Competing institutional logics in talent management: talent identification at the HQ and a subsidiary

Abstract

Talent Management (TM) is a hot topic among both practitioners and scholars, but it still has to overcome some important limitations. Studies have been overly unitarist and managerialist in their orientation, and we still know little about how local contextual factors relate to TM, especially with regards to one of the most critical aspects of any TM system, i.e. talent identification. This research, which adopted a qualitative case study including data from interviews, observations and documents, studied how talent identification unfolded in practice at both the headquarters (HQ) and a subsidiary of a large Swedish organization. By drawing on the institutional logics perspective, we suggest that the way in which organizational actors conduct their talent identification is grounded in the logic they enact and make use of. Attention is thus focused on how the cultural norms, symbols and practices of different institutional orders are incorporated into the identification of talent. Having identified competing institutional logics at the HQ and the subsidiary, we also suggest that this is a credible explanation for the discrepancy between intended and actual HR practices. The findings are in contrast with previous research, which suggests that self-interest, ad hoc approaches, and a lack of skills nested in talent identification are underlying causes of differences in how talent identification is conducted.

Keywords:

• Talent management
• talent identification
• human resource management
• institutional logics
• competing institutional logics
• critical HRM

Introduction

Talent Management (TM) has been growing in significance and gaining interest among practitioners as well as in the scholarly debate (Collings & Mellahi, 2009; McDonnell, 2011), especially so over the past 10 years (Thunnissen, Boselie, & Fruytier, 2013). We have even witnessed an explosion of TM scholarship over the past six years (McDonnell, Collings, Mellahi, & Schuler, 2017).

Despite this explosion, scholarly debate is still limited in three important ways. First, although arguments about the contextual relevance of TM have been made (e.g. Collings, Mellahi, & Cascio, 2018; Thunnissen et al., 2013), and how research has been conducted in various countries, and has thus considered the national context, there is often a lack of contextual sensitivity, local contextual factors in particular (Dries, 2013; Iles, Chuai, & Preece, 2010; Sidani & Al Ariss, 2014). We thus know little about how organizational configurations and institutional contexts impact TM (Al Ariss, Cascio, & Paauwe, 2014; Gallardo-Gallardo & Thunnissen, 2016; Van den Brink, Fruytier, & Thunnissen, 2013). This is a major limitation because organizations, large ones in particular, are often considered multicultural and varied in their institutional contexts (e.g. Gregory, 1983; Phillips & Tracey, 2009), with local internal contextual factors having recently been proposed as equally important to national dissimilarities in explaining how contextual influences affect organizations that adopt HRM practices (see Brewster, Mayrhofer, & Smale, 2016; Delbridge, Hauptmeier, & Sengupta, 2011; Thunnissen & Buttiens, 2017; Tung & Baumann, 2009; van Hoorn, 2018). More empirical research that considers the internal context, and in doing so incorporating an awareness of local contextual factors, is thus needed.

Second, and closely connected, most TM studies have been overly unitarist and managerialist in their orientation, focusing on HR managers and/or top management teams (Collings, 2014; McDonnell et al., 2017; Thunnissen et al., 2013). This limits our understanding of how TM unfolds in practice (Garavan, 2012; Thunnissen, 2016) because organizations are seldom unified actors, instead consisting of multiple internal actors with frequently divergent interests and motives acting in accordance with particular rules, logics and norms (Delbridge, 2010; Delbridge & Keenoy, 2010; Thompson, 2011). To capture the complexity of different perceptions of TM policy and practice, we need to adopt a pluralistic approach by including, besides HR managers and top management, multiple internal actors and line managers in particular (Delbridge et al., 2011; Gallardo-Gallardo & Thunnissen, 2016; Paauwe, 2009; Thunnissen et al., 2013).

Third, and connected to the previous limitations, talent identification, being one of the most critical aspects of any TM system (Boudreau & Ramstad, 2005; McDonnell, Hickey, & Gunnigle, 2011; Mellahi & Collings, 2010), is seldom featured in the current TM literature (Gallardo-Gallardo, Dries, & González-Cruz, 2013; Gallardo-Gallardo & Thunnissen, 2016; McDonnell et al., 2017). The few existing studies of talent identification (e.g. Jones, Whitaker, Seet, & Parkin, 2012; Mäkelä, Björkman, & Ehrnrooth, 2010; McDonnell et al., 2011; Silzer & Church, 2010; Wiblen, Dery, & Grant, 2012) often support arguments claiming that ad hoc, biasing effects, self-interest, a lack of skills and training are all common and serve as explaining factors as regards how talent identification work unfolds (Dries, 2013; Gallardo-Gallardo et al., 2013; Jones et al., 2012; McDonnell et al., 2011; Swailes, 2013). However, and connected to the above-mentioned limitations, these studies often rely on formal descriptions, thus being too simplistic to describe how talent identification works in practice (Collings et al., 2018; Delbridge & Keenoy, 2010; Van den Brink et al., 2013; Watson, 2004).

To truly understand how TM unfolds in practice, we thus need to explicitly consider the local contextual factors, and how the multiple internal actors are shaped by the existing rules, logics and norms. In this regard, the institutional logics perspective (Friedland & Alford, 1991; Thornton, Ocasio, & Lounsbury, 2012) is a promising framework, as it emphasizes the local contextual embeddedness of social and organizational phenomena. Attention is focused on how multiple actors incorporate the cultural norms, symbols and practices of different institutional orders into their thoughts, beliefs and decision-making, as in the case of identifying talent. In this article, therefore, we draw on the institutional logics perspective and thus start to address the limitations mentioned in the TM literature. Based on a case study of a specific talent management practice, i.e. talent identification, at both the headquarters (HQ) and a subsidiary of a Swedish medical technology corporation (Medico), we ask: How is talent identification shaped by institutional logics?

The findings illustrate how organizational actors’ ways of conducting talent identification are grounded in the logic they enact and make use of. Multiple and sometimes competing institutional logics were available at the studied organization, we suggest that this is a credible explanation for the discrepancy between intended and actual HR practices.

We make several important contributions. First, we contribute to the TM literature by responding to the call for in-depth studies (from the perspective of multiple actors) of how talent identification unfolds in practice. By showing the explanatory power of the institutional logics perspective, as regards understanding the motives of various organizational actors when engaging in talent identification, we thus advance theoretical understanding of TM by explaining how local organizational configurations and institutional contexts impact work with, as well as the nature of, TM. By adopting a more critical and contextually-based study, we provide credible explanations as to how and why actors engage with TM practices the way they do.

Second, we contribute to the HRM literature, especially the emerging stream of critical HRM (e.g. Delbridge & Keenoy, 2010; Delbridge et al., 2011; Keegan & Boselie, 2006), and its call for more contextualized and non-managerialist research that gives voice to multiple organizational actors. While much of the contemporary HRM literature has focused on cross-national variations as its main contextual factor (see Almond, 2011; Bévort & Poulfelt, 2015), we highlight the importance of local organizational contexts in our aim of understanding the contextual embeddedness of HR practices. We also emphasize social relationships, with cultural values as root causes of how and why actors engage with HR practices as they do, in contrast to the rational explanation, as adopted in much of mainstream HRM (Delbridge et al., 2011; Evans & Tourish, 2017).

The article is structured as follows: First, a theoretical background is presented emphasizing the TM literature and arguments existing within the critical HRM literature; second, the institutional logics perspective is introduced. Thereafter, the research setting and method are described, followed by the findings. The discussion is then presented, followed by the conclusion accompanied by suggestions for future research.

Theoretical background

The increasing interest in TM is often explained using two underlying ideas; i.e. TM is a source of competitive advantage and thus key to competitive success (McDonnell et al., 2017), and further globalization, coupled with an ageing workforce and falling birth rates, has boosted the war for talent even further (see Michaels, Handfield-Jones, & Axelrod, 2001). TM is often defined as ‘activities and processes that involve the systematic identification of key positions which differentially contribute to the organization’s sustainable competitive advantage, the development of a talent pool of high-potential and high-performing incumbents to fill these roles, and the development of a differentiated human resource architecture to facilitate filling these positions with competent incumbents and to ensure their continued commitment to the organization’ (Collings & Mellahi, 2009, p. 304). This definition of TM, using talent as a high-potential and a high-performer, is more in line with the exclusive approach to TM, whereby some people are seen as more talented than others (Gallardo-Gallardo et al., 2013). This exclusive approach has been criticized, however, for promoting inequality among employees (see e.g. Swailes, 2013), thus an inclusive approach whereby all employees are seen as possessing certain strengths, and are consequently also seen as potential talents (Thunnissen & Van Arensbergen, 2015), has therefore been suggested. Thus, a key debate in the TM literature is whether TM is, or should be, an inclusive or an exclusive approach (Thunnissen, 2016).

Despite the advances concerning a number of debates in the TM literature, there are, as mentioned in the introduction, a number of important limitations. In this article, we build on certain arguments put forward in the critical HRM literature, as well as in the previous TM literature, in order to address these limitations. First, we build on the arguments that acknowledge differences between the rhetoric of formal HR policies and the reality of what unfolds in practice (e.g. Grant, 1999; Khilji & Wang, 2006; Legge, 1995; Nishii & Wright, 2008). In the contemporary TM literature, however, little conceptual and empirical attention is paid to the differences between intended and actual TM practices, and to the underlying causes (see, for example, Collings et al., 2018; Meyers, van Woerkom, & Dries, 2013; Skuza, Scullion, & McDonnell, 2013; Sonnenberg, van Zijderveld, & Brinks, 2014; Thunnissen, 2016; for exceptions). In a recent conceptual paper, for example, Collings et al. (2018) developed a theoretical framework describing how global talent management (GTM) links to organizational performance. The authors stressed the importance of considering the factors determining what is actually implemented at the subsidiary level, rather than HQs intended practices. Among the limited empirical evidence, Thunnissen (2016), for example, has shown how, regardless of formal protocols, a discrepancy between intended and actual TM practices was prevalent, as well as how factors at the institutional, organizational and individual levels were causing this variance. A similar discrepancy was found by Van den Brink et al. (2013) in their study of recruitment and selection practices regarding junior and senior academic talent in the Netherlands. Despite the protocols formulated, and the rules set out, by HR managers, providing steps and guidelines for decision-makers, these protocols were implemented differently.

Thus, taking seriously the argument that intended and actual HRM practices can diverge means that we need to move beyond much of mainstream thinking in HRM (Paauwe, 2009). Instead of a managerialist orientation that relies on formal descriptions, often from the HR perspective, taking a more critical perspective (e.g. Delbridge & Keenoy, 2010; Delbridge et al., 2011) entails TM researchers focusing on how TM is adapted to fit local contexts and to include the organizational, institutional and cultural factors that play an important role in this (Alvesson, 2009). We thus need to reconsider the notion of context, and to extend it beyond how it has been used in the contemporary HRM literature, especially the transfer literature regarding MNCs (e.g. Ahlvik & Björkman, 2015; Ahlvik, Smale, & Sumelius, 2016), which has focused on comparing institutional arrangements between countries and has thus considered cross-national variations to be the main contextual factor (see Almond, 2011; Bévort & Poulfelt, 2015). Thus, critical HRM is helpful in redirecting the focus towards everyday practice (Watson, 2004) and the local organizational context in which the negotiated nature of HR practices unfolds (Delbridge et al., 2011; Jenkins & Delbridge, 2013; Thompson, 2011).

Second, and closely connected, since the intended HR practices are often implemented by actors other than the designers or decision-makers themselves, and since those practices that are implemented often unfold differently to the initial intention (Khilji & Wang, 2006; Nishii & Wright, 2008), we logically build on arguments highlighting the fact that, frequently, multiple actors with contrasting goals are present in organizations (Boselie, Brewster, & Paauwe, 2009; Delbridge & Keenoy, 2010). Despite the way in which a more pluralistic TM approach has been suggested (e.g. Collings, 2014; Collings et al., 2018; Gallardo-Gallardo & Thunnissen, 2016; Sonnenberg et al., 2014; Thunnissen et al., 2013), sensitivity to the potentially conflicting interests of the different internal actors, at different levels, is largely neglected in empirical TM research (Huang & Tansley, 2012; Iles et al., 2010). One exception here is Wiblen (2016), who showed how organizational actors drew upon different understandings when identifying talents. Two different talent identification discourses were therefore present at the studied organization.

When following these arguments, an initial starting-point is to move beyond the unitarist approach, whereby the organization is treated as a unified actor with aligned goals and interests, as is dominant in the mainstream HRM research (Brewster et al., 2016; Delbridge et al., 2011; Paauwe, 2009; Watson, 2004) and much of the TM literature (Collings, 2014; Gallardo-Gallardo & Thunnissen, 2016; Sonnenberg et al., 2014; Thunnissen et al., 2013). However, if a pluralistic model is to be taken seriously, TM researchers will not only need to recognize multiple actors, and how they potentially have different perspectives on TM but also to acknowledge the possibility of potentially legitimate conflicts of interest arising (Delbridge & Keenoy, 2010). Here, the critical HRM literature is helpful as it questions the legitimacy of managerial interests, and expresses concerns that a pluralistic model would still risk paying attention to powerful actors when defining whose claims are legitimate and urgent (Alvesson, 2009; Janssens & Steyaert, 2009). The main concern is, thus, that prioritization will continue to be ascribed to a dominant, economic understanding that privileges more powerful actors (the principal) with the aim of getting other self-interested actors (the agent), acting aligned with the former’s intention. This dominance of unitarist and economic logic exaggerates the instrumentality and predictability of rational action on the part of the actors (Bowen & Ostroff, 2004; Delbridge & Keenoy, 2010; Keegan & Boselie, 2006) and underestimates the local institutional environment and the social relationships, with cultural values and behavioral norms, in which organizational actors are embedded (Delbridge et al., 2011). Thus, instead of assuming that people are motivated by individual self-interest, mainly in the form of economic calculations, and focusing on how to minimize shirking behavior on the part of an agent, as stressed in previous studies of talent identification (e.g. Jones et al., 2012; Mäkelä et al., 2010; McDonnell et al., 2011; Silzer & Church, 2010) and in the mainstream HRM literature, using theories based on economics, e.g. agency theory and rational choice theory (Eisenhardt, 1989; Jensen & Meckling, 1976); if we aim to develop TM, we need to examine more deeply, in practice, the root causes to how and why actors engage differently with TM practices (see Evans & Tourish, 2017; Watson, 2004).

An institutional logics perspective (Friedland & Alford, 1991; Thornton et al., 2012) is therefore appropriate to our study as it emphasizes the local contextual embeddedness of social and organizational phenomena, in a non-determinist and non-functionalist way. Attention is focused on how multiple actors incorporate the cultural norms, symbols and practices of different institutional orders into their everyday activities regarding thoughts, beliefs and decision-making, e.g. when identifying talents. As organizational actors may incorporate competing belief systems which shape behaviors and practices (Pache & Santos, 2013), this is a particularly well-suited perspective when it comes to supplementing the more traditional unitarist approaches, when explaining the role of agency, how and why actors engage with TM as they do, and thus also when it comes to advancing our understanding of why intended purposes are differently instantiated and enacted.

Institutional logics

Introduced by Friedland and Alford (1991), the institutional logics perspective is a key tenet of institutional theory (Goodrick & Reay, 2011) and has become a widely used analytical tool for analyzing institutions at various levels (Thornton et al., 2012). Thus, considering that it is one of the most important theoretical perspectives in management and organizational theory (Martin, Siebert, & Robson, 2016; Reay & Jones, 2016), the scant reference to it by HR scholars (for exceptions, see, for example, Almandoz, 2014; Alvehus, 2018; Bévort & Poulfelt, 2015; Martin et al., 2016), and TM scholars (Thunnissen & Buttiens, 2017), is surprising.

Defined as socially-constructed sets of material practices, assumptions, values and beliefs that shape cognition and behavior (Thornton et al., 2012), institutional logics guide us in how to act in particular situations. Early proponents highlighted institutional logics at the societal level, proposing certain ideal types of logics, e.g. the market, the state, the family, religion and democracy (for an overview, see Thornton et al., 2012). Later proponents argued that logics also exist at the industry and organizational field levels (Goodrick & Reay, 2011). These field level logics are historical contingencies that contribute to variations in practices (Thornton et al., 2012), triggering organizational change (Greenwood & Suddaby, 2006). Research into field-level logics has also emphasized the replacement of logics, whereby some features of a current logic are exchanged for one or more other logics (Thornton & Ocasio, 1999), the mechanism of blending, whereby logics are transformed by combining the dimensions of diverse logics, and the segregation of logics, whereby different logics emerge from a previously-shared common origin (see Thornton et al., 2012).

While acknowledging these insights, previous research streams regarding institutional logics have been criticized for neglecting the micro-foundations of institutional logics (Thornton et al., 2012), thus ignoring how logics are used on the ground (McPherson & Sauder, 2013). Following this criticism, there has been increasing interest in understanding how institutional logics unfold in practice, and in the locally-constructed manifestations (Besharov & Smith, 2014). It is specifically this line of research we draw on in this article. The focus is on how actors engage with logics during their day-to-day work, since this is the level where ‘overarching sets of meaning and normative criteria become encoded in “local” logics that are manifested in rituals, practices and day-to-day behavior’ (Greenwood, Raynard, Kodeih, Micelotta, & Lounsbury, 2011, p. 334). Institutional logics are seen as emergent and derived from empirical study, rather than given or decided upon a priori (Currie & Spyridonidis, 2016).

From studies of everyday organizational activities, the seemingly abstract concept of institutional logics does, in fact, determine everyday work, as these shape the rules of the game (Dunn & Jones, 2010). Institutional logics act as frames of reference that condition actors’ decision-making and their motivation to act (Thornton et al., 2012); thus, the bounded intentions of the actors are circumscribed by their institutional embeddedness (Martin et al., 2016). Institutional logics constrain the interests, values and strategies of actors (Greenwood & Suddaby, 2006), thus shaping both individual and organizational practices because they represent sets of expectations regarding social behavior (Goodrick & Reay, 2011). It has also been shown that multiple logics are actualized or manifest in practice, a process often associated with contestation and conflict, coexistence and logic blending (McPherson & Sauder, 2013). For organizations, the process of dealing with multiple logics is complicated, and the general view is, even though other logics exist, it is the dominant logic that guides action (Lounsbury, 2008).

Logics may also be segmented, with different groups operating according to different logics (Reay & Hinings, 2009). Actors belonging to an organizational group will often closely adhere to that group’s primary logic (McPherson & Sauder, 2013; Pache & Santos, 2013). Actors thus represent, and import into an organization, the meanings and norms of the logics to which they have primarily been exposed (Greenwood et al., 2011). The more varied the composition of the site, in terms of professional and institutional domains, the more likely it is that logics will be used during micro-level interactions in order to achieve local organizational goals (McPherson & Sauder, 2013).

Individuals are often aware of the differences in the cultural norms, symbols and practices of the different institutional orders, thus encountering institutional alternatives (Besharov & Smith, 2014). This familiarity with the different logics may dissolve differences and contribute to mutual adjustments (Pache & Santos, 2013). A mutual adjustment between multiple logics is argued to only be possible when the logics used by powerful actors are not threatened. This is because when powerful actors’ interests are threatened, such mutual adjustment dissolves, and only limited agency is granted to less powerful actors by more powerful ones (Currie & Spyridonidis, 2016). This was shown among layers, where their practices were questioned, and became defensive in maintaining their own practices, then starting to justify their own logic by legitimizing their own practices by means of disrupting the alternative logic (Smets & Jarzabkowski, 2013). Actors use the reinforcing properties of co-mingling logics to legitimize their own practices (Swan, Bresnen, Robertson, Newell, & Dopson, 2010). Powerful actors may also exercise compliance pressure in order to maintain their dominant logic (Pache & Santos, 2010).

Methodology

The aim of this study was to describe and analyze how talent identification is shaped by institutional logics. Since we wanted to provide a deeper understanding of this phenomenon on the ground, in practice, a case study methodology was adopted (Flyvbjerg, 2006). This method was additionally appropriate since the field of TM, and especially the practice of talent identification, is a novel research area wherein few theoretical frameworks have been defined, allowing themselves for detailization and careful empirical texting (Eisenhardt & Graebner, 2007).

The case company (Medico) was selected as a research site because we suspected that it was of special interest to our research question. Due to the many acquisitions made by Medico, we suspected that different types of local cultures, including norms and values for example, would meet up. We expected that the HQ, due to having activities mainly within Finance, Human Resources, Corporate Communications and Brand Management, would adhere to some kind of corporate culture. In contrast, we expected that the subsidiary would adhere to some kind of engineering or production culture, since it had mainly been employing engineers and had been involved in Production, Supply-Chain and R&D activities for a long time. Medico was also selected because the TM system had been introduced and launched quite recently, and we expected that TM activities would not be routinized but would consist of negotiations and controversies, necessitating the actors to articulate the reasons for their behavior.

It is important to mention that we did not use ‘pattern matching’ in order to compare the actual data with ideal types of logics (see Reay & Jones, 2016). Instead, our aim was to capture the nuances of localized practices by analyzing qualitative data using a bottom-up inductive approach. This entailed patterns associated with the logics emerging inductively from the data, and then being used to understand how sets of behaviors or practices reflect the influence of the specific guiding logics. To accomplish this, we grounded our insights and abstractions in the context using quotes, observations and thick descriptions.

Data collection

Data collection (see Table 1 for a summary) took place during 2015 both at the headquarters (HQ) and at a subsidiary of a large Swedish organization (referred to as Medico) operating in the medical technology industry. We collected data via 23 in-depth interviews with HR managers, HR specialists, line managers and non-managerial employees. The HR managers and HR specialists were interviewed in order to gain a top-down perspective and to understand the intended talent identification practice. The line managers were interviewed because they are often the ones responsible for the initial part (i.e. evaluating performance and constructing local talent pools) of talent identification. It is, therefore, a major limitation to neglect the role of the line manager, as has been the case in many previous studies of talent identification since he/she is needed in order to advance our understanding of how things actually unfold in practice.

Table 1. Summary of data collection.

Interviews	Observations	Documents
HQ: HR managers (#6) Subsidiary: HR specialists (#4) Line managers (#8) Employees (#5)	Informal discussions/meetings. Product demonstration/presentation. Eating lunch with the employees.	TM-guide. Core-values documents. Potential identification guides. Performance appraisal forms. Yearly manager/employee discussion forms.

HR personnel and line managers were selected on the basis of their level of involvement in TM activities. Our goal was to gain access to personnel working with TM on a daily basis. This was achieved by discussing potential interviewees with HR managers responsible for TM at both the HQ and the subsidiary. To further widen our understanding and to adopt a more pluralistic perspective (Thunnissen et al., 2013), we also interviewed non-managerial employees to whom TM activities are directed (Huang & Tansley, 2012). Medico distinguished between talents and non-talents, with all the employees (apart from one) whom we interviewed being identified as talents and selected using the criterion that all three functional areas (production, supply-chain and R&D) were represented. All the interviews were semi-structured, lasted between 60 and 90 minutes, and included questions about the interviewees’ work in general, their perspectives on what a talent was, and their experiences, interpretations and understandings regarding talent identification practice. They were audio-recorded, and the recordings were transcribed verbatim.

We also conducted observations (Czarniawska, 2004), of informal meetings and discussions between employees of Medico. We spent half a day on four occasions at the HQ and the subsidiary, thus enabling us to observe relationships in a natural setting. These informal meetings and discussions did not have a formal agenda, instead consisting of small talk, about how the participant viewed the talent identification process in general, as well as some of its consequences. At the subsidiary, we were also shown and had demonstrated to us how the products were designed and how they functioned. These observations were helpful as regards gaining a sense of the culture at the HQ and the subsidiary. On these occasions, we took notes and wrote field stories based on these notes. We refer directly to some of our observations in the Findings section (with reference to field notes); however, the observations also broadly confirmed the impressions gained during the interviews.

In addition, we also collected TM-related documents and review matrixes involved in the practical work of identifying talent, e.g. performance and potential identification guidelines, as well as material for yearly manager-employee interviews. These documents provided us with important background information and supplemented the interviews and observations.

Data analysis

A constant comparative analysis (Glaser & Strauss, 1967; Strauss & Corbin, 1998) was used to inductively analyze the field material. We initially engaged in open coding of interview data (Strauss & Corbin, 1998), which yielded multiple first-order categories. These categories faithfully adhered to interviewees’ terminology, and focused on their interpretations of the talent concept, their experiences and views of the talent identification process, and their perceived differences between the HQ and the subsidiary. Observations and documents were then coded and compared to the interviews. In this way, the interviewees’ accounts were triangulated (Silverman, 2006) by means of cross-checking with documents and observations. At this stage, our analytical strategy shifted to axial coding (Strauss & Corbin, 1998), where we compared the first-order categories with each other in order to search for potential connections. We subsequently reduced the first-order categories to second-order themes by clustering them in relation to their focus. The commercial and sales functions, for example, characterized a leadership/management idea. In contrast, the importance of the technical aspects and a deep knowledge of the products characterized an engineering idea. When analyzing the identified themes, and the descriptions of how talent identification was conducted, it became evident that the HQ and the subsidiary were both highly involved in identifying talents but differed in their understanding of how to do this. This constant comparing of material allowed us to discover relevant theoretical concepts that were useful to address in our study. After it became evident that the values, norms and attitudes seemed important as regards how to view talent, and the work of talent identification, we identified concepts from institutional theory, especially the institutional logics perspective, as being useful. Following our more frequent iteration between field material and theory, we thus distilled the second-order themes into two aggregate dimensions (Gioia, Corley, & Hamilton, 2013) which we labeled the business logic and the engineering logic. These two logics were inductively derived, underpinning our analysis of talent identification at Medico. Figure 1 provides an overview of the data structure (Gioia et al., 2013). Our Findings section is the result of moving to a higher level of abstraction and is presented in line with the relevant analytical themes mentioned.

Figure 1. Illustrating the analysis process and moving from a lower level of abstraction to more aggregate theoretical dimensions.

Medico and its TM system

Medico is a Swedish organization in medical equipment and systems which is knowledge-intensive, and world-leading. Its offerings contribute to innovation and technological advancement in the field of healthcare, the life sciences and care of the elderly. Medico employs over 15,000 people in more than 40 countries, turning over approximately EUR 3.1 billion in 2016. Its HQ is a modern office building in Sweden’s second city. It employs mainly executives with functions in Finance, Human Resources, Corporate Communications and Brand Management. The subsidiary is located in Sweden’s first city and was acquired by Medico in 2000. It employs mainly engineers and has been developing medical technology products, e.g. operating tables and ventilators, for more than 150 years. The subsidiary takes pride in its engineers, and their technical competence, perceiving them to be costly to lose and replace. Its activities lie mainly in production, supply-chain and R&D.

During recent years, Medico has been experiencing stagnant growth, with most high-level managers and business unit CEOs having been recruited externally, indicating that Medico did not have the ability to identify and develop talents internally. An important part of Medico’s revival, therefore, was developing a TM system to ensure that talents could be identified, thus viewing talent as a strategic resource and a source of competitive advantage. External recruiting was also considered time-consuming and costly compared with internal recruitment.

Launched in 2013 by HQ, the new global TM system consisted of a structured and standardized process for identifying and developing talents. The principles of the system were documented in a 104-pages TM guide, with the goal being all subsidiaries carrying out TM along similar lines. While the system was being developed by HR at HQ, respondents at HQ were generally commenting that subsidiaries were being allowed some degree of input into the system. This was mainly visible in the way the performance ratings were being ‘owned locally’ by line managers (as one HR manager at HQ put it). A cornerstone of the TM system was the Performance Development Dialogue 3 × 3 matrix (PDD), in which the line manager rated employee performance using two dimensions (‘what’ and ‘how’) on three-point scales (see Figure 2). The ‘what’ dimension captured the fulfillment of agreed-upon objectives while the ‘how’ dimension captured adherence to Medico’s cornerstone behaviors and core values when fulfilling objectives. The scale consisted of the grades: needs improvement (NI), meets expectations (ME) and exceeds expectations (EE). An NI grade entails employees having to improve and line managers having to draw up an improvement plan. An EE grade entails line managers having to challenge the employee with more difficult projects and having to draw up a development plan. An ME grade does not require line managers to take any specific actions. Based on the two dimensions, and the three-point scale, employees were then placed on a nine-box grid, where the top right corner was the talent box. As the identification of talents was based on the employees’ locations on the nine-box grid, the TM approach was thus exclusive in which only a selected few were considered talents.

Figure 2. The performance development dialogue (PDD) matrix.

Once the line manager has completed the PDD, the rating is then calibrated jointly with a fellow manager (at a ‘buddy’ meeting) and with a senior manager (at a ‘grandfather’ meeting). These meetings are characterized by the ‘buddy’ and the ‘grandfather’ asking questions and challenging the line manager’s evaluations and ratings. The PDDs are then sent to HR at the subsidiary, which, together with senior managers, evaluates them and constructs new talent pools to be used during career development discussions. The importance of the PDDs is enhanced by their use in salary negotiations and career development interviews at the subsidiary. The benefits of being identified as talent at the subsidiary were, for example, increased rewards and salaries, and opportunities to attend development courses. These courses could be related to management for non-managers or to expert courses in order to deepen engineering expertise. In some cases, it is also possible to take advantage of internal job rotations, advancing either vertically into more senior roles or horizontally into other areas. Although career opportunities are often limited to the subsidiary, it happens that talents also benefit from job rotations abroad to other subsidiaries.

In addition to filling out PDDs, HR at HQ also requires line managers at the subsidiary to compile a high-potential-talent list every year. In this list, there are additional criteria, apart from performance, which line managers need to consider, e.g. mobility (the ability to relocate), potential (the ability to advance and not just to perform at the current level) and leadership and commercial abilities (including strategic focus, presentation skills and driving innovation). Ratings for these additional dimensions are intended to guide the identification of talents for higher-level management positions, as well as positions beyond the subsidiary. The talent identification process at Medico is summarized in Table 2.

Table 2. The talent identification process at Medico.

Steps	Timing of step	Location	Main actors	What is done
Step 1: Performance evaluation	Once a year (end of March)	Subsidiary	Line manager Colleague line manager Senior manager	Rates employees based on the PDDs. Challenges the rating in a ‘buddy meeting’. Challenges the rating in a ‘grandfather meeting’.
Step 2: Local talent reviews and rewards	Once a year (end of May)	HR division at subsidiary	Senior manager and HR manager	Assess current talent pools and based on the PDDs, adjust according to newly identified talents. Plan compensation, salaries, pay by performance and development courses.
Step 3: Initiating high-potential talent pools	Once a year (end of July)	Subsidiary and HR division at subsidiary	Line manager HR manager	Rates employees according to additional HQ criteria (as in high-potential talent list). Compiles PDDs and high-potential talent lists; sends to HQ HR.
Step 4: Constructing high-potential talent pools	Once a year (end of Sept./Oct.)	HQ	Executive vice president and HQ HR	Evaluates and constructs high-potential talent pools; used for career development.

Findings

To understand how talent identification is shaped by institutional logics, we start by identifying the logics available during everyday interactions, by which the organizational actors are guided in their work (McPherson & Sauder, 2013). We found two distinct institutional logics invoked by talent identification at Medico, i.e. the business logic (prevalent mainly at the HQ) and the engineering logic (dominant at the subsidiary). Table 3 summarizes the keywords and representative quotes pertaining to the two logics. We end the Findings section by describing how the two logics not only were different but also, to some extent, competing.

Table 3. Illustrates keywords and representative quotes of the business and engineering logic.

Logics and keywords	Representative quotes
Business logic: Leadership, sales, commercial, strategic, mobility, broad	‘It is not enough to have worked in one niche-you need to have worked within several functions, often sales or other commercial functions, and unfortunately, yes, you must have succeeded in these jobs’ (HR Manager, HQ). ‘We ultimately have a view of what talent and potential look like… here’s an overweight of people working in commercial functions, so sales, country manager’ (HR Manager, HQ). ‘Mobility is an important criterion.(..)it is possible to develop if staying, but then there are limited opportunities to develop’ (HR Manager, HQ). ‘We are starting to introduce willingness to relocate as one of the critical elements in our talent assessment’ (HR Manager, HQ).
Engineering logic: Engineers, Expert, technical, deep knowledge, products	‘The engineers have a strong position here.’ (Line Manager, subsidiary). ‘We are good at technical aspects.’ (Line Manager, subsidiary). ‘We have important superstars that are experts with deep engineering knowledge about specific areas that also are our core practice. This aspect is always coming up when talking about and working with talent identification.’ (HR Specialist, subsidiary). ‘We have special products here, with special areas in which expert knowledge about developing products become especially important.’ (Employee, subsidiary). ‘…we constantly need to develop new products as we are in the healthcare industry. Our technical experts can therefore not be neglected because they are so valuable for us.’ (Line Manager, subsidiary).

The business logic

During both interviews and observations, actors at the HQ frequently discussed the pressures they had faced delivering on sales targets, due to increasing competition. Even though new and better products had to be engineered, developing the best products was still not enough; they also had to be sold:

We worked a lot with product development, but then we said we have too many engineers and we’re making too many road-maps. We need people with a bigger customer- and market perspective… We didn’t want to have so many product specialists only talking about product qualities, we wanted people talking about growth, how much, what margins we have on our products, and how much we can develop our different markets. What is it the customer wants and needs? These values impact how we work with TM (HR Manager, HQ).

A business culture was thus invoked, which also seemed to affect the work with talent identification. For example, in their talent list, they valued experience of the sales and leadership fields. The list was described thus:

…containing a majority, even a preponderance of people working in commercial functions…These people were, for example, sales or country managers. (HR Manager, HQ)

The criteria for assessing talents mainly consist of eight leadership behaviors, and these consist of the relevant business and leadership capabilities (HR Manager, HQ)

When discussing talent at the HQ, there was almost an absence of references to engineering. Managers instead based the how dimension on the leadership competencies that the HQ was looking for in a talent. This was done by evaluating employees as regards their ‘leadership, business skills, and project management.’ (HR Manager, HQ).

When invoking this business logic, actors at the HQ were thus valuing certain aspects of their view of a talent. From an HQ perspective, there thus seemed to be a clear definition of what a talent was, and also of how talent identification ought to be conducted. This notion was strengthened by the way the HQ had formulated its TM guide, where leadership was mentioned over 100 times. This can be compared to engineer/engineers not being mentioned at all, experts being mentioned six times and the word technical being mentioned 11 times. Being a knowledge intensive organization, relying on engineers to develop products, it was surprising to sense how the TM guide was largely neglecting the engineers.

The guide also said, regarding instructions when performing talent assessment, that managers should ‘utilize the eight leadership competencies as part of the assessment’. Thus, HQ managers emphasized a business logic in which ‘talents are very much about sales behavior, sales performance and how to drive that forward’ (field notes). This was further supported by comments regarding what a talent was:

Who are we identifying as talents? We’ve made it clear that having a business-oriented mind and leadership competencies is the most important criterion…Taking responsibility, cooperating with others, and communicating – that’s talent. (HR Manager_, HQ)

A talent needs to be focused on monetary performance…you can also put them anywhere in a domain outside of their own expertise and they’ll succeed…(HR Manager, HQ).

As seen from the above quote, and from informal discussions, the business logic also focused attention on the importance of moving across niches and trying out ‘uncomfortable areas’ (field notes). Lacking mobility could have a detrimental effect on being identified as a talent, while being mobile could move an employee into the talent pool. Illustrating the importance of this mobility, two HR managers explained things thus:

Mobility is an important criterion. We look for people who either have broad experience or who can gain that experience, and that requires mobility. And we’ve had many people who were unwilling to move and who then were not identified as talents. (HR Manager, HQ).

In France, we have an extremely strong sales manager who is really great but only speaks French. He’ll never be a top talent because we can’t move him. (HR Manager, HQ).

As well as emphasizing a clear definition of what a talent was, and how to conduct talent identification, when enacting the business logic, HQ actors also legitimatized themselves by highlighting the limitations of the way managers at the subsidiary identified talents:

A manager might identify a co-worker as hands-on, a doer, - and then they [the line managers] know things will happen. And that is a great attribute, but then you lose the strategic capability or to think in advance and to have a plan, and things become very operational in the here and now. (HR Manager, HQ).

As this was communicated to the subsidiary, it seemed important for the HQ not only to put forward its ideas but also for the subsidiary to recognize and understand this. How the subsidiary responded, and what it valued in its context, is further described in the following empirical sections.

The engineering logic

In contrast to the HQ, actors at the subsidiary frequently discussed the importance of their engineers in terms of ‘the ones doing the real work, developing products’ (field notes). The work of talent identification seemed to be driven by this engineering culture, described as follows:

We’re a technically-oriented and knowledge-intensive company. The engineers have a strong position here. (Line Manager, subsidiary).

When invoking this engineering logic, attention was focused on what mattered at the subsidiary, which also effected talent identification:

I identify talents who can develop and become very good experts, or material specialists. (Line Manager, subsidiary)

The assumptions, values and beliefs present in the engineering logic thus shaped the cognition and the behaviors of the actors at the subsidiary:

Researcher: Has your view of talent and your approach to identifying talent changed since you entered this [Medico’s] culture?

Line manager: Yes, I’d say so, unfortunately I’d add. I’d rather emphasize more leadership and a broader ability in terms of the areas you’re not an expert in. Then you can be more of a leader and a coach, and work with a greater number of aspects. However, I get affected in various forums, where a lot of it’s about making technically important decisions. Then I notice that’s what’s being requested, and then it’s also what I start asking for when searching among my employees. The technical [skill] is important, and it’s what I look for in a talent. I know that these people find it easier to advance, as technical skills are a part of our culture.

In this way, engineers were seen as constituting an important part of the culture of the subsidiary, thus also being seen as talents. Interestingly, despite the experience of a practical reality in which leadership skills were more relevant, the present engineering logic impacted and shaped the nature of talent identification in terms of redirecting the focus to technical skills.

During our observations, we also noted how the proponents of this logic often mentioned their innovative products, e.g. operating tables and ventilators, in terms of being the ‘gold standard in the industry’ (field notes). In addition to the informal meetings, when we were shown around in the showrooms and had how the products worked demonstrated to us, it also became evident that the actors at the subsidiary were proud of their products and engineering accomplishments. One line manager reflected on this as follows:

It’s very nice to be working for a globally leading company in med-tech. We’re very good at the technical aspects and developing products. (Line Manager, subsidiary).

They were also proud of what their products were doing for other people:

Our products help people and we bring value to their lives, both patients and caregivers. And in that I take great pride. (Talent employee, subsidiary).

It’s something special to be working with something that helps other people…We have the best products! And I take great pride in working with them. (Non-talent employee, subsidiary).

Being emotionally attached to their products, the actors at the subsidiary almost seemed to forget that these had to be sold. The priority was developing the best products and remembering what they were doing for sick patients. When asked how employees of the subsidiary experienced and perceived what the HQ was looking for, and valuing, in a talent, they often answered in following ways:

I feel they choose the ones showing leadership competencies. (Talent employee, subsidiary).

A lot of the focus has been on sales, and then this turns into what they value, clearly. The best sales people are the ones who are identified, and if you compare with the engineers who are not identified as often. They’re not as visible as the sales people. (Talent employee, subsidiary).

However, leadership competencies were seldom valued by the actors at the subsidiary, and thus they started encountering an institutional alternative. This not only made them conscious of the HQ’s way of identifying talents but also of defending their own practice.

We can’t just have a lot of leaders. We also need to develop our engineers who can then develop our products. (Line Manager, subsidiary).

Talent identification and competing logics

The sections above indicate that the way in which actors perform talent identification is grounded in their respective institutional logics. As these logics were different, in effect multiple in the way they stressed different aspects, the way talent identification was conducted also differed between the HQ and the subsidiary. Table 4 summarizes in which way the two logics were different, and to some extent competing.

Actors and logics, and Representative Quotes of how they were competing

TABLE 4. Highlights actors, logics and representative quotes of how the business and engineering logic were competing.

Actors and logics, and Representative Quotes of how they were competing
Actors and Logics	Representative quotes
HQ-Business logic	‘You will go further with your personality than with technical skills.’ (HR Manager, HQ). ‘… the right behaviors. The rest you can learn. So behavior are definitely more important than actual skills. Everyone can learn these, or almost everyone. But not everyone can be taught how to be leaders.’ (HR Manager, HQ). ‘They [criteria] better fit a professional and managerial population than a production and technical population. Typically you would see a higher degree of proportion of that, you know talent ambitions in a professional setting than in a blue-collar direct labor setting’ (HR Manager, HQ). ‘We are a business-driven company, and that reflects how we identify our talents. We identify our talents from managerial potentials and to a certain extent also from sales.. A manager [at the subsidiary] might identify a co-worker as hands-on, a doer…but then you lose the strategic capability or to think ahead…’ (HR Manager, HQ).
Subsidiary-Engineering logic	‘Suddenly we have to talk in totally different terms. And our organization [subsidiary] is not used to that… Now we apparently need more business people.’ (Line Manager, subsidiary). ‘Not everyone is socially competent, and it should be space also for another type of talent. A well-functioning TM system takes that into consideration, and should not be done with a taken-for-granted idea that we should have a sales person, or a leader or so. There must be a local flavor as well, that takes our technical specialist into account.’ (Line Manager, subsidiary). ‘You do not have to earmark for a managerial career, but you can also involve and develop the engineers.’ (Line Manager, subsidiary). ‘At the headquarters you often think of talent as potential leaders, and making career as a manager but someone also has to work.’ (Employee, subsidiary).

As already mentioned, this difference was also noted by the actors involved, as shown by one line manager at the subsidiary when explaining what she thought the HQ was looking for in a talent:

What is being asked for is people with business minds who can lead and transform people….The people that the organization [HQ] wants …must have this thinking. (Line Manager, subsidiary).

Some of the managers at the subsidiary also expressed their disagreement regarding the way in which the criteria focused on a certain type of employee, while neglecting another type:

Engineers don’t seem to be as valuable here as in other organizations, and this has the effect that a lot of the profiles being asked for by the Headquarters, not everyone will be recognized in these. The picture becomes skewed when focusing on business thinking, something that the engineers can’t always identify with. (Line Manager, subsidiary).

Employees noted this difference, even arguing that it was in competition with what they valued at the subsidiary:

Whenever they, at the Headquarters, discuss talent management, they only do it with regard to leadership talents, and that part is important but not the only important thing, and it gets narrow-minded. The technical and expert aspects we don’t even pay attention to, or manage. (Talent employee, subsidiary)

Several employees thus had difficulties identifying with the view of talent as managerial talent, often distinguishing themselves from managers; ‘I believe the best engineers are not good managers. Engineers are not good managers because we’re so odd’ (Interview with non-talent employee). It was not only the leadership-related aspects that felt unfamiliar at the subsidiary but the actors also experienced the mobility criteria as being unfamiliar:

There was a drop-down menu showing different attributes and then you’d fill in different individuals. Then you’d notice what the person behind the system wants, and what I especially noticed last time was the criteria regarding wanting to take jobs internationally. Movement and the like came up and that was something I hadn’t been thinking about myself when working with talent. (Line Manager, subsidiary).

Despite the written instructions on how to identify talents largely based on, the eight leadership competencies, there were also strong views at the subsidiary as regards talents greatly being engineers and technical experts, and thus rooted in the engineering logic. Actors at the subsidiary thus tried to defend their own practice:

At the Headquarters, you often think of talents as potential leaders, and making a career as a manager, but someone also has to work. Other perspectives must be considered than just matching people to become CEOs. Not everyone wants to be a manager, and we need to be successful in helping people to make a career in their expert role since they’re very important. (Line Manager, subsidiary).

By defending its own way of identifying talents, the subsidiary also expressed a strong resistance to starting conducting talent identification in the manner prescribed by the HQ. This resistance was illustrated by one HR specialist when explaining how they put their engineers forward:

When we sent in the lists of talents to the Headquarters, the names on them weren’t just managerial potentials, some of them were also very important technical experts. (HR Specialist, subsidiary).

Line managers not only identified potential managers as talents, but also introduced a ‘local flavor’ by putting engineers forward. This position was supported by the HR specialists at the subsidiary, who similarly stressed the importance of engineering capabilities in being identified as a talent. The subsidiary thus realized it would need to work in ways that made more sense. It adapted and altered the ways of working with talent identification to a greater extent, matching the talent identification practice with its own logic. Thus, even while acknowledging that there was a business component, the subsidiary also privileged the engineering side and this illustrates how it chooses its predominant logics.

Some of the identified talents at the subsidiary, which names were sent on lists to the HQ, began to be talked about. They often had leadership potential which was one of the main aspects HQ was looking for. The idea was then to develop stretch projects to test talents. However, most of the time there was a ‘glass ceiling’ at the subsidiary. Interviewees from the subsidiary, for example, expressed their frustration at continuously being questioned, leading to further irritation and a lack of trust:

You need to have some trust in the organization inasmuch as we’ve identified an employee who we see as a talent, and we shouldn’t in detail, be asked to justify why we think that, not on this level (referring to the detailed criteria), then you won’t have any trust in your management. (Line Manager, subsidiary).

Given this discrepancy and frustration, it became surprising to hear that work and responsibility regarding the TM platform, particularly the talent identification practice, was largely assigned to the subsidiary. Managers explained this as being due to the various mergers Medico had undergone over the years, and how line managers had been closest to the employees and thus knew them best. One HR manager reflected on what this entailed:

I think it’s a strength in one way. I love people on the ground who take on responsibility. I think it’s in our DNA. (HR Manager, HQ).

This was at first described as unproblematic; in the way the process was implemented and put in place in the organization, as the following quote illustrates:

The process is up and running at all subsidiaries and there’s no problem in that. They know what to do and we get all the templates back. It’s quick, after launching talent management, to make the [local] managers aware of what to do. (HR Manager, HQ).

Empowering the subsidiary also brought complexities and ambiguities regarding how talent identification unfolded in practice, complexities not mentioned initially. Many HQ managers indicated challenges in how the initiative was eventually used, despite stating their love of local initiatives. This indicated that the process was more complex than first thought. Disagreement regarding how talent identification ought to be done originated from both the subsidiary and the HQ. How employees were identified and put forward locally was seen as competing with what the HQ favored. As two HR managers put it:

Often, we don’t agree with the assessments made locally…We ultimately have a view of what talent and potential look like. (HR Manager, HQ).

If you took the data directly from the local site [subsidiary], then you’d have 35% talents, and that would be great. But this isn’t accurate. (HR Manager, HQ).

The quotes thus illustrate the notion of competing values, and a potential conflict regarding what the HQ and the subsidiary valued.

Discussion

This article sets out to answer the following research question: How is talent identification shaped by institutional logics? To begin with, by moving beyond the managerialist and unitarist approach, the findings show that two different institutional logics are available at the studied organization; i.e. a business logic (dominant at the HQ), whereby actors facing increasing sales pressures drew upon a business culture, and an engineering logic (dominant at the subsidiary), whereby actors focusing on the importance of developing products drew upon an engineering culture. Despite a few interviewees referring to both logics, the use of certain logics was highly tied to specific categories of people, or members of particular groups, at the organization (McPherson & Sauder, 2013; Pache & Santos, 2013).

With respect to talent identification, the findings also showed that the way in which actors identify talents is grounded in their respective institutional logics. In this way, the institutional logics are used by actors to achieve organizational and individual goals – we also showed how actors used these to shape and make talent identification decisions. When invoking the business logic, the actors mainly drew on aspects such as, mobility, sales, commercial and leadership competencies when conducting talent identification. When invoking the engineering logic, the actors instead drew on technical, expert and specialist competencies when conducting talent identification. Since these logics stress different aspects, the way in which talent identification was conducted also differed between the HQ and the subsidiary. Our findings also showed that institutional logics not only shape the talent identification practice but also affect the talent concept itself, as well as the answer to the question ‘talent for what?’. Competing institutional logics can thus translate into competing definitions of talent. A talent was both a potential leader and an engineer, suggesting that talent identities are embedded in logics and provide the contexts for decisions and outcomes (Thornton & Ocasio, 1999). We highlight the importance of considering talent as a contextualized and multidimensional concept, and not always ascribing it the meaning of being a leader or a manager. In doing this, we add empirical support to recent conceptual arguments (e.g. Collings et al., 2018), stressing the importance of considering talent beyond senior organizational leaders and HQ employees in general, which are often assumed to be in focus of TM. We thus also add to the central inclusive-exclusive debate, by stressing an additional inequality risk in using the exclusive approach, when not considering the multiple internal actors’ potential legitimate understandings (Delbridge & Keenoy, 2010) of the question ‘talent for what?’.

From an economics-based perspective, as adopted by much of the mainstream HRM and TM research (see Delbridge et al., 2011; Thunnissen & Buttiens, 2017), organizational actors are expected to exert a degree of agency in order to maximize individual self-interest, and to not merely be passive respondents to institutional forces (Martin et al., 2016). In contrast, we have illustrated how the HQ’s and the subsidiary’s self-interested intentions were most often bounded by their dominant institutional logics. By drawing on the institutional logics perspective, our findings thus indicate that actors belonging to an organizational group will closely adhere to that group’s primary logic (McPherson & Sauder, 2013). We suggest that the way in which talent identification unfolds in practice is a result of cultural forces exerting an influence on everyday organizational behavior, confirming the importance of considering both the local organizational context and social relationships, as stressed in the critical HRM literature (e.g. Alvesson, 2009; Delbridge et al., 2011; Watson, 2004). Thus, instead of relying on other potential impacts, e.g. instrumentality and rational action (Delbridge & Keenoy, 2010; Keegan & Boselie, 2006), this demonstrates, consequently, that the institutional logics constitute a key contextual factor impacting the nature of TM in organizations.

There are a number of important implications of this finding. We add empirical evidence that matches previous arguments, mainly conceptual, to be found in both the HRM and TM literature (e.g. Collings et al., 2018; Khilji & Wang, 2006; Nishii & Wright, 2008; Sonnenberg et al., 2014; Thunnissen, 2016), concluding that intended and actual practices may diverge. We agree that this divergence can partly be explained by the different interests of the involved actors (Thunnissen, 2016); however, when considering everyday practices and the local organizational context (Delbridge et al., 2011; Watson, 2004), we also extend this argument by offering an alternative explanation and also why interests differ. It is often assumed that the way in which people define and identify talent is shaped by the multiple actors involved in a TM process. However, rather than being driven mainly by self-interest or a lack of training, our findings suggest that the different interests of the actors can instead be explained in terms of being rooted in different institutional logics. Actors represent, and import into an organization, the meanings and norms of the logics to which they have primarily been exposed (Greenwood et al., 2011). In contrast to rational choice theory, and the importance of sufficient and correct training, knowledge and information, we argue that the bounded intentions of the actors are circumscribed by their institutional embeddedness (see Martin et al., 2016). Actors may still be driven by self-interest, but under the influence of the institutional arrangements. This finding is especially interesting for the field of TM since it could explain why studies have reported a discrepancy between the intended and actual practices regardless of formal protocols, rules and prescriptions. The explanation for this is not that these practices are implemented by actors other than decision-makers per se, but the presence of multiple and competing institutional logics on the different organizational levels. This further highlights the importance of considering organizations’ local institutional contexts when aiming to study how talent identification and TM in general, unfold in practice. This is the case because it is here that the negotiated nature of these practices unfolds (Delbridge et al., 2011; Jenkins & Delbridge, 2013). We have thus shown the importance of considering the multiple internal actors’ acting in accordance with particular logics and norms (Delbridge & Keenoy, 2010; Delbridge, 2010; Thompson, 2011).

We have shown how multiple institutional logics were in existence and affected how the work of talent identification, and the concept of talent, unfolded in practice. It was also important to address the issue of how these logics were dealt with, especially since, at times, they were seen as competing. With increasing intensification, in terms of attempting to structure and standardize the talent identification process, the involved actors also became more familiar (Pache & Santos, 2013) with each other’s logics. For example, since leadership competencies were seldom used by the actors at the subsidiary, instead being encountered via directives issued by the HQ, they thus started to encounter an institutional alternative (Besharov & Smith, 2014). In this way, the TM actors were aware of the differences in the cultural norms, symbols and practices of the different institutional orders (Besharov & Smith, 2014). However, rather than dissolving differences (Pache & Santos, 2013), this familiarity made the logics segmented as the different groups continued to operate according to their own respective logics (Reay & Hinings, 2009). This segmenting was enabled by means of some dimensions of work being reflected in the business logic, while other parts were reflected in the engineering logic. The subsidiary defended its own practice in accordance with its dominant logic, as awareness of the HQ’s alternative and competing logic became greater. Although other logics existed, it was the dominant logic that guided action (Lounsbury, 2008), affecting the strategy and structure of the respective site. The attention of the decision-makers (talent identifiers) was thus focused on the issues that were consistent with the respective logic (see Thornton et al., 2012).

Faced with increasing pressure to deliver on sales targets, the HQ even exercised compliance pressure (Pache & Santos, 2010) through demands it put on the subsidiary to identify talents. The instructions and intended practice initiated by the HQ continued, however, to encounter resistance, and adapted in order to make sense in everyday practices at the subsidiary. The HQ’s attempt to use institutional logics as tools for reaching agreements with the subsidiary, and for maintaining its legitimacy (see Swan et al., 2010) by indicating the limitations of how the subsidiary identified its talents, thus did not work. Instead, given the increased frustration and irritation of being questioned, the subsidiary did not reconfigure the conflicting logics to be complementary, instead further demarcating them.

As already mentioned, our findings support previous HRM and TM studies (e.g. Boselie et al., 2009; Delbridge & Keenoy, 2010; Farndale, Pai, Sparrow, & Scullion, 2014; Gallardo-Gallardo & Thunnissen, 2016; Sonnenberg et al., 2014), highlighting how multiple actors with different perspectives are present in organizations. More importantly, we also take a step further toward increasing current understanding of the importance of using a pluralistic TM approach (Collings, 2014; Collings et al., 2018; Thunnissen et al., 2013; Thunnissen & Van Arensbergen, 2015), and the potential gap between intended and actual TM practices. Much of the mainstream HRM and TM research has tended to assume that not only are people mainly instrumental, self-interested and rational, but also that the focus ought to be on how to minimize the shirking behavior of divergent actors (Delbridge & Keenoy, 2010; Keegan & Boselie, 2006). For example, by communicating, educating and even controlling TM actors in intended practices, it is believed there will be an increase in the alignment between the intended and actual practices (e.g. Collings et al., 2018; Jones et al., 2012; Mäkelä et al., 2010; Silzer & Church, 2010). However, in contrast to these arguments, we have shown how increased familiarity with each other’s logics did not decrease the institutional distance, instead causing the actors to start defending their own practices even further. As there was no unique alignment between the two logics (Voronov, De Clercq, & Hinings, 2013), we cannot assume that an increased awareness, interaction and attempts at alignment, through compliance pressures for example, would lead to the convergence of practices. One implication of this finding is that informing and educating local TM actors in intended practices may have the opposite effect, i.e. increasing the discrepancy even further. As our findings have shown, the social relationships with the cultural values seem more important than the rational and instrumental motives, and the important thing is thus to consider the organizational institutional and cultural differences (Alvesson, 2009) in order to understand the root causes of how and why actors engage with HRM and TM practices. This is further suggested to increase the chances of striving for alignment between different institutional logics.

Conclusion

At first glance, this could be seen as a straightforward story about a classical agency issue in which the subsidiary ‘did its own thing’, leading to misalignment between the HQ’s intentions and the subsidiary’s actions. However, if we wish to move beyond our understanding of the HQ-subsidiary relationship as a principal-agent structure, as rooted in, for example, agency theory and rational choice theory, and as adopted by much of the mainstream HRM and TM research, the story unfolds somewhat differently. By drawing on the institutional logics perspective, and on arguments put forward in the emerging critical HRM literature, we have illustrated how the way in which organizational actors conduct talent identification is grounded in what logic they enact and make use of, something which contradicts arguments about being driven by pure and individual self-interest based mainly on financial calculations. We also highlight how multiple and sometimes competing institutional logics were available at the studied organization, causing a divergence between intended and actual TM practices.

In doing this, we make several contributions. First, we contribute to the TM literature by answering the call for in-depth studies of how talent identification unfolds in practice (e.g. Gallardo-Gallardo et al., 2013; Mäkelä et al., 2010), and by introducing the institutional logics perspective as an explanation of how the intended and actual TM practices can vary. By relying not only on formal descriptions given by HR but also incorporating other internal actors, e.g. line-managers and employees, we are advancing knowledge of talent identification and progress toward a more realistic and in-depth understanding of how it unfolds in practice. In contrast to earlier arguments focusing on ad hoc, lack of skills and training when explaining how intended and actual practices diverge, the institutional logics perspective puts forward logics, rules and norms, in effect the institutional context, as explaining factors to why actors perform talent identification as they do. By showing the explanatory power of the institutional logics perspective in understanding the motives of various organizational actors during talent identification, we thus advance the theoretical understanding of TM by explaining how local organizational configurations and institutional contexts impact the work and nature of TM. By adopting a more critical and contextually-based study, we answer the call (e.g. Thunnissen & Buttiens, 2017) to provide underlying explanations as to how and why actors engage in TM practices as they do.

Second, we contribute to the HRM literature, especially the emerging stream of critical HRM (e.g. Delbridge & Keenoy, 2010; Delbridge et al., 2011; Keegan & Boselie, 2006), and its call for more contextualized and non-managerialist research that gives a voice to multiple organizational actors. More specifically, we contribute to existing studies that are focused on unpacking the notion and the role of context in HRM (Brewster et al., 2016; Jenkins & Delbridge, 2013), using the institutional logics as explanatory leverage in order to shed light on how internal actors are embedded in local (intra)organizational arrangements, and how this impacts work on HR practices. While much of the contemporary HRM literature has been focused on cross-national variations as its main contextual factor (Almond, 2011; Bévort & Poulfelt, 2015), we point to the need for greater attention to local organizational contexts in order to understand the contextual embeddedness of HR practices. We also provide a deeper examination and understanding of how organizational and employee goals and interests can diverge, i.e. through the way in which organizational actors enact and make use of different institutional logics. We thus engage with the critical HRM stream (e.g. Delbridge et al., 2011; Evans & Tourish, 2017; Watson 2004) by questioning the exaggeration that people are motivated by self-interest mainly on the basis of economic calculations, instead advancing social relationships, with cultural values and behavioral norms, as root causes of how and why actors engage differently with HR practices.

Our study shows that standardizing a talent identification practice is not as easy as it seems. It is not enough, or even helpful, to inform and educate local managers in intended practices. Increasing familiarity may even cement divergences all the more. Managers need to understand and navigate the multiple institutional contexts of various organizational sites, and to strive to align these. In order for TM practices to be strategic, they need to consider the possibility of multiple strategies within the same organization, or at least invite multiple organizational actors in when designing intended TM policies. Managers also need to challenge and question their own practices, as well as the underlying assumptions upon which these are based. This is believed to offer management better support in its balancing of organizational and employee goals and interests.

Our study has a number of limitations, offering avenues for future research. We adopted a single case study, something which limits the generalizability, and thus caution should be applied if and when attempts are made to transfer conclusions to other contexts. However, the in-depth understanding we have obtained as regards how talent identification unfolds in practice, and the influence of the local institutional context, may not be possible using large quantitative studies. Future research might thus be able to take our insights further and investigate talent identification in other contexts, especially in public and non-profit organizations since these are often described in terms of being based on different values and norms than private organizations. It would also be interesting to understand the impact of institutional logics in highly professionalized contexts, involving e.g. healthcare professionals and lawyers. We have focused on talent identification, but we would also expect institutional logics to shape other TM practices, and we would thus urge future research to advance our understanding in this regard.

Acknowledgments

The author thanks Editor Michael Dickmann, and the anonymous reviewers for their insightful comments and suggestions throughout the review process. The author also gratefully acknowledges Evelina Börjesson for her assistance during the data collection and Niklas Egels-Zandén for his helpful comments on earlier drafts of this article.

Disclosure statement

No potential conflict of interest was reported by the author.