The impact of social capital-based selection practices on talent management: Evidence from telecom MNCs in Bangladesh

The purpose of this paper is to explore how social capital-based selection practices influence talent management (TM) in multinational companies (MNC). To date, little is known about how this relationship fulfills business priorities. The study collected empirical data through in-depth interviews with 12 managers from three telecommunication firms in Bangladesh. Our study found a significant influence of social capital-based selection practices on TM. We found that the relational and cognitive dimensions of social capital were the major drivers of selection practices that influenced a talent search process. Trust and cultural norms especially within networking relationships through shared narratives, codes and language, and vision affected TM involving selection practices. Therefore, the role of existing human resource management (HRM) system for selection practices was found inadequate without the integration of social capital. We report that upper-level managers have the capability to shape and implement a TM system that benefits candidates with network ties of informal social relationships, which consequently marginalized formal HRM practices. HRM system is only an administrative function alongside the dominant social capital-based TM selection practices. For practical implications, upper-level managers used social capital-based selection practices beyond HRM systems that were deemed pragmatic in selecting the right candidates for positions in organizations favoring intra-group homogeneity.


Introduction
Since proper talent management has been perceived as a solution to human resource (HR) challenges (Beechler and Woodward, 2009), selection practices have become imperative functions of human resource management (HRM) in the quest for talents to achieve person-organization fit (Collins and Clark, 2003;Collins and Smith, 2006;Lado and Wilson, 1994;Mahmud et al., 2012;Van Maanen, 1977). Nowadays, this concern regarding HR Challenges is true when filling most positions in organizations (Dreikurs et al., 2013). In this light, there is an ongoing debate about how selection practices can be more dynamic and apt in achieving greater person-organization fit. Accordingly, Augoustinos et al. (2014) argued that selection practices should be socially embedded in choosing suitable candidates-it should be based on close interpersonal relationships during the search for talents. More specifically, employers are keen to select candidates who possess common social traits (Rothstein and Jackson, 1980). Similarly, Louis et al. (1983) argued that most innovative organizations prioritize social interactions when selecting the right candidates. Although a review of literature reveals that the field of talent management (TM) is gaining momentum in research (Crane and Hartwell, 2019), in-depth investigations of how socially driven selection practices facilitate TM have yet to be conducted. Furthermore, research that addresses talent selection of multinational corporations (MNCs) from the perspective of social capital theory is somewhat lacking.
It should be noted that this type of research related to socially driven selection practices in TM was undertaken to a limited extent in the European context, which seemed to have provided impetus to garner more similar research in developing countries, particularly Bangladesh (e.g. Hosen, et al., 2018;Islam, et al., 2018;Mohaimen, 2013;Uddin and Arif, 2016). Thus, the present study aims to answer the following research question: How do social capital-based selection practices facilitate TM in telecommunication (telecom) organizations despite having formalized HRM systems in Bangladesh? The research objectives of this study are (a) to explore the process of selection practices from social capital perspectives in the light of formalized HRM system and (b) to examine how social capital-based selection practices facilitate TM in telecom organizations in Bangladesh. This study espouses social capital as the theoretical lens to evaluate relationships between social capital-based selection practices and TM.
Although social capital and TM are indirectly intertwined (Collings and Mellahi, 2009;Thunnissen, 2016), how can they be managed properly in the selection practices? If a wrong candidate is selected, it is more likely to adversely affect the existing workforce (Dreikurs et al., 2013). In addition, the new recruit is more likely to underperform and might subsequently leave the position, restarting the entire selection process until a suitable candidate is found. In this context, this paper details how HR department executes talent acquisition with the approval of top management for the implementation of social capital-based selection practices. The remainder of the paper is structured as follows: the next section reviews a conceptualization towards multidimensionality of social capital, selection practices, TM, integration between social capital and TM, the benefits and risks associated with social capital-based selection practices and TM, before we present our research methodology, findings, and discussion. We conclude with a brief discussion on this study's limitations and future research.

Social capital
Social capital, consisting of norms and informal values, drives group members to be subjectively obligated to share with one another in their networks. Social capital is instrumental in shaping relationships to make organizations work effectively (Adler and Kwon, 2002;Coleman, 1988;Fukuyama, 1995;Nahapiet and Ghoshal, 1998). Building interpersonal relationship is deemed as the essence of social capital theory in ensuring performance outcomes. The theory also describes the environment as embedded with goodwill, trust, and rapport, yielding positive outcomes through the development of stronger connections within social networks (Adler and Kwon, 2002;Maurer and Ebers, 2006). In this light, Adler and Kwon (2002) suggested two perspectives stemming from social capital, namely bridging and bonding relationships.
The perspective of bridging relationship concentrates on information, resource, brokerage, and control that benefit from weak ties and redundant relationships especially those that have disconnected entities (Burt, 1992(Burt, , 1997Granovetter, 1973Granovetter, , 1983Knoke, 1999;Steier and Greenwood, 2000). Contrarily, the perspective of bonding relationships is inclined towards producing shared norms, values, and cooperation from strong ties embedded in multiple relationships (Coleman, 1988;Fukuyama, 2003;Granovetter, 1973). Researchers evaluate social capital in several contexts being differentiated at the national (Fukuyama, 1995), strategic (Ireland et al., 2002), and organizational levels (Cohen and Prusak, 2002). Subjectively, researchers discerned a need for in-depth examination into the multiple dimensions of social capital in its developmental approach to improve employee commitment and performance (Adler and Kwon, 2002;Arregle et al., 2007).

Multidimensionality of social capital
In the past, social capital was studied for its attributes to unique resources that are embedded in relationships among network ties (Camps and Marques, 2014;Inkpen and Tsang, 2005;Nahapiet and Ghosal, 1998;Putnam, 1995;Sanchez-Famoso et al., 2014). These attributes have been streamlined into three dimensions: structural, relational, and cognitive. The structural dimension of social capital encompasses both informal and formal networks (Kaasa, 2009;Nahapiet and Ghoshal, 1998). Informal networks are developed through interpersonal relationships and become reciprocal among friends, relatives, colleagues, neighbors, club members, etc. On the contrary, formal networks allow members access to associations and voluntary organizations (i.e. professional groups, religious bodies, cultural groups, alumni associations, consulting service providers, etc.). Both networks are actively involved in providing support and communication channels for exchanging information between groups of mutual interest (Kaasa, 2009).
The relational dimension among member develops gradually through interactions that are based on respect and friendship which have a direct influence on the behavior of members in a network. The relational dimension anchors on personal relationships that can fulfill social motives such as sociability, approval, and prestige (Nahapiet and Ghosal, 1998). The cognitive dimension relates to resources that exhibit shared codes and language, shared narratives, and shared goals (Nahapiet and Ghosal, 1998). Therefore, such shared resources enable members to achieve their objectives.
The structural and relational social capital significantly integrated knowledge in project management context, but not cognitive social capital (Ekemen and Şeşen, 2020). This highlights the priority of structural and relational social capital over cognitive social capital in knowledge integration which can be applicable in HR context. Han et al. (2020) examined the relational characteristics built upon social capital theory that facilitate knowledge-sharing relationships. Interestingly, the structural dimension of social capital was central to creating networks for knowledge sharing. In particular, their research revealed that trust and bonded network ties were the significant influencing factors behind knowledge sharing.

Selection practices
Numerous research papers have found that selection practices are crucial to facilitate TM from the traditional HRM perspectives. Although selection practices are designed as a legal process from HRM perspectives, actual TM process requires integrating social perspectives apart from the HRM system (Barber, 1998;Collings and Mellahi, 2009;Thunnissen, 2016). Interestingly, Van Maanen (1977) critically identified "personorganization fit"-oriented selection process to be most sustainable in serving the short-and long-term interests for both employee and the organization. Selection practices that facilitate organizations should be judicious in extracting talents from a pool of potential candidates to fill vacancies by side-stepping the traditional HRM system. Accordingly, Thunnissen (2016) advocated proper employment planning and forecasting that can enable organizations to conduct successful selection. Barber (1998) found three phases of selection process which are generating applicants, maintaining applicants' status, and influencing job choice decisions. Researchers explored the need for new approach to selection practices to attract applicants. In a similar vein, Rynes et al. (1991) argued that employers need to act wisely in considering a wider range of possible objectives to ensure effective selection practices. In instituting selection objectives, organizations should focus on measuring post-hiring outcomes (Breaugh, 1992), such as the selection cost, the speed of which jobs are filled, the number of individuals hired, and/or the diversity of new employees.
Employers are often not concerned with these post-hire outcomes (Williams et al., 1993), but are rather interested in pre-hire outcomes such as the number of individuals applying for a position based on the quality of their credentials, diversity, and the number of individuals who accept job offers are tenured. If an employer showed interests in the more proximal outcomes of selection process, its strategy development would be to focus on how to accomplish them. If selection objectives have not been achieved, it would be difficult to develop a sound selection strategy (Breaugh and Starke, 2000;Rynes and Barber, 1990).

Talent management (TM)
According to Scullion et al. (2015), the term "talent" implies either "genius" or "highly gifted" with extraordinary insights and critical thinking in getting the job done. Iles et al. (2010) claimed that TM emerged as a different approach to managing people compared to existing HRM functions in organizations. TM is more than a subtle repackaging of old techniques and ideas with new labels. Therefore, TM has distinctive approaches that lead to organizational success through competitive advantage. Specifically, TM is the prime driver of attracting and retaining talents which works well with certain sections of the workforce that contribute to the core business value chain. Collings et al. (2011) found the TM concept in European context to be quite differentiated from the North American context. In the North American context, TM is seen as a management "fad" or "fashion" that differs from the HRM system. Although HRM may integrate management techniques (e.g. Total Quality Management, Business Process Reengineering, Just-In-Time) in its application, HRM with TM has a selective focus on attracting and retaining talents. Furthermore, TM has an organizational-focused competency development view which is a firm-level perspective on improving competency by retaining talents (Iles et al., 2010).
A somewhat different analysis by Lewis and Heckman (2006) presented TM as being mostly similar to HRM which involves activities such as recruitment, selection, training, and development appraisal. Basically, scholars merely replaces the word "human resource" with "talent" and is considered a mere relabeling or rebranding of HR practices to enhance credibility and to show novelty. Moreover, CIPD (Chartered Institute of Personnel and Development, 2007) described TM as a conceptual understanding of how HR professionals define and recognize talent. This is contrary to Iles et al. (2010), which saw TM as a strategic management of talent flow. The aim of this strategic management is to select the right persons from the pool of candidates and to assign them the right positions at the right time based on the firm's strategic business objectives. Creelman (2004) claimed that TM is not perceived as a set of topics, but instead as a perspective or mindset that support the idea of selecting the most critically talented individuals through social networks who play a pivotal role in the successful implementation of corporate plans. Cappelli (2008) claimed that TM concepts are drawn mostly from operation management and logistics theories, and not from marketing theory. Because TM is dependent on talent pipelines which are mostly internal to the organization, TM focuses on the competence development of individuals by means of managing the flow of talents. Korzynski et al. (2020) claimed that attracting new talents to a firm requires social activities stemming from the corporate brand and the media influences that would mobilize applicants and prospective employees to the firm's vacancies. In particular, HR plays a crucial role by leveraging employees who promote the corporate brand to help firms attract new talents to the firm's vacancies. Similarly, Blass et al. (2008) emphasized TM to be more functional than HRM in terms of leadership development and succession planning. TM is more of a collective approach of social networks directed towards selecting, retaining, and developing personnel as talents for organizational objectives.
Streamlining talents who can enable organizations to meet their current and future business priorities creates a platform for growth and for meeting new opportunities (Anand, 2011). Superior talents would contribute to exceptional business performance (Taleo, 2010). Collings et al. (2019) claimed that the adoption of a global, multidomestic, or transnational strategy to facilitate global talent management by multinational enterprise at the headquarter level significantly influences the performance of the enterprise. Furthermore, it was viewed that vertical fit stems from higher-level factors such as individual performance and humancapital resources and how these factors are transformed to global TM. Based on the resource-based view (RBV) theory, the effective and successful implementation of TM strategy could result in unique internal competencies of the firm when people are highlighted as the most valuable resource who are rare (unique), non-substitutable, and inimitable due to their adeptness in certain areas of expertise (Kabwe and Okorie, 2019).

The integration of social capital and TM
Some works have made theoretical contribution to the study of TM (Asag-Gau and Van Dierendonck, 2011;Bethke-Langenegger et al., 2011;Cappelli, 2008;Collings and Mellahi, 2009;Groysberg, 2010). Many organizations face the challenge of creating a blueprint for their talent pools (Iles et al., 2010) to develop person-organization fit in workplaces (Van Maanen, 1977). Thunnissen (2016) identified that there is significant dearth of theoretical formulation that integrates social capital and TM. However, TM as a rapidly growing field is addressed in HRM literature facilitated by social capital through its networking relationships (Collings and Mellahi, 2009). Social capital and its three dimensions (e.g. structural, relational, and cognitive) have been reported to be vital in facilitating the TM processes of attracting, assessing, developing, training, and retaining employees (Ahmadi et al., 2012). Leana and Van Buren (1999) argued for the indirect impact of TM on the development of knowledge management through the use of activities that improve social capital. This is because social capital is considered an organizational asset inherited from social relations and networks. Similarly, Jones (2008) showed that TM, in the context of HRM literature, leads to social capital and thereby enhances knowledge management resulting in sustained competitive advantage for firms. Accordingly, Harsh (2009) claimed that knowledge is basically linked to human actions and it is context-specific, relational, dynamic, and humanistic. As such, knowledge creation will be unattainable if the complex nature of human interactions is perceived wrongly. For the retention strategy, organizations actively harness talents to manage knowledge by facilitating and boosting social capital. Therefore, organizations can obtain capabilities to affect social capital through TM processes (Jones, 2008).

Benefits and risks associated with social capital-based selection practices and TM
The direct benefit of social capital is the facilitation of the right information for the focal member. Social capital facilitates access to broader sources of information and improves information quality, relevance, and timeliness. According to Jones (2008), social capital develops employee interactions and harnesses processes that eventually disseminate learning to one another through building and reinforcing trust, communication networks, and norms among members. These actors are seen to facilitate TM through social networks. Therefore, TM is an implementation of systems or strategies conducive to developing improved processes that are intended to increase workplace productivity for the purpose of attracting, upgrading, retaining, and utilizing personnel with the requisite skills and aptitude capable of meeting current and future business needs. Subsequently, an increased organizational competitiveness can be attained by implementing welldesigned TM processes that reduce employee turnover and amplify job satisfaction and commitment.
Despite having numerous benefits embedded in social capital, critics found apparent risks in social capital information. Building social capital requires substantial investment to establish and maintain relationships. In the case of selection practices, social capital might influence them with regard to facilitating TM. In this light, less qualified people might be selected due to their networking relationships with focal actors who are in charge of final selection and employability. Social capital is seen as an expensive investment and may not be cost efficient in certain situations (Adler and Kwon, 2002). It is often found that social capital builds strong solidarity within group members. This might over-embed the actors in the relationship. Such over-embeddedness reduces the flow of new ideas into the group, resulting in parochialism and inertia (Gargiulo and Bernassi, 1999).
Despite these criticisms, the extant literature depicts social capital as a valuable resource that influences TM practices through emphasizing and facilitating social networks at managerial levels. Besides, Leana and Van Buren (1999) have demonstrated the positive effects of social capital effects on career development and work-related outcomes. Similarly, it was found that social capital plays a vital role in offering a cozy environment in workplaces for sustainable employee development and retention (Coleman, 1988;Dress and Shaw, 2001). However, scholars such as Luthans and Youssef (2004) argued that organizations need to develop latent resources through social capital to sustain competitive advantage. Yet, the literature gaps still remain unexplored and empirical investigations are needed to reveal how social capital-based selection practices influence TM in relation to a qualitative approach.

Research Methodology
This study employed a qualitative case study with cross-case analysis as qualitative outcomes provide useful truth (Faizan et al., 2019). Stebbins (2001) suggested that the exploratory nature of qualitative study is useful for the extraction of unforeseen elements from the phenomenon while doing a case study research with cross-cases analysis. A case study research evolves replication logic (Glaser and Strauss, 1999;Yin, 1989), which not only ensures the confirmation of emerging evidence independently (Eisenhardt, 1989;Lewis-Beck et al., 2003) but also draws validity through generalizing contextual research (Yin, 1989). Therefore, involving in-depth interviews with participants are crucial in obtaining a deep understanding of the context. There are no alternatives to in-depth interviews in capturing personal experiences, life stories, feelings, and contextual information (Hennink et al., 2020). In-depth interviews would also enable self-reflection and introspection during data analysis (Patton, 2002). As such, the cross-case study was conducted on three MNCs within the telecom industry in Bangladesh.
There are pertinent reasons for carrying out this research. First, the demand for talent research in telecom organizations has risen from 2015 to 2020 due to the increasing numbers of mobile subscribers in Bangladesh (Lucini and Hatt, 2014). The number of mobile subscribers has exceeded 165.337 million by the end of March 2020 (BTRC, 2020). This makes Bangladesh an important hub for MNCs to operate successfully in mobile telecom despite shortages of skilled managerial employees. Second, the employee selection practices were designed based on the existing HRM system where organizations faced extreme difficulties in identifying candidates for better person-organization fit (Rima and Islam, 2013;Van Maanen, 1977). Lastly, this research is exploratory in nature and therefore unforeseen aspects and rich data can be captured using a qualitative approach (Creswell, 2018).
The current research adopted the interpretive philosophy with an inductive approach. Literature on managerial perspectives over the last five decades has included this research approach. In addition, this study used the "socio-anthropological" research paradigm that is aligned with the epistemological stance involving a critical relativist ontological stance within a qualitative research methodology.
Furthermore, this study used the triangulation approach as part of ethical considerations for the benefit of cross-case analysis. The triangulation approach is useful to verify and authenticate information, including participants' background. Indeed, it is an effective way of facilitating the inclusion of the right kind of participants with substantial experience, knowledge, and skills pertinent to departmental talent search through selection practices. Systematic observations and interviews were conducted in this study.
The model of qualitative data analysis (QDA) espoused in this research is coherent with Berente's et al. (2019) suggestion involving an iterative, recursive, and holographic process. The noticing process in the current study involves making observation, taking down field notes, conducting interviews, collecting documents, etc. The codes utilized were in line with the theoretical framework exhibited to illustrate meanings. For the purpose of triangulation, two researchers sorted the information into codes separately and explained their findings. Raw data twice was codified twice, as shown in Figure 1.

Figure 1
The qualitative methodological process

Source: Own elaboration
However, the data were codified at large under the lens of social capital that affects selection practices. In turn, the data derived from TM in practice were codified both under social capital and TM process. Through this, two clusters of codes were generated. The thinking process involves closely examining, seeking for similarities and dissimilarities, and raising questions regarding the phenomena in the data (Strauss and Corbin, 1998). The double codification was useful to possibly contrast between the two clusters of codes. We evaluated and qualified social capital codes applicable for every participant to observe homogeneity in managerial groups on one hand, and observed relationships between social capital-based selection practices and TM on the other. Finally, in relation to how TM inspires a talent pool, we observed how selection practices are shaped by relational, structural, and cognitive social capital influencing TM. Our QDA led by systematic observation allowed us to record and capture narratives, codes, and complimentary data that stemmed from the individuals' in-depth interviews that lasted over a reasonably lengthy period.
Based on thematic analysis, we developed themes from the data. The thematic analysis also enabled us to ameliorate the sequential flow and structure of the interview process. We adopted Attride-Stirling's (2001) proposal for a thematic data analysis framework pertaining to systematic data analysis, which facilitated researchers to arrange verbatim data into meaningful and manageable portions. Hence, we generated codes and themes to elucidate the relationships of the contents through the same  Haque and Aston's (2016) "funnel approach" to reveal unexplored facts from the research phenomenon. Verbatim transcriptions of 12 interviews were checked twice when data was entered into text files separately.

Data collection and sampling
Three telecom organizations were chosen because they are private MNCs and share similar business operations. Since our study aimed to identify the influence of structural, relational, and cognitive social capital over internal selection practices within intra-group homogeneity in organizations, we needed to gain access to a generously small group inside organizations to gauge homogeneity. In a discourse with a Chief Executive Officer (CEO), he mentioned that the internal selection at the operational level is managed mostly by mid-level managerial group (MLMG). MLMG is involved in internal selection either through mobility of talents or to fill vacancies and the group would collectively facilitate socialization and foster training and development for new recruits. Accordingly, for our research, we selected both MLMG and upper-level managerial group (ULMG) as the unit of analysis. ULMG was considered to reduce response bias.
Data was collected data between 2018 and 2019. MLMG and ULMG were drawn from various departments, namely finance, HR and administration, corporate affairs, information technology, sales, infrastructure development, procurement, commercial division, and network quality. To select our cases, we employed a theoretical-based sampling method as adopted by Eisenhardt and Graebner (2007) that explores the phenomena from the managerial level in these telecom organizations. This enabled us to see how social capital is related to HRM functions, especially internal selection, in facilitating TM. Thus, the empirical data collected through in-depth interviews were from the managerial employees. Since the research focus is nested towards the selection practices of HRM process, we managed to capture interviews on this basis to access those who voluntarily agreed to participate. Some departments such as marketing, operations, logistics, etc. could not be covered due to their tight schedule and also their repugnance towards interviews.
The purpose of this study was to investigate the organization processes that facilitates the TM process through selection practices. We needed to position ourselves in close proximity to the participants for these interviews to capture relevant data without hindering their usual modes of operation. We collected primary data from in-depth interviews, involving a semi-structured questionnaire with open-ended questions that allowed participants to discourse freely without disruption. We developed the questionnaire based on the tripartite framework of social capital presented earlier which included questions on social capital, selection practices, and TM. The first part of the interview questions was related to social capital and its structural, relational, and cognitive dimensions. The second part of the interview questions was directed towards selection practices, in line with Chen and Huang (2009) and Chuang et al. (2013). Finally, TM was explored based on the current state of organizational practices which were supported by extant literature that exhibits indirect effects of social capital on TM.
We interviewed the participants outside their offices like at cafés so that the participants would feel at ease and answer as naturally as possible. This type of setting is suitable for the participants to speak eloquently and for us to make sense of the participants' subjective understandings (Alvesson and Sköldberg, 2000). We also conducted non-participant observation to supplement the primary data. The CEO and senior level staff of each organization were also interviewed. In addition, we gathered secondary data from the organizations' websites, conducting case studies that could be of importance. Table 1 represents the anonymous identities of the telecom MNCs, the regional locations, nature of business, total number of in-depth interviews, and the designation of the participants. Managers from ULMG and MLMG were selected based on age, level of education, and years of work experience. Managerial groups' characteristics are presented in Table 2. According to Francis et al., (2010), a minimum of 10 responses is sufficient to reach the conclusion of saturation point. In this study, we obtained 12 interviews resulting in a total saturation point in which 9 interviews were sufficient to draw logical conclusions. To represent both ULMG and MLMG equally, we ensured the sample size of each group to be six participants (Haque et al., 2018). This equal representation would also reduce response biasness as suggested by Faizan et al. (2019).

Data analysis process
To analyze the data and develop themes, we followed four steps. First, interview and archival data was transcribed and transformed into text files. Second, every transcription was assessed and given keywords through initial codes which were used by participants or subjects of the archives. Third, initial codes were converted into themes in the light of theoretical concepts. Finally, themes were presented through a diagram showing their relationships. As aforementioned, this study's is on three telecom organizations and we conducted cross-case analysis for case-based generalization. For the purpose of data processing, this study used the qualitative software NVivo version 11. The outcome of the interviews was submitted to the participants for review and approval and was used for this analysis.

Findings
The findings revealed that HRM practices aimed to internally employ people of higher quality and to facilitate in creating a skilled workforce for an organization. For that reason, managers were keen to prioritize internal selection and socialization to fulfill vacancies and to promote young and creative talents in organizations using social capital.
The internal search for talents was carried out within known network members. From the interviews, the participants claimed that there was social capital, through its relational and cognitive dimensions, integrated in their selection practices to support TM apart from the existing HRM system: "It is difficult and to some extent, extremely time consuming to find talents from external sources to adjust to our needs, so we've gone far more to develop it in-house, creating junior positions and trainees and enhance their leadership role in the organization."-HR and Administration Director, from ULMG of Company A "Basically speaking, social relations among people and selection of talents are co-related and both of them are interdependent. In this way our company recognizes TM by means of this kind of correlations especially inhouse."-CFO, from ULMG of Company C However, the Senior Manager (Corporate Affairs) from MLMG of Company A observed that: "As far as our understanding is concern, TM is still deemed as selection, training and development, retention, incentives and promotion for own networked people." The aforementioned statements clearly indicate that social networking relationships are of central considerations for selecting in-house talents from the organizations. Similarly, Adler and Kwon (2002) argued that the power structure stemmed from social networks often shape selection process. For that reason, network members appear in the form of dominating leaders who substantially approved the final selection of candidates.
"Like-minded people are our best choices, because they are aware of what we want them to do."-CCO, from ULMG of Company C

"I need an excellent and steady team for long term. For this purpose, the attraction and retention policy which has become a key issue should be addressed prior to selecting talents. If you have a sustainable attraction and retention policy for talents, it matters a little to those talents who were selected from your social informal connections despite a trivial offer was made to them. It would be a win-win situation if social relationship sustains between employers and employees."-CEO, from ULMG of Company A
However, the Senior Manager (Revenue Assurance) from MLMG of Company A stated: "Our company's management intentionally selects talents from its own networks to serve its own purpose. Although our company conducts some entry level exams for selection purpose as a placating measure in order to legalize selection process. Actually, fellow candidates for talent hunt are pre-decided by our top management." The aforementioned statement unambiguously deliver the message that social networks considered selection procedure as strategic and critical in nature where personal preferences of network members were pre-chosen. Therefore, candidates were nominated based on social relationships. These names were provided to HR in order of choice and then assessed for eligibility of the positions offered.
"We have a total game plan. We arrange our nominated players who will play according to our proposed plan. The only thing we do is that we recommend to the referee so that we can finish our game smoothly with triumph."-Head of IT, from ULMG of Company B "Our company pays a great deal of attention towards TM. Because, TM is the final output that is originated from socialized networks of people who are highly connected with one another."-Senior Manager (Infrastructure Development), from MLMG of Company B Besides, Senior Manager (Network Quality,) from MLMG of Company C observed that: "TM development scheme is organized around our company from the perspective of social orientation among people within highly connected networks." These statements provide evidence of how candidate screening processes are a mere placating measure for HR to assess the suitability of pre-selected candidates. Although this systematic procedure is in place, interviews were only granted to those on the managers' priority list. In this regard, Senior Manager (Human Resources) from MLMG of Company C stated: "Our role happens to be as facilitators or promoters other than involvement in the processes due to our sparser network connection. We conceived selection practices as routine work where we are only involved in executing the process after being approved by top authority. They are all in all and densely connected with each other." However, Senior Manager (Network Quality) from MLMG of Company C observed that: "Bonded networks and socially connected people are the answer for selecting talents in our organization." "You can't always find what is called talents via advertisement during the time of your need, what you need to do is to search talents near surroundings of your neighborhood."-Senior Manager (Revenue Assurance), from MLMG of Company A From these statements, it is implied that there are denser networks connected to each other. Subsequently, nominated candidates are derived from such networks and when selected, their profiles will be prepared and submitted to the respective division heads. HR would then look for a suitable vacancy for the candidate. In this regard, Sales Director from ULMG of Company B said: "Despite challenges, we have to unfold window of opportunities for the recruits so that they can evolve proper systematization of our existing set up. It is very important for new assignments of candidates to fit well with the positions, otherwise continuous improvement can't be ensured to sustain against our competitors." "Simply speaking, by TM we understand it as an efficacious means to achieve the concept called person-organization-fit. To obtain that fit-ness, our company is committed to getting the right people into right places from social connectivity and making sure that right people and places are well matched."-Senior Manager (Procurement), from MLMG of Company B

"TM is the product obtained through the selection of qualified and valuable people in organizations. We truly care about TM that requires networked corporate family members for its continuous support and development"-Senior Manager (Corporate Affairs), from MLMG in Company A
Participants trust the selection process to deliver the right candidates to contribute necessary firm-level performance. However, HR still receives both positive and negative outcomes from the selected candidates. Candidates who accept the offer of the new position would be notified and served with an employment contract. Having reviewed the terms and conditions, candidate would sign the employment contract. While talking about the terms and conditions of the employment contract, Senior Manager (Human Resources) from MLMG of Company C stated: "Although we use reasonable terms and conditions of employment, we always face criticisms from fellows as if we look down to them. We simply comply with what the management wants and that's why we often become the objects of acerbity." "Our company successfully launched TM program mainly on the basis of two considerations. First, when a company gradually becomes larger in size, it requires trustable appropriate manpower (e.g. talents) to ensure person-organization fit. Second, we realize the difficulty of getting trustable talents. They only can be selected from own sources rather than disconnected networks in order to fill vacancies we need."-CEO, from ULMG of Company A Despite all criticisms, contracts were signed and appointment letters were given to selected candidates. HR would then take the initiative to facilitate the socialization of these new recruits with others in the organization and arrange training for the recruits for further skill development. Therefore, such informal connections in network ties that explicitly play a role in internal selection practices facilitate the building of talent pool inside the organization. This establishes the importance of cognitive and exclusively relational social capital as a primary feature in the selection process in the TM of these MNCs. In this light, an empirical model is delineated that presents how TM is facilitated by the pursuance of social capital-based selection practices. Figure 2 represents our empirical model developed through data analysis.

Internal selection through the utilization of social capital
Telecom MNCs in Bangladesh have informal social systems to select promotable employees or fill vacancies internally whenever possible. Figure 2 presents the many steps taken in an internal selection process where informal social relations dominate the process in these organizations. The Senior Manager (Human Resources) from MLMG of Company C stated: "For different levels of positions, internal selection method is used to search for candidates through social networks." "Non-exempt positions are treated as lower-level jobs involving manual and even clerical duties with minimum wage paid on hourly basis and overtime provisions as per fair standard act while exempt is considered for positions at the middle and upper level."-Senior Manager (Revenue Assurance), from MLMG of Company A

Figure 2
Managing talents through social capital-based selection practices "When internal selection takes place, the hiring manager who considers candidature relies on supervisors' recommendations of better qualified candidates which are influenced by whether these candidates appear to have informal social relationships with executives at upper levels." The recommendations by supervisors discriminated candidates based on their informal social relationships, specifically the relationship with upper management. This is considered a normal practice as implied in the statement. The names of these candidates would then be proposed and submitted to HR for data screening. However, anonymous managers from the middle echelon cautioned about the drawbacks embedded in such a selection system which prioritized informal social relationships of candidates-they might accept incompetent candidates while disregard qualified ones who really deserved such positions.
The CFO from ULMG of Company C also mentioned that: "It is not at all an established truth that candidate selected from informal social connection appears incompetent or undesirable. (1) These candidates may be good candidates whose potentials may not be justified properly or overlooked, (2) A supervisor may hoard his or her qualified candidate because supervisor prefers the candidate to stay in the department and therefore recommends desirable position internally rather than elsewhere." This suggests the superiority of such a subjective selection process to fit the right candidates in the right positions. "Hoarding" and discrediting a formal system implicitly suggest the preference for intra-group homogeneity reinforced by shared norms, languages, and codes (the cognitive dimension of social capital). In fact, as a post-hoc justification, some ULMG members suggest the need for the use of a skill inventory that encompasses a systematized method capable of validating the qualified candidates. Nonetheless, after thorough screening of candidate's data, HR separated their profiles into two sets-unacceptable and acceptable.
"Those who were approved unacceptable at the rudimentary stage by HR, they were advised to enhance certain skills and qualifications and then encouraged to apply again. And those who were successful and approved acceptable were then processed with the permission of the top management to conduct interview."-HR and Administration Director, from ULMG of Company A In this situation, the top management's approval to conduct the interview requested is based on implicit preferences. When permission is granted by the top management, HR precedes a long series of steps to determine the fitness of the candidature for the opening of the positions. Accordingly, Senior Manager (Corporate Affairs) from MLMG of Company A mentioned that: "This depends on the implicit choices that contain two perspectives. Firstly, the top management usually avoids granting permission to call for an interview if the selection of candidate is made from a sparser network. Secondly, the reverse case is true when the selection of candidate is made from a denser network; permission is usually granted to conduct the interview." Preference seems to be assigned to candidates from familiar networks which was mentioned as "denser network". It appears that the culture of norms in the network relationships, and shared narratives, codes, and language among network members stemming from relational and cognitive dimensions influence top management decisions in selection practices. This priority for candidates from familiar network connections becomes common practices as implied in the statement. Subsequently, with the granted permission, HR prepares and submits the candidates' profiles to the respective department, which would then conduct interviews to fill their vacancies by selecting suitable candidates. In this light, the CCO from ULMG of Company C stated: "Two scenarios appear after the interview. Firstly, candidates who are rejected are provided feedback to the candidates' source division heads by HR. Simultaneously, they are notified with specific reasons why they are rejected and also recommended to enhance areas of individual's skills which has current deficits or shortcomings in order that one can adjust with the requirements of the respective division which currently needs people to run their departmental functions. Secondly, selected candidates are notified by the selection division and then HR determines a suitable opportunity for the selected candidates to locate them in positions." The remarks suggest that the traditional HRM system, coupled with the subjective selection process, ran concurrently in performing administrative functions to offer available positions that contain job responsibilities to the successful candidates. However, HR and Administration Director from ULMG of Company A observed that: "A number of candidates refuse to accept the positions offered while some welcome the offer and provide their positive consents. Meanwhile, after getting the positive consents, the selection division notifies HR to prepare employment contract. Accordingly, HR issues new employment contract with all details of official code of conduct and emoluments against the job responsibility." Upon receiving the said employment contract, the candidates would finally confirm their willingness to work in the new position and sign that new contract. This entire process was formulated based on the cognitive and exclusively relational social capital especially in connection to the informal setting within the network ties in a denser network.

Discussion
The paper has explored how social capital-based selection practices facilitate TM in the context of an emerging country like Bangladesh. For the telecom organizations in this study, the TM process was managed effectively through the selection practices of the HRM system in a way that merges the different dimensions of social capital especially when there is trust. This was found to be important in implementing selection practices. TM practices appeared to operate differently compared to normal HRM practices in these organizations. This is not a simple repackaging of old techniques and ideas of HRM with TM as a novel brand label. TM is not viewed as another endeavor by HR personnel to ground their legitimacy, reliability, credibility, and status, but as offering different approaches that were useful to facilitate selection processes in service-oriented, innovative organizations like these telecom organizations.
Most participants expressed their views about the TM process which is not supported in literature related to traditional HRM system. TM process within HRM system was found to be integrated with a social perspective that gave an edge to candidates from their networking relationships. Our model has demonstrated the importance of social capital-based selection practices of denser managerial groups. This paper offers a dynamic perspective to the selection of candidates by aligning social capital-based view stemming from social networking relationships. We have illustrated how social capital is embedded in network relationships and how it has impacted selection practices in order to facilitate TM.
Preferences were assigned to candidates in the selection process who have network relationships among managerial employees. This was found relevant in our analysis and at times appeared powerful in ensuring that desirable candidates were selected to fit required positions. This relational dimension of social capital shaped the selection practices. The participants believed such selection processes promote TM by subjectively ensuring better person-organization fit. Basically, social capital-based selection practices of TM in the three telecom MNCs helped to ensure intragroup homogeneity within the firm.
All the while, the TM processes were integrated within traditional HRM systems. This theoretical analysis indicates that selection practices need to be adjoined with an understanding of social capital for effective TM. These findings suggest that intra-group homogeneity that maintained networking relationships, implicitly supported by shared norms and languages, is common in the formal TM process in Bangladesh.
The three dimensions of social capital derived from the literature review were all in existence in the TM selection processes of the MNCs in this study. These dimensions were behind the facilitation of the selection practices of talents from within the organizations rather than from external sources. In this process, social relations are the drivers pushing towards finding talents for the organizations. For example, the relational dimension with its implicit trust and obligation factors was the preferred criteria in the selection practices dominant in Company C. The selection practices recognized social in-house co-relations among network ties as a legitimate approach to TM. It was also revealed from Company A that networked people through selection practices benefitted from TM in terms of training and development, retention, incentives, and promotion. Similarly, the practices in Companies A and C suggested that talents from informal connections are well-regarded by the top management in terms of favoring such inclusion as valid criteria for granting interview approval in the selection process.
Like-minded people were considered as best choices by top management because, coming from familiar networks, they were assumed to share similar norms, codes, and languages, and were therefore expected to be easier in being shaped and then utilized to achieve what top management required. Further to this, selecting "exclusive people" as talents was seen purposefully done in order to facilitate their own network ties. As such, in Company A, the entry-level exams as a selection method was a placating measure to legalize the process. This was similarly the case in Company B where talents were candidates pre-chosen from talent pools originating from social networks within formal connections to top managers in the firm.
Besides, in Company C, while the selection processes were critically reviewed to elucidate the facilitator's role in performing the routine work of processing the selection of candidates, the key decision-makers were top managers. The top management is predisposed to consider social networked people as potential candidates who will fit well with the organization to attain organizational objectives. Furthermore, Company B, while inclined towards using favorable terms like "smart workforce", took a more precise viewpoint on TM as organization-fit which stemmed directly from selection practices favoring candidates from network ties closely related to family members of top management.
Middle managers were often subjected to acerbity by selecting talents not based on the prescriptions guided by the top management. Yet the TM programs were designed to reflect on the conditions upon which trust issue and person-organization fit (Van Maanen, 1977) essentially dominate selection practices. For example, as the company grows in size, it requires trustable talents to ensure person-organization fit. Company A identified the difficulty of getting trustable talents from disconnected sources. Rather, trustable people are deemed only available from its own social networks. Similarly, regarding internal selection, Company C also saw that candidates were nominated based on preferred social networks for non-exempt employees who are eligible for overtime pay. The selection practices of TM were seen as filling "exclusive positions" and upper-level managers were the major contributors of manpower to fill these vacancies. However, the selection practices of exempt employees' positions were conducted at times drawing from disconnected networks subject to the unavailability of candidates from the denser network.
In this connection, there were similarities to Company B which found that internal selection practices could not guarantee better qualified candidates. Therefore, social affiliations with the upper-level managers appeared as very crucial for talents to move further ahead in their career. The candidates affiliated with the upper level were approved as acceptable to take up new positions. If a candidate belongs to the denser network within the upper-level management, she or he has greater opportunities to become the leader in succession planning. However, Company C tended to critically view candidates selected based on informal social relationship. Such critical view may result in the misconception of incompetence in such candidates. As a result, supervisors seldom want to keep such candidates long in their department and instead recommend them to positions elsewhere.
The first contribution derived from our findings provide evidence for how networking relationships of social capital influenced selection practices. This type of social capital-based selection practices highlights the importance of managerial groups that take action to facilitate TM. Prior studies have neglected network-based selection practices that were dependent on the relationships of managers. Our empirical analysis indicates that social capital plays a greater role in shaping selection practices by using the networking relationships of managers to encourage TM. The originality of this research stems in respect to the focus on social capitalbased selection practices among managers.
The second contribution is devoted to the balanced approach involving networking relationships of relational and cognitive dimensions of social capital explicitly, and structural dimension implicitly, in the selection process which functioned effectively alongside the traditional HRM system which were neglected in prior studies. This study shows how the selection practices of TM were driven by social capital in which the cognitive dimension imbued in the socialized relationships prioritized intragroup homogeneity. The approach that suggests social capital largely affects TM by means of selection practices is novel in the literature.

Conclusions, Research Limitations, and Future Research
We confirm through our in-depth qualitative analysis that social capital-based selection practices have a significant impact over TM in telecom organizations. We found that relational and cognitive dimensions of social capital are the major drivers facilitating selection practices to influence the talent search process. This is usually the case when trust and cultural norms within networking relationships are embedded in shared narratives, codes and language, and vision. Therefore, the role of the existing HRM system for selection practices was found inadequate without the integration of social capital especially relational and cognitive dimensions that largely impact TM.
The findings of this study offer important practical implications for TM. Our analysis suggests that socially oriented selection practices apart from existing HRM system are useful for TM in the telecom MNCs as innovative service organizations. Telecom organizations need to employ people in accordance with organizational fit. Among distinctive managerial groups homogeneity, ULMG has been mostly engaged in networked relationship and they prefer and show a keen desire towards selecting candidates from well-connected network ties. This group exhibited their autonomous power over the lower layers of management. MLMG has been less empowered to influence selection practices in order to facilitate TM because their power has been uniquely marginalized by ULMG.
However, socialization within network ties is significant to build relationships through trust, norms of culture, obligations, and identification in determining effective TM. At ULMG, strong networking relationships with shared codes and language, vision, and narratives among members have significant influence over TM processes. Our analyses showed that intra-organizational social capital stemmed from social networking relationships and ties significantly influences managers to adopt social capital-based selection process to promote talent management. It is evident that top managers in Bangladesh telecom MNCs greatly rely on social networking relationships and ties while selecting talents.
Based on these findings, we suggest that organizations should prioritize social-capital dimensions that ensure "person-organization fit" type of selection practices as part of TM. The inclusion of social capital dimensions within an existing HRM system would significantly improve the working efficiency and competitiveness of organizations especially for service innovation firms. In addition, service innovation firms such as telecom organizations should focus on developing further diversified and heterogeneous network groups in relation to the top management, because higher diversification and group heterogeneity can improve personorganization fit (Van Maanen, 1977). Additionally, creativity and innovation especially for service-oriented firms require social support. The management of such service innovation firms will benefit from "social orientation program" such as social events, social collaborations, information exchanges workshops, networks building and extension gathering, community of practices, etc. The purpose of these is to encourage the building of social networks and strengthening of network ties.
Besides, it would be of practical benefits for individual employees to find ways to socialize with other managers. It would be useful for management practitioners to espouse the importance of informal relationship towards social capital-based selection practices. Social capitalbased selection practices suggest novel perspectives on integrating such TM mechanism within HRM system effectively. For this endeavor, managers should consider their social network relationships as well.
This research has few limitations. First, data were collected through a small sample of specific MNCs from the telecom industry of Bangladesh. Findings of cross-case analysis are limited to represent only one single industry rather than other sectors. Second, further replication studies are necessary in other firm populations and in other organizational cultures where cultural disparities of HRM system shapes selection practices (Aycan, 2005). Third, industry-specific focus has been accomplished in this study but holistic effects through panel data that ensure rigorous evidence in order to constitute a firmer basis for this type of exploratory research is needed as future direction for research. Fourth, this study could benefit by including experts-driven focus group discussion using the Delphi technique. The Delphi technique could facilitate in-depth exploration of the research phenomenon from the expert perspective.