The devolution of HR Tasks & Responsibilities to Line Management

(1)

THE DEVOLUTION OF

HR TASKS &

RESPONSIBILITIES TO

LINE MANAGEMENT

Jeremy Bernard Buitenhuis, 10278664 7 december 2017

Executive Programme in Management Studies – Strategy Track University of Amsterdam

(2)

1 Statement of Originality

This document is written by Student Jeremy Buitenhuis (10278664), 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)

2

Abstract

This study focusses on the transfer of HR responsibilities from HR professionals to line managers, also known as ‘the devolution of HR tasks & responsibilities to line

management’. In practice line managers didn’t seem too happy with taking on HR responsibilities. Through this study I try to understand how devolution affects line managers. The main research question was:

Does devolution of HR tasks to line managers increase their effectiveness?

This quantitative study explores devolution within a social firm through a literature review, followed by a theoretical model. A survey sent to 110 line managers within the social firm. The results were analyzed using a regression analysis. The findings show that HR role ambiguity is a predictor for line management effectiveness.

Furthermore, this study contributes to existing research by providing a new definition of devolution, a cohesive model of devolution that can be built upon and the creation of

variables that can be used in future studies. This study also adds to existing research from Gilbert et al. (2011) and stimulates the further use of role stressors in business related research.

(4)

3

List of tables & figures

Figure 1: Model of impact of devolution of line manager effectiveness ...27

Table 1: Correlations & Reliability scales ...37

Table 2: Descriptive Statistics...37

Table 3: Regression results for Line manager effectiveness ...44

Table 4: Regression results for Line manager effectiveness score ...45

(6)

5

Introduction

After I graduated in HR, I started working as an HR administrator in an internet company. During my time there I learned a lot from my manager and HR colleagues as they were very experienced in operational HR. One of the things they taught me was the importance of empowering line management. In their view, empowerment meant that the line

managers would take responsibility for all employee related issues. Whether it be

managing the vacation schedule or taking the lead in disciplinary matters. To my surprise, I realized that line management wasn’t all too happy with having that responsibility and often complained about the amount of time HR responsibilities took. However, at that time, I didn’t stand still to ponder why.

After a few years I moved on to another company in a different sector – it was a social employment firm. Due to changes in the social employment legislation, the company was forced to reduce its costs, which I’ll get in to later. One of the things they changed, was the HR organization: they increased the HR responsibility of line managers in order to cut costs by reducing the number of HR personnel, organize the HR processes in a more efficient manner and position HR to be a strategic partner. Again, after a few months in the ‘field’ it became apparent: line management again didn’t seem too happy about it. Their primary reason for disliking the new situation was the amount of time that their new HR

responsibilities required and how they struggled with meeting their performance targets. This made me wonder: does the increased responsibility hamper the ability of a line manager to perform? And if so, does that impact the organization in a negative way? As I was in need of a research topic for my thesis, this ‘empowerment’ of line management really piqued my interest. I found it strange that HR specialists were very content with this

(7)

6 idea and line managers weren’t. I had worked in the Dutch military and the concept of leadership was drilled into my head. Most military organizations are hierarchies. Officers and Non-Commissioned Officers (NCO’s) are trained in all aspects of leadership: they are taught to take care of their team and achieve results through the use of their team. Thus, I believed that being a team leader in a regular company meant taking care or being

responsible for your teams wellbeing and general employment affairs. So why wouldn’t line management be happy with the opportunity to really manage and lead a team?

Since I wanted to know more about this phenomenon, I decided to focus on this subject for my thesis. After reading a few research articles I learned that this phenomenon of

empowering line managers or increasing their HR responsibilities was named: the devolution of HR tasks and responsibilities to line management.

Following the strategy track of business studies, I wondered what the impact of devolution on the firm could be. Does it provide any strategic value for the firm? From a contingency point of view, the most efficient form of an organization is dependent on its context (Blok, 2013). If so, under which circumstances would it make sense for an organization to devolve HR tasks and responsibilities to line management? And what would contribute to a

successful implementation of devolution? In the literature review I explore the first question and through the survey I try to answer the last two questions.

Most literature on devolution is written from an HR perspective (Gilbert, de Winne, & Sels, 2011). Neglecting the organization or line managers involved. That is why I’ve chosen to focus this study on the line manager. Another fact is that most studies are either

qualitative case studies or surveys completed by high level HR professionals, so this study contributes to existing research by performing a quantitative survey focused on the line manager.

(8)

7 I try to understand the relation between devolution and line manager effectiveness through the use of the attribution theory. Attributions are individuals’ explanations for the causes of their successes and failures (Martinko et al., 2011). Understanding how these explanations affect devolution, could provide valuable insights on devolution in practice.

As I’ve mentioned before I started working for a company who had just devolved HR tasks and responsibilities. The company is named Pantar. In order to increase the

generalizability of the study, it would be best to send the survey to a large number of companies and line managers, however this could potentially lower the strength of the study as there is no solid and tested model. It would also take a lot of time to find the appropriate companies and respondents. That’s why I’ve chosen to perform the survey within Pantar. This way I can test the model and provide a more solid base for future research.

The above leads to the main research question:

Does devolution of HR tasks to line managers increase their effectiveness?

In this thesis, I start by explaining what devolution is, why organizations devolve HR tasks and responsibilities and who are involved with devolution. During this literature study the hypothesis are introduced. The literature study ends with a summary of the hypothesis and a proposal for a model combining various elements from the literature. In the next

paragraph I explain the research method, including the sample and a detailed explanation of the variables. Here I also explain the statistical procedures used in this study.

The second part of the thesis begins with the statistical results and reviews the outcomes of the various hypothesis. I then draw my conclusions based on these results, followed by a discussion and my recommendations for future research.

(9)

8 For now, I would like to thank my readers, for taking the time and read this thesis. It has been a trying endeavor to go through the various research steps and finish it. I’ve learned that quality is achieved through collaboration and that such collaboration can come in many different forms.

My partner who at times dragged me to the computer to work continue my work. My own team manager who kept supporting me even though time kept passing by. Old friends that sparred with me on the statistical procedures and new friends that pointed me in the right direction. I hope this thesis reflects not only my own efforts, but their efforts and support as well.

(10)

9

Pantar

Pantar is a non-profit organization in Amsterdam. A few years ago there was legislation on how to provide work to people who, due to physical or psychological reasons, have

difficulties finding a job. This was called ‘social work’ and people who qualified for social work could apply for a job at so-called social work organizations. Pantar is one of these social work organizations in the Netherlands who provides work to people living in Amsterdam and Diemen (Pantar Corporate website, 2017).

In 2015 the laws on social work changed and people no longer received a social work qualification. The new laws wanted to increase the responsibility of regular firms by encouraging them to create jobs for people that previously qualified for social work. This change had a huge impact on social work organizations who depend on the subsidies they receive for each employee. The municipality of Amsterdam lowered the subsidy per employee and where their previously was a steady stream of new applicants, there no longer were joiners with a social work qualification. Instead, people joined with a new qualification, which came with a subsidy which was a percentage of their annual income based on the work they should be able to do when compared to a regular employee. For example, a new joiner that would work as a cleaner would come with an indication stating that the new employee could do 60% of the work compared to a regular employee and so Pantar would receive a subsidy of 40% of the annual gross income of that employee (Pantar, 2013).

These changes meant that Pantar had to change their business model in order to remain financially solvent. This resulted in a reorganization of the staff departments, which also affected HR (Pantar, 2013). In order to do the same amount of work with less people,

(11)

10 management decided to transfer a part of the HR responsibilities to line management and give HR a more advisory role.

Pantar has around 3000 employees and is divided in four different units: Production, Services, Public Area Services and Secondment services. The management organization consists of team leaders, department heads, business unit managers and the directors, which are around 130 people. In addition, there are ca. 200 staff employees working in HR, finances, communication, facility, etc. (Appendix 1 Pantar Organogram).

Pantar’s main client is the municipality of Amsterdam and they maintain public locations, rural areas, provide postal, catering and cleaning services. They also have other clients and perform manual production activities such as packaging at the Pantar office location or directly at the location of the client. Following the new laws on social work, Pantar no longer just provides work, they focus on training and developing the necessary skills of their employees in order to qualify for regular work (Pantar Corporate website, 2017).

(12)

11

Literature review

Theory on devolution

Blok (2013) notes that very little HR literature uses academic theory. Indeed, in my literature review I didn’t come across any references of models or theory related to devolution. However, devolution can also be seen as decentralization. As early as 1945, Hayek wrote about how decision-making power should be either centralized or

decentralized based on how knowledge is utilized within the organization. So it is

interesting to learn that elements from organizational literature aren’t used in conjunction with a HR perspective. Strikwerda (2003) wrote about what parent boards need to devolve to subsidiary board in order to cope with the changing economic environment. Although this is not the same as devolving HR tasks and responsibilities, the underlying principles might be applicable in both cases. Meaning that HR should consider their own organizational structure before devolving responsibilities to line managers and how they can best facilitate conditions for success of such devolution. Such support could be given through the use of platforms (Zuboff, Maxmin, & Hutton, 2004).

The emergence of individualism and changing economic circumstances place an emphasis on knowledge and the knowledge worker. A firm’s intangible assets, including tacit

knowledge, are provided by these skilled workers. They provide the firm the capabilities it requires to gain a competitive advantage. Thus, when considering devolution, the question is how to best facilitate these knowledge workers so their knowledge can be retained, shared, distributed and used towards achieving the company strategic goals.

(13)

12 In the case of Pantar knowledge plays an interesting role. In the current situation there is no specialized workforce or does it utilize the knowledge of its employees in order to gain a competitive advantage. When considering the employees at the lowest level this makes sense as they are more focused on survival instead of self-determination (of course there are exceptions). However, this does not apply to line managers and staff specialists. They’ve accumulated knowledge and experience over the years that might be utilized. At this moment in time there are no clear signs of how this knowledge put to good use within the organization or how it contributes towards gaining a competitive advantage.

What is devolution?

The devolution of HR responsibilities from HR managers to line managers is both a growing and global trend (Perry & Kulik, 2008). In the 90’s strategic integration was one of the key selling points of the HRM concept seeking integration with business strategy

(Cabral-Cardoso, 2004). Devolvement was also growing more popular as it emphasizes the short line of communication between the line manager and the employee (Budhwar, Strategic

Integration and Devolvement of Human Resource Management in the UK Manufacturing Sector, 2000). There is a trend of devolvement happening in most European countries, which could indicate that organizations consider devolvement to be beneficial to the organization (Cabral-Cardoso, 2004).

Budhwar (2000) defines devolution as the degree to which HRM practices involve and give responsibility to line managers rather than HR specialists. Most authors describe

devolution in a similar matter (Gollan, Kalfa, & Xu, 2015) (Brewster, Comparative HRM: European views and perspectives, 2007). Krulis-Randa (1990) in (Harris, Doughty, & Kirk, 2002) expands the definition by describing the role of the HR function: HR specialists need

(14)

13 to support and facilitate line management with their HR responsibilities and not control them. Another definition is given by Torrington & Hall (1996) in (Brewster, Comparative HRM: European views and perspectives, 2007), who define devolution as: “Devolution is indicated by the co-determination of HR strategy with the line management rather than being the province solely of the personnel people.” This definition adds to the others, because it mentions HR strategy and by doing so it places devolution within a broader organizational context.

The word devolution is a combination of ‘evolution’ and ‘devolve’. This relates to the view of the HR function evolving itself as a strategic partner and in that process certain HR

activities are moved to line managers. The role of HR as a strategic partner is lacking in the definitions mentioned above and therefore I propose a new definition:

Devolution is the process of redistributing HRM tasks and responsibilities between line managers and HR specialists, with the aim of integrating HR strategy with business strategy.

Reasons to devolve

According to Budhwar (2000), a direct line between line manager and employee affects efficiency and motivation because: a) local managers can respond to local problems and solutions faster (Perry & Kulik, 2008) b) certain issues are too complex for top managers to comprehend c) it can help to reduce costs (Harris, Doughty, & Kirk, 2002) d) it can help to prepare future managers (by allowing middle managers to practice decision making skills). Renwick (2000) adds: e) increase local management accountability, and f) introduce policies more appropriate to the locality.

(15)

14 With the exception of the reason that certain issues are too complex for top managers to comprehend, the list sums up most apparent reasons for devolution. However, it seems too easy to say that certain issues are too complex for top managers to understand. Top

managers have received more training (such as traineeships or management development programs), higher base education and have more experience in terms of years spent as a business leader. Thus I believe it less probable that they cannot understand certain issues. Budhwar (2000) assumes there is a positive relationship between integration and

devolvement, because as HRM becomes more and more strategic, personnel specialists will have less time for the traditional routine type of HR activities. These activities will then be devolved to, and performed by, the line managers. There is partial support for his theory, but he concludes his introduction by stating more empirical studies are needed on this topic.

His mixed-methodology approach of survey data and interviews showed no clear distinction or relation between integration and devolvement, except that British managers feel the need to increase both. His study was conducted in a specific sector and its conclusions are therefore not generalizable, it does help to get a better understanding of the concept of devolution.

Renwick (2000) contradicts Budhwar’s views by stating that the devolvement of HR tasks and responsibilities to line management is the result of the evolving function of line managers and line manager behavior as they became responsible for: a) setting objectives b) organizing c) motivating and communicating d) measurement and e) people development (Drucker, 1974 in Renwick, 2000). It seems that the line manager plays a more important role within the organization (Holden & Roberts, 2000) (Renwick, 2000). For this study it is important to note a different viewpoint on why devolvement is occurring in the first place:

(16)

15 to organize the HR and line function in such a way that it provides the most added value in reaching company goals and gaining a competitive advantage.

Another reason mentioned in the literature is that HRM responsibilities are shared between HR specialists and line managers to better integrate HR and business strategy (Harris, Doughty, & Kirk, 2002). Devolution is seen as a vital part of Strategic HRM (SHRM). SHRM focusses on several issues, including the fit between HR practices and organizational strategic goals, the integration of HRM in organizational strategic management, the involvement of the HR function in senior management teams, the

devolvement of HR practices to line managers and the value that is added to organizational performance by HRM. (Andersen, Cooper, & Zhu, 2007) (Gollan, Kalfa, & Xu, 2015). To further integrate HR in strategic decision making, the responsibility of routine execution and administration of the HR practices should be delegated to line managers, as they have direct and frequent contact with employees and a capacity to understand, motivate, control and respond quickly to employees (Andersen, Cooper, & Zhu, 2007) (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015). In addition, devolution also contributes to the effective implementation of HR strategy through line managers (Andolsek & Stebe, 2005).

The reasons for devolution seem to involve different sides of the organization. It is theorized to positively impact line managers, realize cost reductions and stimulate business strategy integration. However, there is little empirical evidence of these effects. Furthermore, although devolution is expected to positively impact line managers, its purpose is to help better the strategic HR role within organizations. In order to learn how devolution affects line managers, our first hypothesis is:

(17)

16

H1A: The (amount of) devolution of HR tasks and responsibilities has a negative influence on line management effectives.

Not all organizations devolve HR tasks and responsibilities to line management. Organizational factors believed to influence the decision to devolve are: a) size of the

organization, b) the extent of unionization, c) strategic role of HR specialists, and d) location of collective bargaining as well as the strategic nature of the HR department (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015). Although it is unclear why, Brewster et al. (2015) found that organizations with a HR department involved with strategic decision making are less likely to have devolved HR responsibility to the line. This could indicate that devolution might not contribute to the strategic position of HR at all.

HR & Organizational performance

Organizational performance is defined as: ‘the comparison of the value produced by a company with the value owners expected to receive from the company’, this can be defined in HR-related outcomes such as turnover, absenteeism, job satisfaction, commitment, etc. or financial outcomes such as profits, sales, return on assets or capital market outcomes’ (Darwish & Singh, 2013).

It is estimated that a firm’s intangible assets, e.g. human capital, information capital and organization capital, may represent over 75 percent of its value (Kaplan & Norton, 2004). In order to create value from intangible assets, they need to be aligned with strategy. Kaplan & Norton (2004) wrote that activities such as HR, should be integrated with the

organization in order to maximize their contribution towards the value creation process. The amount of overlap and integration between intangible assets make it impossible to express the ‘tangible’ amount of value that is contributed by HR or other intangible

(18)

17 activities. Although Kaplan & Norton (2004) do provide a tool that helps to quantify the contribution of the individual and helps with the human capital and strategic alignment. According to Andersen et al. (2007) there is a positive relation between strategic HR alignment and the perceived financial performance of the firm. In general, there is a belief that HR is connected to firm performance, but it is unclear how (Gollan, Kalfa, & Xu, 2015). A possible explanation is given by Sanders & Yang (2016) who wrote that HR can

contribute to organizational performance by motivating employees to adopt desired attitudes and behaviors that, in the collective, help to achieve the organization’s strategic goals. It can do so by helping to create an organization image with which employees can identify. Considering human nature, identification inspires individuals to collaborate and accomplish feats not possible for the individual (Simon, 2002).

Besides financial firm performance, employees and intellectual property can give organizations a competitive advantage (Lawler III & Mohrman, 2003). In a survey conducted by Darwish & Singh (2013) among HR directors, the directors state that they believe that employees are key resources who can enable businesses to gain a competitive advantage.

Often line managers function as a first point of contact for employees. This means their effectiveness or performance can have a direct impact on their employees’ motivation, turnover, absenteeism and an indirect impact on organizational performance.

Who are the main players in devolution?

There are two important players involved with devolution: the HR department and its specialists and the line management within an organization. Other stakeholders such as the board of directors are important when it comes to making the decision to devolve or not,

(19)

18 but are not concerned with daily operational HR tasks and responsibilities. Employees are seldom mentioned in literature about devolution.

The HR department or HR in general are a much debated subject in research and

management literature (Cappelli, 2015). Not always seen as a positive element within the organization. HR is accused of having no connection with the business, they are slow to act and they create unrealistic policies (Whittaker & Marchington, 2003). With the integration of HR and business strategy, devolution could be an effective transformation strategy that HR could use to increase the image of HR among line managers (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015).

However, when it comes to devolution the HR department plays a crucial role. As HR tasks and responsibilities are transferred, the HR department is need to also transfer the

knowledge of such duties. In addition to knowledge transfer they need to change their service model to facilitate, support and give advice to line managers (Harris, Doughty, & Kirk, 2002). As the gap between HR and business strategy shortens and devolution

increases, HR takes on the role of strategic partner and coordinator for the activities of line managers (Cabral-Cardoso, 2004). In order to be successful HR needs to develop new competencies such as policy creation, e-HRM, and expert advice & consultation (Whittaker & Marchington, 2003). Unfortunately not all organizations provide the HR support needed by line managers to successfully accomplish their HR responsibilities (Perry & Kulik, 2008). Based on this the second hypothesis is:

H1B: The influence of devolution on line manager effectiveness is more negative among managers who receive a high amount of personal HR support.

(20)

19 In the case of labor legislation, HR plays a very important role to protect the organization from costs incurred by errors or lawsuits (Harris, Doughty, & Kirk, 2002). Harris et. al. (2002) mention that an increase in employment legislation also demands increase in

knowledge and expertise from those who are handling employment issues. It is difficult and potentially costly to transfer such knowledge to line managers. In order to prevent such costs, HR creates rigid employment processes and monitors these processes intensely. As a result HR seems bureaucratic, inflexible and doesn’t seem to care about operational

problems (Harris, Doughty, & Kirk, 2002). In that case , it may seem wise keep the

responsibility for these issues within a trained and specialized HR department (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015).

The (role of the) Line manager

The other important players within devolution are the line managers. Line managers play a central role within the organization (Holden & Roberts, 2000) (Renwick, 2000). They

organize the work, implement strategy, manage disputes and motivate staff (Holden & Roberts, 2000). Because line managers are concerned with implementing the corporate strategy, they play a very important role in the change process. They are tasked with relaying the order from above and receiving the feedback from the operation. Because of their ‘stuck in the middle’ position they may experience tension, conflict and stress. Holden & Roberts (2000) note that because of this, they experience difficulties with performing effectively, especially in conjunction with the devolved HR functions. Holden & Roberts (2000) wrote that line managers became more and more responsible for recruitment, disciplinary functions, most training and development issues, performance appraisal, health and safety monitoring and record keeping.

(21)

20 Definitions found described line managers as: ‘a manager who deals with any of the

functional areas other than HR’ (Srimannarayana, 2010), ‘covers those people in the

management who are directly responsible for production or services’ (Brewster, Brookes, & Gollan, 2015) and ‘the first level of management to whom only non-managerial employees report’ (Gilbert, de Winne, & Sels, 2011). It is interesting to see that these definitions would lead to small differences in the target group, such as excluding HR or staff departments. In essence any manager, whose HR tasks and responsibilities increase, could be more or less effective as a result of the change in amount of HR tasks and responsibilities. Depending on the size of the organization, there could be a several layers of management who are not considered part of the senior management or strategic apex. Pantar for example, has such an additional layer of management. These ‘department heads’ are responsible for

production and manage the line managers directly supervising non-managerial employees. As they are not part of the strategic decision making process this layer of management is also considered part of line management in this study.

Not all line managers have the necessary skills and competencies to perform their HR duties (Whittaker & Marchington, 2003). According to Perry & Kulik (2008), the skills and competencies of a line manager are an important factor in devolution. Training helps to build a competency base for line managers which they can utilize to perform their HR tasks successfully. Insufficient training offered to line managers greatly undermines the capacity of line managers to perform HR activities effectively (Andersen, Cooper, & Zhu, 2007).

H1C: The influence of devolution on line manager effectiveness is positive among managers who received a high amount of HR training.

When untrained even the most routine HR tasks, such as writing letters or taking the appropriate first disciplinary steps, can become very time consuming. In practice, line

(22)

21 managers are often caught between their operational and people management

responsibilities. Especially when it comes to prioritizing between the two, operational demands are regarded as more important by line managers, which resulted all too often in a minimalist approach towards people management issues (Harris, Doughty, & Kirk, 2002). According to Whittaker & Marchington, as long as people management goals are not

embedded within the performance targets of the line managers, people management responsibilities will always be second (Whittaker & Marchington, 2003) (Perry & Kulik, 2008). However, lack of training and competence could potentially result in a

disproportionate part of their time spent on HR tasks.

H1D: The influence of devolution on line manager effectiveness is lower or neutral among managers who already spend a high amount of time on HR tasks.

Besides managing their employees, line managers also implement policies designed by HR (Harris, Doughty, & Kirk, 2002). Some of these policies are meant to make sure that internal policy meets the external regulatory requirements. Yet, the difference between designed and implemented policies affects value creation as true implementation may differ from the intended implementation (Renwick, 2003) in (Gollan, Kalfa, & Xu, 2015). In

addition to said difference, Perry & Kulik (2008) mention that line managers seldom apply HR policies in a consistent manner.

Policy implementation increasingly goes through software, especially HR-software. According to Conway & Monks (2010), problems with policy implementation may occur when line managers are not involved with policy creation and only responsible for policy implementation. For line management involvement in HR to be successful, there are three important factors to consider: line management needs to understand the rationale of their

(23)

22 involvement, belief in the value of their involvement and HR role clarity & capability

(Conway & Monks, 2010).

‘It is difficult to see how managers, whose responsibility is to get things done through others, could not have HRM responsibilities’ (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015). Although this seems logical, there are a lot of unknowns when it comes to the details of that responsibility. Yet, managers acknowledge that HR is an important part of their role (Conway & Monks, 2010). Line managers have always had HR responsibilities (Gilbert, de Winne, & Sels, 2011). Solving HR problems at a lower organizational level, created more responsible line managers and improved the efficiency of employees (Perry & Kulik, 2008). If devolution is welcomed by experienced managers, because it provides more responsibility and decision making power, it should have a positive effect on their effectiveness. Leading to the next hypothesizes:

H1E: The influence of devolution on line manager effectives is positive for line managers who have a high amount of experience as a line manager.

H1I: Devolution leads to increased feelings of responsibility & accountability which in return leads to higher line management effectiveness.

The staff-line relationship

When personnel specialists try to change their role it could lead to conflict, especially when senior management views the role of HR differently (Renwick, 2000). However, Renwick has a mostly negative view on the relationship between HR and line management. He believes that this is caused by HR not delivering or adding value (Renwick, 2000). These sentiments tend to pop up every now and then, with arguments such as HR makes us

(24)

23 perform tasks we dislike, like documenting problems with employees or tells we can’t hire who we want (Cappelli, 2015).

The negative views are offset by progressive views such as more integrated approach of personnel management. In which HR is treated as a shared responsibility between line managers, HR professionals and employees (Srimannarayana, 2010), (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015). In such a relation HR is seen as a strategic partner that can contribute to the development and implementation of corporate strategy (Lawler III & Mohrman, 2003). Especially in organizations that focus on competencies and knowledge management, HR is more likely to be a strategic partner (Lawler III & Mohrman, 2003). Being a strategic partner and devolving HR tasks to line management requires trust and HR tasks need to be transferred without HR adopting a controlling role (Lawler III & Mohrman, 2003).

HR information system

A large part of the devolved tasks has an administrative nature. However, the impact of modernization and digital processes is rarely discussed in devolution literature. This is especially remarkable as more and more HR tasks are being digitalized through digital HR systems (Blok, 2013). Time potentially being a bottleneck for both HR specialists trying to adopt a more strategic role and line managers coping with the added strain of HR tasks and responsibilities. Much administration can be done through Management Self Service and Employee Self Service. This could save time of the HR specialist and line manager (Lawler III & Mohrman, 2003).

In his book about Self Service Centers Strikwerda (2010) discusses the case of KPN. KPN changed the HR service structure when they needed to severely reduce costs. Switching to a

(25)

24 e-HRM system enabled them to achieve significant cost reductions through the removal of paper messages, manual calculations and physical filing. It also provided line managers with access to relevant information about their employees (Strikwerda, Shared Service Centers II, 2010).

A sound digital HR infrastructure should lead to higher line management effectiveness, thus leading to the next hypothesis:

H1F: The influence of devolution on line manager effectiveness is positive among managers who are satisfied with the HR information system.`

HR role stressors

Role stressors, or HR role stressors, find their origin in role theory. Unclear role

expectations or incongruities, cause negative attitudes or bad performance of those who fulfill the role (Gilbert, de Winne, & Sels, 2011). Gilbert et al. (2011) wrote: “An

organizational role can be defined as the set of activities that are expected to be performed by an employee occupying a certain position in the organization (i.e. the role incumbent) (Kahn et al., 1964)”. Various studies portrait line managers who mention that that workload is increased and has become more complex. According to Gilbert et al. (2011) there is not a lot of research about these role stressors and how they are related to

devolution. This study could contribute towards gaining a better insight in that relation. In their study, Gilbert et al. (2011) research the impact of different antecedents (such as HR support, HR task requirements and line manager competency) on HR role ambiguity and HR role overload. They do so based on the idea that role stressors lead to negative work attitudes, less wellbeing and reduced individual performance, role stressors might impact the implementation of HR by the line manager (Gilbert, de Winne, & Sels, 2011). HR role

(26)

25 ambiguity means: “uncertainty about the scope activities and responsibilities that are expected in the HR role” (Gilbert, de Winne, & Sels, 2011). HR role overload is defined as: “Being unable to complete all assigned HR tasks effectively due to time limitations” (Gilbert, de Winne, & Sels, 2011). Problems with role overload emerge when HR tasks are added to already overburdened work commitments (Conway & Monks, 2010).

The negative impact of HR role stressors on line managers is based on earlier studies of role theory. In order to add to existing research, in this study the relation between HR role stressors and line manager perceptions is tested:

H1G: Devolution leads to increased feelings of HR role overload which in return leads to lower line management effectiveness.

H1H: Devolution leads to increased feelings of HR role ambiguity which in return leads to lower line management effectiveness.

Attribution theory

At the core of line manager effectiveness lies the behavior of the line manager and its impact on situational outcomes. For example, poor team member performance might be addressed through punishment or reward, which in return may or may not reinforce the desired performance. Attributions are individuals’ explanations for the causes of their successes and failures (Martinko et al., 2011). These insights could help explain how different factors influence the success or failure of devolution in practice. Also as the

attribution theory has not been widely used in organizational science (Martinko, Harvey, & Douglas, The role, function, and contribution of attribution theory to leadership: A review, 2007), this research could contribute to the organizational science community by exploring the attributions line managers make about the impact of devolution on line manager effectiveness.

(27)

26

Research question and hypotheses

Main:

 Does devolution of HR tasks to line managers increase their effectiveness?

Hypotheses:

 H1A: The (amount of) devolution of HR tasks and responsibilities has a negative influence on line management effectives.

 H1B: The influence of devolution on line manager effectiveness more negative among managers who receive a high amount of personal HR support.

 H1C: The influence of devolution on line manager effectiveness positive among managers who received a high amount of HR training.

 H1D: The influence of devolution on line manager effectiveness is lower or neutral among managers who spend a high amount of time on HR tasks.

 H1E: The influence of devolution on line manager effectives is positive for line managers who have a high amount of experience as a line manager.

 H1F: The influence of devolution on line manager effectiveness is positive among managers who are satisfied with the HR information system.

 H1G: Devolution leads to increased feelings of HR role overload which in return leads to lower line management effectiveness.

 H1H: Devolution leads to increased feelings of HR role ambiguity which in return leads to lower line management effectiveness.

 H1I: Devolution leads to increased feelings of responsibility & accountability which in return leads to higher line management effectiveness.

(28)

27

Conceptual model

(29)

28

Method

This section details the base and structure of this empirical study. It begins by explaining the choice to conduct a case study. Then it moves on to data collection methods for the literature review and the sample selection. It finishes with a thorough review of the chosen variables and the intended statistical approach to test the hypothesis. Both the Dutch and English versions of the questionnaire can be found in the appendix chapter 2 and 3.

Case study

A case study is an approach capable of examining simple or complex phenomenon, with units of analysis varying from single individuals to large corporations and businesses; it entails using a variety of lines of action in its data-gathering segments and can

meaningfully make use of and contribute to the application of theory (Cresswell,2007; Yin, 2003a) in (Berg, 2009).

A large-scale survey would be ideal to gain a broad perspective on the subject of devolution. Most research on devolution relied on case study research designed to identify the costs and benefits of devolution (Perry & Kulik, 2008), so a large scale survey would contribute to the field of research. However, such a survey requires identifying firms who have recently devolved HR tasks and responsibilities. In order to get the required response needed for statistical testing, a large number of companies need to be identified. For the scope of this study it is important that it can be completed within a limited time. Therefore, the study of a single firm is preferred.

The devolution of HR tasks and responsibilities is a complex research subject. There are many variables involved and a lot of different angles that can be investigated. For example:

(30)

29 under which circumstances may a firm chose to devolve HR tasks & responsibilities to line managers. Another angle could be the characteristics of a successful implementation of devolution within an organization. A case study is perfect for examining complex research subject and can be used in both theory building and theory testing (Berg, 2009).

A drawback is that the findings cannot be generalized, so this research can be viewed as a first step in developing a theory concerning the impact of devolution on line managers. Future steps would definitely include multiple case studies and/ or a large scale survey. Most research on devolution starts from an HR perspective and neglects line views and opinions (Gilbert, de Winne, & Sels, 2011). Another way to contribute to this field of

research is by focusing mainly on the line manager perspective. This case study studies the effects of various aspects of devolution on line management effectiveness by sending a questionnaire to all line managers within one organization.

Data collection method

To gain a better understanding of devolvement the first step was to review relevant

literature. The digital library of the University of Amsterdam was searched to find relevant articles. The following keywords were used: HR, Human Resource Management, line

management, line work, devolving, devolvement, HR to line. Recommendations made by the database based on articles selected have also been included in the search.

The search was set to find articles published between the present and the year 2000, so to ensure information was relevant within the context of time. Most articles will include summaries of relevant information from before 2000. It is expected that research conducted in the last seventeen years would give a more actual view of the devolvement of HR

(31)

30

Sample

Pantar has a total of 132 people working in managerial positions. The sample consists of line managers who meet the definition provided by Brewster et al.: ‘Line management covers those people in the management who are directly responsible for production or services’ (Brewster, Brookes, & Gollan, 2015). There are 110 team leaders and department heads who are responsible for directly supervising employees who work in production or services and all of them are included in the sample.

The rest of the managerial positions are held by staff managers, senior management and the managing board. They are excluded from the sample, because devolution impacts them less. The majority of line managers have a span of control between 20 and 45 employees. As you move up in the hierarchy the span of control becomes smaller as managers have around five department heads. It’s logical to assume that line managers spend more time on HR tasks compared to staff managers, senior management and the managing board as they manage larger teams.

All respondents received a digital email invitation with a link to the survey. The email addresses have been provided by Pantar. Since Pantar is going through a strategic transition at the moment, respondents might feel that this study could affect them in a negative way. Therefore, the invitation letter explains that all data is treated as

confidential. Additionally, management and leadership are only given aggregated data which cannot be translated to individuals. The Dutch versions of the email correspondence has been added to the appendix chapter 4.

(32)

31

Measurement of variables

Translation, back-translation procedure

As the primary language of the respondents is Dutch and the study is written in English, the questionnaire needed to be translated. To avoid changes in the true meaning of the items, a translation- back-translation procedure was used. In this procedure a third person translates the Dutch questionnaire back to English. Any differences found have been discussed and, if a consensus was reached, altered in the final Dutch version of the questionnaire.

(Amount of) Devolution of HR tasks and responsibilities

In order to measure the current state of devolution within Pantar, the item “In the last five years has line management involvement in people management activities within Pantar: increased, decreased or stayed about the same.” was added, based on the similar item in the Perry & Kulik (2008) study.

The extent of the devolution of HR tasks and responsibilities is measured using an index constructed from 27 practice HR tasks (Budhwar, Evaluating levels of strategic integration and devolvement of human resource management in the UK, 2000) (Gilbert, de Winne, & Sels, 2011) mentioned in devolution literature and the extent to which line managers indicate their responsibility: Fully responsible, Main responsibility but consult with HR, Shared responsibility with HR, HR has primary responsibility, Not applicable/ no response and I don't know. As according to Gilbert et al. (2011) and Cascón-Pereira et al. (2006) the amount of devolution is caused by the number of tasks devolved. All items for which line managers feel responsible are summed, because in general a task is deemed devolved once

(33)

32 there is a shared responsibility between line management and HR (Gilbert, de Winne, & Sels, 2011).

Personal HR support

This variable was measured using a five-item scale (Cronbach α = .89) used by Gilbert et al. (2011). The items ask about the amount of support received by the line manager from the HR department, for example: “The HR department supports me in executing my HR tasks”. All items were indicative for personal HR support. A 7 point Likert-scale ranging from was used, ranging from 1 (strongly agree) to 5 (strongly disagree), ‘not applicable’ and ‘I don’t know’.

Amount of HR training

To measure the level of HR training a self-designed six-item scale (Cronbach α = .86) was used. The items were based on the HR competence variable used by Gilbert et al. (2011). Additionally, three items were based on Takeuchi’s (Takeuchi et al., 2007) scale. All items were indicative for the amount of HR training. Again, the same 7 point Likert-scale ranging from Personal HR support was used.

Time spent on HR tasks

A single item ‘Please specify the average amount of time you daily spend on HR tasks’ was used to measure the variable. Respondents are asked to fill in the average amount in hours (per day).

HR information system

HR information systems are seldom discussed in devolution literature. As a result, a five-item scale (Cronbach α = .78) was devised based on a thorough search of five-items in an online

(34)

33 database (INN constructing theories, integrating research, 2017). Most items didn’t fit the question ‘How do line managers perceive the use of the HR information system?’. The final set of items was constructed in line with the other items in terms of wording and phrasing. All items were indicative for the HR information system. Also the 7 point Likert-scale was again used.

Experience as line manager

Another single item ‘Please specify how much years of experience you have as a line manager’ was used to measure this variable. Respondents are asked to fill in their experience in years.

HR role overload

This variable is measured in the same way as in the study conducted by Gilbert et al. (2011). It consists of a five-item scale (Cronbach α = .90) and also uses the 7 point Likert-scale. An example item is ‘My HR tasks lead to a work load that is too heavy.’ All items were indicative for HR role overload.

HR role ambiguity

Another variable used in the Gilbert et al. (2011) studies is HR role ambiguity. It is measured using a four-item scale (Cronbach α = .86). The items mainly ask about the clarity of the HR tasks and the role of the line manager in performing these tasks, e.g. ‘I know what my HR responsibilities are’. All items were indicative for HR ambiguity. The 7 point Likert-scale was used again.

(35)

34

Sense of responsibility & accountability

Another hypothesis explores the relation between what line managers see as their

responsibility and how that impacts their effectiveness as a line manager. Unfortunately, not a lot has been written about that relation in devolution literature. In order to explore the relationship a six-item scale was made. Again the online database (INN constructing theories, integrating research, 2017) was consulted to find items matching our question ‘How responsible or accountable does a line manager feel considering his role?’. No items were found and items were created based on that question. During the reliability analysis the Cronbach’s α for all items was .60, the analysis showed that by deleting ‘As a line

manager I am accountable for the mistakes of my team.’ The Cronbach’s α would rise to .69. As there is no guide or other research to consult, the decision was made to delete the item. In the new scale, excluding the question ‘As a line manager I am accountable for the mistakes of my team.’, all items were indicative for the sense of responsibility & accountability. A 7 point Likert-scale ranging from was used, ranging from 1 (strongly agree) to 5 (strongly disagree), ‘not applicable’ and ‘I don’t know’.

Line manager effectiveness

First an item asking for an overall rating of the own effectives was used. Second, a four-item scale (Cronbach α = .76) measured the variable. Line manager effectiveness is derived from a few related to reasons to devolve: a) the decision making abilities of the line

manager (Renwick, 2000) (Perry & Kulik, 2008), b) local line manager responsiveness (Budhwar, Evaluating levels of strategic integration and devolvement of human resource management in the UK, 2000) (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015) and c) the employee bond

(36)

35 (Brewster, Brookes, & Gollan, The institutional antecedents of the assignment of HRM responsibilities to line managers, 2015). As I did not find a similar scale, I searched for items in the online database (INN constructing theories, integrating research, 2017) that resembled the reasons mentioned above and constructed my own variants on those items. The final scale uses a 7 point Likert-scale. All items were indicative for line manager effectiveness.

To gain a better understanding of what attributions line managers make about the various variables, a matrix was added that included with all variables. Line managers need to indicate if they believe a variable makes a large positive contribution, small positive contribution, neutral or no contribution, small negative contribution or large negative contribution.

Results & Analysis

Over a period of three weeks an online survey was sent out to 110 line managers within Pantar. A total of 54 line managers responded. This response may result in low predictive power of the statistical analysis. A sample size calculation at the 95% confidence level and a confidence interval of 10, suggest that a minimal sample size of 52 is needed.

After reaching the minimal required sample size, the data was examined using the Statistical Software Package for Social Sciences (SPSS). The first step was to recode the variables. ‘Not applicable’ and ‘I don’t know’ have been recoded as missing values.

Respondents that only completed the first questions and didn’t finish the questionnaire have been deleted. HR role overload was reverse coded as it was the only negatively formulated variable.

(37)

36 Devolution strength has been recomputed in a new variable using the method described by Gilbert et al. (2011). This means that each of the 27 individual items related to the presence of devolution has been recoded into a new dummy variable. The values that include full or partial responsibility of line management have been recoded into 1. Responses based on full HR responsibility, ‘not applicable’ or ‘I don’t know’ were recoded into 0. The next step was to compute a new variable called ‘Devolution Strength’ which sums together all 27 items for each individual respondent. The number represents the total score of devolution: a higher number represents a stronger presence of devolution and a lower number indicates a lower presence.

The next step was a scale reliability analysis to determine the strength of the variables. The results are shown in brackets in Table 1. The scales of ‘Sense of responsibility &

accountability’ and ‘Line manager effectiveness’ are good as their Cronbach’s α is close to .7. However, the scales for the other variables were considerably higher: ‘Personal HR support’ and ‘HR Role overload’ both have a Cronbach’s α close to .9, whilst ‘Received HR training’ and ‘HR Role ambiguity’ have a Cronbach’s α of .86. ‘HR information system’ scores in between with a Cronbach’s α of .78. After the analysis, new scales were computed based on the mean average.

(38)

37

Table 1: Correlations & Reliability scales

Once all variables were computed a descriptive overview including histograms was

produced using SPSS, see Table 2. Examining the histogram’s showed outliers on ‘Tenure’ and ‘Time spent on HR tasks’, upon inspection it showed the values were higher than what one would logically expect: spending more than ten hours per day on HR tasks or spending more than 40 years at the company. These values were removed.

Table 2: Descriptive Statistics

Age (yrs) Tenure (yrs) The amount of devolution Personal HR support Received HR training HR information system Time spent on HR tasks (hrs) HR role overload (R) HR role ambiguety Amount of experience as a line manager (yrs) Sense of responsibil ity & accountabi lity Line manager effectivene ss Line manager effectivene ss score (1 to 10) Age (yrs) -Tenure (yrs) .31*

-The amount of devolution -.02 .13

-Personal HR support -.18 .02 -.04 (.89)

Received HR training -.19 -.06 -.15 .51** _(.86)

HR information system .01 .06 .00 .10 .13 (.78)

Time spent on HR tasks (hrs) -.12 -.04 -.20 -.11 .16 -.21 -HR role overload (R) -.14 -.30* _-.18 _.35** _.44** _-.01 _.17 _(.90)

HR role ambiguety -.10 -.15 -.03 .53** _.44** _-.01 _.09 _.31* _(.86)

Amount of experience as a line manager (yrs) .62** _.29* _.02 _.00 _-.11 _.12 _-.04 _.09 _-.14

-Sense of responsibility & accountability -.01 -.14 -.01 .11 .17 .36** _-.08 _-.10 _.10 _.06 _(.69)

Line manager effectiveness .06 -.07 .04 .15 .27* .07 .16 .27* .46** .02 .24 (.76) Line manager effectiveness score (1 to 10) .06 -.03 .13 -.25 -.29* _.09 _-.18 _-.22 _-.57** _.19 _.01 _-.30*

-Note: N=54 (Valid N=48)

*. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed).

Pearson Correlations

N Minimum Maximum Mean Std. Deviation

Statistic Statistic Statistic Statistic Statistic Statistic Std. Error Statistic Std. Error Skewness Kurtosis Age (yrs) 54.00 30.00 64.00 52.74 9.07 -1.12 .32 .36 .64 -3.46 .56 Tenure (yrs) 53.00 .50 37.00 13.16 9.80 .66 .33 -.57 .64 2.01 -.88 The amount of devolution 54.00 15.00 27.00 22.04 3.11 -.33 .32 -.63 .64 -1.02 -.99 Personal HR support 54.00 1.00 4.00 2.42 .74 .12 .32 -.44 .64 .36 -.68 Received HR training 54.00 1.33 4.67 2.70 .77 .43 .32 -.26 .64 1.31 -.41 HR information system 54.00 1.67 4.33 2.68 .59 .59 .32 -.02 .64 1.81 -.04 Time spent on HR tasks (hrs) 48.00 1.00 8.00 3.46 1.69 .63 .34 -.01 .67 1.83 -.01 HR role overload (R) 54.00 1.00 5.00 3.36 .89 -.37 .32 -.29 .64 -1.14 -.45 HR role ambiguety 54.00 1.00 5.00 2.43 .75 .83 .32 2.10 .64 2.56 3.28 Amount of experience as a line manager (yrs) 54.00 5.00 40.00 22.31 9.16 .04 .32 -.68 .64 .11 -1.06 Sense of responsibility & accountability 54.00 1.00 3.40 1.95 .49 .38 .32 .20 .64 1.16 .32 Line manager effectiveness 54.00 1.00 3.25 1.79 .50 .37 .32 .10 .64 1.15 .16 Line manager effectiveness score (1 to 10) 54.00 4.00 9.00 7.30 1.18 -.68 .32 .25 .64 -2.10 .39 Note: N=54 (Valid N=48)

z-score

Descriptive Statistics

(39)

38 In order to verify normality of the data, the skewness and kurtosis for all variables was checked. ‘Age’ showed a substantial negative skewness (<-1). ‘Tenure’, ‘HR information system’, ‘Time spent on HR tasks’ and ‘HR Role ambiguity’ showed moderate positive skewness (between .5 and 1). ‘Line manager effectiveness score’ showed moderate negative skewness (between -.5 and -1). The skewness and kurtosis of the remaining variables is close to 0 and therefore are considered normally distributed. To check significance z-scores have been calculated for all variables. The z-scores are compared with the known values for the normal distribution (field). This showed that ‘Tenure’ and ‘Line manager effectiveness score’ values show significant skewness (p<.05); however, they are very close to the lower limit (1.96 or -1.96). ‘HR ambiguity’ (p<.01) and ‘Age’ (p<.001) also show significant

skewness. The z-scores have also been calculated for kurtosis, showing significant kurtosis for ‘HR role ambiguity’ (p<.05).

Not all variables are distributed normally. In order to resolve this issue, a data

transformation can be applied in order to normalize the data (Field, 2009). To find out which transformation would be best different transformation were applied: a log10, square root and 1/X transformation. The square root transformation normalized all but ‘Line manager effectiveness score’. After the transformation ‘HR role overload became skewed, so the transformation didn’t solve all problems.

Field (2009) mentions two things that complicate matters: a) to apply the same

transformation to all variables b) reverse the reversed score transformation afterwards in order to avoid issues with the interpretation. The skewness differs for each variable: some are positively skewed; others are negatively skewed. This becomes a problem when

(40)

39 results in negatively skewed data to which the transformation cannot be applied. The same principle goes for reversing reversed score transformations.

He also mentions that transforming data is subject to debate (Field, 2009). Some authors consider the added value to be low and rarely worth the effort (Field, 2009). According to Games (1984) in (Field, 2009) it is difficult to determine normality in small samples, such as ours. Of our variables only ‘age’ showed substantial skewness (<-1). This makes sense considering perceptions about the population made by managers and HR professionals: the population is around 50 years old and predominantly male. The sample accurately reflects this perception. As ‘age’ is considered a control variable the impact on the statistical model should be low. Leading me to believe that transformation could have severe negative

implications and thus should not be considered1_{. For statistical purposes all, but the control}

variables, are considered normally distributed.

Correlation analysis

Before I begin detailing the results of the correlation analysis, I would like to point out that the Likert scale used in this study ranks from one to five. One represents ‘strongly agree’ and the other ‘strongly disagree’. All items are formulated positively and the Likert scales are treated as interval data2_{. I would like to remind the reader that a higher score reflects a}

negative association with the variable and a lower score has a positive association. When reviewing the correlation data this means that a strong positive correlation indicates that the more they disagree on one variable the more they disagree on the other and vice versa. The results from the correlation analysis can be viewed in table 1.

1_{For statistical enthusiasts the descriptive statistics for the transformations have been added to the}

(41)

40 ‘Age’ was significantly correlated with ‘tenure’ (r = .31, p < .05), and ‘amount of experience as a line manager’ (r = .62, p < .01); ‘tenure’ was also correlated with ‘amount of experience as a line manager’ (r = .29, p < .05).

There was a significant relationship between ‘tenure’ and ‘HR role overload (R)’ (r = -.3, p < .05). This means that the longer a person works for the company, the less role overload that person experiences.

‘The amount of devolution’ and ‘Time spent on HR tasks’ showed no significant relationship with any of the other variables.

‘Personal HR support showed a significant correlation with: ‘received HR training’ (r = .51, p < .01), ‘HR role overload (R)’ (r = .35, p < .01), and ‘HR role ambiguity’ (r = .53, p < .01). ‘Received HR training’ is also significantly related to ‘HR role overload (R)’ (r = .44, p < .01), and ‘HR role ambiguity’ (r = .44, p < .01). ‘HR role overload (R)’ and ‘HR role ambiguity’ are correlated as well (r = .31, p < .05).

‘Received HR training’ is significantly positively related to both ‘Line manager

effectiveness’ (r = .27, p < .05) and negatively related to ‘Line manager effectiveness score’ (r = -.29, p < .05). The more dissatisfied a person is with the HR training, the lower the

corresponding effectiveness score is.

There was a significant relationship between ‘HR information system’ and ‘Sense of responsibility and accountability’ (r = -.36, p < .01).

‘Line manager effectiveness’ is correlated to ‘HR role overload (R)’ (r = .27, p < .05) and ‘HR role ambiguity’ (r = .46, p < .01).

(42)

41 ‘Line manager effectiveness score’ shows a significant relationship with ‘HR role ambiguity’ (r = -.57, p < .01) and ‘Line manager effectiveness’ (r = -.3, p < .05). High effectiveness scores, result in lower scores for ‘HR role ambiguity’ and ‘Line manager effectiveness’. The latter is interesting because that relation could mean both measure the same.

Regression analysis

The next step in the statistical analysis is to do a regression analysis and determine the impact of the predictor variables on the outcome variables. There are two outcome variables measuring line manager effectiveness: line manager effectiveness based on a four item scale and line manager effectiveness score that is based on a ten-point scoring system. Although the MANCOVA test is better suited for multiple predictor and outcome variables (Field, 2009), it is not part of my curriculum. Therefore, I’ll run two regression analysis: one for each outcome variable. As there are more than one predictor variables, I’ll be running the multiple regression analysis.

The chosen method of regression is forced entry. The model used in this study is based on existing research and literature, but has not been tested as a whole before. In a hierarchical regression the researcher decides the order in which the predictor variables are entered into the model. This could influence the outcome of the regression. The researcher choices should be based on previous studies. The same goes for the forced entry method, except the researcher doesn’t decide in which order the variables are entered into the model. All variables are entered into the model simultaneously. According to Field (2009) this method is best suited for theory testing (Studenmund & Cassidy, 1987) in (Field, 2009).

Unfortunately, the current sample size seems too small to make accurate predictions. The sample size determines what effect sizes we can measure (Field, 2009). In general, the

(43)

42 bigger the sample the better. Field (2009) mentions a few methods to calculate the

minimum sample size. First, if you wish to test the overall model the recommended formula is 50 + 8k (k stands for the number predictor variables, we have eleven in this regression). Second, in order to test the individual predictors, the formula is 104 + k. So that would mean the required minimum sample size for this regression is 138 to test the whole model or 115 to test the individual predictors. For this study the size of the line manager

population within Pantar is 110. That would mean that even if every line manager at Pantar completed the survey, we would still lack the necessary numbers to perform an accurate/ powerful regression analysis.

When considering different effect sizes (large, medium, small), Field (2009) provides a graph produced by Miles & Shevlin (2001). The graph shows the sample size/ number of predictor variables for different levels of power. Field (2009) summarizes the graph in three rules: 1) in order to detect a large effect, a sample size of 80 is sufficient 2) in order to detect a medium effect, a sample size of 200 is sufficient and 3) in order to detect a small effect, a sample size of 600 is sufficient. This goes for up to twenty predictor variables. For ten predictor variables we need a sample of around 60 cases to be able to predict large effects. Therefore, in order to enhance the power of the regression analysis, I’ve decided to reduce the number of predictor variables used in the regression. The control variables ‘age’ and ‘tenure’ will not be included in the regression analysis. This means that our regression won’t be able to detect medium or small effects, but possibly can detect large effects if present, but I’ll discuss this a bit more in the general discussion.

However, the outcomes are very similar. ‘Line manager effectiveness’ is correlated to ‘HR role overload (R)’ (r = .3, p < .05) and ‘HR role ambiguity’ (r = .52, p < .01). ‘Line manager effectiveness score’ is correlated to ‘HR role overload (R)’ (r = -.3, p < .05) and ‘HR role

The devolution of HR Tasks & Responsibilities to Line Management