Beyond vision : exploring paradox in HR service delivery models

1

Master Thesis

Beyond vision: Exploring Paradox in HR Service

Delivery Models

Ilja Bitterling 10894802

MSc in Business Administration: Track Leadership & Management University of Amsterdam

Supervisor UvA: Dr. A.E. Keegan

Internship supervisor Berenschot: Hella Sylva MSc Submission date: 29-6-2015. Final Version.

2 Statement of originality

This document is written by Ilja Bitterling who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

3

Acknowledgements

First and foremost I offer my sincerest gratitude to my supervisor at UvA, Anne Keegan, who has inspired me immensely to go on this exciting journey into the world of paradox and who supported my unconditionally throughout my whole thesis, and, especially when I started to lose my cool. I would also like to express my gratitude to Berenschot, who gave me the opportunity to contact two of the case companies through their network and many of my HROI colleagues who provided me with insights regarding the complex nature of HR in practice. Special thanks goes to my internship supervisor at Berenschot, Hella Sylva, who joined me on the majority of the interviews and who helped me clarify my thoughts, not only regarding my thesis. Finally, I would like to thank the case companies for giving me the opportunity to conduct my research there and all the interviewees that provided rich, insightful and honest accounts of the intricacies and present-day challenges of the HR models they work with.

4

Contents

Abstract 6

1.

2.

2.1.

2.2.

2.3.

2.4.

2.4.1.

2.4.2.

2.4.3.

2.5.

3.

3.1.

3.2.

5.1.2.

5.1.3.

5.1.4.

5.2.

5.2.1.

5.2.2.

5.2.3.

5.2.4.

5.3.

5.3.1.

5.3.2.

5.3.3.

5

6.

6.1.

6.2.

6.3.

6.4.

6

Abstract

Ulrich’s (1997) influential book Human Resource Champions has defined the overarching paradigm guiding much of the thinking among HRM practitioners over the last two decades. Therefore, the Ulrich-inspired three-legged model of HR service delivery seems pervasive as a general design approach, but not much is known about its use and the dynamics of the model in practice. The aims of the research are to explore evidence of three-legged design models in practice, the reasons for why such HR structures come about and the tensions and paradoxes encountered by actors working within these models on a daily basis. Using a qualitative approach, 20 interviews were conducted with HR directors, managers, professionals and (senior) line managers in 3 large organizations operating within the Dutch energy & utility sector. Findings reveal that the three-legged-stool model is alive and well within these organizations and that improvements to the model are driven by HR’s continuous focus on delivering strategic value. By adopting a paradox lens to systematically examine the tensions inherent in the observed HR structures, I find that HR structures are characterized by three paradoxes: a global stakeholder paradox, an efficiency/flexibility paradox and a power paradox and that the HR function in itself, is, paradoxical. Based on the findings I argue that that the future of HR function design lies in how well HR practitioners can cope with organizational dynamics characterized by shifting boundaries, complexity and business goals that are interrelated, yet contradictory.

7

1. Introduction

Human resource management (HRM) is at the core of any business (Welbourne, 2012). In its broadest definition, HRM constitutes “the management of employment” (Marchington & Wilkinson, 2005, p. 26) and includes “anything and everything associated with the management of employment relations in the firm” (Boxall & Purcell, 2000, p. 184). Over the last several decades, HR academics and practitioners have increasingly focused on how the HRM function can add strategic value to the business (Wright, Snell & Dyer, 2005; Marchington, 2015). Following Boxall and Purcell (2011), strategic here means that the “people management” practices designed and implemented by the HR department elicit employee behaviour that can lead to sustained competitive advantage. Some scholars go even further by arguing that HRM is strategic only when firms’ strategic decisions are finalized after the HR implications are discussed (Boxall & Steeneveld, 1999). Given HR’s legacy of having relatively little influence in strategic decision-making, HR practitioners and researchers alike have long been interested in HR models emphasizing the strategic contribution of HR.

From a historical perspective, commentators have often lamented the “hodge podge nature” of HR activities and the ambiguity of HR practitioner roles is a long-standing debate in the field (see Legge, 1978, Guest and King, 2004). Perhaps as a result of the multifaceted and ambiguous profile of HR responsibilities traditionally seen in the literature on HRM, it is unsurprising that a model promising clarity, simplicity and, most importantly, strategic impact, would gain influence. Even accepting that, the rise of the so-called Ulrich model of HRM organization has taken many, and many HRM academics in particular, by surprise.

In his seminal work, Ulrich (1997) advocated that if HR wanted to overcome its beleaguered reputation of a highly bureaucratic and inefficient organization, it had to shift focus from what HR people do (the practices they employ) to what they deliver. The HR function needed to become strategic, value-adding and business aligned and Ulrich advocated the adoption of shared service arrangements and role specialization (Reilly, Tamkin & Broughton, 2007). What emerged is the so-called three-legged design of HR service delivery which comprises shared service centers delivering administrative support to line managers and staff (e.g. payroll administration and support), centers of excellence providing specialist support in specific HR areas, such as recruitment or training & development, and business partners who would work closely with line managers to ensure alignment between business goals and HRM.

Interest in the Ulrich-inspired three legged stool model of the HRM function has spread itself across the organizational landscape to an impressive extent. Talk of the model is ubiquitous (Robinson, 2006) and, as Marchington (2008: 8) argues: “Ulrich’s work has attracted a mass of publicity and research, and many organisations have adopted the business partner role or some variant of it (Brown et al., 2004; Reilly et al., 2007)”. That said, there is little empirical research into the reasons why organizations might adopt the three legged model, the exact nature of the three-legs (Robinson, 2006) and their relationships with each other, nor especially the challenges HR practitioners and other important HR stakeholders encounter when

8 working within such a configuration (Caldwell, 2003). Furthermore, while the face validity of the model appears to strong, this may have engendered a complacency of kinds among academic researchers where it appears taken for granted that an organization will adopt some kind of Ulrich-type mode and empirical enquiry into variations, exceptions, and diversity of modelling HRM along the three-legged stool kind is very under developed at this time (Reilly et al, 2007; Francis and Keegan, 2006).

I will argue in this thesis that we need better reflective engagement with the kinds of configurations emerging in HRM practice in different contexts. For example, we do not currently know enough about the different reasons for adopting (aspects of) the three-legged-stool model, the contextual challenges for HR actors based on the differences in adoption types, and tensions arising from adoption including paradoxical tensions. A better and more contextually rich understanding of these issues might aid in the practical understanding of the consequences of the three legged stool, and help address important theoretical questions such as are three legs enough (Robinson, 2006), what challenges face actors in systems configured along these lines, and whether paradoxical tensions are responded to in active or passive ways leading to vicious or virtuous cycles (Legge, 1978, Smith and Lewis, 2011).

To that end, this thesis addresses the following main research questions:

RQ: How do organizations configure their HR function and what paradoxes do they face based on the configurations adopted?

To answer this question a number of sub-questions will be answered:

• Do organizations conform to the Ulrich-style three-legged stool type configuration? • How do the configurations they adopt come about?

• What are the key tensions faced as a result of adoption of the identified configuration and are these paradoxical in nature?

In order to answer these questions, I first review the literature on the structuring of the HR function. This review provides a basis for describing the historical development of, and current understanding of, how the HR function might arrange its activities and roles to meet the demands most commentators describe as confronting the HR function and individual HRM practitioners. The literature review provides insight into the main model offered in the academic HRM literature, which is the so-called Ulrich three-legged model. In the empirical part of the thesis, I carry out a systematic analysis of the data on three case studies conducted in one industry to examine the structure of the HR function. I then draw on paradox theory (Smith and Lewis, 2011) to systematically examine the tensions described by case participants interviewed during the study, to understand the tensions arising from the particular ways of configuring HRM activities. I describe the responses to tensions the participants describe. On the basis of a comparison and contrast of the data on each company, I provide insight into the diversity as well as overlap in empirical practice regarding the configuring of the HR function.

9

2. Literature Review

Since the publication of his most famous work HR Champions in 1997, the Ulrich-inspired three legged stool model of the HRM function has gained widespread attention and is the source of much discussion in academic and practitioner circles. However, while the model is pervasive (Marchington, 2015), there is little empirical research into reasons why organizations adopt the three legged model, how organizations implement the model, the extent to which it is adopted wholesale or piecemeal by companies, the exact nature of the three-legs and their relationships with each other, or the challenges HR practitioners and other important HR stakeholders encounter when working within such a configuration. This combination of widespread attention to and discussion of the model, and lack of empirical evidence regarding how it is used and the challenges associated with it, constitute the main puzzle (Alvesson and Karreman, 2007) driving this thesis and the empirical study carried out that is reported here. First though, some background to the model is needed.

2.1. Origins of the Ulrich model

In the years before publication of Ulrich’s (1997) seminal book ‘Human Resource Champions’, there was growing talk about whether the HR function could be fully outsourced or even abolished.

Nestling warm and sleepy in our company, like the asp in Cleopatra’s bosom, is a department whose employees spend 80% of their time on routine administrative tasks. Nearly every function of this department can be performed more expertly for less by others. Chances are its leaders are unable to describe their contribution to value-added except in trendy, unquantifiable and wanna-be terms … I am describing your human resources department, and I have a modest proposal: Why not blow it up?

Stewart (1996), p. 105.

While others have not been as forthright as this, there was a trend of seeing the HR department in ‘crisis’ as HR struggled for legitimacy and status in an era of corporate downsizing (Redman & Wilkinson, 2008). Despite the rising recognition that HR issues are vitally important to organizations, many senior managers seemed to think that personnel management was far too important to be left to the HR department. This had all been fueled by growing concerns among senior management about the quality and responsiveness of in-house HR functions (Greer, Youngblood & Gray, 1999). In fact, HR departments were generally seen as ‘bureaucratic’ and ‘isolated from the outside world’ (Guest & King, 2004). In his scathing analysis of why “actuality has fallen so far below expectation in personnel management,” Skinner (1981, p. 106) argued that the problem is rooted in the problematic role of HRM in corporate decision-making and the challenge of aligning HR strategy with the different needs of the business. As discourse on human resource management sought to overcome the HR department’s poor reputation, the view emerged that HR needed to become more aligned with the strategic imperatives of the organization, with a clear focus on how to deliver value, else it might see a painful demise.

10 In this context, Ulrich (1997) provided a contemporary focus, arguing that the value of the HR function is not defined by what’s inside of it, its practices or activities, but rather by what it delivers to its end users. In this respect, he proposed that the HR function needs to attend to four ‘result domains’, namely 1) the management of strategic human resources, aligning HRM with business strategy, 2) the management of transformation and change, developing organizational capabilities for change, 3) the management of employee contribution, listening and responding to employees and providing them with the resources necessary to perform, and 4) the management of firm infrastructure, constantly improving the organization of an efficient HR administration. The model is depicted in figure 1. These domains are corresponding with four core roles that HR professionals need to adopt, the roles of strategic partner, change agent, employee champion and administrative expert. This framework formed the basis for Ulrich’s ‘from doing to delivering’ argument, which incorporated dominant ‘best fit’ and ‘best practice’ approaches while emphasizing focus on delivering economic value.

In his framework, Ulrich (1997) argues that HR professionals must be both operational and strategic, focusing on the long term and the short term. However, commentators have warned for ‘role ambiguity’ and ‘role conflict’ within the HR profession as different constituencies, such as senior executives, line managers, corporate HR and employees, prioritize different aspect of the role (Buyens & de Vos, 2001; Caldwell, 2003). Caldwell (2003) reports that as HR is tailored to the needs of the business (operational focus), it runs the risk of becoming short term and reactive, which can undermine HR’s potential strategic contribution in the long run. Similarly, Wright and Snell (2005, p. 181) argue that HR professionals “must distinguish between decisions that are driven by the business and decisions driven for the business. A focus on short-term financial returns for fickle investors may be made at the long-term cost of organizational viability.” Moreover, several commentators report that in HR’s eagerness to become more strategic, HR professionals have been encouraged to aspire to the role of strategic business partner, causing their role as employee champion, i.e. making sure that employee concerns are being heard, to shrink (Francis & Keegan, 2006; Brown, Metz, Kregan & Kulik, 2009). As Guest & Woodrow (2012, p. 110) succinctly summarize: “despite the appealing rhetoric, the need to fulfil multiple and potentially competing roles serves mainly to highlight the challenges facing HR managers and outlining the four roles does not in itself resolve the inherent conflicts.” The challenge, thus, lies not in believing the vision, but in putting it to practice, moving beyond the vision.

11 Figure 1. Ulrich's (1997) 4-role typology

2.2. The rise of shared service organizations

In addition to outlining the roles of strategic partner, change agent, administrative expert and employee champion, Ulrich (1995, 1997) set out to describe how HR could best be organized to fulfill these roles. Over the years, HR had been organized in a variety of ways. Some functions were characterized by a highly decentralized structure, with each part of HR looking after a certain business unit or geographical location. Others were characterized by a highly centralized structure, with HR professionals and managers providing service to the various business units or locations from a centrally coordinated center. Looking at the HR department itself, there might have specializations by work area or by employee grade or group (with some looking after clerical staff whereas others would provide services to higher management) (Reilly et al., 2007). There were also several accounts of HR lacking a structure of sorts. For instance, Skinner (1981) described how personnel managers had engaged in many different techniques to acquire an effective, loyal, and committed group of employees and as the HR department continued to pursue all sorts of ‘expertise’, the HR function could resemble what Drucker (1961, p. 243) described as ‘a collection of incidental techniques without much internal cohesion … a hodge podge.’ In the context of subsequent changes, it is also important to highlight that administrative HR tasks were carried out within (local) HR teams.

Within this context, Ulrich (1995) argued that organizations needed to become shared service organizations, comprising shared service centers, centers of excellence and business partners. These seem, on face value, required or valued “parts” of a HRM function, but what does each represent? What problems does it seek to solve? And finally, is there anything missing? The following sections explore in more depth what is known about the three legs as well as raising questions about what aspects of the legs, or the connections, are still, perhaps, problematic.

12

In Shared Service Centers (SSCs), common administrative tasks and activities associated with the HR

function are consolidated and standardized across different business units into one single service center (Maatman, Bondarouk & Looise, 2010). In this way of organizing, business units that used to perform administrative tasks locally, pool their resources and ‘share’ in the service delivery, resulting in cost benefits through economies of scale (Reilly et al., 2007). The main task of a SSC is to efficiently process paperwork and to supply information and advice on HR policy and practice to employees, line managers and HR staff (Ulrich, 1995; Reilly, 2000). Shared service centers rely heavily on IT and E-HRM applications and are characterized by a tiered system, often consisting of on-line support, a call center and specialists that handle inquiries that need special attendance (Maatman et al., 2010). Shared service centers often follow a click-call-face philosophy in which employees and line managers are encouraged to first look for answers to their questions on-line, before calling the service desk or meeting with a specialist. Although resources from the field are shared and bundled, the control over the use of these resources resides with clients and end-users, “the user is the chooser” (Ulrich, 1995, p. 14). In fact, shared service centers resemble sourcing arrangements where business units articulate the “type, level and quality of services they want from the center, at the price they are willing to pay” (Quinn, Cooke & Kris, 2000, p.13). These tend to be formalized within Service Level Agreements (SLAs) in which the terms, against which HR activities need to be delivered, are specified (e.g. Farndale, Paauwe & Hoeksema, 2009). Whereas Oates (1998) suggests that SSC’s could be seen as a form of internal outsourcing, Quinn et al. (2000) argue that SSCs can be seen as an interim step between in-house production and outsourcing. Since the implementation of a SSC involves formalizing the terms of delivery, the availability of such criteria makes comparison with other providers a logical next step.

Centres of excellence/expertise (COEs) combine distributed talent throughout an organization into a

shared service, acting as an interface for line managers and HR staff dealing with complex issues that require specialist knowledge (Ulrich, 1995). Consequentially, HR specialists no longer work in separate divisions but come to together in one central business unit, which responds to all calls for expertise within the organization. Centers of excellence often include global or regional teams, which specialise in a particular HRM discipline, for example recruitment, legal issues or learning (Ulrich, 1995). By creating a shared pool of resources, the organization will become more flexible and dynamic, i.e. resources can be shifted quickly to meet business needs (Ulrich, 1995). These centers of excellence are often also concerned with the design of corporate policy.

Shared service organizations are driven by multiple logics. They are driven by goals of efficiency and cost-effectiveness as well as increased customer experience and increased flexibility in the allocation of resources (Farndale et al., 2009). However, Farndale et al. (2009) warn for the fact that a focus on achieving cost effectiveness can undermine the degree of customer orientation of the HR function and the quality of its services. In fact, the challenge of simultaneously pursuing efficiency (cost effectiveness) and flexibility (customer orientation) has been highlighted across various HRM literatures. For instance, within literature on ‘green HRM literature’ or the literature on sustainable approaches to HRM, Ehnert (2014) highlights the efficiency-substance-paradox. The challenge of the efficiency-substance-paradox lies in

13 focusing not only on financial criteria and the current workforce but also on meeting the resource demands of the future, which requires flexibility and a long-term focus. Similarly, Francis and D’Annunzio-Green (2003) used the term paradox to describe the tension between ‘hard’ and ‘soft’ versions of HRM. Organizations that follow the imperative of standardization (hard HRM) might achieve cost-effectiveness (efficiency) but run the risk of limiting their employees in terms of creativity and commitment. However, organizations that follow the imperative of delivering customer-focused quality might elicit entrepreneurial behaviours (soft HRM) but run the risk of reduced management control. As of yet, it is not clear how the competing forces of efficiency and flexibility play out in the HR shared service structures of today.

Besides shared service centers and centers of excellence, Ulrich (1995, 1997) argues that HR professionals need to become business partners in order to align HR practices with business goals. The aim is that business partners are deeply ingrained within the organization and develop and maintain close relationships with senior managers and specific business units. They carry the main responsibility for translating HR initiatives into business results and “become the dominant interface between shared services and business requirements” (Ulrich, 1995, p. 18). Although Ulrich (1997) originally applied the concept of ‘business partner’ to all four roles (strategic partner, change agent, administrative expert and employee champion), the term business partner has mostly been used in conjunction with the role of strategic partner, i.e. the strategic business partner. “HR professionals have mistakenly defined business partnership as taking place exclusively in the strategic arena, not recognizing the importance of with, and for, employee contribution” (Ulrich, 1997, p. 126). The emphasis on business partners becoming more strategic can be understood if we take into account the ‘strategic fit’ ambitions that have been haunting the HR function and HR professionals for years (Keegan & Francis, 2010). Moreover, a focus on strategic issues presents HR with an opportunity to regain some of its professional legitimacy that has been challenged by the continuous devolvement of HR work to line managers and the outsourcing of peripheral HR responsibilities to specialist organizations over the years. So, given HR’s legacy, its emphasis on strategic contribution only makes sense.

With regard to the structure of the HR function, Ulrich himself did not use the term business partner as a category of HR function design in his original work. The emergence of business partners as the third leg of the stool must have come from HR practitioners and consultant interpreting Ulrich’s original work (cf. Francis, Parkes & Reddington, 2014). In later works (Ulrich & Brockbank, 2005; Ulrich, Younger & Brockbank, 2008), Ulrich incorporated strategic business partners under ‘embedded HR,’ reserving the term business partner to be used in conjunction with different HR roles.

14

2.3. The Three-legged-stool

While references to the three-legged stool are common, precisions regarding the origins of the model is lacking in academic writing on HR function design. In their discussion on the origins of the three-legged-stool approach, Hird et al. (2009) note that the earliest reference they found was in an HR magazine article written by HR consultancy firm Mercer, in 1999 (see exhibit 1).

It is likely that the emergence and further development of these Ulrich-style structures has been led by HR consultants and practitioners, driven by functional concerns about improving the efficiency and quality of HR service delivery (Francis et al., 2014). These improvement to the model are important as Hird et al. (2009) argue that a successful implementation of the three-legged-stool model relies on the organization’s skill in aligning and balancing how shared service centers, centers of expertise and business partners are integrated with each other, which demands an understanding of the model beyond its prescriptive principles. The HR structure must truly reflect the needs of the business model and the line managers within that model, it needs to be ‘fit for purpose’ (Hird et al., 2009, p. 26). To date, few empirical studies have been undertaken to examine if organizations adopting (some variation of) the three legged stool model have practices or processes in place that would constitute the type of balancing or alignment raised by HRM researchers discussing the organization of HRM activities.

Exhibit 1. First mention ‘three-legged-stool’.

The traditional design typically includes a vice president of HR, then a manager of compensation and benefits, a manager of HRIS and payroll, a manager of employment, and so on… However, the emerging model is more like a three-legged stool. One leg of the stool includes an administrative service center, which processes payroll, benefits and the like and focuses on efficiency in transaction functions. The second leg is a center of excellence (or expertise), in which managers and specialists work. These employees concentrate on design rather than transactions and will have line managers as their customers. HR business partners make up the third leg. These are generalists who may report directly to line managers and indirectly to HR. Business partners do not get involved in transactions, but instead act as consultants and planners, linking the business with appropriate HR programs… about 80 percent of businesses today have a mix of the traditional and three-legged stool models

15 It is, at this point, also worth a minor digression to consider the implications of implementing a three-legged-stool model in international contexts. Despite the fact that the Ulrich model seems to have been adopted by many multinational corporations (MNCs) (Hird et al., 2009), the implementation process of a three-legged-model for global organizations does not appear to follow a common path. For instance, Boglind, Hällsten and Thilander (2011) found in a sample of 7 Swedish MNCs, that the adoption of Ulrich’s three-legged model from blueprint to actual practice followed successive translations according to local context. Literature on international HRM has often investigated the (HR) strategies of MNCs through a framework of globalization-localization (Bartlett & Ghoshal, 1989). In order to compete on a global level, MNCs are often pressured to standardize HRM policies and practices across subsidiaries to seek economies of scale or scope (internal consistency) and simultaneously adapt business strategy to changing local demands, e.g. legal requirements (external consistency). However, whether a three-legged-stool model accommodates a structure that balances these competing demands in international settings, is not (yet) clear.

2.4. Three-legged-stool: signals of inherent problems?

The evidence on the uptake of the three-legged stool indicates it is pervasive as a generic design type. However, the trend in HR publications over the last couple of years seems to indicate that the model doesn’t always work, or at least not without tensions. Based on the current literature, I will now highlight three potential problems that can surface in three-legged Ulrich-inspired HR service delivery models.

2.4.1. Operational execution: the fourth leg of the stool?

In a survey among 800 senior HR professionals, Reilly et al. (2007) report that the most common problem encountered in implementing a shared service model is boundary disputes, followed closely by gaps in service provision and communication issues. Upon closer inspection, a central concern of splitting the function into three distinct entities is that it can leave a hole at the very heart of HR operations (Reilly, 2006). This often concerns the execution of operational services in areas in which line managers need most assistance. Examples include handling interdisciplinary issues, attending performance review meetings and processing changes to employment contracts. These operational tasks are usually not taken on by shared service centers since they often require personal attention and sometimes physical presence, centers of expertise are usually limited to an advisory role and are not ingrained enough in the business to be of assistance anyway and line managers often lack the knowledge or resist acquiring it. Reilly (2006, p. 36) refers to this as the ‘polo’ problem where “HR is unable to provide operational support because it’s not a specific part of anyone’s job.” As a consequence, business partners often report that they are “getting drawn into the wrong activities” (Reilly et al., 2007, p. 13) as they find themselves forced to take on operational work that conflicts with their strategic objectives. In a yearlong ethnographic study, Pritchard (2010) found that strategic (business) partners struggled to be released of their previous HR generalist role, but also noted that their continuous involvement in day-to-day HR activities was necessary until new ways of working were fully established. In any way, confronted with role ambiguity, business partners run the risk of taking on more operational work then their role prescribes, causing their strategic objectives to fall by the waste line. The lack of provision in the execution of operational HR services has

16 led some researchers to debate whether an additional leg, concerned with operational execution, would be necessary in order for the model to function optimally (Hird et al., 2009; Reilly, 2006). From the current literature it is not clear, however, whether organizations have elaborated on their three-legged design and if so, whether it has led to less role ambiguity.

2.4.2. Line managers

Although not explicit in the model, the strategic rise of the HR function has raised the question of whether, or not, ultimately, it is line managers who are responsible for putting HR into effect (Kulik & Bainbridge, 2006; Whittaker & Marchington, 2003). Ulrich (1998) argues that line managers have ultimate responsibility for both the processes and outcomes of the organization and thus play the most important role in fully integrating HR into the business. Purcell and Hutchinson (2007) identify line managers as essential to the successful application of HR practices and, in turn, to the way employees evaluate the presence and effectiveness of HRM practices. Using LMX (leadership member exchange) theory and social exchange theory, Purcell and Hutchinson (2007) stress that HRM practices and their effective implementation are strongly linked with line manager leader behaviour and that this mediates employee perceptions of, and reactions to, HRM practices. However, others ask if this is a realistic aim, if line managers have time, competence and opportunity to be the main delivery mechanism for HRM practices, given their short-term focus and business orientation (Whittaker & Marchington, 2003). Moreover, by continuously devolving HRM activities to the line, HR might relinquish an important source of power, that of control over HRM, while still being responsible for the effective management of employees (Sheehan, De Cieri, Greenwood & Van Buren, 2014). As previously highlighted, issues of powerlessness and marginality have plagued the HR function for decades (Legge, 1978; Caldwell, 2003). What are the implications for the credibility and power of the HR function when organizations adopt HR models that encourage the continuous devolvement of HRM activities to the line?

2.4.3. Increasing fragmentation

Another issue related to boundary disputes is the increasing fragmentation of the HR function (Hird et al., 2009). Driven by cost, value and service delivery, the HR function is increasingly forced to be segmented into specialist centers, subsumed by line management or outsourced to external parties. Gratton (2003), professor at London Business School and expert on HRM functional design refers to the increasing fragmentation of the HR function as ‘the humpty dumpty effect:’

While the HR function may at one time have been a rather inefficient, but pleasant and kindly “Humpy Dumpty,” it has now fallen from the wall and shattered into fragments on the hard pavement below. … This fragmentation of the HR function is causing all sorts of unintended problems. Senior managers look at all the fragments and are not clear how the function as a whole adds value

(Gratton, 2003, p. 18)

A consequence of the increasing fragmentation of the HR function is ambiguity surrounding responsibility and accountability. Ambiguity tends to surface especially in interdisciplinary issues that require the

17 involvement of service centers, specialists, line managers and business partners. Take for instance relocation services. Such cases often require the cooperation of various HR entities and the current theoretical model is unclear as to who can be held responsible. Furthermore, even among HR professionals it’s not always clear how responsibilities are distributed or even, where expertise resides, especially in large global organizations where a geographical dimension increases complexity. As for line managers, fragmentation results in maintaining relationships with various HR professionals. As Reilly (2006, p. 36) explains: “Managers may simply become frustrates that there’s no longer a one-stop shop to handle all HR matters. Instead, they must search the internet for basic information, ring a call center when they have a specific query, speak to an expert if the question is complex and take strategic advice from a business partner.”

The three-legged-stool model seems to portray a world without HR-HR and HR-line boundary issues and ambiguities even though empirical studies emphasise that this is the reality in organisations (Marchington, 2015; Gratton, 2003; Legge, 1978). Whereas the face value of splitting up HR activities across various specialist entities might be captures in terms of efficiency, the increased fragmentation of the HR function might also account for more complexity and ambiguity.

2.5. Concluding thoughts

The review presented here suggests that there are various tensions and ambiguities surrounding the adoption of a three-legged HR model. Many tensions resulting from role ambiguity have emerged from the intricacies of the model itself, such as the tensions surrounding the role of the business partner, the increasing fragmentation of the HR function and the continuous devolvement of HR responsibilities to the line. If so, it raises the question about whether the HR structures organizations currently employ in terms of the three-legged-stool, where devolvement trends are clear and boundary issues are not resolved, might account for the tensions between the different elements HR professionals traditionally deal with in doing HRM work. Or rather, is the HR structure organizations employ related to the tensions that actors working within these structures encounter? And, are these tensions inherent to the model? These questions will be considered through my empirical research by drawing on a paradox perspective inspired by recent theorising in organizational studies (Smith and Lewis, 2011). Paradox theory provides an opportunity for understanding dynamic complex organizational settings characterized by tensions and oppositions (Poole & Van de Ven, 1989; Eisenhardt, 2000; Lewis, 2000). Paradox theory is useful for systematic examination of the tensions arising from particular ways of configuring HRM activities and exploring tensions between elements that are related, persistently, over time. Before presenting the results of my empirical research, I will now review the literature on paradox within organizational settings.

18

3. Paradox theory

The notion of paradox has long captured the interest of philosophers, psychologists and more recently, organization theorists (Lewis, 2000). Most definitions of paradox address contradictory elements (tensions) that seem logical in isolation but absurd when juxtaposed (Poole & Van de Ven, 1989; Lewis, 2000; Smith & Lewis, 2011). In one of the most recent definitions, Smith and Lewis (2011, p.382) defined paradox as “contradictory yet interrelated elements that exist simultaneously and persist over time.” Some confusion has arisen from the inconsistent and sometimes synonymous use of the terms paradox, tension, contradiction and duality in different journal articles, however, a detailed exploration of terminology is beyond the scope of this thesis. For current purposes, it suffices to understand that paradoxes are composed of two contradictory elements and that they –unlike dilemma’s or either/or choices –signify two sides of the same coin simultaneously (Lewis, 2000).

3.1. Categorization of organizational paradoxes

Paradoxes emerge from contradictory elements (tensions), which come in many forms. In their categorization of organizational tensions, Smith and Lewis (2011) discern between paradoxes of belonging, learning, performing and organizing.

Paradoxes of belonging arise between the individual and the collective as actors strive for both self-expression and collective affiliation. According to Smith and Berg (1987), such paradoxes are driven and fueled by a self-referential cycle: strong and resourceful groups are formed through team members expressing their individuality, which in turn engenders group conflict. For instance, in their study of string quartets, Murnighan & Conlon (1991) found that actors joined string quartets, rather than large orchestras, to have more impact. However, quartets were successful if they continuously strived to achieve transcendence, integrating their music.

Learning paradoxes emerge from tensions between old and new, such as tensions between episodic and continuous change or radical and incremental innovation. They surface as organizational dynamics change and actors are confronted with tensions between the comfort of the past and the uncertainty of the future (Lewis, 2000). These paradoxes revolve largely around processes of sensemaking and transformation.

Paradoxes of performance arise from the plurality of stakeholders and conflicting managerial demands. Tensions surface as stakeholders seek divergent organizational success, such as an organization’s social responsibility program, in which both financial as well as social goals are to be attained (Smith & Lewis, 2011).

In turn, organizing paradoxes address the conflicting aspects of organizational design (Lewis, 2000). These tensions “depict the intricate nature of performance, empowerment, and formalization as managers seek to encourage commitment, participation, and trust and increase efficiency, coordination and security” (Lewis, 2000, p. 767). As an illustration, effective manufacturing often relies on organizational structures driving efficiency as well as flexibility (Adler et al., 1999).

19 In their framework, Smith and Lewis (2011) argue that tensions operate within as well as between these categories. For instance, organizing and performance tensions surface when the competing goals of driving performance and managing people are juxtaposed. “Focus single-mindedly on delivering profits, and you disenchant your workforce, destroying you capacity to drive needed strategic change. Concentrate solely on employees, and you slide into complacency, eroding your competitive vitality” (Eisenstat, Beer, Foote, Fredberg & Felming Norrgren, 2008, p. 50). As this thesis is primarily concerned with the organization of HR and the roles of HR professionals therein, focus will be on paradoxes of organizing, performing and adjacent tensions.

3.2. Responses to paradox

Responses to paradox revolve around the debate regarding the ontology of paradoxes. Some scholars argue that paradoxes are inherent features of a (social) system whereas others conceptualize paradox as social constructions that arise from people’s cognition and rhetoric. Clegg (2002, p. 2) addresses this division by discerning between representing paradoxes – social constructions of our lived experiences – and materializing paradoxes – residing in the material world, “where vibrant organizations, groups, and individuals change by simultaneously holding the two states.” From a material perspective, organizations and their various subgroups are regarded as complex human systems that are inherently paradoxical (Lewis & Smith, 2011).

In contrast, from a representing perspective, paradoxical tensions are seen as the constructions of dialectical minds. Elements only seem contradictory from the perspective of an observer, when in fact they are part of a larger whole (Ford & Ford, 1994). As Ford & Ford (1994, p. 760) clarify: distinctions “reside in the observer(s), not the observed.” Such social constructions help actors to make sense of complex realities, and once such frames of reference are entrenched, they become highly resistant to change (Lewis, 2000). So, in an effort to understand complex situations, actors tend to place boundaries around certain elements, to see the trees for the wood. But as such boundaries and contradictions become grounded in one’s perception, actors tend to regard their perceptions as distinct and immutable (Lewis, 2000). This ontological discussion has implications for strategies to managing paradox (Lewis & Smith, 2011). The literature tends to distinguish between strategies of acceptance and strategies of resolution.

In their seminal article on managing paradox, Poole & Van de Ven (1989) identified four strategic responses to paradoxical tensions: 1) opposition (acceptance), keeping opposing poles separate and studying their differences, 2) spatial separation, allocating opposing poles across different organizational units, 3) temporal separation, choosing and switching between opposing poles over time and 4) synthesis, adopting a new perspective that accommodates the opposing poles. In this typology, the opposition strategy is one of acceptance whereas the latter three are aimed at resolving the underlying tension.

Researchers calling for strategies of acceptance argue that learning how to ‘live with’ paradox offers a sense of freedom (Schein, 1990). It helps actors to avoid debates and potentially destructive conflicts and focus on their primal tasks. If we return to Mungham & Conlon’s (1991) study on group dynamics we find a relevant example. It turns out that members of string quartets tend to ‘play through’ rather than confront tensions, resulting in high performance and the avoidance of group conflict. Such a strategy can be characterized as passive acceptance. However, other researchers argue that it can be a powerful,

pro-20 active strategy as well. In their action research, Lüscher & Lewis (2008) show that helping managers understand tensions as paradoxical helped them in their sensemaking processes. Managers tend to regard paradox initially as a dilemma. However, as they became increasingly aware of the fact that choosing between opposing poles of a polarity would intensify the needs of the other, they began to adopt both/and thinking instead of either/or thinking, resulting in a more workable situation.

Other researchers, however, call for actors to resolve paradox. Here, resolution doesn’t mean eliminating a tension, but rather “finding a means of meeting competing demands or considering divergent ideas simultaneously” (Smith & Lewis, 2011, p. 386). While Poole & van de Ven (1989) suggest spatial separation, temporal separation and synthesis, these strategies are aimed at managing paradoxical tensions that are inherent to the system, as they involve physical separation of opposing poles and underemphasize cognitive efforts to resolve tensions. However, if we assume that paradoxes are (partly) socially constructed, then apparent contradictions might also be resolved by shifting one’s frame of reference (Smith & Berg, 1987).

In reality, not all actors adopt constructive responses to paradoxical tensions. Lewis (2000) argues that if paradox is treated with ambivalence or repression, actors may become trapped in vicious cycles that perpetuate and exacerbate the underlying tension. A relevant example that has plagued the HR function for years is the vicious cycle caused by the difficulty of identifying the unique contribution of HR in business outcomes. Legge (1978) has argued that since it is often unclear how the personnel department contributes to organizational success and how its unique contribution could be measured, personnel managers – uncertain about what tasks to prioritize (ambivalence) – are tempted to adopt a reactive attitude, waiting for matters to be passed on to them. As a result, the HR department was viewed as being engaged in relatively unimportant, necessary chores, which encouraged line managers to burden the personnel department with a variety of minor issues, spurring the vicious cycle.

Instead of being trapped in vicious cycles, Lewis & Smith (2011) propose a dynamic approach and argue that actors can fuel virtues cycles, which fosters creativity, continuous learning and long-term sustainability, by inherently accepting paradoxes and then vacillating between strategies of splitting/choosing (spatial and temporal separation) and strategies of synergy (accommodating opposing poles). The argument is that when actors assume that paradoxical tensions can and should coexist, they can explore the dynamics between the polarity, which increases their confidence in developing more complex and challenging resolution strategies.

In sum, paradox theory provides a theoretical lens through which tensions and subsequent responses can

be systematically identified and understood. In organizational literature, scholars have discerned between paradoxes of learning, belonging, performance and organizing and actors responding to paradox can enable either vicious or virtuous cycles.

21

4. Methodology

This study examines HR structures and tensions and their possible relationship within the context of several large companies based in the Netherlands. Given that this research aims to shed light on areas in which specific relationships are as yet unclear, I conducted qualitative research in the form of a comparative case study approach and followed Patton’s (1987) proposal to use semi-structured interviews to capture the complexity of the data. The case study approach has been described as “an empirical enquiry that investigates a contemporary phenomenon in depth and within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident” (Yin, 2009, p. 18). In a comparative case study approach, the case organizations are treated as experiments, with the findings of each case helping to confirm or disconfirm the findings drawn from the others (Yin 1994). In line with prescribed methods of comparative case study research (Eisenhardt, 1989; Yin, 1994), cases were selected using a theoretical sampling method and data was analysed using iterative, inductive analysis. The aim of this research was to balance the rich-story logic of each case with well-grounded theory in order to present compelling propositions that are supported by the rich narratives of the case participants (Eisenhardt & Graebner, 2007).

Conducting this study was part of my internship at Berenschot Consulting. It is also part of a larger project examining the intricacies of HR configurations within large organizations operating in the Netherlands. The project is a joint collaboration between Berenschot Consulting and the Amsterdam Business School.

4.1. Sampling method

The research was conducted among three large energy and utility companies based in the Netherlands: EnergyCo, UtilityCo and OilCo (pseudonyms related to the organizations’ core activities were used to represent the organizations for confidentiality). The cases were selected on the basis of theoretical sampling. I focused on case settings and respondents that could be potentially insightful (Eisenhardt, 1989). The three cases included represent contrasting settings within which the organization of HR and associated tensions could be explored. The case organizations all operated in the energy & utility sector, as to control for different institutional pressures that might influence HR configurations across industries (see Farndale & Paauwe, 2007). Thus, differences in configurations can be largely attributed to internal design choices.

4.2. Data collection

The participating companies were contacted through the combined networks of Berenschot and the Amsterdam Business School. I established one contact person per company who facilitated access to the interviewees we selected on a theoretical basis. Data was collected through semi-structured interviews with the following HR stakeholders: HR director on a national level, HR business partner, HR shared

22 services employee and several line managers. By utilizing multiple sources of information, internal validity was strengthened (synchronic primary data source triangulation) (Paauwels & Matthyssens, 2004). Interviews were held at the interviewee’s office location and each interview lasted one hour. Most of the interviews were carried out by two person teams: I would handle the questions, while the other focused on taking notes and raising issues that required clarification (Eisenhardt, 1989).

An interview protocol was designed to unravel the tensions encountered by HR professionals and stakeholders, but we did not ask informants to reflect on specific tensions. The interview covered general subject areas such as the structure and history of the organization’s HR function, relationships with team members, HR professionals and line managers, current and past challenges and in the case of line managers, questions regarding the effectiveness of the organization’s HR service delivery. Given the inductive nature of this research, we encouraged interviewees to wander freely in their answers and asked for clarification when necessary. Interviews were conducted in a single time frame and interviewees were encouraged to reflect on present challenges as well as past challenges, allowing us to collect present as well as retrospective data. In total, 5 interviews were conducted at EnergyCo, 7 at UtilityCo and 8 at OilCo, making for a total of 20 interviews. All the interviews were recorded on audio and then transcribed. An overview of interviews is listed in table 1.

Table 1. Interview data Case study Participants

Number of Interviewees by Function

Company Number of Employees Worldwide (2014) Total # HR director sr. HR manager Business partner HR professional (COE) Employee shared services Manager shared services Line manager UtilityCo 7,000 7 1 2 1 1 2 EnergyCo 2,688 (59,784)* 5 1 1 1 2 OilCo 92,000 8 1 1 2 1 1 2

23

4.4. Data analysis

In order to reduce the enormous volume of data I first drafted detailed descriptions of each case study, describing in broad outline the HR configurations employed and how they came about (Yin, 1994). The second level of data analysis consisted of cross-case analysis to compare systematically the configurations employed with the evidence from each case in order to allow for differences and similarities to emerge (Eisenhardt, 1989).

The emergence of tensions

I openly coded the interviews in Nvivo in order to identify patterns and variance in descriptions of tensions using verbatim indicators such as: tensions, dilemma, balance, duality, yet, but, either/or, one the one hand, on the other hand, etc… In 19 of the 20 interviews I identified references to tensions in how HR is configured.

In a second step I proceeded to pattern-matching (Yin, 1994), wherein I searched for relationships between and among these categories, resulting in several higher-order themes. Following methods of inductive analysis, I allowed concepts and relationships to emerge from the data, rather than being guided by pre-determined hypotheses (Strauss & Corbin, 1990). From this analysis, themes could be captured in three general paradoxes: the global stakeholder paradox, the efficiency/flexibility paradox and a power paradox. Data revealed that the paradoxes were specific to the HR configurations employed, as in, different paradoxes were identified in each case organization. The data analysis procedure that yielded these three paradoxes is depicted in figure 2. In this figure, it is also apparent which paradoxes emerged in which organization.

In a third step, I studied how HR professionals discursively respond to these tensions. Here, I used a framework of acceptance/resolution (Poole & Van de Ven, 1989; Smith & Lewis, 2011) to identify different modes of response to tensions. To this end, I sought verbatims in the case participants’ discourse that indicated a particular response mode.

24 Figure 2. Data structure: three paradoxes

25

5. Results

The purpose of this section is to present the results of the case based empirical analysis of HRM configurations and to interpret the results in terms of paradoxes evident in the three case companies. To achieve this, the section is organized as follows. First, I will describe how the HR function is organized in each case company. Second I describe how this came about and the reasons for the choices we see today. Finally, I provide results on what paradoxes are evident and can be constructed from the case informants’ discourse, based on the interview data. In this I follow the example of Andriopolous and Lewis (2009) on constructing paradoxes from discourse (see also Lüscher and Lewis, 2008).

5.1. EnergyCo

EnergyCo is the largest energy producer and provider in the Netherlands, providing over 2 million private and business customers with gas, electricity and energy services. It is headquartered in ‘s-Hertogenbosch and employs over 2,700 employees. EnergyCo operates several power plants in the Netherlands, which are fired using coal, biomass, wood gas and residual gases and steam from nearby waste-incineration plants and chemical producers. Furthermore, EnergyCo operates 12 windparks and several hydro power plants. In 2009, it became part of a German electricity and gas company (MotherCo) that is currently one of the five leading power suppliers in Europe, with more than 66,000 employees providing over 16 million customers with electricity and over 7 million customers with gas. EnergyCo’s executive board consists of a chief financial officer and a chief commercial officer that report though MotherCo Retail to the executive board at MotherCo.

5.1.1. EnergyCo HR function

EnergyCo restructured its HR function in 2014 and introduced a four-legged-stool, based closely on the work of Ulrich. As such, the case of EnergyCo provides insights of considerable interest for this thesis due to the recency of the restructuring and the conscious choice to model the new HR configuration a la Ulrich (as it were). According to the current EnergyCo HR configuration, HR activities are distributed across four functional entities. These are (1) an in-sourced HR services department, (2) centers of expertise, (3) HR advisors and (4) HR business partners.

HR services handles all of HR’s administrative processes. Most of the correspondence with

personnel is through e-mail. There are plans for employee service system (ESS) and a manager self-service system (MSS) but at the time of the study these were still underway. The HR self-services organization reports hierarchically to their HR services counterpart at MotherCo, which is headquartered in Essen, Germany.

EnergyCo employs two centers of expertise: people development and compensation & benefits. A third COE, HR control, is currently being introduced to help HR with making more fact based decisions. Its core activities have not been specified specifically as of yet, ‘it’s still a rather vague COE at the moment’ (HR director, EnergyCo). The local COEs report to their parent COE in Germany. Present focus is on reducing the high number of employment condition regulations that are making HR policies highly bureaucratic.

26 EnergyCo makes a distinction between HR business partners and HR advisors. HR business partners have a highly specialized role and serve the leadership teams and help businesses realize success through use of human resources. HR Advisors serve middle management and are concerned with operational management such as the management of personnel in- and outflow. HR advisors report to an HR operations manager. HR business partners and HR advisors jointly constitute the local HR

organization, which is coordinated by EnergyCo’s HR director. The decision however to distinguish the

local HR organization in terms of HR business partners and HR advisors is an important design element in the four legged model at EnergyCo.

5.1.2. Emergence of the present HR structure

How did the current EnergyCo HRM configuration come about? The most important point to emerge from the results is that the current model was explicitly and consciously introduced in 2014 as part of MotherCo’s efforts to drive uniformity and efficiency across the support organizations, including HR, of all of its subsidiaries. Besides scale efficiencies, the model was also introduced to improve communication and to support strategic decision making for HR or, in the words of the HR Director, to “increase the effectiveness of internal resourcing, to understand each other better, to speak the same language and to make integrated strategic choices” (HR director, EnergyCo).

With regard to the reasons as to why this specific model was introduced, these seem to be largely institutional in nature, in the sense of driven by mimetic isomorphic institutional pressures (DiMaggio and Powell 1993) or, put simply, the pressure to copy perceived best practices. During the interviews, the Ulrich model was framed as the most logical and perhaps even the only viable option.

So why this particular model? I think it’s a model that is being implemented by many multinationals. You regularly hear about it here and there. I was recently visiting ChemCo and they are literally configuring their HR function in the same way, albeit on a larger scale. I think it’s an accepted model for structuring HR within multinational companies.

(HR director, EnergyCo) It was about time that we used Ulrich to organize HR.

(HR business partner, EnergyCo)

These utterances from the data highlight the extent to which Ulrich’s approach to HR is perceived to be institutionalized and it suggests that adoption is driven by mimetic isomorphic pressures (Dimaggio & Powell, 1983).

With the introduction of this model, some detailed aspects are worth explaining in more detail. Firstly, the adoption of the four-legged model and the separation of HR advisors and HR business partners was implemented to the effect that HR business partners were stripped of their daily operational

27 HR business partners have been stripped of their hierarchical power over HR advisors, so that they can focus more on content and be more strategic.

(HR business partner, EnergyCo)

However, the outcome is that business partners now seem to be forced into a ‘sales’ role in which they have to sell business needs to COEs, HR services and even their own HR organization. Whereas HR advisors used to report to a business partner, they now report to an HR operations manager. The HR business partners have to lobby to have their operational HR work executed.

It’s almost approaching a sales function, it’s always been like that with regard to the business, but now it also requires stakeholder management towards my own HR organization.

(HR business partner, EnergyCo)

What further complicates this situation is that the COEs and HR services are now reporting to the MotherCare HQ in Germany. The result appears to be that this has decreased their ability to understand the relevance of certain issues for the business. As a business partner explained: “Now I have to convince somebody [at the COE] who is thinking: help me remember, which business was this again and what are they doing exactly?” (Business partner, EnergyCo).

5.1.3. Emerging tensions

When analysing the discourse on the challenges emerging within EnergyCo’s HR configuration, two specific themes can be induced that suggest the presence of two core HR elements in opposition and at the same time co-existing simultaneously. These are pressures for localization versus pressures for globalization.

I like to refer to it as ‘building the plane while flying’; on the one hand we are focused on supporting our local businesses, which are under a lot of pressure, and simultaneously we are reorienting ourselves internationally. This combination is an interesting one.

(HR director, EnergyCo)

Case informants stress the importance of both globalization as well as localization. On the one hand, the underlying reasons for emphasizing globalization appears to revolve around increasing the consistent application of policy across businesses.

When you are serving different businesses, you would like to have that employment condition regulations are applied consistently across these businesses. … If you don’t, you run the risk that people start improvising with the conditions, resulting in a number of exceptions and we cannot possibly explain that to each other.