Why are care workers from the global south disadvantaged? Inequality and discrimination in Swedish elderly care work

ABSTRACT

Using quantitative and qualitative methods, this study investigates inequalities in occupational status and wages between native-born and foreign-born employees in elderly care institutions in Sweden. It finds that employees from Africa, Asia and Latin America – the “Global South” – are disadvantaged in both respects. Combinatory explanations of the inequalities are needed. The shorter work experience of foreign-born workers in the care sector plus the lesser value given to educational credentials obtained outside Sweden are among the factors related to human capital theory. Access to less-valuable resources in the workplace social networks of foreign-born employees is related to social capital theory. The processes that result in exclusion from powerful social networks, in turn, are found to be affected by discrimination in the workplace.

KEYWORDS:

• Migrant worker
• workplace discrimination
• wage and status inequality
• Sweden
• social capital
• racialization

Introduction¹

Individuals of migrant background, particularly from the Global South, tend to be over-represented in the lower echelons of labour markets, with lower wages, poorer working conditions and less employment security (Schierup, Hansen, and Castles 2006). Despite variation between countries, this is a basic pattern and one that is also seen in Sweden (Rödin 2011). The aim of this article is to contribute to knowledge regarding the causes of such inequalities in the workplace among employees from various parts of the world – including native-born individuals of native parents as well as migrants – analysing in particular the consequences of individual access to social capital in the workplace.

Unlike many previous studies that analyse the occupational status and wages of employees across the labour market as a whole, this study concentrates on particular workplaces: care homes for the elderly in a major Swedish city. It focuses on the conditions for nursing assistants and assistant nurses. The study contributes to research on conditions for migrant care workers by adding detailed knowledge regarding the complex factors that influence inequality patterns within individual workplaces, drawing on both the literature that focuses on human capital and the literature that highlights social relations in the workplace. In exploring social relations, it makes a novel contribution by empirically measuring the effect of access to work place social capital (WPSC) on wages and positions, which, to our knowledge, has not been previously undertaken. The study contributes knowledge regarding the situation in Sweden, where conditions in the care sector differ from those in many other countries where similar research has been conducted.

The following research questions about the studied workplaces guide the study:

• Is there a gap between native-born and foreign-born care workers in terms of occupational status and/or wages?

• If so, how can the differences be explained?

Below, we first present explanations of working-life inequality in previous research and the particularities of the Swedish elderly care sector. Following this, we present the study, introducing our data, key concepts and measures before presenting the results.

Explanations of occupational status and wage inequalities

Migrants from the Global South receive lower wages than native-born Swedes (le Grand and Szulkin 2002) and tend to be segregated into lower-ranked jobs (Åslund and Skans 2010). The explanations for these kinds of inequalities that dominate international research are related to: (a) human capital, (b) macro-level factors and (c) discrimination and other social relations in the workplace.

Neoclassical economic theories maintain that, as for other individuals, the labour market outcome for immigrants is generally determined by their “human capital”, such as education and labour market experience (e.g. Hammarstedt and Palme 2006; Luik, Emilsson, and Bevelander 2016). However, labour market studies demonstrate that differences in productivity explain only a small part of the income and employment rate gap between natives and immigrants (Altonji and Blank 1999). Thus, human capital approaches must be complemented with other theoretical perspectives.

Other studies focus on macro-level factors, such as the organization of labour markets and immigration policy regimes (Tilly 2011; Scarpa 2016). While important in revealing general tendencies, these studies leave invisible the actual processes and practices that produce the inequalities.

Therefore, social relations in the workplace, not least the relational power of the various actors and groups, have become an increasing focus (Acker 2006; Tomaskovic-Devey 2014). In the workplace, employees are embedded within various networks. Resources that are accessible via these social networks (social capital) have a substantial impact on the position of employees within the organization (Bourdieu 1986; Lin and Erickson 2008; Behtoui and Neergaard 2012). Interactions usually occur between actors with similar resources and lifestyles; therefore, marginalized groups are likely to see their positions in the social structure reproduced. The social networks of stigmatized “racial” and “ethnic” groups provide less access to information and influence because these groups tend to hold subordinate positions (Lin 2001).

In organizational research, formal as well as informal workplace structures are in focus. Formally specified relationships between subordinates and superiors follow the typical organizational system. Informal networks, on the other hand, involve relationships which are “work-related, social or a combination of both” and are usually more extensive than formally specified relationships (Ibarra 1993, 58). In informal networks, individuals seek each other out and develop and expand their networks either for work and career goals or for friendship and social support. According to Ibarra (1993), friendship networks affect decisions and mobilize resources in the organization.

Much of the research about workplace social capital defines this form of capital at the collective level (Avgar 2010). This article emphasizes instead the advantages that individuals can utilize as a result of membership of a group (Bourdieu 1986). One group’s WPSC gain may mean another group’s loss (Erickson 2001) because the benefits of social capital to one group regarding, for example, promotion enable the exclusion of others. In this way, WPSC becomes a central cause of differing outcomes for individuals with similar human capital. Discriminatory relations and processes in the workplace are also of particular relevance to the issue of inequality. We refer to discrimination as the unfavourable treatment of individuals due to their (presumed) belonging to a particular group (cf. Banton 1994). Discrimination includes the more favourable treatment of those who are assumed to belong to the “in-group”, a process by which “advantage is passed along from those who are already in positions of power, authority, and privilege to those with whom they identify” (DiTomaso 2015, 73). Loury (2002) distinguishes between discrimination in contract and discrimination in contact. The first denotes the unequal treatment of individuals in formal situations. The second refers to the unequal treatment of persons in social relations – as when some workers are excluded from friendship networks in the workplace (Loury 2002, 95; Behtoui and Neergaard 2012).

Thus, this study draws not only from the human capital approach but also from research that shows how social relations in the workplace matter. To our knowledge, the impact of access to WPSC on inequality in wages and positions has not previously been directly measured, and we set out to do this here. Among social relations, discrimination is a potential driver of inequality in organizations, not least for unequal access to WPSC. For this reason, we use the concept of racialization, as explained below, and use the qualitative part of our mixed methods study to enhance the understanding of these processes.

Particularities of Swedish elderly care work

In Sweden, elderly care is the workplace for several professions, including doctors, nurses, cleaners, occupational therapists and physiotherapists; however, assistant nurses (undersköterskor) and nursing assistants (vårdbiträden) are the two largest occupations. Elderly care is one of the most gender-segregated sectors in the labour market (Bihagen and Ohls 2007). Almost 93 per cent of assistant nurses and 83 per cent of nursing assistants were women in 2015.² In 2012, the percentage of foreign-born individuals among the general population of Sweden was 15.5 per cent, while foreign-born employees represented 24 per cent of employees in the elderly care sector. The percentage of foreign-born men in this care sector was more than twice as high (15 per cent) as that for native-born men (6 per cent) (Wondmeneh 2013). Among foreign-born female assistant nurses and nursing assistants, approximately half are from the Global North and half from the Global South; among men, the proportion is closer to 25 per cent and 75 per cent respectively (Statistics Sweden [SCB]).

Many studies have demonstrated the inferior position of migrant care workers in the health sector of Global North countries – for example, England and Henry (2013) noted that numerous studies have shown that in the United Kingdom (UK), black African migrant nurses in particular are placed in rankings that are lower than their training and experience would suggest and that they face racism and exclusion. Research has demonstrated that the inferior position of migrant care workers is due to the informal working conditions in this sector (Ngocha-Chaderopa and Boon 2016; Simonazzi 2012). Van Riemsdijk (2010) analyses the racialization of employees with a less secure right of abode, while Lutz (2011) examines gender relations when care work is a negotiation between women as household care work employers and care workers.

The Swedish care sector, however, differs from that of many countries in ways that would suggest a better position for migrant care workers. In a study comparing elderly care policies in 21 European countries, Sweden stands out in several respects. Sweden has the highest expenditure on care for the elderly, the lowest percentage of the population declaring that care should be provided by close relatives, the fewest employed women giving up work for the care of an elderly relative and the highest female employment rate (Naldini, Pavolini, and Solera 2016). Unlike many other countries, residential elderly care work in Sweden is regulated and formalized. In addition, in many Global North countries – such as the UK and United States of America – a shortage of nurses has been met through the recruitment of educated nurses from the Global South (England 2015), while in Sweden this is uncommon: most assistant nurses/nurses from the Global South have immigrated for other reasons, often as refugees or through family reunification. Until recently, this has normally meant permanent residency and access to citizenship.

Thus, the extent of formalization, the organization of care through elderly homes rather than in households and a more secure right of abode for migrant care workers will create better expectations regarding the situation of care workers from the Global South in Sweden compared to many countries studied in the literature on migrant care workers.

Nevertheless, care workers who appear to be “non-Swedish”, according to a number of studies, experience racism or discrimination from some patients (Lill 2007; Sörensdotter 2008). Storm (2018) studied Swedish and Canadian nursing homes and found that all non-white care workers in both countries reported incidents of racist comments from residents.

The study

Data gathering methods

This study used a mixed method approach with quantitative and qualitative data-gathering and analyses. Quantitative analysis was used to find the overarching patterns of inequality and a potential explanation for them, while qualitative analysis was used to develop the explanation.

The quantitative data was collected in 2012–2014 in twelve residential elderly care institutions in a large Swedish municipality. A comparison of the individual characteristics of care workers in this city (education, age, gender, etc.) with employees in the entire Swedish care sector (Wondmeneh 2013) showed no significant differences. The quantitative analysis is based on a telephone survey conducted by Statistics Sweden (SCB) in the spring of 2013, with employers providing the contact details of their employees. Each telephone survey lasted 30–35 min. Further information (respondents’ age, educational level, gender, income and birthplace) was obtained from the register databases of SCB. Of the total population surveyed – 510 individuals – 304 (61 per cent) answered. Non-respondents did not significantly differ from the final sample in terms of age, gender or region of birth. In this study, 289 employees are included – mainly those who work directly in the care of elderly people in the wards (15 deleted individuals had administrative positions).

Participant observations were conducted between winter 2013 and autumn 2014 on three different wards at two elderly care institutions, one of which was run by the municipality, and the other was privately run. One researcher spent seven to eight days on each of the three wards. Field notes were taken and then coded with the qualitative data analysis software NVivo 11 for the themes of central importance to our research questions.

During the same period, semi-structured individual interviews on various issues, such as workplace routines and social relations, were conducted with 30 employees, the selection of whom was based on a stratified sample encompassing native and foreign-born men and women and privately and publicly employed workers. While more than half of the interviewees worked on the same three wards mentioned above, the remaining interviewees worked in other units/departments. Among the female interviewees, ten were foreign-born. Five of the interviewees were men, only one of whom was a native of Sweden. The interviews lasted 60–90 min, and the transcripts were coded in NVivo 11.

Key concepts and measures

The concept of racialization is used in the literature to capture the socially constructed notions of “race” that enter into social relations (Solomos 2003). The term as used here refers to processes in which not only phenotypical traits but also perceived cultural differences are used to categorize groups of people as “Other” while simultaneously constructing a “We”. In this process, people are categorized according to their “physical characteristics, culture, and historical domination and oppression, justified by entrenched beliefs” (Acker 2006, 444). As previous research has revealed, workers from different parts of the world risk different degrees of racialization in Sweden. These results mirror studies regarding other countries, such as Van Riemsdijk’s (2010) study, which found that Polish nurses in Norway were partially included at their workplaces, whereas other migrants were more fully excluded and were considered less “like us”. A survey by Mella and Palme (2013) shows that, among Swedish natives, there is a high degree of social acceptance of immigrants from Northwest Europe and North America, followed by immigrants from Eastern and Southern European countries, while people from non-European countries are perceived as the most different and dissociated. According to Myrberg (2010), when native respondents hear “the word invandrare [immigrant]” (2010, 59), they are more likely to think of people who have migrated from Asia (including the Middle East), Africa and Latin America and much less likely to think of people from Western European countries. This ranking of people of distinct origin coincides with studies of self-reported discrimination in Sweden: in a survey, only seven percent of migrants from “the West” but 38 per cent of respondents from “the South” reported experiences of ethnic discrimination during the previous twelve months (Myrberg 2010, 67).

For this reason, we compare the experiences of four groups of employees based on their country of birth. The first group consists of workers born in the “Global Northwest” and is mostly made up of individuals from Western European countries (except Sweden), North America, Australia and New Zealand. The second group comprises those born in other European countries, mostly those of Eastern Europe. The third group includes employees born in the “Global South”, i.e. in African, Asian or Latin American countries. Although there is not a wholly consistent correspondence between individuals who tend to be racialized and their continent of origin, this categorization is adequate for our research purposes. Our reference group contains employees born in Sweden with two native-born parents. In light of Sweden’s immigration history, this means that the control group includes only or almost only “white” individuals.³

The main outcome variables of our study are the occupational status and the wages of employees. To measure occupational status, we use the prestige rating of Svensson and Eriksson (2009), in which a nurse ranks higher than an assistant nurse who in turn ranks higher than a nursing assistant. We also considered status differences within the same job, distinguishing assistant nurses/nursing assistants who are assigned as “team leaders” or who have the important task of scheduling. The measure wages used the current monthly wage based on register data and information from employers (adjusted to full-time equivalent, pre-tax earnings per month).

The human capital of respondents was measured as employees’ years of schooling and their educational level, which was obtained from register data. We also asked the respondents whether their highest credential was obtained in Sweden, as previous studies have demonstrated that the economic rewards that migrants receive for credentials obtained outside their new country of residence are lower than those received for credentials from their new country (le Grand and Szulkin 2002). Furthermore, we asked about the total number of years of employment and the number of years of employment in the care sector.

To explore social relations in the workplace, we measured resources in an individual’s social network (social capital) through the “name generator” method (Lin 2001). Respondents were asked to think about

the people who are important for you at work. They are those you feel the greatest confidence in at your workplace. You can rely on their advice when there is a problem. How many such persons are there at your workplace?

Furthermore, respondents were asked to specify the gender, country of birth, highest educational level and job of their two closest contacts at work. The number of contacts and the prestige scores (Svensson and Eriksson 2009) of the occupational status of these two contacts constitute the main basis for measuring the WPSC of the employees. A factor analysis was performed on these variables, and a factor score was constructed with differential weights assigned to them.

Another measure for social relations is employees’ views of their immediate line manager, which is based on a four-answer scale in the survey questions: “I trust my senior managers” and “I have a good relationship with my manager”. The variable measuring an employee’s relationship with their manager is constructed using these two indicators through an exploratory factor analysis.

Other control variables in this analysis are gender (male = 1), sector in which the employee works (private sector = 1), job security (permanently employed = 1) and employees’ number of children. Including two strongly correlated variables, namely age and labour-market experience, in the same regression model causes the problem of “multicollinearity”. Therefore, we used only the experience variable.⁴ In estimating the individual wage and job-status equations, we also controlled for whether some care institutions stand out from the rest due to their uniqueness in particular aspects. Regarding this, we did not find a statistically significant coefficient in any of the cases.

Results

Descriptive data

As shown in Table 1, employees born in Africa, Asia and Latin America (the Global South) are, on average, younger. Their educational attainments are approximately the same as those of native-born employees and lower than those of other foreign-born groups. Due to their age, their total job experience is less than that of other groups. The specific job experience in the care sector for individuals from Eastern European countries and from the Global South is significantly lower than that of others, which might indicate that they changed their occupational career after migration to Sweden (one of our interviewees described how she had changed her profession to be able to get a job). As is also evident from Table 1, there is differential access to WPSC for the different groups. However, the only statistically significant disparity is between those born in the Global South and all others.

Table 1. Summary of sample characteristics (% or mean).

Origin	Native-born	North West	Eastern Europe	Global South
Number and (percent)	192 (66,4)	8 (2,8)	13 (4,5)	76 (26,3)
Age	46,62	55,63	47,38	43,58
Gender-male (%)	11	25	15	12
Years of Education	12,73	13,87	13,62	12,63
Highest education in Sweden (%)	100	50	31	58
Job experience general (years)	28	36	28	25
Job experience care sector (years)	20	27	14	12
Permanent tenure (%)	91	100	100	91
Workplace social capital (mean)	,175	,230	,145	−,013
Private sector (%)	12	13	15	17
Income/months, adjusted for working hours (SEK)	27,301	27,509	25,450	24,038
Assistant nurses (vårdbiträde + undersköterska) (%)	70,3	37,5	61,5	93,4
Occupational Status	37	43	36	32
Good relationship with superior (mean)	3.4	3.5	3.5	3.1
Superior trustable (mean)	3.4	3.7	3.4	3.0

Another feature of the descriptive data is the significantly lower mean wage (adjusted for working hours) and occupational status of individuals from the Global South. These individuals had both lower formal positions (e.g. they were under-represented among nurses) and lower-status assigned positions within the sector as assistant nurses/nursing assistants (e.g. they were under-represented as team leaders).

The number of employees with a permanent job is high in our sample because the lists of employees provided by their employers almost exclusively included individuals with permanent tenure. Yet, our participant observations show that the percentage of those without permanent tenure (normally those born in the Global South) is higher than on the lists. It is possible that, in the case of part-time positions, we have the same kind of bias in our data.

Respondents were then asked about the impact of having a native (compared to a foreign) background on their prospects for promotion, better working conditions and a higher wage. Twenty-six percent of native-born and 43 per cent of foreign-born employees maintained that the former are at an advantage. Fewer than two percent stated the opposite, while others said that background makes no difference.

Respondents were asked to rate the extent to which their “wage corresponds to their efforts in the workplace” and their “duties correspond to their qualifications”. Relative to others, individuals born in the Global South perceived a lower correlation between their wages and their efforts as well as between their duties and their qualifications (not shown here, but available from the authors on request). Below, we examine whether employees from the Global South are, in fact, worse off concerning their occupational status and wages and what explains any gaps. We used Ordinary Least Squares regressions with the current occupational status and wage of the worker as our dependent variables, controlling for the set of control variables presented above.

The occupational status gap

Table 2 reports the regression coefficients for the equations that estimate the association between the control variables and the occupational status of employees. In Model 1, employees from the Global South have lower positions than others. In Model 2, employees’ education is the most decisive factor determining a person’s position in the workplace hierarchy, while general work experience and experience in the care sector in particular show no significant impact. Model 3 reveals that those with Swedish credentials obtain higher positions than those with diplomas from other countries. The gap between individuals from the Global South and native-born individuals decreases significantly due to the impact of this variable.

Table 2. Associations between individuals’ characteristics and their occupational status, OLS unstandardized coefficients (standard coefficients in parentheses), N = 289.

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Gender (male)	−0.43 (−0.014)	−1.29 (−0.043)	−1.07 (−0.036)	−1.146 (−0.039)	−1.07 (−0.036)	−1.011 (−0.034)
Migrant background
North West	6.24 (0.105)	2.69 (0.046)	4.16 (0.070)	3.906 (0.066)	(0.066)	3.57 (0.060)
Eastern Europe	−0.59 (−0.013)	−3.55 (−0.076)	−0.713 (−0.015)	−0.922 (−0.020)	−0.76 (−0.016)	−0.79 (−0.017)
Global South	−5.06** (−0.229)	−4.78** (−0.217)	−2.83** (−0.128)	−3.05** (−0.138)	−1.21 (−0.055)	−0.925 (−0.042)
Education (years)		3.39** (0.608)	3.32** (0.596)	3.35** (0.601)	3.00** (0.539)	2.97** (0.534)
Work experience, general (years)		−0.026 (−0.030)	−0.078 (−0.089)	−0.065 (−0.074)	−0.041 (−0.047)	−0.051 (−0.058)
Work experience, care sector (years)			0.09 (0.109)	0.085 (0.099)	0.066 (0.076)	0.079 (0.092)
Highest education in Sweden			3.27* (0.123)	3.25* (0.123)	2.92* (0.110)	2.97* (0.112)
Number of children				0.380 (0.025)	0.541 (0.035)	0.512 (0.033)
Private sector				2.152 (0.076)	0.798 (0.028)	0.673 (0.024)
Permanent tenure				1.204 (0.035)	0.726 (0.021)	0.674 (0.020)
WPSC					2.89** (0.297)	2.59** (0.266)
Relationship with manager						1.05** (0.107)
R^2 adjusted	0.055	0.425	0.438	0.440	0.513	0.522

Note: **denotes significance at 1% level and * at 5% level.

In Model 4, the control variables show no significant effect. In Model 5, we observe a significant and positive association between WPSC and occupational status. Those with more resources in their social networks have higher positions. What is remarkable is that by including this variable in Model 5, the gap between employees born in the Global South and native-born persons is no longer significant. We included this variable to understand whether the association between workplace relationships and the occupational status of employees is the same for native-born individuals and individuals from the Global South. The significant decrease in the position gap indicates that workplace relationships do mediate part of the differences in position between the two groups. In other words, employees in the former group not only have less WPSC (see Table 1) but also have lower benefits from this form of capital. The statistically significant coefficient of the interaction effects between being born in the Global South and WPSC confirms this (available on request).

In summary, employees born in the Global South have a lower occupational status partly because of their lower educational level and partly because of the lesser value accorded to their non-Swedish credentials. Fewer resources in their workplace social networks (less WPSC) and lower benefits from this kind of social capital explain another part of this gap.

In Model 6, we investigated the importance of a worker having a good relationship with his or her immediate superior and found that having a good relationship is positively and significantly associated with a worker’s position, even after control for educational level and WPSC.

The wage gap

The dependent variable in this section is employees’ wages. The variable is log-transformed; hence, the regression coefficients can be read as the percentage change in income with changes in the covariates. Table 3 reports the regression coefficients for equations in which the associations between the control variables and the (natural logarithm of) employees’ monthly wages have been estimated.

Table 3. Associations between individuals’ characteristics and (LN) wages, OLS unstandardized coefficients (standard coefficients in parentheses), N = 289.

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Gender (male)	−0.006 (−0.011)	−0.009 (−0.016)	0.000 (−0.001)	−0.001 (−0.002)	0.000 (0.000)	0.001 (0.002)
Migrant background
North West	0.014 (0.012)	−0.034 (−0.031)	0.001 (0.001)	−0.002 (−0.002)	−0.002 (−0.001)	−0.007 (−0.006)
Eastern Europe	−0.063 (−0.071)	−0.089* (−0.101)	−0.012 (−0.014)	−0.020 (−0.022)	−0.017 (−0.020)	−0.018 (−0.020)
Global South	−0.122** (−0.293)	−0.114** (−0.273)	−0.059** (−0.143)	−0.064** (−0.153)	−0.036 (−0.087)	−0.032 (−0.076)
Education (years)		0.030** (0.288)	0.028** (0.265)	0.028** (0.269)	0.023** (0.219)	0.023 (0.215)
Work experience, general (years)		0.002* (0.108)	0.000 (−0.010)	0.000 (−0.004)	0.000 (0.018)	0.000 (0.009)
Highest education in Sweden			0.081** (0.162)	0.080** (0.160)	0.075* (0.150)	0.076** (0.151)
Work experience, care sector (years)			0.003** (0.204)	0.003** (0.189)	0.003* (0.170)	0.003** (0.183)
Children				−0.001 (−0.002)	0.002 (0.006)	0.001 (0.004)
Private sector				0.047 (0.087)	0.026 (0.049)	0.024 (0.045)
Permanent tenure				0.033 (0.050)	0.026 (0.039)	0.025 (0.038)
WPSC					0.044** (0.237)	0.039** (0.212)
Relationship with manager						0.016 (0.088)
R^2 adjusted	0.074	0.148	0.184	0.185	0.230	0.235

Note: **denotes significance at 1% level and * at 5% level.

In Model 1, gender and country background are controlled for. The only statistically significant variable is being from the Global South. These individuals have, on average, a wage that is 12.2 per cent lower than that of native-born employees.

In Model 2, a higher level of education and greater general job experience have a positive and significant impact on employees’ wages. Introducing these control variables increases the wage disadvantage of care workers born in Eastern Europe (compared to native-born workers) from a non-significant 6.3 per cent to a statistically significant 8.9 per cent. The wage gap between care workers from the Global South and our reference group (native-born with native-born parents) decreases slightly from 12.2 to 11.4 per cent.

In Model 3, we include a dummy variable which indicates whether the employee has completed their highest level of education in Sweden, and the individual’s work experience in the care sector. The outcome of including the first variable confirms the results of previous studies: the economic rewards that employees receive for credentials obtained in Sweden are higher than for credentials obtained outside the country. Concerning the second variable, the coefficient for work experience in the care sector is positive. At the same time, by including this variable, the general work experience variable is no longer significant. This means that a person’s work experience in the care sector generates a higher wage, irrespective of how long the individual has been active in the labour market. This has negative consequences for some migrants, who may not have been able to find a job in their previous occupational sector. Introducing these two variables decreases the wage gap between employees from the Global South and native-born employees from 11.6 to 5.9 per cent.⁵

In Model 4, the impact of the variables having children (interruption in career), working in the private (versus public) sector and having a permanent position are examined. None of these variables show a significant effect. With regard to having children, women and men are separated in the analysis.

In Model 5, we included the WPSC of employees. There is a positive association between this variable and wages, even after controlling for educational level and work experience. By including the WPSC, the wage gap between native-born employees and employees from the Global South ceases to be significant. Thus, the lower wages of the latter group (after controlling for human capital variables) can be partly explained by their lesser access to WPSC. In addition, the statistically significant coefficient of the interaction effects (between being born in the Global South and WPSC) once again shows that the benefit of WPSC is lower for employees from the Global South compared to others. Disparities in access to and benefit from WPSC between care workers from the Global South and native-born workers demonstrate that the lack of extensive and influential social connections at the workplace can explain an important part of the wage and position divergences between these groups.

In the last model, we examined the effect of employees’ relationships with their immediate superior on their wage – this association is not statistically significant.⁶

In all the regression models (Tables 2 and 3), multicollinearity was controlled by ensuring that the variance inflation factor was always less than 2.

Social relations at the workplace: Digging deeper

As our analysis of the quantitative data demonstrates, human capital variables (i.e. lower educational level and the lesser value accorded to the non-Swedish credentials) explain a part of the salary – and status disadvantages of employees born in the Global South, while less WPSC explains another part. In the qualitative part of the study we dig deeper, in order to explore why employees in this group have fewer resources in their workplace social networks (less WPSC) and lower benefits from this kind of social capital.

To learn more, we now turn to our participatory observation and semi-structured interviews. We focus on aspects of workplace social relations that have a negative effect on the opportunities for these employees to collect WPSC. The basis for the analysis consists of the sections in our interview transcripts and field notes that were coded as being about or relating to employees being from different parts of the world.

An initial observation is that the terms “Swedes” and “immigrants” were frequently used in the workplaces in our study (and in many other Swedish workplaces, Boréus and Mörkenstam 2015) to distinguish between employees. The line between the two categories is blurred, but it is obvious that, in many contexts, this is neither related to citizenship nor how long a person has lived in Sweden but rather expresses a feeling of difference between “us” and “them” (Myrberg 2010; Boréus 2013, 299). Some of the native-born employees in our study not only differentiated between “Swedes” and others but also expressed a sense of distance from foreign-born colleagues:

… there are cultural differences. […] deep inside you, if you’re born in another country […] we don’t need to say that much to each other, so to speak, those of us who are Swedes. We understand each other anyway. (Informant 13)

Some native staff not only emphasized differences between themselves and their foreign-born colleagues but also problematized the latter group. Concrete problematization was often related to an alleged poor knowledge of Swedish among their foreign-born workmates:

Yes, my God, yes! You ought to know that the elderly care sector is the slop pail for those you don’t know where to place. They wouldn’t fit into any other industry. […] with this language comprehension, how the hell would it work? (Informant 27)

Although Informant 27 is likely to be expressing her frustration with the general treatment of the care sector here, her negativity towards foreign-born colleagues is also apparent. Some interviewees made other kinds of negative generalizations:

… there are almost only foreign workers here. […]. They’re not as outgoing as us Swedes, perhaps wanting to meet outside work, go out for a meal, go to the theatre, go dancing or, for example, go on a cruise together. Then they would need permission, I suppose, from their husband and relatives too. (Informant 4)

Informants 15 and 28 also made sweeping comparisons between Swedes/us and their non-native workmates, who were problematized in different ways, while informant 27 claimed that her workmates from different countries “look down on each other”, creating tension in the group.

The quotes above are drawn from interviews with native-born employees. No doubt their colleagues from the Global South are all too aware of the existence of negative feelings and prejudice. Some of them described their experiences of discrimination – for example, related to their use of the Swedish language:

Sometimes you sit and read reports on [the IT platform]. If someone has made a small mistake in writing, a spelling mistake … ‘Ha ha ha’, they laugh. […] I’ve told them that Swedish isn’t my native language. I try every day to improve. […] We all try. But it’s not easy. You get very hurt, you know. They say: ‘the foreigners’. I hate that word. Sure, I’m a foreigner but they don’t need to point that out all the time, or they say: ‘You have strange names’. (Informant 24)

Informant 11, from an Asian country, told a similar story about being laughed at when making an error in speaking or writing. The researcher also observed a situation when Informant 10, from Latin America, had her Swedish corrected in a not-very-friendly way. When the correct expressions were not used, Informant 10 remarked: “They think you’re stupid” (field notes, 12 November 2012). As several studies have demonstrated (Boréus and Mörkenstam 2015; Thuesen 2016), language tends to become an issue in multicultural workplaces. Although we found no significant correlation between self-estimated Swedish skills (see endnote 5) and status or wage, Swedish that is not spoken the way native speakers speak the language could be used as a negative marker in the sometimes hostile creation of “us” and “them”.

We were also told about experiences of how native-born employees in superordinate positions discriminated (Informants 2 and 29). In one ward, there was much commotion stirred up by a new team leader who was considered racist by several of our informants. She was accused of refusing to give foreign-born employees extra working hours when they asked for them, of using racist slurs and of favouring “Swedes”. The native-born manager of the ward also brought the aforementioned team leader up in the interview:

… we had a person, she’s been on several wards and […] there was some trouble, she has acted negatively towards our immigrant staff […] And then all of a sudden she got another position on another ward. Well, this woman is beaming now. You can see she’s enjoying herself now […] It works well. It feels good. (Informant 8).

Thus, although well aware of the problems this particular person had caused, such as acting “negatively towards our immigrant staff”, this manager considered the promotion to team leader a successful move. She had apparently not checked what consequences this change had had for the “immigrant staff”, which shows a lack of understanding or consideration regarding the problem of discrimination. The final quote is from an employee born in the Global South, who refers to the existence of two separate groups at her workplace:

A group of Swedes and a group of foreign [employees]. When I and [two other foreign-born employees] were involved in this conflict, the Swedes took advantage of it. (Informant 29)

The negative and prejudiced attitudes expressed by some native employees, the discriminatory behaviour of different team leaders, the hostility and scorn shown towards colleagues who could not speak Swedish in the way that natives could, the allegations of discrimination in other respects and the lack of forceful anti-discriminatory action by managers all make it easier to understand why employees from the Global South could have difficulty in becoming part of well-resourced networks dominated by natives. The interview guide included no explicit questions on racism and discrimination; the quotes above are all drawn from answers to other questions. Explicit questions about discrimination or racism would have been likely to have resulted in more descriptions of discriminatory incidents. Discrimination is thus at least part of the story: these employees do not always get invited in, might not get accepted when they try to take part or might be put off from even trying. Whether employees from the Global South are discriminated against for being perceived as culturally different, because (in some cases) they are visible Muslims or because (in some cases) they are darker than most Swedes, we do not know. It is likely that these grounds for discrimination add up for this group of employees.

Conclusions

The aim of this article was to add to the knowledge about workplace inequality between native and foreign-born employees. We set out to determine whether there was a gap in occupational status and wages between native and foreign-born care workers in elderly homes and, if so, how the differences could be explained. In line with previous research, we found that, after controlling for human capital variables, those from the Global South, who experience the highest risk of racialization in Sweden, have statistically significantly lower occupational positions and lower wages than others.

Human capital had a positive and significant impact on wages and position. Credentials obtained in Sweden paid off more than those obtained in other countries and work experience in the care sector was significantly more valuable than general work experience. This partly explains the subordination of migrant care workers who have work experience in their countries of origin, sometimes in higher-status employment. Another explanation relates to workplace relationships. Employees from the Global South had lower positions and lower wages partly because of their reduced access to WPSC and fewer benefits from it; furthermore, they reported poorer relations with their managers than other employees, something that partly explained their lower positions.

Our qualitative data elaborated upon these explanations: several employees from the Global South reported experiences of discrimination from both co-workers and superiors. We understand this to be at least a partial explanation for their poorer access to WPSC. This result corresponds to those of Jönson and Giertz (2013), who found that care workers from outside the Nordic countries ran a higher risk of perceiving that they were not appreciated by their co-workers. However, in contrast to Jönson and Giertz, who answered no to the question “Are they disadvantaged?”, we state “Yes they are” when referring to employees from the Global South. These employees also feel that they have fewer opportunities to influence their working conditions than others (Behtoui et al. 2017).

The results indicate that further research on inequality in organizations (inequality regimes; see Acker 2006) is needed). Such research is necessary to grasp how the patterns of inequality found in research considering the labour market is produced in organizations. First, the need to consider the multiple factors and processes involved in reproducing the subordinate positions of employees from the Global South in the workplace should be recognized. Second, the effect of access to WPSC must be corroborated by further research. Third, an important next step is to learn more about how WPSC is accessed and by whom, which will require qualitative data at a micro level.

Even in Sweden, with a more formalized care sector and a less precarious situation regarding the right of abode for migrant care workers than many Global North countries, employees who tend to be racialized experience harsher conditions. This is a serious societal problem and an issue for further research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Vetenskapsrådet [grant number Dnr 2010-1499].

Notes

1 The four authors of this article have each contributed to the same extent to the project and to the article.

2 https://skl.se/ekonomijuridikstatistik/statistik/personalstatistik/personalenidiagramochsiffror/tabellerkommunalpersonal2015.8832.html.

3 Although small numbers of employees in this study were born in the Global Northwest or in Eastern Europe, we kept them separate to study the potential differences between migrant workers from different parts of the world. Non-significant statistical differences in salary and status between workers from Northwest and Eastern Europe and the reference group might be a consequence of the restricted number of individuals from the first two groups in our data.

4 In addition to age, the length of time that a foreign-born employee has been in Sweden is another factor that may affect his or her salary and job status. However, due to the strong correlation of this variable with other employee characteristics in our models (the length of a person’s general labour market experience; work experience in the care sector; whether an individual's highest education was completed in Sweden), an inclusion of the variable ends in a high variance inflation factor (VIF > 10) of the coefficient, which indicates a multicollinearity problem. Consequently, length of time in Sweden was excluded from our estimations.

5 We asked all employees to self-estimate their knowledge of the Swedish language. The statistical non-significance of this variable did not affect the other coefficients of Model 3 in Tables 2 and 3 (details available on request).

6 To compare the impact of human capital and WPSC independently, we ran two separate regressions with variables indicating each of these types of capital. Human capital and WPSC explained 18 and 13 percent of the variation in salary, respectively.