Patton, 2015). The four researchers worked face-to-face during the data analysis phase for one week. During this time, we used the reflective discussion to delve more deeply into the similarities and differences across countries and asked ourselves the questions "why" and "how" there were similarities or differences across the career narratives.
The final phase we followed was to create an audit trail that helps to make connections between data, codes, categories and themes (Lester et al., 2020). Table 1 shares the first-order concepts, second-order themes and aggregate dimensions developed from coding the 41 interviews. We used axial coding to make connections between those patterns (Strauss and Corbin, 2015), enabling us to engage in a cyclical data loop to compare and contrast data, apply data reduction and consolidate data (Williams and Moser, 2019). This process aligns with the principles of the Gioia method (Corley and Gioia, 2004). Richardson et al. (2022) recommend including tables that show the coding process of qualitative analysis to ensure transparency of the undertaken approach. We also highlight the intelligent career component (knowing why, how or whom) most associated with the themes across each level. In this way, the study is both multilevel (macro, meso and micro) and multi-layered (knowing why, how and whom). Table 2 presents the empirical validation of Table 1 and displays the themes which were pertinent across each of the participant's narratives. Here, we focused on which themes we found in each of the participant's narratives rather than the strength of particular themes for particular participants. This was in line with the aims of the paper to develop an empirically validated multilevel framework for understanding the broad range of factors which influence migrants' careers, where that framework may be used to inform skilled migrants, practitioners and policy makers on the most effective career supports. These factors are unpacked further in the Findings section.

CDI
Findings Table 1 and Table 2 present the aggregated findings of how macro, meso and micro factors contributed to the careers of the skilled migrants who were interviewed for this study. In this section, we present our findings under the headings of the macro, meso and micro levels, respectively, sharing examples from the total sample to highlight career-influencing factors at different levels. Following this, we present a single case of one participant (Carmel, a 56-year-old Irish woman in the USA) to highlight the interconnection of structure and agency across and within multiple levels within an individual career story.

Macro level factors
At the macro level, pull factors like career and education opportunities, perceived ease of immigration and integration and lifestyle attractions presented by host countries (knowing why) were the most important factors contributing to launching and maintaining the skilled migrant participants' careers overseas (40 of 41 participants cited this as an important factor). We found subtle but important differences between our study participants in the USA, Australia and Ireland on some of the macro-level factors. For instance, more participants in the USA highlighted career, educational and economic opportunities in the host country (knowing why/knowing how) as important, while more Australia-and Ireland-based participants alluded to socio-cultural attraction factors: It was almost a certainty that I was going to move to the U.S. because at that point there wasn't really a big software industry in India and even the companies that were hiring, they were pretty much looking for the people that graduated at the top of the game. (Bill, Indian male in USA).
In contrast, Ama (below) mentioned the cultural diversity in Australia. That said, our sample of skilled migrants in the USA also often highlighted the fact that the USA is a land of migrants, making it easier for them to integrate. We also found that the Australia-located participants were more positive about the macro-level migration policies and pathways in Australia to obtain permanent residency.
We thought it was very easy to get migration status here because we are qualified accountants. In 2006 the skill was in demand, so you got 25 marks (points), so you could easily get it (visa). (The USA) was more difficult . . . we only thought Australia is good. Canada didn't even cross our minds. (Sanka, Sri Lankan male in Australia).
On the other hand, 15 of 41participants cited push factors such as poor conditions in their home country that influenced their migration decision and career.
There was a social and economic crisis in Argentina. There were no jobs and I was just at the end of my course. (Celina, an Argentinian female in Ireland).
The differences in the macro-level push and pull factors may depend on how the skilled migrants and their families make migration-related decisions that provide them with the greatest benefit and satisfaction. The motivations can range from macro-level context suitability for career development and quality to their overall life (O'Connor and Crowley-Henry, 2020). However, it is essential to note that the decision-making of skilled migrants related to macro-level factors can be bounded by the available information at a given time (Giddens, 1986). Therefore, the macro-level context is vital in better understanding the push from the home country and/or the pull to a host country for skilled migrants. It is important The multilevel intelligent career framework both for attracting and retaining skilled migrants, given the intense competition for skilled migrants globally and their aspirations for careers.

Meso-level factors
Meso-level findings indicate that in all three countries, employers' policies towards integrating skilled migrants in the workplace, such as support for career change and development, the provision of mentoring and flexible work played a significant role in supporting our sample's careers: The very first time typically if you're coming into a company that is big enough, they have somebody who actually helps you out . . . Although I came in by myself with my family, I was . . . the next day when I came to the office and . . . there were 15 other people as well and the entire administrationso social security, opening a bank account, somebody actually walked us through it. (Sam, Indian male in USA).
They were offering foreigners free English classes for a year and a half for a couple of days a week, with a very good teacher for two evenings a week for free. I think they saw the potential of very highly qualified people coming from everywhere, and if they provided them with their English skills they would be valuable employees for [organisation]. That was really nice. (Anthony, Spanish male in Ireland).
Participants also highlighted how they enhanced their knowing how and knowing whom at the professional level in the host country using other forms of meso-level institutions: I started going [to] all possible networking, including the Chamber of Commerce and other networking groups and fundraising activities. That took a lot of my after-work hours getting me involved in some of these things, so I built that network over time, but it was hard work. (Nick, Sri Lankan male in Australia).
The xxx job came through the Lithuanian professional network. There was somebody there, that girl who worked in xxx I knew, and I sent her my CV and I met her on my LinkedIn. I am very active on LinkedIn in terms of adding contacts, people I know and people I thought would be useful. (Isla, Lithuanian female in Ireland).
In addition, university clubs and associations can provide the vernacular and cultural knowhow to newly arrived migrants: They were small clubs and communities that helped you understand slang, which I thought was helpful, and then you understand what people are talking about. Must understand slang, because then you know what people are talking about and it took me time to understand that. You know what, it makes you feel included. (Mani, Sri Lankan female, Australia).
The role of organisations and intermediary bodies in supporting skilled migrants' orientation and adjustment to the host country is underlined across our sample.

Micro-level factors
We found that micro-level factors were the most influential in determining the skilled migrants' career orientation (24 of 41 participants). For instance, objective career success (like pay and career advancement opportunities, cited by three-fifths of the sample) was proactively sought by the participants (micro knowing why/knowing how): Throughout my life I keep upgrading myself . . . I went to school part time, I did certification, I learned new things, I keep upgrading myself which is very important; otherwise, they become difficult as you become more senior. (Fred, Pakistani male in USA).
I was always very ambitious. I always wanted to move and make more money and be independent. (Luke, Italian male in Ireland). CDI However, all participants cited subjective factors such as work-life balance, job satisfaction, achievement and the love of learning as being relevant to their career orientation in Australia (knowing why and how): I was very open and frank: this is my family and family is the most important thing, more than anything else. I could get a job, not as good as that is, but my top priority is my family. And we had that conversation and my manager was from US; he said "Ok, let's have a look around for another opportunity" . . . I was lucky to be given another opportunity, smaller scope, smaller team, less responsibilities and more importantly less travel, and that is better. (Victor, Vietnamese male in Australia).
Human capital know-how (such as technical skills and multilingual abilities) and personal characteristics and attributes (including resilience and hard-work ethos) were vital microlevel factors that contributed to their career orientation: It's amazing how much we learn when we are living abroad because we are challenged, I think, all the time. (Bella, Brazilian female in Ireland).
In terms of finding a job, I didn't have much difficulty. One of the reasons is, I had some good background and education, so it worked out well. (Fred, Pakistani male in USA).
What really assisted me is my ability to adapt to things. (Janine, Malaysian-Chinese Australia).
Social capital (knowing whom) factors, such as networking with peers and people outside the organisation were also frequently mentioned as important. The assistance of close ties (spouses, parents and in-laws), especially with childcare, was reported in some cases as vital career support: Because the money was very good it was worth it for him [her husband] to stay here and do his work from the phone. He would get the kids to school, then go in his office and do his thing, and then leave in time to pick them up . . . It was great and without him we couldn't have done it and he was always ready to say "okay, you go, I'll stay. (Carmel, Irish female in USA).
Relational support in the host country contributed to career success for more participants (39/41) than transnational support from immediate relatives like parents and in-laws (13/41), with some exceptions, such as Samantha: That was a great help otherwise I think I could not have advanced my career because the reason I could go back to work, having two younger kids at that time, is my parents and my husband's parents . . . came every other year until two years ago. (Samantha, Sri Lankan female in Australia).
Our analysis shows the prevalence of different knowing why, knowing how and knowing whom factors at the macro-, meso-and micro levels, which helped (or hindered) the skilled migrants' career orientations. Moving to illustrate the complex agentic navigation of multilevel structures and influences on skilled migrants' careers in the host country, we present a case from our study, Carmel, who, at the time of the interview, had lived in the USA for over thirty years.
Carmel was a 56-year-old Irish mother of four with an Irish husband, working as a registered nurse in a hospital. Her career had been influenced within and across the macro, meso and micro levels. At the macro level, she emigrated to the USA aged 22 when "there was very little work in . . . nursing" in her home country due to the poor economic circumstances in Ireland at that time and the superior economic situation in the USA (macro level knowing why). In addition, she was motivated to come to the USA "to investigate another country" (micro level knowing why) and had moved as a newly-wed. Her husband "had wanted to come here [USA] forever". This strong micro knowing whom was tied to her husband influenced her career choices and the decision to remain indefinitely in the USA, as "He is happy here".
The multilevel intelligent career framework A mix of weak and strong social contacts at the meso and micro levels aided her in establishing her career initially ("somebody who said you should try her, . . . so it was reaching out . . . One night, I went to her house, she was a recruiter. Gave her all my paperwork . . . and she hired me"). This knowing whom encounter (weak tie) was the genesis of her working life in the USA. Relational support (micro level knowing whom) in the host country then aided her career progression as she juggled her work with different childcare support from different babysitters (weak ties), although she admits "it was difficult to get babysitters; it was difficult to find somebody who you think was reliable enough to watch your children". Carmel also relied on her spousal support (strong tie) to maintain and progress her career by working opposite shifts to her husband ("I worked nights, he worked days").
The interconnection between Carmel's micro knowing why (financial motivation while juggling her family responsibilities), knowing how (her skills as a registered nurse, which were sought after at the macro and meso levels) and knowing whom (private clients she knew from organisation-based work (meso-knowing whom) informed her subsequent career opportunities and choices. Her meso-knowing why (relating to demand for her skills) allowed her to negotiate part-time work, knowing how (training provided by the hospital helped keep her skills current) and knowing whom (a decades-long relationship with the hospital employer where she still works) all coalesced in her story. Carmel's work pattern moved between agency work in nursing and working as a care assistant, to working as a registered nurse in the hospital, to working part-time in the hospital (and doing supplemental agency work) and then moving to work full-time in the hospital "because I needed the benefits" (family health care plan; meso-level knowing why). She summarised her career choices as evolving to accommodate her family and children and her career ambitions at the micro level, supported by the meso-and macro-level contexts. In reflecting on her career, Carmel noted that: "It evolved . . . we took the opportunities that came" and "that I am a much better independent nurse than what I could have been allowed to have been in Ireland".
"Took the opportunities that came" in Carmel's case summarises well the particular flavour of agency she employed in her careerflexible, adaptive and opportunistic reaching out for support wherever possible. Figure 2 summarises Carmel's career story using our multilevel intelligent career framework.
Carmel's storyas for each of our skilled migrant participantsis very individual to her personal circumstances, with varying priorities at different periods of time, highlighting the myriad of influences shaping her career orientation and the dynamic nature of career selfmanagement over time and space. She exercised agency to navigate different structures at different levels adaptively and creatively, given the choices available at those times, to follow the best career for her circumstances. Table 2 illustrates these multilevel influences across all our samples.

Discussion and contribution
Unpacking the influences on careers and career orientation is complex, given the diversity of factors that shape careers (Crowley-Henry et al., 2019;Zikic et al., 2010). Scholars have adopted integrative pluralistic approaches that bridge the macro-micro divide (Rousseau, 2011) or the meso-micro divide (Zikic, 2015). We propose our multilevel intelligent career framework as a more integrated framework that encompasses intelligent career theory and structuration theory across the three levels. Based on our cross-country qualitative empirical findings from career stories that unfolded over time, we argue that our multilevel framework more fully depicts the myriad of structures that impact career pathways, while not overlooking the individual's agential potential to shape their careers.

CDI
In our conceptualisation of a multilevel career framework (see Figure 1), we have drawn on the underlying premises of structuration theory (Giddens, 1986) which highlights the constraints that social structures can place on agency and shows how individuals can negotiate their pathway within the norms of those social structures; and on intelligent career theory (Arthur et al., 1995) which focuses on the factors of motivations (knowing why), competences (knowing how) and social capital (knowing whom). Our exemplar vignette of "Carmel" shows how in practice these theories offer both a more holistic framework of lived experiences of career influencing factors and underlines the individual negotiation of structures and processes resulting in a heterogeneous individualised career orientation. Our data table (Table 2) based on the full sample's career stories also shows the influence of intelligent career factors operating at the three levels. For instance, as shown in Table 2, 40 of 41 participants shared that macro-level career/education and economic opportunities in the host country (knowing why) were important to them, 27 of 41 said employer support for career change (meso knowing how) had been important to them, while 40 of 41 participants said at micro-level personal attitude and characteristics (knowing why/how) were the main influencing factors for their career success. The differing ability and experiences of skilled migrants in navigating the social structures they encounter in their international careers show the recursiveness of structures and agency in determining how particular individual careers unfold in a host country.
Theoretical and practical significance Our multilevel framework, with intelligent career theory as the backbone, allows us to understand the complexity and dynamic nature of factors influencing international careers. The framework suggests an element of empowerment for skilled migrants and organisations  The multilevel intelligent career framework seeking to facilitate their successful careers. It does this by unpacking and identifying the agentic factors at the different levels around career capitals, which they can mobilise to counter structural and other constraints hindering their progress. For example, a skilled migrant who experiences blocked career paths may tap into mentoring opportunities through their professional association or seek alternative employment which offers career mentoring (knowing whom/meso level).
Extending intelligent career theory to the macro and meso domains allows us to capture the wider social structures and contexts within which careers unfold and to which skilled migrants respond with their own personal strategies. Our study suggests that skilled migrants' careers are both constrained and facilitated by the social structure(s) and (contexts) within which they unfold. However, even with limited control, skilled migrants in our sample employed agentic responses toward opportunities they perceived, for example, at the macro or meso levels. Such insights are important both theoretically and to inform practical action and strategies. They illuminate the important social factors at the organisational, individual and family levels (such as mentoring, spousal support and the support of in-laws and parents) that cushion the challenges that skilled migrants may experience in their early years of migration.
At a practical level, our multilevel framework can help organisations, policymakers and skilled migrants by showing how intelligent career concepts can interact with structural factors and personal agency to achieve desired career outcomes. For example, in the post-COVID environment, macro factors such as closed borders impeded some skilled migrants' careers by preventing transnational mobility. However, the reopening of international borders allowed a freer flow of skilled migrants and family members, which our research suggests helps skilled migrants to combine family responsibilities and careers. Similarly, the agentic responses of some skilled migrants in our sample, by harnessing the knowing whom (through their social networks) and knowing why (seeing the big picture) helped a number of them to successfully navigate career bottlenecks and setbacks. Knowing the importance of social networks and mentoring (the knowing whom) to skilled migrants' careers, organisations and practitioners at meso and macro levels can provide tailored programmes for networking and mentoring opportunities. As earlier mentioned, this is the case for TRIEC (Toronto Region Immigrant and Employment Council) in Canada, which has organised mentoring programmes for skilled immigrants for more than a decade, leading to their inclusion in local professional employment (https://triec.ca/). Even after finding their first professional position, ongoing mentoring programmes can assist skilled migrants to navigate career challenges and opportunities successfully. Government agencies at the macro level can provide financial incentives for organisations such as employers and education institutions at the meso level and state/provincial agencies to collaborate and invest in such programmes. Such programmes might be better targeted to particular ethnic neighbourhoods or regions where migrant settlement and integration can be more impactful (Ng et al., 2019). Furthermore, government agencies can provide targeted research grants to institutions that specialise in migrant and career-related research and/or fund projects to initiate social networks, mentoring and sponsorship of skilled migrants, building in evaluations of such interventions.
Spreading the knowledge that such networks are needed, that human capital alone is usually insufficient for career success, is an important career resource that policy makers can make available through social media, professional associations, ethno-specific organisations, etc. These insights and resources can inform career development programmes, mentoring schemes and work experience opportunities offered by professional bodies and organisations that employ skilled migrants.
At the micro level, another area where policy can assist skilled migrants is for organisations to understand and support skilled migrants' needs for culturally appropriate CDI childcare, which may also require flexible work arrangements. At the macro level, national policy makers can take into account the importance of family migration in this context, as in-laws and grandparents are often the first childcare resource skilled migrants call on to facilitate their career transition and success. Greater opportunities for family reunions or long-term family visas would help in this regard.
While the framework was abductively developed in our study with skilled migrants, the essence of the framework may be considered for all working populations, domestic or migrant. Therefore, our multilevel intelligent career framework has practical significance in providing research-informed themes concerning structures and processes at different levels which influence individual career orientations.
Limitations and future research opportunities Our study was exploratory in nature and based on a relatively small sample of skilled migrants based in three countries. Participants tended to be more successful skilled migrants who had achieved a career path of some kind. As such, they do not represent the broader population of skilled migrants, many of whom face under-utilisation and under-employment in their host country (CEDA, 2021).
Our sample stemmed from diverse home countries and cultural backgrounds, but we do not deeply interrogate this cultural context in our analysis. Instead, our paper focuses on presenting the usefulness of a multilevel intelligent career framework to explain more fully the career influences for skilled migrants in host countries. We call on future research to delve further into the relevance of home country cultures in explaining and possibly shaping the career orientations of the migrants from those countries.
A further limitation is the lack of formal policy analysis to corroborate some of the macro factors skilled migrants reported influencing their decisions to launch and maintain a career in their host country. Further studies could explore country comparisons in greater detail to document push and pull factors to migrating and remaining in those countries over time. Meso factors such as organisational and professional association policies were also not directly interrogated.
However, despite these limitations, we suggest that the data from our participants provide a strong justification for examining the respective career influences across multi levels of analysis using intelligent career theory and structuration theory. More research using our multilevel intelligent career framework for more targeted and larger samples would provide a richer knowledge base around these key influences on skilled migrants' careers.

Conclusion
Our research contributes to the field of skilled migrants' careers by proposing an intelligent career framework incorporating structures across multiple levels and highlighting the role of individual agency in career self-management (Hirschi and Koen, 2021). The multilevel intelligent career framework presents career behaviours at the individual level as situated within and informed by the broader contexts and structures which recursively and relationally impact on each other. It represents a more accurate depiction of the CSM of individuals in often unfamiliar macro environments and related meso contexts; of cultural novices intending to pursue their careers in their new host countries.
Our framework extends extant intelligent career theory to include a more contextual focus, answering several calls from scholars to incorporate context more deeply into career studies. Rather than a polarised depiction of skilled migrants' careers as successful or failed, our study highlights the more nuanced and dynamic structures that influence individual careers in different ways. While in this paper we focus on the multilevels, we call on other The multilevel intelligent career framework studies to incorporate more closely the temporal dimension pertaining to the respective structural influences at different time periods, and relative to individual circumstances at particular times.
By drawing on structuration theory, we can explain how skilled migrants' careers unfold heterogeneously as they negotiate a system of structures across multilevels that influence their ability to shape their desired career paths. Despite existing literature suggesting that skilled migrants encounter negative career ramifications post-migration, our analysis shows that skilled migrants are also agentic and resilient in negotiating the social structures and processes they face.