10. Skill Requirements and Employment of Immigrants in Swedish Hospitality

Structural change in labour markets due to migration are characterised by a more diverse workforce in terms of workers’ national origin, their reasons for immigration, level of education and qualifications. At the same time, skills requirements for jobs are changing with an increasing share of service sectors jobs, changes exacerbated by the Covid-19 pandemic’s effect on physical workplaces, and rapidly expanding online delivery of goods. For many years, the share of permanent jobs has also been in decline in most developed countries (Thornley, 2006), making migrant workers more exposed to changes in job content and skills required to handle those jobs. As of 2022, workplaces increasingly require skilled workers to handle more complex and interactive tasks. Employees must have sufficient skills to become a part of labour force within the changing requirements of jobs due to technological change like digitalisation (van Laar et al., 2017).

National labour markets throughout Europe and North America has also become more diverse in terms of workers’ national background (Kazlou & Urban, 2023). In this chapter, we study Sweden, where, as of 2022, more than a fifth of the active labour force is born abroad, a vast increase since the 1990s. Furthermore, historically, migration has tended to consist of more homogenous groups arriving in intervals due to host country labour shortages or due to conflicts in a specific country or region. However, migrants are increasingly diverse in terms of country of origin, reasons for immigration, age, gender, level of education and skills (Meissner, 2015). These migrants join the host country labour force as students, as labour migrants, as kin to current residents and as refugees fleeing conflict and persecution in their home countries.

This chapter probes the questions of why and how immigrant employees are overrepresented in the hospitality industry compared to other low-skilled industries. We examine labour market segmentation across immigrant groups compared to native workers in Sweden. We focus on the hospitality sector, which employs an increasing share of the immigrant workforce (Åslund et al., 2014; Přívara & Kiner, 2020; Sönmez et al., 2020). The chapter investigates the sorting of immigrants and natives into jobs with different levels of skill requirements and according to their ethnicity¹ (Ponomareva et al., 2022) within the three comparable low-skill industries: hospitality, retail and construction. Jointly, we examine how different ethnic groups may be sorting into different sectors (Åslund et al., 2014). We also probe how the skill composition of service jobs shape ethnic labour market segmentation among male and female workers as well as foreign-born and natives.

This chapter combines detailed data on occupations and workers in the Swedish hospitality, construction and retail sectors matched with data on the skill requirements of these jobs based on standardised measures from the United States, thus providing a unique opportunity to study the sorting of immigrants by skills and ethnicity into different jobs and the increasing stratification of the labour market.

Using Statistics Sweden’s Longitudinal Integrated Database for Health Insurance (LISA) and occupation-education information at the individual level in 2012, we have access to comprehensive information, including the demographics and education of both immigrants and natives in different occupations. By matching the individual data from LISA with data on required skill characteristics for different occupations according to the International Labour Organization (ILO) and O*NET (the Occupational Information Network), we suggest the fundamental premise that skill level requirements differ among jobs. Type of jobs are reported according to the Swedish Standard Classification of Occupations (SSYK), which is fully harmonised with the ILO’s occupational classification categories.

Combining data on skills requirements from the O*NET with individual-level data on the entire Swedish population during 2009–2012 allows us to analyse occupational employment as changes in demand for skills and the matched supply of immigrant and native employees. We select three industries that are similar in level of skills and that are generally considered as low- to medium-skilled sectors exhibiting fair entry rates for immigrant workers but differ in structure of specific skill requirements: hospitality, construction, and retail.

Figure 10.1 shows that immigrants in the hospitality industry in Sweden in 2012 had lower wages (log-wages are on the vertical axis on the left box plots in Fig. 10.1) compared to natives, which can partly be explained by the fact that they have lower levels of education (the right box plots in Fig. 10.1).

Two box whisker plots. a, the graph plots the social skills of natives and immigrants in hospitality, construction, and retail. b, the graph plots the technical skills of natives and immigrants in hospitality, construction, and retail sectors.

Stark evidence of opposing differences in composite skill levels across individual jobs in hospitality among immigrants and natives is shown in Fig. 10.2. The required level of social skills is much lower while the level of technical skills is much higher for those jobs occupied by immigrants compared to natives. At the same time, skill requirements are equally distributed in the retail and construction industries (Fig. 10.2), with the slight exception of jobs held by native workers in the construction sector having slightly higher requirements for social skills. In comparison to hospitality, neither retail nor construction exhibit such obvious differences among natives and immigrants in sorting to jobs requiring social or technical skills.

Two box whisker plots. a, the graph plots the social skills of natives and immigrants in hospitality, construction, and retail. b, the graph plots the technical skills of natives and immigrants in the hospitality, construction, and retail sectors.

Figure 10.3 shows details for selected social (persuasion, negotiation and service orientation) and technical skill requirements for specific jobs by native and immigrant workers (equipment maintenance, equipment selection and installation; see Appendix for definition of the skills). Overall, Fig. 10.3 shows sharp and striking differences in employment patterns among immigrants and natives—immigrant workers are more often employed in jobs requiring lower level of social skills and a higher level of technical skills. These results are interesting and can reveal some risk baring by immigrants due to substitutions of routine tasks by machines within digitalisation (van Laar et al., 2017), while jobs with higher requirements for social skills, frequently highlighted as important in people-to-people interaction and service-intensive work (Clark, 1993; Ivanov, 2020), are increasingly in demand in hospitality. The construction and retail industries do not have such striking differences in the sorting of immigrant and natives by skills.

Together, the graphs in Fig. 10.3 highlight the starkly different ethnically based job segregation in the three sectors studied, where immigrants and natives are found in jobs with very distinct skills requirements in the hospitality sector, but not in the construction or retail sectors.

Table 10.2 shows the sorting of immigrants into the three selected industries by country of origin in 2012. Immigrants from Turkey comprise more than 25% of all immigrants employed in hospitality, followed by individuals from Iraq (13.6%) and Syria (8%) in our sample. Immigrants from Poland are overrepresented in the construction industry (19.7%), followed by representatives from Finland (17.1%) and former Yugoslav countries (8.7%). The retail industry represents a more balanced case, where immigrants from the former Yugoslavia are the largest group (15.7%), followed by immigrants from Iraq (10.3%) and Finland (7.5%).

Figure 10.4 reveals a clear pattern of segmentation by country of origin and gender in the three industries and shows the diversity regarding residence permit status in Sweden. Figure 10.4 show different ethnic and gender groups within the three industries (on x-axis: N is number of observations for specific ethnic groups, both genders and S are a share of an ethnic group in the total number of employed immigrants in the industry; the y-axis shows the number of observations of immigrants by ethnic group and gender). In the hospitality industry (Fig. 10.4), male immigrants from Turkey, who mostly arrived as refugees or for family reunification reasons, dominate. The group “student male migrants” comprises the largest share of immigrants from the next largest ethnic groups in hospitality (Iraq, Syria, Lebanon and Iran). Surprisingly, we note smaller shares of immigrants employed in hospitality who arrived in Sweden as refugees or for family reunification. A very small share of female immigrants from these countries are employed in hospitality and they arrived in Sweden mainly as students or for family reunification.

Three stacked bar graphs. a, the graph plots migrant males and females in hospitality in Turkey, Iraq, Syria, Lebanon, and Iran. In all the countries, males record the maximum immigration. The highest immigration of males is from Turkey. b, the graph plots migrants in construction from Poland, Finland, Yugoslavia, the Soviet Union, and Iraq, the highest number of immigrants is from Poland. c, The graph illustrates the number of immigrants in the Retail industry from F. Yugoslavia, Iraq, Finland, and Iran. Among female immigrants, the highest number of individuals studying are from F. Yugoslavia

Male labour migrants more often work in the construction industry in Sweden (Fig. 10.4). Student migrants from the former Yugoslavia mostly work in Swedish construction. In contrast to the hospitality and construction sectors, immigrants are more balanced by gender in the retail industry. Immigrant students dominate retail.

Overall, our descriptive results indicate that immigrant workers in the service-intensive hospitality sector are overrepresented in jobs requiring hard technical (routine) skills, while native workers in the same sector are more often found in jobs with higher requirements for social skills. Since routine-based technical skills related to machinery operations, monitoring and control are skills highly susceptible to automatisation, it adds vulnerability to immigrants’ employment in the hospitality sector.

In this chapter we set out to examine why some categories of immigrant workers are more likely to find jobs in hospitality but not in similar low-skilled industries like retail and construction. Using detailed data on the skill characteristics of jobs within the three sectors of hospitality, construction and retail, we probe the extent to which these differences may be driven by differential skill requirements as well as by ethnic sorting into specific jobs and sectors. Our findings suggest that skill-biased change and ethnic segmentation (based on country of birth) provide insights into the patterns of immigrants sorting into jobs and sectors. The fact that mostly native women and immigrant men are employed in hospitality suggests that more vulnerable groups find employment in this industry. We look deeper into the trend to understand what skills are typically required in hospitality and not in other low-skill industries, finding that immigrants are more often employed in technical skill routine jobs in hospitality. Immigrant men from Turkey and the Middle East have a higher probability to start and be employed in hospitality sector. The hospitality sector also provides jobs for native younger women in jobs requiring social skills even though they commonly earn less than natives in other industries.

After the radical halt of Swedish refugee immigration in 2015 and labour migration on the decline, it is notable to examine the labour market segmentation of migrants occurring during the decades leading up to this shift and what they can mean for the future. The Covid-19 pandemic of 2020 has radically disrupted the working conditions in the hospitality industry with an initial shock of layoffs and workplace closures, a decline in tourism which may be once again on the upturn, and a more long-term shift towards online retail and digital provision of services and food deliveries. Many jobs are moving from in-person service to a background and on-distance work. At the same time, working conditions in hospitality has been noted as unattractive, and employers are struggling to fill vacancies for new positions and (re)recruit staff after the pandemic lockdowns of service establishments have ended. The hospitality sector is a frequent “entry path” for individuals with limited labour market experience and highly dependent on both young workers and migrant workers. The bi-annual work quality surveys conducted by Statistics Sweden notes that hospitality workers are almost twice as likely to experience risks of lay-offs if they complain about working conditions as workers in retail or construction, and four times as likely to work despite being ill for fear of losing their job (Statistics Sweden, 2022). What can we expect in terms of ethnic segmentation of jobs and shifting jobs tasks in the nearby future? On the one hand, one would expect that shortage of service workers would improve wages and working conditions for those applying to such positions. However, the proliferation of non-regulated and illegal workers and low unionization rates among young workers and immigrant workers in hospitality (Bender, 2023, this volume) means that employees bargaining position is much lower than in other industries. The ongoing digitalisation and closure of workplaces after the pandemic may also lead to increased requirements for technical skills for tasks like operating hotel services and customer service desks and systems. These changes do offer potential for immigrants as it can decrease requirements for communication and language skills, however, these are also positions that can be readily offshored.

Most likely, the ethnic segmentation of labour is likely to persist, but may shift as work tasks and related skill requirements in jobs change and become more technically oriented. In our study, we find the hospitality sector to be ethnically dominated by immigrants from Turkey and Middle Eastern countries, while the construction sector is dominated by immigrants from Poland, showing clear ethnic sorting between industries. While refugees and family migrants from Turkey more often find jobs in hospitality, surprisingly, more students from other Middle Eastern countries are employed there. The construction industry is dominated by Polish and Finnish male workers, which might create barriers for other ethnic groups looking to find jobs in the industry. Similar ethnic sorting of immigrants is also reported in Norway (Friberg & Midtbøen, 2018, 2019) and in the United Kingdom (Ram & Smallbone, 2002; Thiel, 2010; Vershinina et al., 2018). According to labour market segmentation theory, immigrants are more likely to sort into routine technical skills jobs in hospitality based on their human and social capital characteristics, including unrecognised or lower levels of education, or previous work experience in a similar industry. Immigrants are also prevented from sorting into social skill jobs because of poorer host language abilities (Dustmann, 1999). Further, immigrants who enter specific industries and workplaces often refer co-ethnic workers to job openings via their social networks (Andersson et al., 2014). Ethnic labour market segmentation (Noel, 1968; Esser, 2010; Thiel, 2010; Haller et al., 2016) explains why some ethnic groups of immigrants are overrepresented in specific industries or niches. Native workers may be unwilling to take on certain jobs not simply because they generate low pay but also because they confer low status (Leontaridi, 1998). Immigrants, however, may be less picky due to a so-called dual frame of reference and limited options in host-country labour markets (Friberg & Midtbøen, 2018). This willingness will often be interpreted as a sort of skill or “work ethic” (Piore, 1983). Nevertheless, neither labour market segmentation nor skill-biased technological change theories can individually explain the sorting evidence by skill and ethnicity in the low-skilled industries. Thereby, the allocation of immigrants in host-country labour markets is closely linked to processes also defined in literature as categorical inequality (Massey, 2007), with labour market hierarchies tending to reflect the social structure of the societies within which they are embedded (Waldinger & Lichter, 2003). However, this social structure also opens opportunities for network-based recruitment in labour market segments that are socio-ethnically segregated. Employers might also use their employees as referrals and personal contacts to reduce the costs of finding good matches (Holzer, 1987; Montgomery, 1991). Employees might also be sorted into different firms based on their language skills; in our case, the Polish language might be a working language in some construction companies in Sweden. Because of language similarities, immigrants from Ukraine and other former Soviet Union countries might join the Polish dominance in the construction industry in the future, a trend we also observe in the data in 2012 (the former Soviet Union group is among the largest five groups in the construction sector). Several existing studies (e.g. Wilson & Portes, 1980; Hellerstein & Neumark, 2008) indicate that employers prefer to recruit employees who speak the same language. This implies that immigrants are recruited by firms where fluency in the native language is of less importance or where many of the employees speak one common languages. For front-line service workers in hospitality and retail, language requirements and communication skills are likely to remain imperative.

Policymakers can benefit from this study by considering nuanced sorting of immigrants with different origin background into jobs with social or technical skills in different industries, which might require adjustments in skill formation and education policy interventions for specific groups of immigrants. Further, supporting the skills of immigrants seeking to establish themselves on the job market should consider facilitating the acquisition of language and interpersonal skills in addition to formal training and education. For these purposes, practical job training in service intensive workplaces may be equally important to formal language training. Authorities should also consider the rapid digitalisation of both retail and services, and the opportunities inherent in equipping prospective service sector workers with training in system operations, database and customers support systems, for better access to better jobs in the future.