Careers advancement of women: Applying a multi-level relational perspective in the context of Turkish banking organisations

This article investigates the question of why women remain underrepresented in management positions despite their increasing presence in the Turkish banking sector context. By using an integrative framework that brings together the relational and institutional perspectives, this article seeks to understand how macro-level contextual influences reflect on meso-organisational processes and practices, which in turn influence women’s micro-individual career trajectories. This article explores the careers experiences of 49 female banking professionals from three distinct banking organisations (Turkish, western and Islamic). The findings reveal that women’s limited representation at managerial level is an outcome of the interplay between macro-, meso- and micro-level issues. At the macro-level, the article highlights religion-based fragmented social structure as a key factor in the socio-cultural context of Turkey. At the meso-level, the analysis of the three different cases identifies factors relevant to both western and Middle Eastern societal contexts. At the micro-level, the analysis highlights how women’s perceptions of macro-level barriers change depending on their organisational context, emphasising the interrelatedness between these three levels. The article contributes to women’s career literature by providing the Turkish perspective and introducing novel insights into the interplay between societal forces, organisational practices and individuals’ perceptions.


Introduction
This article explores the multilevel interconnected barriers that influence female managers' career advancement in the Turkish banking sector and compares findings across different organisational structures. The majority of the literature on women's careers is heavily concerned with understanding the barriers to women's career advancement in western European-and US-based organisations, which do not necessarily correspond to the experiences of women living outside of these contexts (Joshi et al., 2015). Recent studies conducted in non-western contexts have shown that women continue to experience significant challenges in their careers owing to local cultural practices (Ali and Syed, 2017). Furthermore, relatively little research has been conducted on the organisational practices, including human resources management policies within developing regions, such as the Middle East (Metcalfe, 2007). Given the gradual increase of women in middle-and senior-level management positions in these regions, exploring women's careers may widen the existing empirical scope beyond the western context.
There has been a growing call from career researchers for a greater focus on multilevel research on women's careers in non-western contexts in order to reveal the interplay between societal forces, organisational practices and individuals' perceptions and motivations (Alvesson and Billing, 2009;Kossek et al., 2017). Although some studies have begun taking a multi-level perspective on women's careers, they do not take into account the possibility of divergences amongst organisations within the same societal context (Ali and Syed, 2017;Jamali, 2009;Pryce and Sealy, 2013). Thus, it is uncertain whether different organisational cultures may coexist within the same socio-cultural environment.
Turkey is strategically positioned between Asia and Europe and has historically acted as a bridge between the East and the West. Furthermore, Turkey stands out as an atypical Muslim country as it has a secular constitution and a long tradition of modernisation since its establishment in 1923 (Esmer, 2008). As a result of this modernisation project, Turkey has made significant progress in adapting to the western business context (Daller and Yildiz, 2006). On the other hand, local cultural values still play an important role, particularly in the more rural eastern parts of the country (Aycan, 2001;Marcus et al., 2019). Thus, the choice of Turkey provides the opportunity to explore its distinctive hybrid organisational and societal context, which is relevant to the research on women's careers in both the western and eastern countries.
With the notable exceptions given above, the literature has primarily focused on either micro-or meso-level factors while giving limited attention to the broader institutional features related to women's career advancement. This article seeks to extend our knowledge of women's careers by exploring women's work and social relations within three distinct type of organisational hierarchies of power and authority (i.e. Islamic, Turkish and Western Banks) in Turkey, one of the few countries that accommodate secular values as well as strong religious and patriarchal norms. By doing so, it addresses barriers to women's career advancement at the micro-(individual), meso-(organisational) and macro-(societal) levels, and explores whether different organisational experiences can coexist within the same national context by looking at women's career advancement within three different types of banking organisations.

Contextual perspectives on women's careers
In the career literature, there is a broad consensus amongst scholars that women's careers may unfold differently in non-western contexts owing to differences related to national and cultural factors. Although this article focuses on career advancement, defined as promotions to higher levels in a management hierarchy (Lyness and Judiesch, 2008), it recognises that careers are a multi-level phenomenon (Iellatchitch et al., 2003) and acknowledges that the broader structural and situational factors within the external 'macro' environment may influence organisational careers and individuals' perceptions and motivations about their careers and advancement.
A growing number of studies have shown that the societal context plays a significant role in the career experiences of women in more traditional societies. These studies have argued that our understanding of careers has been largely informed by the individualistic and de-contextualised studies developed in western countries. Thus, they may not correspond to the career experiences of women in other national contexts owing to the significant influence of social and cultural factors on values, attitudes and life experiences. Hutchings et al. (2012) found that, although both North American and Middle Eastern women considered work-life balance and societal expectations to impact their careers, Middle Eastern women considered 'discrimination based on the culture' and the 'lack of protective legislation and government support' as additional barriers to their careers. Other studies have also found that patriarchal national culture and the endorsement of societal sex role stereotypes often lead to gender bias and labour market segregation in more traditional societies. For example, studies conducted in China as well as the UAE have found that women's career development was highly dependent on their social status and family connections, which was restricting their career alternatives Omair, 2010;Tatli et al., 2017).

Differences between male and female managers' careers
To explain women's underrepresentation in managerial positions, some researchers argued that women's and men's career patterns were significantly different and that women defined career success more holistically than men and pursued unique career development models that were part of a larger web of interconnected issues regarding work and private life (Mainiero and Sullivan, 2005;Powell and Mainiero, 1992). Other researchers focused on the gender-based barriers imposed by external factors affecting women's career experiences (Kossek et al., 2017;Lyness and Thompson, 2000). In particular, studies conducted overwhelmingly in western countries showed that although men and women had similar career aspirations, equally prioritised promotions (Eagly, 2013;Kirchmeyer, 2002;Ng et al., 2005) and demonstrated similar leadership skills (Eagly and Karau, 2002), women were likely to experience additional barriers to their careers compared with men. Both perspectives are useful for the purposes of this research, as they collectively provide insights into the distinct features of women's career advancement, as (i) women may follow unique career models that do not necessarily follow a traditional linear career path and (ii) women may experience different and intensified multi-level barriers to their career advancement compared with men. It is proposed that women's careers and their social relations at work will be influenced both by their own values and perceptions and by the diverse and distinct individual, organisational and societal factors that influence their careers.

Barriers to women's careers advancement in the Turkish banking sector: Applying a multi-level relational perspective
There has been a significant amount of research in the western literature on identifying the organisational and psychological mechanisms to explain women's career advancement to managerial roles. However, in many parts of the world, women continue to experience overt and visible bias in addition to other important contextual challenges related to religion, culture and local business practices (Joshi et al., 2015). In particular, women in Middle Eastern countries often experience additional challenges owing to cultural, patriarchal and familial pressures (Metcalfe, 2007). Despite advancements in the literature, little is known about how broader factors stemming from the national context, such as the cultural norms and local laws and policies related to gender equality, influence society's perception on women's careers and how this in turn affects female managers' career advancement (Joshi et al., 2015;Kossek et al., 2017;Metz and Kulik, 2012). Neither individual-nor organisational-level studies address the distinct macrolevel factors that make up the national context. As a result, the substantive literature on women's managerial advancement primarily presents the separate pieces of the puzzle, rather than providing a comprehensive understanding of the different levels of influences. Recognising this need, some scholars have identified factors that may be linked to the status of women in organisations and confirmed the impact of national culture and social norms (e.g. Banihani and Syed 2017;Ituma and Simpson 2009;Metcalfe, 2008). Furthermore, understanding career experiences requires a multi-level analysis crosscutting individual, organisational and institutional influences and their interdependent relationship (Iellatchitch et al., 2003). To achieve this objective, this article draws upon Scott's (1995) three institutional pillars to demonstrate how cognitive-cultural, normative and regulatory factors impact women's careers and the barriers they face. According to this perspective, institutions are highly influential factors that help explain both organisational and individual-level behaviours, owing to their capacity to either enable or restrict their activities. A growing number of studies in a range of countries such as Lebanon (Jamali, 2009), Pakistan (Ali and Syed, 2017) and the UK (Pryce and Sealy, 2013) have begun to take a broader look at the multi-level conditions that affect the career patterns of women. One such perspective is provided by Syed and Özbilgin (2009). They argued that single-level conceptualisations conceal the gendered processes at macro-, meso-and micro-levels and thereby fail to capture the interplay of structural and agentic levels concerning equal opportunity. In order to develop a more holistic and contextual understanding, this article adopts a relational and multilevel perspective to explore the individual, organisational and macro-contextual barriers that affect women's career advancement. Such a perspective offers a richer understanding that takes into consideration the subjective experiences of women as well as the structural and institutional influences that impact their career reality. At the macro-national level, the relational framework refers to the national-level policies and legislations that influence women's careers opportunities (see Figure 1). The two internal smaller circles represent the meso-and micro-levels, which are embedded in the macro-level context. The meso-level considers the organisational processes related to gender equality and explain how macro-level inequalities may reflect on organisational hierarchies. Finally, the micro-level considers issues related to subjective experiences of individuals with respect to career aspirations. The overlap between the two sub-domains (meso-and micro-) represents the interconnectedness between them (Syed and Özbilgin, 2009).
In light of this context, this article explores the overarching research question: What are the key national (macro-), organisational (meso-) and individual (micro-) level barriers influencing women's career advancement in the Turkish banking sector?
Modern banking in Turkey began with the establishment of public banks after the founding of the Turkish republic in 1923 (Culpan et al., 2007). Until 1980, the Turkish economy was heavily state-regulated and international capital movements were restricted. In 1980, the economic structural adjustment program was adopted that relaxed the entry barriers to the financial services sector, resulting in a significant expansion with new foreign entrants from the western and Middle Eastern countries (Özbilgin and Woodward, 2004). The Turkish banking sector continues to attract both foreign and domestic investors (Sümer, 2016). The interest-free banking model started in Turkey in 1983. At the time, these banks were referred to as 'special finance houses' (SFHs) as the constitution does not allow banks to be labelled as 'Islamic'. SFHs were not able to work under the same legal and operative framework as other commercial banks because of concerns raised by state bureaucracy, which was strongly committed to secular values. Following the 2005 Banking Act, SFHs were given the same status as commercial banks and were re-named as 'participation banks' (Yilmaz and Güneş, 2015). The six participation banks currently operating in Turkey have attracted large sums of savings from religious individuals as well as Gulf funds, typically through private sector investment loans (Aysan et al., 2018;Pehlivan, 2016). Despite the growing attractiveness of participation banks, they constitute a relatively small segment of the Turkish banking sector, as their market share at the end of 2017 was only 6.3% (TKBB, 2020).
The limited number of empirical studies conducted in the Turkish banking sector demonstrate that the traditional organisational career path is still relevant for large hierarchical organisations, including banks (Çakmak-Otluoğlu, 2012). These studies provide evidence that many organisations, especially in the public sector and finance, continue to have a structure based on bureaucracy where managerial and professional employees are able to hierarchically progress in a single organisation over many years (Çakmak-Otluoğlu, 2012). On the other hand, according to the Turkish Statistical Institute (TUIK, 2020), the banking and finance sector in Turkey has one of the highest proportions of female employment. Nevertheless, women continue to experience inequalities and barriers in the Turkish banking sector. Gender-biased organisational culture and practices remain to be by far the greatest problems for women's career advancement (Karataş, 2013;Minibaş-Poussard et al., 2011;Özbilgin and Woodward, 2004). In their study of HR managers, Günlük-Senesen and Özar (2001) found that female employees were considered by employers to be more suitable for jobs in the lower levels of the organisational hierarchy because they were perceived to be more obedient, patient and tolerant than male employees. Özbilgin and Woodward (2004) found that in Turkey, the managerial elites' perception of the predominantly male 'ideal worker' created significant barriers to women's career advancement. Furthermore, studies on female managers' careers in participation banks are virtually non-existent in Turkey. However, research on female managers in participation banks remains imperative to understand the key problems women encounter during their careers in such conservative work environments.
Although several scholars have argued for multilevel research on women's managerial career advancement (Kossek et al., 2017;Pryce and Sealy, 2013;Tharenou et al., 1994), past studies in banking or finance sectors either used samples from a single organisation (e.g. Lyness and Judiesch, 1999) or did not incorporate macro-level factors (Metz, 2003). By adopting a multi-level perspective, the study aims to provide a holistic insight into the barriers to women's careers in the Turkish banking sector and extend earlier research by emphasising the contextual influences of Turkey, one of the few Muslim countries that accommodate secular values as well as strong religious norms. Furthermore, it has been suggested that different managerial/organisational cultures (i.e. prevailing managerial values, shared beliefs, norms and assumptions) may coexist within the same socio-cultural context (Aycan and Kirmanoglu, 2007). For example, in the Turkish banking sector the organisational culture of western banks may be different from those of Islamic banks, as the managers of Islamic banks may hold more traditional and conservative values. This, in turn, may affect women's career progression in such organisations. However, the potential coexistence of different managerial/organisational cultures within the same sector and societal context remains underexplored. The article addresses this by exploring women's career advancement in three types of banks operating in Turkey (i.e. Turkish, western and Islamic).

The institutional perspective
The institutional perspective has been used in various social-science-related disciplines, including sociology, political science and economics (DiMaggio and Powell, 1991). Institutional theory is mainly concerned with how organisations survive and secure legitimacy by adapting to the regulatory, social and cultural pressures deriving from the external environment. In his well-known formulation of the three institutional pillars, Scott (1995) encompassed various institutional perspectives and combined the key regulative, normative and cultural-cognitive elements that were previously identified as the crucial features of institutions within a single framework. Scott defined institutions through three main pillars, which were the 'regulative, normative and cultural-cognitive elements that together with associated activities and resources provide stability and meaning to social life' (Scott 2014: 56). These three pillars are not mutually exclusive but rather are interconnected. Together, they establish a foundation for a social order and exert pressure on both individuals and organisations through societal expectations, laws and regulations (Yang and Konrad, 2011).
The cultural cognitive pillar refers to the socially constructed idea of reality through the shared beliefs of individuals coming from similar national/cultural backgrounds. Examples of cognitive structures can be ethics and culture; beliefs that are taken for granted within a society. The normative pillar refers to the 'normative rules that introduce a prescriptive, evaluative and obligatory dimension into life' (Scott, 2014: 64). Examples of normative institutions include the social and professional norms shared by the society (Ituma and Simpson, 2007). The regulative pillar refers to 'the capacity to establish rules, inspect others' conformity to them, and, as necessary, manipulate sanctions-rewards or punishmentsin an attempt to influence future behaviour' (Scott, 2014: 59). Scott (1995) argued that organisations became more similar (or different) depending on the pressures of these three structural pillars. Building on DiMaggio and Powell's (1983) work, he contends that if organisations are subjected to similar pressures, they may develop similar administrative structures, resulting in organisational homogeneity. This article also considers similarities and differences between organisations operating in the same sector within a single country and explores the extent of organisational variation within that sector.
A growing number of studies have utilised institutional theory to explain the distinctive cultural, normative and legislative barriers affecting women's careers. For example, Ituma and Simpson (2009) found that Nigerian women suffered from cognitive perceptions related to having secondary status in the society and less competency for jobs requiring technical abilities. Reflecting on the regulatory pillar of institutional theory, Thams et al. (2018) found that women's presence on corporate boards was highly dependent on the existence of diversity policies. Furthermore, Fernando and Cohen (2014) found that women in Sri Lanka experienced significant normative barriers when they tried to use their individual agency in their careers. Extending these studies, this article draws on and contributes to Scott's (1995) framework to explore the extent to which women's individual career trajectories reflect the societal conventions embedded in institutional settings, as studies suggest that individual career choices are dependent on wider contextual factors. The following section describes the Turkish context using the lenses of these three pillars.

Institutional theory in the Turkish context
Achieving gender equality has been a national policy of the Turkish state since the 1920s, and a number of legal and institutional arrangements have been adopted for this purpose (Hatipoglu, 2014). However, there remain significant barriers to the career advancement of Turkish women. In particular, traditional gender roles and ineffective implementation of existing legislation remain as the key obstacles to achieving gender equality in Turkey. Despite the legal and governmental efforts to increase female employment, the participation rate for women in the Turkish labour force is significantly lower than that for men, with the majority of women either not working or working informally without access to state benefits such as minimum wage or unemployment insurance (İnan and Aşık, 2015). Women's restricted access to the labour market remains as one of the key challenges for Turkey's economic and social development.
In the past decade Turkey has undergone major legislative changes to increase women's social, political and economic participation. Major legislative changes have also taken place in the field of women's employment through the Labour Act, which became effective in 2003, and prohibited all kinds of gender discrimination between the employer and employee (Kabasakal et al., 2004). For example, employers are prohibited from terminating a contract owing to their employees' gender, marital status, pregnancy or family responsibilities (Kabasakal et al., 2004;Sural, 2008). The labour law also includes measures concerning maternity leave and childcare. Sixteen weeks of paid maternity leave are available to employed women, which is higher than the standard 12 weeks according to the International Labour Organisation. Female employees can also use paid leave for breastfeeding children under the age of one. With the same regulation, employed parents are entitled to work part-time until the child reaches primary school age. In turn, key normative norms affecting women's career advancement in Turkey are likely to be traditional gender roles and the negative perception towards female managers. The literature on Turkey emphasises the prevalence of traditional gender roles and the significant impact of cultural expectations on women's self-perceptions and lifestyle choices (Healy et al., 2005;Kabasakal, 1998;Zeytinoglu et al., 2001). In particular, empirical evidence suggests that women's compliance with their traditional gender roles (i.e. homemaker), which reduces their chances of securing post-primary education and undermines their future career prospects, is greater in rural areas and patriarchal families (Rankin and Aytaç, 2006). However, compliance with traditional values is also evident for the Turkish urban middle class (Aycan and Eskin, 2005).
Despite legal efforts to create gender equality, gender-biased organisational cultures and practices stemming from negative attitudes towards female managers result in discriminatory organisational practices including recruitment and promotions (Burke et al., 2008;Özbilgin and Woodward, 2004). In a study on Turkish sports organisations, Koca et al. (2011) found that both male and female employees had a general preference for male sports managers; men held more traditional views of gender roles and were more sceptical of women's competencies as managers. With respect to cultural-cognitive factors, the Turkish society overwhelmingly demonstrates collectivistic patterns. Many aspects of Turkish organisations reflect the society's collectivistic nature (Fikret Pasa et al., 2001). These aspects include HR practices such as recruitment practices, where kinship and interpersonal relationships are critical. Studies have shown that common Turkish leadership behaviour involves a paternalistic and nurturing style, where managers pay attention to employees' professional and personal matters and expect loyalty and obedience in return (Aycan, 2006). However, little is known regarding how this leadership style impacts women's career advancements in particular.
Furthermore, the increasing support for conservative values within Turkish society over the past decade has sparked considerable debate in the popular media as well as among policy makers and the academic community. Kalaycioğlu (2007) identified the three core values of Turkish conservatism as religion, traditionalism and nationalism. Many scholars have expressed their concern that the frequent emphasis on women's 'traditional' role by right-wing conservatives threatens the achievement of gender equality in Turkey (Arat, 2010; Çitak and Tür, 2008). Conversely, it is argued that there are potential liberating effects of Islamic values on women from lower socio-economic and religiously conservative backgrounds in Turkey (Erman, 2001;Meyersson, 2014). However, the debate on whether the rising conservatism in Turkey restricts or liberates women is, as yet, unresolved.

Research methods
The research question of the study is exploratory as it studies how Turkish female banking professionals give meaning to their career advancements based on their real-life experiences. A qualitative case-study research design was chosen to gain insights into the participants' multiple identity roles and social relationships within and outside of organisational hierarchies. By looking into three different organisational contexts, the researcher had the opportunity to gain an in-depth understanding of both the overt and subtle ways in which women's career advancement opportunities were constrained by societal and organisational factors and how they responded to them (Bryman and Bell, 2015;Tlaiss, 2014).
The three selected types of case study banks together represent the entire set of conventional and non-conventional banks that offer services to individual consumers in Turkey. Yin (2014) contended that, compared with single-case designs, multiple-case designs were often compelling and more robust. Relatedly, Eisenhardt and Graebner (2007) explained that by making comparisons through several cases, multiple cases permit broader explanations for research questions. The decision to choose the cases from the banking sector was primarily for three reasons. Firstly, the unique context of the Turkish banking sector made it a suitable field for investigating more subtle forms of gender discrimination. Turkish banks are generally considered family-friendly (Culpan et al., 2007) and are known to provide the most generous work-life balance policies in the Turkish private sector (Capital, 2015). However, female banking professionals are mostly represented in junior and mid-level jobs. Secondly, the stratified structure of the Turkish banking sector enabled the researcher to explore three distinct cases within the same national context and examine the varying implications of organisational culture and practices for the career progression of female managers. Thirdly, although previous research has suggested that participation banks were now more likely to hire female labour to meet the high demand for educated employees (Metcalfe, 2007), there has been very limited research on the career experiences of women working in such contexts.

Data collection
This study used two main types of data collection methods -primary and secondary data. The secondary data informed the design of the interviews and also supported the analysis phase. For secondary data collection, the publicly available data published by the Turkish Statistical Institute was used as a reference for the general employment situation of women in the Turkish labour market. Furthermore, survey data provided by the Ministry of Family and Social Policies and the World Values Survey was used to understand the society's views on women. Statistical data provided by the Turkish Banks Association (TBB) was also used to document the number of women working in banks. Furthermore, additional data were collected from the banks via company publications such as annual reports and websites, as well as interviews with the female HR managers who also participated in the main study.
Interviews were conducted with 49 female banking professionals from the three banks operating in Turkey through personal networking of the author and via snowballing methods. The sample included female banking professionals from a variety of grades with 5+ years' experience, to allow sufficient time to experience issues related to career advancement. Previous research has suggested that women at the middle ranks of organisations are exposed to significant discrimination in terms of promotions (Yap and Konrad, 2009) and receive fewer development opportunities (Hoobler et al., 2014). Thus, this study focused extensively on female middle managers and drew upon the barriers to their career progression. The snowballing process involved asking interviewees to recommend other potential participants within and outside their organisation who may have different career experiences (Cohen and Arieli, 2011). This strategy was particularly useful owing to the difficulty of obtaining formal research permission from banks in Turkey, which would otherwise limit the participation activity. The profile of the interviewees is provided in Appendix 1 (online).
During the interviews, all participants were assured that the research process would be anonymous by using pseudonyms and that their responses would be treated with confidentiality. Semi-structured interviews were used in order to encourage the interviewees to express their thoughts openly. The questions were designed to allow the interviewees to reflect on their perception of themselves, their organisation and the general society.
The interviews were conducted in person and lasted between 50 and 90 minutes. In order to minimise bias, the researcher used open questions and probing techniques and avoided leading the conversations during the interviews (Easterby-Smith et al., 2002). The interviewees were offered to review and verify their transcripts to make the research findings more reliable and valid (Ryan and Bernard, 2003). The interview guide is presented in Appendix 2 (online).
The next section will provide an overview of each bank case, with details related to organisational practices based on interviews with HR managers and secondary data.

Case study 1: The Western Bank
One of the research sites for this study was a western private bank, which has been operating in Turkey for more than a decade. The bank will be referred as the 'Western Bank' to maintain its confidentiality for this study. The bank was one of the 10 largest operating banks in Turkey. In total, out of 564 mid-level managers, 192 were women (34%) and out of 73 senior managers, 24 were women (33%).
The bank had a wide range of selection processes for hiring employees, depending on the requirements and seniority of the role, including telephone interviews, aptitude and English tests for more junior positions, and assessment centres and simulations for more senior roles. Regarding promotions, the bank's annual report stated that 544 (323 men, 221 women) employees were promoted in the headquarters, 554 (237 men and 317 women) in the branches and 45 (23 and 22) in the regional offices. The bank had various training programmes. The most prestigious one was the Leadership Development Programme, designed for high-potential employees. When asked about the gender breakdown of the programme, the HR manager who also took part in the main study was not able to give specific percentages but said that gender imbalance was not an issue in their training programmes. Regarding work-life balance practices, employees were able to work outside the office for 2 days a week. Also, they were able to determine their office hours. For expectant mothers, the bank offered one day off each month to use for their doctor appointments.

Case study 2: The Turkish Bank
The second research site for this study was a large privately-owned Turkish bank. The bank will be referred as the 'Turkish Bank'. According to the data provided by the Turkish Banks Association and the bank's annual report, women made up 51% of its employees, which was equal to the sector average. The Turkish Bank had a similar percentage of female employees (51%) and, compared with the Western Bank, it had a higher percentage of female mid-level managers (45%). However, the share of women dropped significantly at the senior level (13%).
There were three main career routes in the bank: auditor, officer and management trainee/specialist. The HR manager did not provide the gender breakdown of the recruits, but stated that specialist and officer positions were balanced gender-wise and that the auditor position had only a few women. According to the HR manager, the main reason for this was the nature of the auditor's job, which was stressful and required constant travelling and therefore was less attractive for women. For each career route, the Turkish Bank had standard promotion criteria for middle management (e.g. work experience, performance, competency-based interviews, written exams), which was announced transparently within the internal HR system. The promotion to more senior roles was less clear and depended on a range of criteria including tenure, performance evaluations and available headcounts.
The Turkish Bank provided its employees with more career development opportunities compared with the Western Bank. This was mainly because the Turkish bank relied less on external hires and prioritised developing its own employees. After each promotion, employees would go through another three-week training based on their needs. These programmes included a broad spectrum of classroom-based teaching, e-learning and internships in various departments. Similar to the Western Bank, the Turkish Bank focused its work-life balance practices around maternity benefits and the execution of government regulations.

Case study 3: The Participation Bank
The third research site for this study was a participation bank. The bank will be referred as the 'Participation Bank'. According to the annual report, the Participation Bank had the smallest share of female employees (19% in total, 5% in senior management). Very few existing female senior managers were outliers with exceptional educational backgrounds and extensive work experiences.
Similar to the Turkish Bank, there were three main career routes in the Participation Bank: auditor, officer and management trainee/specialist. The interview process differed depending on the career route and covered a few stages, including assessment centres, aptitude tests and one-to-one interviews with the heads of HR and related departments. The bank decided on the promotions to mid-level positions based on the employees' performance evaluation feedbacks on competencies and measurable targets. Employees who had relevant work experience (over 5 years) and consistently met performance targets were invited to the Leadership Training Programmes, which consisted of intensive online and in-class training. Similar to the Western and Turkish Bank cases, the worklife practices in the Participation Bank revolved around maternity benefits. However, unlike the other two banks the Participation Bank stated in its annual report that they viewed motherhood as 'sacred'. This religious emphasis on motherhood coupled with the high turnover rates that the bank experienced with its female employees after childbirth led the bank to become one of the first in the sector to provide flexible work hours to female employees with pre-school-aged children. The bank also implemented a flexible working policy where women were allowed to work from home or part-time until their children reached the age of 5.

The data analysis procedure
The study applied thematic analysis, which is a frequently used approach by qualitative and interpretivist researchers to explore the different interpretations and meanings attached to a phenomenon (Saunders et al., 2016). The thematic analytic process involved six interconnected phases: (i) familiarisation with data, (ii) generation of initial codes, (iii) searching for themes, (iv) reviewing themes, (v) defining and naming themes; and (vi) writing up the findings (Braun and Clarke, 2006). Thematic analysis involved encoding the qualitative data and organising the responses (codes) into a series of sub-themes that can be incorporated into broader themes. The coding strategy allowed the researcher to use in vivo codes (the actual words used by the participants as a source of code), a-priori codes (terms used in existing theory that capture the narration of respondents) and labels developed by the researcher to best describe the unit of data that emerged from the interview (Saunders et al., 2016: 582). This allowed the researcher to capture the concepts used in the existing theories, as well as explore new themes that emerge during the interviews (Graebner, 2009).
The data were analysed in two stages through an iterative process of going back and forth between the codes and the emergent themes (this iterative approach was followed in both stages). In the first stage of the study, the researcher applied line by line coding techniques in two rounds to ensure consistent and rigorous coding practices (Saldana, 2013). The transcripts were in Turkish, but the coding process was executed in English, as the software did not permit data analysis in the language of the interview. While coding the data, both deductive and inductive approaches were used. The inductive component of this study allowed themes to emerge from the data based on the participants' experiences, whereas the deductive component reflected on the conceptual framework of the study (Fereday and Muir-Cochrane, 2006). Therefore, certain themes were apparent before the analysis (i.e. societal gender roles, organisational policies related to gender equality). During this iterative process the researcher constantly reviewed the codes and themes involving the constraints faced by female managers and their careers experiences. This resulted in new themes and codes that emerged from the data (i.e. female modesty, self-gendering and the religious-secular divide). The identified codes were then collated and merged into hierarchical themes, which allowed the study to gather and categorise all relevant data from bottom-up (Berber and Acar, 2020;Braun and Clarke, 2006). An experienced researcher, who is also fluent in both languages, reviewed the coding and rechecked the translations to verify and refine codes and themes as well as to identify any deficiencies in the analysis.
In the second stage of the analysis, a graduate student was introduced in order to independently code the data and strengthen the credibility of the study. The independent coder produced an additional set of codes. The researcher and the independent coder then reviewed the analysis to generate an agreed set of codes, which were then grouped into the emergent themes. This process was repeated for each three banks (see Appendix 3, online).

Findings
The findings of this study revealed how meso-organisational-level processes (i.e. organisational culture, promotion and career development policies) and micro-individual-level career experiences were situated within macro-environments and therefore could not be considered in isolation. Informed by institutional and relational perspectives, the findings have been distilled into three categories, examining the interrelated multi-level issues Turkish female banking professionals experience in their careers. A complete list of emergent themes, listing barriers and number of cases, is shown in Appendix 4 (online). The next section first uncovers the key macro-level barriers identified by the interviewees, then turns to the meso-level barriers and finally looks at the micro-level analysis of barriers.

Macro-level barriers to women's career advancement
At the macro-level, the findings revealed how women's perception of macro-level barriers in Turkey changed depending on the type of bank they worked in, reflecting the complex and intertwined labyrinth of organisational challenges they experienced during their careers (Eagly and Carli, 2007). Regarding the macro-national barriers, the interview data identified two specific factors for the Turkish context. Each factor will be explained in detail below.

Patriarchal norms and the religion-based fragmented structure of the society
All three banking cases confirmed the collectivistic (Hofstede and Hofstede, 2005) and low-gender-egalitarian (Kabasakal and Bodur, 2008) characteristic of the Turkish culture. Interviewees from all three banks described how the traditional gender roles promoted the role of women mainly as wives and mothers, thereby constituting a barrier to women's career advancement in Turkey. However, the findings also revealed that women experienced the effects of patriarchal ideology differently depending on the type of bank they worked in. This variance may have been driven by the on-going religious-secular divide in Turkey, where women who come from secular backgrounds separate themselves and their social circles from the emerging middle class that endorses conservative and patriarchal values (Bugra, 2014). In particular, the interviewees in the Western Bank did recognise the prevalence of gender-role stereotypes in the Turkish society. However, they tended to distance themselves and their organisation from these normative values, suggesting that they were more relevant for the newly emerging middle class, which endorsed more conservative values: If we look at the Turkish average -not in the modern part of Istanbul though -female managers experience serious difficulties in demonstrating themselves as competent. That is why if we set aside the niche environment that I am part of, I think women are significantly disadvantaged in Turkey. (P13, 31, mid-level, Istanbul, The Western Bank) Female managers in the Turkish Bank also indicated that they endorsed secular values. However, they were concerned about the growing levels of conservatism and patriarchy in the country (Esmer, 2012) and feared that Islamic ideology could challenge their future career opportunities as well as lifestyle preferences. In contrast to the Western Bank, the Turkish Bank employed individuals from a diverse range of socio-economic backgrounds and had a much more diversified client base. Hence, this growing concern could be related to the participants' social interactions with male managers. While describing the male dominant organisational culture of the organisation, a respondent provided an interesting insight that suggested that the limited number of female senior managers in the bank was the result of the endorsement of traditional gender roles within the society: I would say that the organisational culture is now more patriarchal, resembling the Turkish society. Just like the society, you see that the current management is more male dominated, which is reflected in their day-to-day language, their attitudes and their preferences for promotions. There are only a few women in senior management and there is no visible commitment to increasing their numbers. (P32, senior level, Turkish Bank) On the other hand, female interviewees in the Participation Bank positioned themselves and their social circles within the societal context endorsing such patriarchal values. In the Participation Bank case, socio-cultural barriers were perceived as a current social reality that was influencing their careers more directly. In particular, women experienced intensified role incongruency between their domestic and work roles, as stereotypically they were expected to prioritise domestic responsibilities over their work duties, especially after having children: As a female professional you may be working for more than 12 hours a day but still hold the main responsibility for family issues. You must think of your children's meal, you have to plan the hours your children will leave and return from school. This societal expectation also reflects on your husband's views as well. Yes, my husband supports me working but I also have to plan everything in advance. When women manage both their work and non-work lives successfully then they earn the respect of the society. (P41, senior level, Participation Bank) Although debates around Turkey's religious-secular divide frequently arose during the interviewees in the Western and Turkish Banks, it was only in the Participation Bank that women perceived it as an actual barrier restricting their career advancements. Interviewees explained how their career opportunities within corporate organisations were confined to the Participation Bank as they were subjected to subtle discrimination outside their organisation. This discrimination often stemmed from scepticism towards their level of competence or commitment. The fact that these women were wearing headscarves was perceived as an indicator of their lack of interest in building a career: I haven't experienced any discrimination here because it is a Participation Bank, but when I go to conferences, people sometimes approach me differently, with prejudice or reservations, not because of my gender per se but because I wear a headscarf. I manage to break the stereotype after showing my expertise in the field but I often experience this scepticism especially in the networking events that I attend the first time. (P42, mid-level, Participation Bank) Overall, the interviewees from the Participation Bank revealed that besides being disadvantaged by their gender, these women experienced additional barriers to working in 'secular' work environments owing to their religious attire and lifestyle. To avoid any potential uneasiness or discomfort for both themselves and potential employers, most female interviewees chose to apply only to conservative workplaces, although they were very much aware that by doing so they were restricting their career options.

Macro-level economic uncertainties
Interviewees from all three banking cases perceived macro-level economic uncertainties in Turkey as a risk to their careers, as they limited their organisations' potential for growth and therefore restricted their career opportunities. However, it was only in the Western Bank that female managers confirmed macro-level economic uncertainties as a barrier currently affecting their career advancement, owing to the bank's heavy reliance on western investments and capital inflows. The interviewees from the Western Bank detailed how the current volatile economic outlook of the country resulted in the bank's decision to downsize, which ultimately restricted the number of available positions and consequently intensified the competition amongst the employees: The Global HQ decides to grow in local markets based on the foreign direct investment the country is able to attract . . . Currently, developing markets -including Turkey -is perceived as unstable. Even when business goes well in Turkey, if things aren't so good over there, this impacts our Turkish operations too. This year, there were less promotions and we witnessed a few closures in departments and branches. (P4,Western Bank) Interestingly, the interviewees at the Turkish and Participation Banks did not raise similar concerns. This may have been owing to the differences in the banks' capital structures as well as target customer segments. The Turkish Bank was domestically owned, and generated a large proportion of its profits from individuals and SMEs (small-and medium-sized enterprises) based in Turkey. As a result, the bank may have been less affected by the global macro-level economic uncertainties: It is difficult to assess Turkey's current economic and political conjuncture. I don't really think about how these uncertainties would impact my career. The Turkish banking sector has a very strong structure where our bank holds quite a central place. If I somehow lose my job that would imply that there are bigger problems in the country to worry about. (P38,Turkish Bank) Similarly, the Participation Bank mainly focused on the modest savings of individual customers and SMEs from the Anatolian region of the country and was recently receiving significant government support owing to the growing popularity of its interest-free products: I feel really comfortable in the current conjuncture. Currently, there is both political and legal support towards the expansion of Participation Banks' operations. I am hoping this will have a positive impact for me in the future. It might increase competition and company transfer opportunities. (P48,Participation Bank) These differences between the three banks may have influenced how women perceived macro-level economic uncertainties' direct impact on their careers.

Meso-level barriers to women's career advancement
The interview findings across the three bank cases revealed a mix of meso-level barriers found in western and Middle Eastern contexts, reflecting Turkey's geographical and cultural blend of the East and the West.
Three key meso-level barriers emerged from the interviews in the three bank cases. The sections below explain each factor in further detail.

Implicit and explicit bias against female managers
All three banking cases in this study confirmed the existence of implicit bias against female managers, to varying degrees. The research also found indications of explicit bias against female managers in the Participation and Turkish banks. Explicit bias was less apparent in the case of the Western Bank. Nevertheless, male management remained the norm at the Western Bank, despite large female representation (Heilman, 2012). For promotions to senior positions, a recurrent theme was the need to be more like the male managers. However, the issue was not of gender but rather the ability to demonstrate masculine and aggressive qualities, as a feminine management style was perceived to be not suitable for the competitive nature of the organisation. However, most women struggled to do so while also trying to maintain a sense of authenticity: The environment is really competitive and aggressive. All of the women in top positions are really masculine. These women lost their identity. You cannot survive by being naive. There are plenty of male managers who swear. They expect you to be like them. I don't want to force myself to become masculine. I want to climb the ladder by being myself, but that doesn't seem quite possible. (P10,Western Bank) The narratives in the Turkish Bank revealed the persistence of both implicit and explicit biases towards female managers. With respect to the implicit biases, the findings revealed that female managers were subjected to intensified role incongruency (Eagly and Karau, 2002). Women were often sidelined for senior roles and withheld from challenging tasks that were essential for their career development: There is a perception that men are better in high-qualified tasks. Even if no one says this, I feel it. I witness this especially when my male colleague does the same job as I do. I receive more operational, labour intensive jobs, while my manager gives my male colleague tasks that require more thinking. (P35,36,HQ,Turkish Bank) The narratives from the Turkish Bank, an organisation that was known for its high female ratio and female-friendly work policies, also revealed the existence of blatant forms of discrimination towards women. The bank's policy to announce the available branch manager positions separately for men and women and its tendency to send male branch managers to conservative cities were also examples of explicit bias, deliberately confining women's career advancement opportunities within a narrower scope: The branch manager headcount is a major problem. It is very difficult to get promoted nowadays . . . You could decide to go to other cities to get promoted -men usually do that -but this is very difficult for women, especially given that most of the headcount outside big cities are allocated to men. (P38,Turkish Bank) The interview findings from the Participation Bank also revealed both implicit and explicit biases in a number of different ways. Regarding the implicit biases, the interviews confirmed that the organisational expectancy from women to comply with 'female modesty' was acting as a barrier to their career advancement (Syed et al., 2005). The interview findings also demonstrated how, in line with the organisational culture's emphasis on motherhood, male supervisors refrained from giving challenging tasks to female employees, which would involve travel and overwork, to make sure they did not neglect their family responsibilities. This, in turn, impeded women's career advancement, as they could not access critical development opportunities: If there is a job or project that requires overwork, they [supervisors] think, let us not give this to that woman -poor her, she has a family. This might seem nice at first, like positive discrimination, but actually they are holding women back by making us do the small things. Often men get promoted because they are given the chance to work on those projects. (P42,Participation Bank) With respect to the more explicit forms of gender bias, most of the interviewees at the Participation Bank explained that their organisation intentionally did not place women in critical roles owing to its cultural alignment with the traditional gender roles of women. If women still wanted to build a career in this environment, they had to prove that they significantly outperformed their male counterparts, which highlights the prevalence of the double standards in performance criteria within organisations (Rosette and Tost, 2010): 'Women have to make three times the effort of men to get promoted. The profiles of the very few female department heads all have amazing educations and careers. Most male managers are not half as talented as these women' (P43, mid-level, Participation Bank).
Overall, whereas the narratives from the Western Bank suggested that implicit bias in the form of double standards was more prevalent, the narratives from the Turkish and Participation Banks revealed the existence of both implicit and explicit biases against female managers. This difference could be related to the gender-neutral and transparent policies of the Western Bank driven by its Global HQ.

Lack of transparent criteria for promotions
Echoing previous studies conducted in Turkey, this study found that the promotion processes at the Western Bank were very much standardised and gender-neutral (Culpan et al., 2007). On the other hand, the interviewees at the Participation and Turkish Banks acknowledged the lack of objective criteria for senior roles as a major barrier to their career advancement, as this led only those with access to the influential organisational networks to be promoted: For mid-level positions, you need to succeed in written and oral exams. They are tough but at least there is clear criteria. After you pass these exams by putting in so much effort, you are likely to stay where you are, because for more senior positions it's all about who you know. I think this is really unfair. (P18, mid-level, Turkish Bank) I don't think I can get promoted further in this bank. I possibly will retire in this position. Will I become a senior manager? I would like to but after a certain point, relations become important, which is rather tricky. We work in a heavily male dominated organisation. (P48,Participation Bank) The interviewees perceived this to be very problematic, as men heavily dominated the senior positions in both organisations. Many respondents were distressed because, despite being recognised for their good performance, they knew it was unlikely for them to get promoted. The interviewees acknowledged informal networks as significant facilitators of career advancement. However, female access to such networks was perceived to be particularly difficult owing to male colleagues' preference to network and work with each other. This finding coincides with the concept of 'homophily', which is based on the premise that individuals prefer to work with similar others (Ibarra, 1993). The interviewees suggested that 'homophily' among male employees ultimately affected promotion decisions, as male managers felt more comfortable working with colleagues more like them: 'I think that men have a significant advantage over women with respect to promotions. This is because they tend to support each other and make sure their buddies aren't left behind' (P39, mid-level, Turkish Bank). Similarly to the Turkish Bank, male homophily emerged as a common barrier at the Participation Bank. A respondent explained how it was against the social norm for men to socialise with women, preventing women from utilising key organisational networks: During lunch breaks my previous manager would invite my male peers to join him and leave me out. I guess he didn't feel comfortable. Maybe he thought people wouldn't approve. When he left work for client meetings, he wouldn't tell me where he was going. He would tell them instead. Maybe he felt if he were to tell me, it would be as if he was reporting to a woman. (P41, senior level, Participation Bank) Overall, the absence of objective criteria for senior-level promotions and the intense male homophily culture in the Turkish and Participation Banks marginalised female managers as 'tokens' and intensified the extensive bias against them, resulting in significant barriers to their career advancement.

Norm of overwork and stigmatisation of family-friendly policy users
The narratives from the Western Bank suggested that it stood out across the three banking cases, with its competitive and 'long-hours' culture. This theme did not come up in the other two cases, possibly because it was in direct contrast with the 'family-like' cultures of the Turkish and Participation Banks. Nevertheless, in Turkey, work-family issues continue to be overwhelmingly framed as 'female problems'. However, success in the Western Bank depended significantly on dedicating long hours to work and outsourcing family responsibilities. As a result, although the organisation provided flexible work opportunities, female managers were stigmatised for using such policies and were often perceived to be less dedicated and therefore less promotable. Thus, the female managers in this study considered that the only way to succeed in the Western Bank was to be willing to sacrifice family obligations, despite the significant social role of being a mother in the Turkish context: The narratives also highlighted inconsistencies between the Western Bank's work-family policies and practices. The female managers explained that the users of familyfriendly policies were being perceived as less dedicated and therefore less promotable: I used to work with this really irritating manager. He used to say, 'Women cannot raise children and have a successful career at the same time. Children become so important to them that they opt-out of their careers.' Although, I despise him very much, this is something I can't get out of my head. Children are important to women. I breastfeed every three hours, how can I have an ambitious career? (P12,Western Bank) The findings revealed that this managerial-level bias had a strong influence on the self-efficacy of the interviewees and resulted in questioning their own abilities and career potential.

Micro-level barriers to women's career advancement
The overall analysis showed that micro-individual-level processes were influenced by both meso-level practices and processes as well as broader macro-level societal norms and expectations. By highlighting the interplay between institutions and agency (Scott 2014), the interview findings support Syed and Özbilgin's (2009) relational framework which integrate multiple levels of analysis to understand the contextual nature of gender. In accordance with the institutional perspective (Scott, 2014), the interviews confirmed that the female managers in all three cases were subject to similar cultural-cognitive, normative and regulative institutional barriers. However, the impact of these barriers depended on the organisational context. Two key themes emerged from the interviews regarding micro-level barriers, as discussed below.

Gendered role perceptions and self-fit
The interviewees at the Western Bank confirmed that the norm for successful management style in the banking sector was overwhelmingly masculine and agentic. Many female managers stated that they preferred to work with a male manager, thereby demonstrating subtle biases related to the 'think manager, think male' paradigms (Schein, 1973). By evaluating female leadership through a stereotypically gendered lens, many female managers were unconsciously 'gendering' themselves: I wish my manager was a man. My manager is one of those women who became more masculine in time. She is difficult to handle; too ambitious, aggressive . . . She acts quite manly. Since she competes in a world dominated by men, I believe she chose to compete with them by becoming manly. (P6,Western Bank) Whereas the discussions at the Western Bank revolved around reluctance towards female managers endorsing an assertive leadership style, at the Turkish Bank, many interviewees did not think women were fit to demonstrate the masculine traits required for senior roles. Many interviewees stated that men were more suitable for senior positions as they believed that by nature women were emotional, and paid too much attention to detail, reflecting the role incongruity between the female gender and managerial positions (Eagly and Karau, 2002): 'Men are a lot better in dealing with stress, and, therefore, they are more suitable for leadership roles. Women are much weaker' (P40, mid-level, Turkish Bank). The findings from the Participation Bank revealed that the female managers held the strongest views of traditional gender roles among the three banking cases. Many interviewees believed that women by nature were less suited for leadership, owing to stereotypically feminine qualities such as being too emotional. Also, most respondents felt that female managers in general had low self-confidence and therefore were not fit for leadership: Being a manager requires staying calm, being self-confident, repressing emotions and taking a more rational approach. By nature, these qualities are predominantly found in men. Women are created to be more emotional. Therefore, it is harder for women to advance in their careers. (P45,Participation Bank) Having low self-confidence, in particular, was a recurring theme across female managers in the Participation Bank. The expectancy from women to demonstrate inhibition and modesty in their social interactions (Syed et al., 2005), prevented many of them from demonstrating the masculine qualities that were required to advance in their organisational hierarchy. Overall, interviewees from all three banks revealed that female managers internalised traditional gender roles to varying degrees. However, the intensive gender-biased culture towards female leadership and the lack of transparent criteria in the Participation and Turkish Banks may have strengthened female managers' own gendered role perceptions and lowered their aspirations for senior roles, signalling an interplay between meso-and micro-levels.

Work-family roles
Starting with the Western Bank, almost all interviewees highlighted the incompatibility between the 'ideal mother' and 'ideal worker' roles, as suggested by previous research (Pas et al., 2014). The interviewees overwhelmingly considered that the only way to succeed in the Western Bank was to be willing to sacrifice family obligations, despite the significant social role expectation for being a mother in the Turkish context: If as women, we decide to become mothers and have a career at the same time, you cannot come up with excuses like, my child is sick; I won't be able to come to work today. This is just unacceptable. Today your child can be sick; the other day it's your husband and then you . . . The working week is only 5 days . . . No one should have such a luxury . . . I didn't have such a luxury. I always had a 24/7 nanny. It is really difficult to live with a stranger, but I had to do this. In my job, I have to work until 21:00 and I need someone to take care of my child. (P3,Western Bank) In contrast, the interviews at the Turkish Bank portrayed a very different picture: most interviewees were satisfied with their work-life balance and felt that they were able to continue their career without having to neglect their families. Female managers attributed this positive experience to the family-friendly organisational environment.
The women at the Participation Bank were overwhelmingly family-centric. Thus, sacrificing family obligations for their career progression was not really an option. The majority of the interviewees perceived motherhood with a religious sense of purpose and fully supported the societal expectation to uphold childcare and household chores as their main responsibility: I believe being able to give birth is a sacred mission that has been given to women. If I were to choose between managing the world's largest company vs. raising a child, I would definitely choose my child. Therefore, women should not reject having children and obsess with being more successful than men or becoming a CEO. A woman's priority should be to give birth to a child and raise them to become a healthy and respectable individual. (P49,Participation Bank) However, these family-centric tendencies overwhelmingly hindered women's career advancement, as their organisation was more likely to value work-centric employees who were able to work for long hours and travel.

Discussion and conclusion
This article investigated the multi-level career barriers of female managers in the Turkish banking sector in order to understand the underlying mechanisms behind their limited representation in the organisational hierarchy. Although there is little doubt that women's careers are determined by a complex set of factors, the literature on women's career advancement remains heavily dominated by single-level and de-contextualised research affording limited attention to the macro-level forces operating behind mesoorganisational and micro-individual processes (Joshi et al., 2015;Metz and Kulik, 2012;Syed and Özbilgin, 2009).
This study marks one of the first multi-level analyses of women's career advancement, based on a comparison of three distinct types of organisations with different cultures and practices. In doing so, this study responds to calls in the gender and management and women's careers literatures to consider the interplay between the macro-mesomicro linkages that contribute to women's representation in organisational hierarchies (Kossek et al., 2017). Although there has been a growing interest in using multi-level models of women's career barriers, they often tend to focus mainly on organisational-, group-and individual-level processes (Hogue and Lord, 2007;Ragins and Sundstrom, 1989). The existing western-and US-based studies that incorporate societal contexts mainly focus on cultural values related to gender stereotypes and national-level equal opportunity and work-family policies (Kossek et al., 2017;Lyness and Grotto, 2018;Lyness and Judiesch, 2008). Extending these studies, this study has revealed how factors such as political ideology and the history of intergroup relations may also determine the gender relations within a national context (Syed and Özbilgin, 2009).
Significantly, this study is the first to identify religion-based fragmented social structure as a macro-level barrier influencing women's career advancement in the socio-cultural context of Turkey. Furthermore, the findings of this study differ from other context-specific studies by revealing how gender patterns can play out in culturally different ways even within a single context (Ali and Syed, 2017;Ituma and Simpson, 2009;Metcalfe, 2007). The findings revealed that women experienced the effects of patriarchal ideology differently depending on the type of bank they worked in. It was suggested that this variance may have been driven by the social fragmentations within the Turkish society based on religious values.
To explore the extent of institutional barriers and their impact within the boundaries of organisations, this study examined three distinct organisational cases. By doing so, the study revealed the existence of organisational heterogeneity rather than homogeneity in the Turkish banking sector. At the meso-level, the analysis found a combination of factors such as male homophily, double standards, intensified role incongruency, female modesty and discrimination. These barriers resonated with studies conducted both in western and Middle Eastern contexts (e.g. Ali and Syed, 2017;Eagly and Karau, 2002;Hekman et al., 2010).
Furthermore, the growing number of studies focusing on 'second-generation' or subtle forms of gender bias suggest that the narrative in the western/US contexts has moved beyond looking into overt bias to covert bias (Ibarra and Petriglieri, 2016). Similarly, studies conducted in Turkish corporate organisations have detected subtle rather than overt biases towards women in management (Burke et al., 2006;Culpan et al., 2007;Özbilgin and Woodward, 2004). However, the narratives in this study revealed that women continue to experience explicit bias in the forms of discrimination and double standards. By doing so, it responds to calls to take a closer look at gender processes within organisations (Joshi et al., 2015). At the micro-level, the study contributed to our understanding of the top-down societal and organisational influences on women's individual careers (Lyness and Grotto, 2018). In particular, the analysis highlighted the intricate links between these three levels by demonstrating how women's perceptions of macro-level barriers changed depending on their organisational context. This article also contributes to institutional theory in several ways. Institutional theory remains underdeveloped within career studies despite its importance in influencing women's career experiences, particularly in a non-western context. Insight from institutional theory is particularly beneficial for women's careers as it offers a multi-level approach to understanding the regulative, normative and cognitive factors affecting women's career advancement (Ituma and Simpson, 2009). This article extends institutional theory by introducing a new and focused perspective to understand the multi-level barriers women experience while advancing in their careers. In particular, this article contributes to the literature on women's career advancement by examining how macro-level institutional influences impact on organisational processes and practices and how they in turn influence women's career trajectories. This integrated framework highlights the embeddedness of women's careers within national contexts by demonstrating the importance of institutions in influencing women's careers while also capturing how these factors and the way in which women respond to them may vary depending on the type of organisation.
Second, this article revealed the existence of intra-sectoral variations regarding women's career opportunities by documenting how different organisational cultures and practices co-existed within the Turkish banking sector, depending on the organisations' capital structure (i.e. western, Turkish or Islamic). When dealing with institutional challenges, researchers often assume that organisations within an industry/sector will react the same way (DiMaggio and Powell, 1983). The findings of this study challenge this assumption. Aycan and Kirmanoglu (2007), who previously compared the managerial values of members of two business associations (one close to secular ideology, the other close to Islamic values and principles) in Istanbul, found no significant differences. The findings of this study are in contrast, and demonstrate how forces such as religious ideology may result in the co-existence of different organisational cultures in a single national context. This is a notable finding as it demonstrates how organisations can absorb or reject macro-level influences to varying degrees and how the female managers' career perceptions are significantly affected by their organisational context.
Third, the influence of Islamic culture on women's careers is substantial, although the literature in this area is still in its infancy (see exceptions: Göle, 1996;Metcalfe, 2007). By including the career experiences of Turkish female managers in a Participation/Islamic Bank, this research contributes to the limited knowledge on how Islamic values impact organisational policies as well as individuals' career perceptions. Specifically, the interviewees showed that female managers were in constant search of reconciling their religious as well as professional identities. Furthermore, although the significant growth of Islamic banking had created important career opportunities for them, they were also constrained by significant challenges within and outside their organisations. The interviews in the Participation Bank were considered imperative for this research, as it gave 'voice' to a particular group of women who have been overwhelmingly overlooked within the existing literature of women in management (Gunz and Mayrhofer, 2007).
The findings of this study have several important practical implications for policy makers, industry practitioners, academics and other stakeholders who are seeking institutional solutions to help women overcome the multilevel issues they experience in the workplace. For policy makers and regulatory authorities in Turkey, the study findings reveal the existence of both overt and subtle discriminatory practices within organisations, thereby signalling the need to enforce laws that close the promotion gap between men and women. Also, adopting policies that enforce private-and public-sector employers to report earnings and promotion gaps can further encourage the organisations to structure their career development opportunities to objectively promote the merit of their employees (Klarsfeld et al., 2012).
In the context of industry practitioners, the banking associations can provide statistics on how each member organisation is progressing in eliminating gender inequality, which would also encourage data transparency. Secondly, the study participants highlighted the need for more organisationally sponsored day care, to support universal access to childcare across the workforce. Thirdly, and specifically for international organisations, the findings indicate that the HRM practices within a country are to some extent context specific and, therefore, multinational organisations should take into account the cross-cultural differences while designing their diversity policies and tailor them to overcome any specific values that normalise gender hierarchies (Groutsis et al., 2014;Tatli et al., 2017).
Lastly, the HR managers are advised not to adopt a 'one-size-fits-all' approach (Rofcanin et al., 2018), but rather tailor practices and policies to counter the barriers experienced by their female employees. For example, the HRM departments of the Participation and Turkish Banks should pay considerable attention to implementing diversity training initiatives that target prejudices towards female leadership and create structures that facilitate transparent criteria for promotions and reduce the token status of women. On the other hand, the implications suggest the need to change the longhours norm in the western banks and restructure the HRM practices towards a familyfriendly work environment where both male and female employees are encouraged to use family-friendly policies.
This study has some limitations, which also provides interesting avenues for future research. Firstly, the sample of female managers at the Participation Bank was relatively small compared with the two other cases. This was primarily owing to the very few women working in mid-and senior-level positions in these banks. Second, some women may have self-selected a certain type of bank because of their distinct organisational processes. Third, owing to the institutional approach informing this study, the findings may have given the impression that the participants had little or no agency. Although the majority of the participants shared some degree of feeling powerless, the narratives of some senior-level women highlighted the importance of proactively negotiating for advancement. Although understanding such individual strategies was beyond the scope of this article, it presents an intriguing avenue for future research. Also, I acknowledge that my subjective interpretations and perceptions have an impact on how the interviews unfold, and constitute an important part of undertaking a qualitative inquiry. Thus, this can be considered as a limitation as well as a strength of the article. For a more comprehensive picture of women's career experiences, future research should also consider including facilitators of their career advancement. Moreover, studies that explore the contextual factors that influence women to adopt particular leadership styles would improve our understanding, particularly in the nonwestern context. Finally, the research suggests that there is scope for a greater focus on the intersectionality of religion and women's work and gender identity. Various Islamic practices and traditions exist amongst Muslim majority countries (Metcalfe, 2007), and therefore understanding religion's implications for gender equality issues in the workplace may play a vital role in integrating more women into the labour force in these national contexts.

Funding
The author received no financial support for the research, authorship and/or publication of this article.

Supplemental material
Supplemental material for this article is available online (Appendixes 1-4).