Constructions of the ideal elite employee: A content analysis of job advertisements for positions in the Norwegian upper class

Previous research has shown that the upper class has a high degree of self-recruitment. Simultaneously, research on job recruitment suggests that there has been an increase in the importance of ‘soft skills’. This article investigates the connection between these two phenomena by looking at elite employers’ constructions of ‘the ideal employee’ and how this may contribute to class reproduction. By analysing 150 advertisement texts linked to positions in the upper class and fractions within it (the cultural, professional and economic upper classes), it explores which qualifications and characteristics are required. The analysis indicates that ‘soft skills’ play a substantial role in recruitment in all upper-class fractions, suggesting that ‘hard skills’ do not suffice to achieve an upper-class position. Indeed, the fact that many of the personal characteristics are linked to modes of being typical of an upper-class habitus makes it harder for people from lower in the class structure to match the description of the ‘ideal candidate’. Furthermore, this article finds the following three distinct employee types: the authoritative leader is idealized in the economic fraction, the dedicated worker in the professional fraction and the coaching leader in the cultural fraction. This article also points to examples of gender stereotyping that may function as contributing factors in reproducing occupational segregation.


Introduction
The upper class has a high degree of self-recruitment.Achieving a position of employment at the top of the hierarchy is thus more likely if someone is born in the higher classes (Flemmen et al., 2017;Friedman et al., 2015;Gulbrandsen et al., 2002;Hansen, 2014;Savage et al., 2013;Toft, 2018).This remains the case even as we have witnessed a mass increase in people from different class backgrounds completing higher education (Reay et al., 2005;Tapper and Palfreyman, 2005).This indicates that something beyond higher education is considered when employment positions in the upper echelons of the labour market are filled.To understand more about the processes leading to this skewedness, this article investigates what employers emphasize and value when recruiting employees to such positions.Investigating recruitment from a class perspective provides insight into how jobseekers originating from the upper levels of the class structure tend to be favoured in hiring processes for top jobs and, by the same token, how jobseekers originating from lower in the class structure tend to be excluded from them.More generally, studying recruitment provides insight into how the allocation of social benefits and resources is assessed, justified and legitimized in the upper echelons of society.
This article looks at recruitment data from upper-class positions in Norway.It explores the initial barrier that many jobseekers encounter when looking for employment -publicly available job advertisements.These texts tell us something about what employers are looking for in a candidate and they indirectly describe employers' understandings of the ideal employee.Class analysis has mainly focused on hierarchical differences but to understand the nuances of recruitment, it is also important to study how classes can be internally differentiated.This article thus explores 150 advertisement texts for positions in the Norwegian cultural, professional and economic upper classes (50 texts from each class fraction).There are the following two main research questions: (1) which qualifications and characteristics are emphasized and valued by which types of employers and (2) are class fraction, sector and gender composition in industry aspects that affect upper-class job advertisements, and if so, how?
Research on recruitment has mainly been quantitative, focusing on discrimination based on gender, ethnicity, disabilities, mental health, personal contacts and sexual orientation (e.g.Bertrand and Mullainathan, 2004;Bjørnshagen andUgreninov, 2021a, 2021b;Granovetter, 1995;Roth, 2006;Tilcsik, 2011).Aspects related to class, culture and lifestyle have been largely overlooked.There are some important exceptions to this; however, Rivera's (2015) study of recruitment and promotion practices in US elite companies and Friedman and Laurison's (2019) study of class barriers in UK elite occupations.These studies explore recruitment in countries with large class divisions and there seems to be a gap in the research concerning the way in which class inequality is reproduced in upperclass occupations in the more egalitarian Scandinavian countries.With the insights from cultural class analysis (see Devine and Savage, 2004) as a backdrop, this article examines how culture and lifestyle aspects can be relevant in our understanding of the job market.

Mass higher education and its impact on recruitment
The process of recruitment can be understood as a social closure mechanism and the closure of mobility opportunities has been highlighted as crucial for class formation (Giddens, 1973).One way of reaching upwards mobility is through higher education.In the second half of the 20th century, we witnessed the rise of mass higher education (Scott, 1995;Tapper and Palfreyman, 2005).In 2019, 34.1% of the Norwegian population had attained higher education, compared to 7. 4% in 1970(SSB Statistics Norway, 2019).In Norway, as most universities and higher education are free and open to all, the rise of mass higher education can be a great equalizer and stepping stone to upwards mobility.Nevertheless, Wiborg and Hansen (2018) have found family background to have had a declining impact on educational attainment in Norway, but an increasing impact on income and wealth.Furthermore, the upper echelons of the Norwegian class structure retain an important core of inheritors (Flemmen et al., 2017;Gulbrandsen et al., 2002;Hjellbrekke et al., 2007;Mastekaasa, 2004;Strømme and Hansen, 2017).Previous research has also shown that there are class-based differences in income (Friedman and Laurison, 2019) and in the temporal aspect of remaining in an upper-class position (Toft, 2018).
This rise of higher education also seems to have had an impact on and has partly shifted the way employers consider and recruit employees.Brown and Hesketh (2004) argue that the rise of mass higher education has caused a 'legitimation crisis' for the workforce.The increase in students in higher education, they argue, has placed in doubt the previous belief that there was a limited pool of talent available for both higher education and managerial positions.In this new competition for top positions in the workforce, they suggest that what separates the winners from the losers is now based more on social and cultural experiences and the way jobseekers manage their employability.Following this argument, who you are now matters as much as what you know.To gain access to these jobs with their tough entry demands, jobseekers now have to rely increasingly upon their 'personal capital' -a combination of hard skills (e.g.credentials, work experiences) and soft skills (e.g.social skills, appearance).The emphasis on soft skills in recruitment implies that the recruitment process has become increasingly personalized, as the 'whole person' is exposed to it and evaluated.This does not mean that grades, credentials and work experience have decreased in importance, rather that the boundaries between education and everyday life are blurred.This allows room for highlighting extracurricular activities and similar as competences benchmarked as indicative of elite employability.The importance lies not only in which experiences the candidates possess but also in how well they package these experiences to elite employers -creating a 'narrative of employability' (Brown and Hesketh, 2004).
A number of researchers have found an increase in the importance of soft skills in recruitment (Brown and Hesketh, 2004;Calanca et al., 2019;Rivera, 2012).This also allows an opening for, and corresponds with, the homophily principle suggested by Lazarsfeld and Merton (1954).This principle states that similar people attract and breed connections at a higher rate than dissimilar people do.This shift in recruitment casts a shadow on the meritocratic ideal, where goods and positions are distributed according to abilities and qualifications.According to Brown and Hesketh (2004), 'one of the principles of the meritocratic competition was the creation of a competition played out in purposebuilt areas (schools, colleges, universities) where social differences were to be removed to reveal differences in innate character and ability ' (p. 36).Opening up the recruitment process to the entire 'self' suggests that the importance of meritocratic achievements may be on the decline.It also suggests that mobility opportunities may be decreasing.

Investigating recruitment and job advertisements
Even though we know that the upper class has an important core of inheritors, we do not know much about how these closure mechanisms take place in job recruitment.There are also a few empirical studies analysing whether and how different steps in recruitment processes may contain class bias.Class-based discrimination in recruitment has thus been relatively underexplored (Jacobs, 2003).
Previous research on recruitment has discovered differential treatment or discrimination occurring based on their specific focus (e.g.gender, ethnicity, disabilities).Studies of labour markets also tend to focus on low-waged rather than high-paying jobs (Rivera, 2015).In Norway, researchers have studied justice and discrimination in recruitment processes, mainly focusing explicitly on the effects of ethnicity, immigration and gender (e.g.Midtbøen, 2015;Petersen and Togstad, 2006;Rogstad and Sterri, 2016).
There are a few studies of recruitment processes in terms of class, culture and lifestyle differences.Some important exceptions are Rivera's (2015) study of elite recruitment in professional service firms in the United States and Friedman and Laurison's (2019) study of 'the class ceiling' in the United Kingdom (see also Ashley and Empson, 2013;Ingram and Allen, 2019;Jackson, 2001;Jacobs, 2003;Koivunen et al., 2015).Rivera (2015) finds that elite employers typically recruit from the top schools in the United States -with clear class-biased consequences.In addition, Rivera (2012) shows that employers base their hiring decisions on what she calls 'cultural matching' -cultural similarities between applicants and employers as a mechanism for assessing candidates to hire.Friedman and Laurison (2019) discover a substantial class pay gap within British elite occupations.They also reveal a 'class ceiling', while detangling how people from working-class backgrounds have a harder time reaching the top in the organizational hierarchy compared to their colleagues from upper-class backgrounds.Their findings reveal some important mechanisms for such class bias; being educated at a prestigious university, being able to rely on 'The Bank of Mum and Dad' and having a sense of 'fit' in an elite work environment.These mechanisms all favour employees from upper-class upbringings.
These studies offer important findings for understanding social closure in American and British upper-class positions.Although these findings reveal mechanisms for understanding class bias and mobility opportunities in a broader sense, they are not necessarily transferable to the Scandinavian countries.An important finding in the studies mentioned earlier is connected to education (top firms recruit from top universities).Comparatively, the Scandinavian countries are more egalitarian in the sense that they lack some barriers that exist in the United States and/or the United Kingdom, for example, high tuition fees and clear hierarchies between universities.Norwegian universities do not have the same elite status and most universities are public and free for everyone.While hierarchies exist between universities and university colleges in Norway, they are less pronounced and institutionalized than in the United Kingdom and the United States.This article aims to uncover what upper-class employers claim to be looking for in an ideal employee and whether or not we find class-biased markers in advertisement texts that can explain some of the self-recruitment in the top echelons of the Norwegian class structure.By looking explicitly at internal class differences in the upper class, one can examine in a more nuanced way whether upper-class recruitment is internally differentiated.
Studies investigating job advertisements have so far appeared in a limited body of sociological literature, and research on job advertisements typically has a gendered focus or is based on a managerial or human resources (HR) perspective.Studies have found differences in both framing and wording in job advertisements as following gendered constructions (e.g.Askehave and Zethsen, 2014;Gaucher et al., 2011).Furthermore, the class and inequality dimensions to this approach have been explored to a much smaller extent.Studies that do investigate skill requirements, as communicated in job advertisements, have found that personal and interpersonal skills -or soft skills -are of great importance for employability (e.g.Calanca et al., 2019;Grugulis and Vincent, 2009;Jackson, 2001;Kuokkanen et al., 2013).A direct link has been found between the skills most frequently mentioned in job advertisements and the skills deemed most important by employees in the specific fields (Bennett, 2002).In addition, job advertisements shape our perceptions of employee ideals, and through repetition, they create a discourse of these ideals that is mediated to jobseekers (Kuokkanen et al., 2013).This makes job advertisements a valuable component in the investigation of the skills employers require and prefer (Tan and Laswad, 2018).
Studies using job advertisements as research data have varied in method.Several studies are based on large-sample quantitative data and others are based on a smallersample qualitative content analysis.This article uses content analysis based on a mediumsized sample, following other studies that have done the same.This article follows these earlier studies by focusing on skill demand and gendered wording, but also adds to our knowledge of job advertisements and recruitment; by studying advertisements for upper-class positions specifically, this study contributes to the existing knowledge of job advertisements and recruitment by providing insight into what constitutes elite employability.It also adds to our knowledge of upper-class recruitment, as previous studies in this field have overlooked job advertisements as research data.

Data and methods
The empirical analysis is based on data from a content analysis of 150 job advertisement texts in Norway.The texts were obtained from the most comprehensive overview of positions vacant in the country, found on the website of the Norwegian Labour and Welfare Administration (nav.no).The texts were gathered during a 3-month period, starting 4 June 2020.This study explores positions in the Norwegian upper class and is based on the Oslo Register Data Class scheme (ORDC; Hansen et al., 2009).ORDC draws on Bourdieu's (1984) notion of the social space where classes are differentiated by two dimensions.The vertical and hierarchical dimension differentiates by capital volume (economic and cultural) and separates the lower and the upper classes.The horizontal dimension differentiates by capital composition: a cultural fraction (primarily cultural capital), an economic fraction (primarily economic capital) and a balanced fraction (balanced capital profile, mainly professional occupations).The 150 advertisement texts collected consist of 50 sampled and strategically chosen texts from each of these class fractions (see Table 1 for further sampling strategies).Both class analysis and research on recruitment have focused on hierarchical differences or classes as uniform entities.More recent studies have, however, found clear internal variations within the upper class (e.g.Flemmen et al., 2017;Toft, 2018), thus illustrating the fruitfulness of studying intraclass variations.ORDC allows us to do so, thus providing an opportunity to investigate potential fraction-specific nuances of recruitment.
All of the 150 advertisement texts were imported into NVivo and created as unique cases.In order to easily examine connections in the empirical data after coding, different case attributes were created: (1) class fraction (cultural, professional, economic), (2) gender composition in industry (male-dominated, female-dominated, equally represented) and ( 3) sector (private, public, organization).
The coding in NVivo was performed in two phases.Phase I involved open coding, consisting of making codes based on meaning emerging directly from the data (e.g.'strategic', 'flexible', 'IT skills').This phase was connected to the first part of the research question where I considered in detail which qualifications and characteristics were emphasized.The coding was therefore closely linked to the specific characteristics mentioned in the texts.This phase resulted in 137 codes.Each code and its content were reviewed, and any mismatched coding and/or mistakes were corrected.Phase II involved axial coding, where open codes and their flat structure were rearranged into a more hierarchical structure.This phase resulted in four main codes connected to the qualifications and characteristics emphasized: (1) education (e.g.'higher education', 'MA'), (2) work experience (e.g.'management experience', 'experience from transformation processes'), (3) other qualifications (e.g.'language skills', 'understanding of public administration') and (4) personal characteristics (e.g.'results-driven', 'networking skills').One additional fifth code was created ('inclusion at the workplace'), including information about whether or not the text mentioned anything explicitly about inclusion at the workplace (e.g.'we encourage qualified candidates to apply regardless of gender, functional ability and cultural background').The second phase of coding also included the merging of similar codes and the relabelling of a number of the codes.
The advertisement texts used as empirical data in this study are publicly available information.There are some limitations to such data.There are many procedures involved in a recruitment process, both formal and informal, and these texts only show one of the first steps.Furthermore, positions vacant in the private sector do not have to be publicly advertised.Since job advertisements are publicly available information with a known sender, this might place pressure and expectations on companies to appear inclusive and open-minded.Firms also have to act in accordance with Norwegian law.One implication of this is that if we do find markers of class bias in these texts, there is reason to think that this may affect the rest of the recruitment process too.

Findings
As described in the 'Data and methods' section, four main codes were developed to illustrate which main qualifications and characteristics were desired by employers: education, work experience, other qualifications and personal skills.These codes were found in most advertisements (76% or more).Personal skills were the only category that was included in 100% of the cases.Figure 1 illustrates the relative importance of each of these codes for each class fraction.*The percentages refer to the proportion of texts in the specific class fraction that includes the code mentioned, relative to the coded material for the fraction in total.
As Figure 1 illustrates, personal skills accounted for most of the characteristics demanded by employers (62% of the coded material overall).Personal skills are most frequently mentioned in the economic fraction and least in the cultural.Higher education was required in 80% of the cases, but only made up a small proportion of the overall employee description.Education is referred to most frequently in the professional fraction and least in the economic.Work experience was required in 91% of all cases, but least in the professional upper class (78%).Other qualifications make up a substantial part of the characteristics required (especially in the professional fraction) but accounted for only 18% of overall coding.
Advertisement texts explicitly mentioning inclusion as a value were more highly differentiated and systematically split by class fraction and sector.Inclusion in the recruitment process and/or at the workplace in general was mentioned most frequently in the cultural fraction (62%) and professional fraction (60%).The economic fraction mentioned inclusion in only 20% of cases.Inclusion was mentioned by far the most frequently by public sector firms (72% of cases) compared to private sector firms (12% of cases).Private sector firms in the cultural fraction mentioned inclusion practices most frequently (33%) but this was almost non-existent in the economic (3%) and professional (0%) fraction.

Ideal employee descriptions
To gain a sense of how employers' requirements and desires for an ideal employee cluster in each class fraction, similar codes are grouped together.These groups of codes are presented as different themes.The themes and the specific codes are listed according to frequency and only the most frequently mentioned codes are listed below.Together, these themes illustrate an ideal employee for each specific upper-class fraction.This way of grouping ideal employee requirements is inspired by Askehave and Zethsen (2014).
Economic fraction: the authoritative leader.The following six recurring themes have been identified to describe the ideal employee in the economic fraction: (1) Education and professional experience.Employees are in most cases required to have some form of higher education.Furthermore, having several years of relevant experience is desired in most advertisements; (2) Leadership qualities.All positions in the economic fraction are leadership positions, and this is clearly reflected in the advertisements.This is most frequently referred to as possessing 'leadership skills' or having previous experience from leadership positions.Employees are also supposed to be confident (in a leadership position) and have the ability to motivate; (3) Goal oriented and results driven.A significant proportion of the advertisements required employees to be driven by results and/or goals.By the same token, they are expected to have an ability to improve and develop (both their team and themselves), possess implementation skills and be strategic; (4) A 'gogetter'.Several recurring adjectives are used to describe the ideal employee as a 'gogetter': being ambitious, an achiever, driven, on the offensive and innovative; (5) Sociable cooperator.Being social and cooperative is highly valued by these advertisements.Possessing good networking skills is especially important, but the ideal employee is also expected to be enthusiastic and energetic, social and sociable, and to have cooperation skills and communication skills; (6) A hard worker.Finally, being hard working is highly valued in this fraction.This is typically described as having a high capacity for work and being structured, but also as being analytical and vigorous.
The ideal employee in this fraction can be termed the authoritative leader (Goleman, 2000).The ideal leader type in the economic fraction has many classical leadership traits -confidence, vision and ambition.There is also a more social component to this leader type.The authoritative leader motivates employees and is aware of the importance of cooperating well with them.
Professional fraction: the dedicated employee.The following five recurring themes have been identified to describe the ideal employee in the professional fraction: (1) Education and professional experience.Having completed a higher education is explicitly underlined in most cases, as is specialization in some cases.Having previous professional experience from the specific field is desired in most cases but is less of a requirement than in the other two fractions; (2) Sociable cooperator.Being both sociable and cooperative are especially valued in this fraction.Ideal employees are also described as being enthusiastic and energetic, to possess communication skills, while being flexible, inclusive, unifying and service-oriented; (3) 'Personal aptitude'.Over half of the advertisements in this fraction explicitly mention that they emphasize 'personal aptitude'.Typically, this term is not elaborated on in the texts themselves; (4) Productive and selfsufficient.The ideal employee is expected to manage a heavy workload and have a high work capacity.He or she is also expected to be independent, structured, solution-oriented and goal-oriented; (5) Responsible.Advertisements in the professional fraction emphasize the skill of being responsible.Connected to this is also the ability to be empathetic, caring and confident.
The ideal worker in the professional fraction can be termed the dedicated employee.Employees in this fraction are expected to be dedicated to the workplace and the profession, both in terms of being hard working and a responsible employee overall.The dedication to the workplace is also illustrated in the social expectations, as employees are expected to devote enthusiasm and service to the job.
Cultural fraction: the coaching leader.The following five recurring themes have been identified to describe the ideal employee in the cultural fraction: (1) Education and professional experience.Having attained higher education is a prerequisite in most cultural upper-class positions.Furthermore, most positions require significant experience from the relevant field; (2) Sociable networker.Having networking skills is one of the most highly ranked personal skills in the cultural fraction.Networking skills are mentioned substantially more frequently than in the professional and economic fraction.Ideal employees are also expected to be social and sociable and possess cooperation and communication skills.Being energetic and enthusiastic, flexible and trustworthy is also frequently mentioned; (3) Abilities for improvement.An ability to improve, both oneself and one's employees and/or colleagues, is a highly valued skill.This is typically mentioned in the advertisements as having development or implementation skills and being motivating; (4) Goal-oriented.Being goal-oriented and results-driven is an important skill in the cultural fraction.By the same token, ideal employees are expected to be structured, strategic and ambitious; (5) Work-specific personal skills.The ideal employee in the cultural upper class is expected to possess a variety of job-related skills.Most typically, having a high capacity for work and possessing analytical skills are mentioned.Other traits are also valued, such as being vigorous, independent and having personal aptitude.
The coaching leadership style (Goleman, 2000) is centred on developing others.This leader type also shares some similarities with the authoritative leader, like being goaloriented, but it is more oriented towards a common goal and developing employees and the workplace.
As the descriptions illustrate, education and previous experience are valued across all class fractions.Furthermore, personal characteristics or 'soft skills' are most dominant in all three fractions.Applicants' social skills are also of great value in all the fractions.Approximately, one-third of advertisement texts across all class fractions require applicants to be sociable.At the same time, there are also significant fractionspecific differences.In the economic upper class, possessing leadership qualities and being goal-oriented and a 'go-getter' are more typical than in the other fractions.The professional upper class especially values 'personal aptitude', being productive and responsible.Finally, in the cultural fraction, networking skills are mentioned in more than two-third of the texts and substantially more frequently than in the professional and economic fraction.This fraction also places much emphasis on abilities for improvement and being goal-oriented, thus more similar to the economic than the professional fraction.

Gender and social class
Gender aspects seem to be of importance for understanding advertisement texts in the upper class.Workplaces explicitly mentioning inclusion in their advertisement texts demonstrate a substantial gendered dimension.In the female-dominated sectors, inclusion is mentioned in 74% of cases, compared to only 4% of the male-dominated ones.As previously mentioned, private sector companies (that are male-dominated) explicitly refer to inclusion in only 12% of cases, compared to 72% in the female-dominated public sector.Therefore, the gender differences can partly be explained by differences in employment by sector.Still, the percentage is considerably lower in the male-dominated sectors, suggesting that gender representation by sector has an impact on valuing inclusion as well.Another difference with regard to gender-specific sectors is the requirement for leadership experience.Leadership experience is required in 73% of the male-dominated sectors, compared to 38% of the female-dominated ones.This can partly be attributed to the fact that leadership positions are more frequent in the male-dominated private sector and in the economic upper class in general than in the female-dominated public sector and especially in the professions.However, these findings do illustrate a gendered dimension too.
In addition, the empirical data show connections between typical gender ascriptions and gender dominance in the sector.At a general level, the term 'personal aptitude' is far more highly valued in the female-dominated sectors (54%) compared to the male-dominated ones (12%).Table 2 shows in more detail specific personal character traits required by the different sectors.
As Table 2 shows, there is a clear connection between gender ascriptions and gender representation by sector.Typical male ascriptions are more frequently mentioned in male-dominated sectors and vice versa.This feature is also seen in the distribution between the private and public sector, whereas the male-dominated private sector most frequently mentions male ascriptions and vice versa in the female-dominated public sector.

Discussion
Returning to the research questions, which qualifications and characteristics are emphasized and valued by which types of employers?Are class fraction, gender composition, industry and sector aspects that affect upper-class job advertisements, and if so, how?By looking at the prevalence of the personal characteristics required (62% of all criteria coded), it is clear that personality traits are of great importance in recruitment processes in the Norwegian upper class.The empirical data show that personal characteristics are included in all the cases and social skills in 97% of the cases.Comparatively, Jackson's (2001) study of British job advertisements found that only 50% of the cases from the managerial and professional class requested social skills or personal characteristics.This suggests that personality traits are of no less importance in more egalitarian countries than in countries with greater inequalities.Jackson (2001) argues that the very existence of personal characteristics in these texts partly contradicts the meritocracy thesis because many of the character traits are ascribed and can rarely be achieved.This focus on ascribed characteristics makes people from certain class backgrounds more likely to The percentages refer to the proportion of texts in the specific sector that includes the code mentioned.Sectors coded as 'equally represented' are not included.
match the employer's version of an 'ideal employee' than people from other class backgrounds (Brown and Scase, 1994).Furthermore, this produces room for several kinds of class biases in the recruitment processes (Jackson, 2001).The findings illustrate three distinct types of fraction-specific employees.While the economic fraction values a classical authoritative leader, the cultural fraction prefers a leader who overall is more dedicated to developing and implementing changes for employees and the workplace.The professional fraction values a dedicated employee.While these three differ in style, they also embody similarities; they all have high expectations of their employees, well beyond their expertise and hard work.
The findings in this article show that the somewhat ambiguous term 'personal aptitude' is especially prevalent in the professional fraction.It is difficult to set objective criteria as to what defines or constitutes 'personal aptitude'.Therefore, one can argue that the focus on 'personal aptitude' is in many cases of a more subjective nature.Evaluating and assessing a candidate's 'aptitude' or 'sense of fit' might then be up to the recruitment panel.According to the homophily principle, subjective assessments of job candidates will increase the likelihood of upper-class positions being filled by people from upper-class backgrounds.From a Bourdieusian perspective, a matching of habitus or a 'cultural matching' can be expected between the candidate and the company (Rivera, 2012).Based on this, the more the recruitment process is based on unstructured, informal and/or subjective evaluation, the more potential there will be for unequal treatment of the candidates and for personality preferences and class-based discrimination.This is also supported by previous research on recruitment (Ashley and Empson, 2013;Friedman and Laurison, 2019;Ingram and Allen, 2019;Rivera, 2015).A similar logic can be used about the requirement to be sociable.Some of the specific personal characteristics required by the advertisements also demonstrate connections to class habitus.Several advertisements prefer candidates who encompass 'leadership skills', who have 'leadership potential' or who have leadership experience.While not excluding candidates from lower-middle or working-class backgrounds, the upper-class habitus can arguably encompass and encourage leadership and self-confidence.As Jacobs (2003) puts it, 'the very nature of a dominant economic/social elite is that they dominate positions of leadership'.A similar notion can be presented with regard to character traits such as being ambitious, having drive, being results-oriented and strategic.These personality traits have been shown in this article to be especially highly valued by the economic upper class and these are values previous researchers have connected to an 'upper-class lifestyle' (Ingram and Allen, 2019;Rivera, 2015).
Networking skills -including having a large network and being a good networkerhave been shown to be of importance in the economic upper class, but especially in the cultural upper class.This indicates that social capital may be of greater importance in the cultural fraction compared to the other fractions.As 74% of the cultural upper-class cases are from the public sector -where one might think that networking and profits are of less importance than in the private sector -this is rather surprising.Regarding the recruitment process, the criterion of networking skills -or social capital -can have a negative impact on occupational integration by class.Emphasizing networking skills, or social capital, will favour those who know key players in the relevant field or those who have the means or preconditions to get to know them.Previous research has shown that people from the highest classes possess the highest levels of social capital, both with regard to formal and informal networks (Pichler and Wallace, 2009).Because of this, people from upper-class backgrounds are given an advantage in these recruitment processes.The inclusion of networking in advertisements can thus produce room for a wide range of possible class biases.
Gendered dimensions are also shown in this article.The gendered ascriptions shed light on specific, potential barriers to women entering male-dominated sectors and vice versa, also resonating with a quantitative study of UK advertisements (Calanca et al., 2019).Previous research has shown that women are less likely to think that they match or fit with the job description when the advertisement includes traditional male ascriptions.Furthermore, they are also reluctant to apply for these jobs (Askehave and Zethsen, 2014;Born and Taris, 2010;Gaucher et al., 2011;Taris and Bok, 1998).Calanca et al. (2019) have found that gender stereotypes in job advertisements are connected to salaries.Soft skills that corresponded to traditionally female ascriptions were predominantly associated with wage penalties, whereas soft skills corresponding to male ascriptions predicted wage premiums.The gender stereotypical language found in the Norwegian job advertisements might then illustrate how gender and social class are interconnected -a well-established interconnection in gender research (e.g.Reay, 1998;Skeggs, 1997).
Finally, the aspect of mentioning inclusion at the workplace varies by class fraction, sector and gender composition.While public companies have stricter rules to follow in being inclusive in recruitment, there is nothing to prevent private sector firms from highlighting inclusion in their advertisements.The lack of this may be a sign of inclusion being of less importance in the sector, at least compared to the importance of recruiting 'the best candidate for the job'.It is also worth noting that the mere act of inserting 'inclusion' into an advertisement does not automatically mean that the workplace itself has an inclusive working environment.It may, however, give various demographic groups the impression of being 'suitable for the job' and therefore make them more likely to apply in the first place (see especially Taris and Bok, 1998).

Implications
The findings from this study contribute to our knowledge of recruitment and class bias.They have several implications, methodically, theoretically and in terms of recruitment.Methodically, this study illustrates how one can utilize publicly available information that is not generated for research purposes.These sources can be understood as particularly 'authentic', as they are published independently and are not research driven.By using texts like these, one can learn about how workplaces construct their ideal employees.Furthermore, since we have found several openings for class bias in these texts that must abide by Norwegian laws, class bias is likely to be taking place later in the more informal parts of the recruitment processes as well.
Theoretically, the findings imply that the cultural aspect of class analysis has a place in research on the workforce, recruitment and discrimination.By including cultural and lifestyle aspects in studies of recruitment, we can gain a broader understanding of the mechanisms at work when a new employee is chosen.In addition, this study shows that the class fractions differ in the character traits required in their ideal employees, illustrating the fruitfulness of the horizontal dimension of Bourdieu's notion of the social space.
Finally, these findings have several implications that may be of importance for recruitment and human resources management.Workplaces should be aware that the focus on personal characteristics, social skills and 'personal aptitude' opens up for a wide range of possible class biases, and this may influence who applies for the job and who is ultimately given a job offer.In addition, these personal characteristics are harder to discern from a CV, leaving more room for decisions based on personal evaluations and 'gut feelings'.This amplifies the chance of choosing a candidate who is similar to oneself, but not necessarily the best qualified candidate.This suggests that people from the lower echelons of the class hierarchy might encounter a second barrier after achieving higher education and a potential mechanism for social closure in the upper class.Lowering the number of ambiguous and/or vague personal characteristics required, and upping the focus on more tangible qualifications, might be a fruitful solution to this.

Conclusion
This article has examined advertisement texts for job positions in the Norwegian upper class and has shown that there are aspects of these texts that open up for a wide range of potential class biases.The contributions of this article can be summarized as follows: (1) In addition to qualifications, higher education and previous experience, personal characteristics play a substantial role in constructions of the ideal elite employee in all class fractions.Almost two-third of the job criteria required consisted of personal characteristics.This indicates that qualifications, experience and higher education do not suffice to achieve an upper-class position.2) There are internal class differences in the job criteria requested and in the emphasis on different aspects of candidates.The economic fraction dominates the criteria that previous studies have linked to an 'upper-class lifestyle' (e.g.being confident, leadership potential, having drive) and the economic fraction is also the one that mentions inclusion the least.However, the cultural fraction most frequently requires 'networking skills' (i.e.social capital).The term 'personal aptitude' is mentioned by more than half of the advertisements in the professional fraction, significantly more frequently than in the other two fractions.This term can give workplaces and recruitment personnel a great deal of leeway in choosing a candidate as they see fit.It can also make the recruitment process less transparent and harder to understand for the candidates, as the term is quite vague.(3) Differences in upper-class advertisements vary systematically in terms of several aspects, and this analysis has shown that class fraction, gender composition and sector are all relevant for understanding the potential closure processes these may lead to.In addition to class fraction, as discussed earlier, gender composition by sector seems to be of great importance.This article has shown several examples of gender stereotyping, which may be a contributing factor in the reproduction of occupational segregation (Askehave and Zethsen, 2014;Born and Taris, 2010;Gaucher et al., 2011;Taris and Bok, 1998).
While these findings illustrate important aspects of recruitment in the upper class, there are some limitations to these data.First, this study is based on a relatively small N.While this allows us to acquire in-depth knowledge about advertisements in different class fractions, it limits our possibilities for generalization.Second, the focus in this study is limited to the advertisement texts alone.As this is just one of the first steps in a recruitment process, we cannot know if and how these advertisements are of importance later in the process.However, as we do know that the wordings in advertisements affect who applies, it is an important aspect to study in-depth.It should also be noted that it is not necessarily the case that workplaces make the potential for class bias consciously or intentionally.Rather, these can be understood as unconscious processes that cause inequality to be reproduced over time.

Figure 1 .
Figure 1.Qualification and skill demand, by class fraction.

Table 1 .
Sample of employment positions.
a Many positions in this class fraction do not specify income.Some online searching was thus needed to ensure that the texts included actually made the positions upper class.This was done by examining the income of the previous incumbent of the position, what other employees in management positions in the same firm earned or what others with similar titles earned in similar firms.The income typically varied between 1.5 and 2.5 million NOK (€148,000-246,000).With other allowances included (e.g.bonuses), many of the positions have an actual income of closer to 4-5 million NOK (€393,000-492,000).

Table 2 .
Personality traits and gender representation by sector. a