How change approaches and team identification influence the level of resistance to change of employees during a reorganisation.

(1)

How change approaches and team identification influence the level of resistance to change of employees during a reorganisation.

Master thesis, MScBA, Specialization Change Management University of Groningen, Faculty of Economics and Business

June 20, 2013

Siemen Duursema Studentnumber: 1766961

Peizerweg 28a

(2)

2

How change approaches and team identification influence the level of resistance to change of employees during a reorganisation.

Abstract

(3)

3

1. Introduction

In recent years, governmental organisations have come under increasing pressure from policy makers, community members and other stakeholders to improve the quality and cost effectiveness of services and the management systems which support them (Packard, Patti, Daly & Tucker-Tatlow, 2012). Since the start of the economic crisis and following the global recession, the national governments are forced to make cuts in their budgets. As a result of this, governmental organisations have undergone changes and reorganisations. This is also the case in the organisation that is central to this paper, namely the municipality of XYZ in The Netherlands.

1.1. The case

1.1.1. The need for change

In 2010 the city council decided that the organisation should improve the service it provides (such as policy making and advice development) to the council, its internal client. In their

document concerning the organisational development (Bestuursopdracht

organisatieontwikkeling, 2010), the council describes the need for “a more effective, efficient and transparent organisation and services”. Starting points (Bestuursopdracht organisatieontwikkeling, 2010) are for example:

• Improvement of the quality of digital services;

• continuous monitoring of the effectiveness and efficiency of the services provided; • the breakdown of compartmentalization and making the process of policy making

more transparent;

• enhancement of comprehensive control within the organisation.

A re-organisation was initiated that reshapes and relocates the whole organisation and its members. The municipality of XYZ is an organisation with 750 employees where people are used to work for several years in the same job function, same department, same place and with the same colleagues. Before the reorganisation there were 15 departments placed in three divisions as shown in figure 1. Between the three divisions, the execution of the work differed using different work processes, ways of communicating and with their own vision. Between the divisions some amount of rivalry existed with one division, the urban development division, being more dominant than the city management division and the public service division.

(6)

6

employees are fired and rehired, some in new positions but all in the new units. The city council decided that there were not going to be any forced layoffs until 2014.

Figure 1: organisation chart of the municipality of XYZ

1.1.2. Implementing the change

This paper solely focuses on the reorganisation of the A unit and the B unit. The reason for this is that these were the first units to start with the reorganisation. The reorganisation for the other five units is planned in the near future. The B unit is composed out of four departments namely the three Management and Control (MC’s) departments and the Policy and Control department. The three MC’s used to be distributed over the three divisions and are now reorganised based on their functionality. For example: the legal teams of the three MC’s (each MC had is own legal team) are now merged into one team being a part of the new B unit. The A unit is composed out of two departments; namely the Strategy and Development department and the Policy and Projects department.

At the start of the reorganisation of these two units there were no unit managers appointed. To avoid suggestions and rumours about who were going to become the new unit managers after the change, external change agents were hired to lead the change. This meant that the change agents would prepare and make the first steps for the change. Their task was focused around the reorganisation. The operational activities that had no overlap with the activities related to the reorganisation were delegated to the team leaders. When the unit managers were appointed the change agents would leave the organisation. Another reason to appoint an external change agent was that it provided the management team with more time to find the right person for the job.

The main task of the change agents was to construct a development plan for the next phases of the change and deal with the newly formed units. This means that, although formally the units were developed, it was the job of the change agents to bring these units together successfully and increase the cooperation within the units.

The implementation of the organisational development had a somewhat flexible character. Flexible because, although the aim of the change was to form new units in order to improve the service delivered to the city council is stated, the process of handling the change to form

(7)

7

the new units is to be chosen by the change agent. This resulted in two distinctly different ways of handling the change.

As mentioned before, the units B and A were the first to start with the reorganisation. The reason for this is that there was no start date on which the reorganisation for all units starts. It depends on the urgency but also, as the programme manager organisational development describes it, “at the right moment for change” when the change agent starts with the organisational changes. Finding the right moment for the change was a wish from the top management. The “right moment for change” can be described here as a moment where parts of a new unit (the departments) are undergoing change. For instance, the unit B was the first unit to be reorganised because three out of four team leader positions were vacant. This gave the management the opportunity to fill these positions with team leaders who supported the reorganisation and had the capabilities to lead a team during the change process. Thus appointing new team leaders provided the right moment for change here. For the A unit, the right moment to change came when two out of three director positions became vacant.

1.1.3. Lessons learned from the change

Now, after these two units are formed and the change agent’s left, it catches the eye of the management team that there are differences in outcomes. In two expert interviews with the programme manager organisational development, two differences in results of the reorganisation emerged.

The top management, the city manager and the director, held interviews with the employees of the two units. They observed that in the B unit there is more resistance as a result of the reorganisation and that there is less unity within the newly formed unit then there is in the A unit. For example: the employees in the B unit rejected the development plan (one of the main tasks of the change agent) whereas the development plan of the A unit was accepted and supported directly after it has been introduced.

Next to this, after the change agent of the B unit left the organisation the team leaders of that unit received feedback from their employees. Their employees described feelings of uncertainty, anger and distrust. The team leaders notified the top management about these feelings. These reactions differed from the reactions within the A unit, where the employees were largely positive based on interviews they held with some of these employees.

(8)

8 1.2. Management question and objective

As described above, the management of the municipality wishes to know what the differences in outcome are between the two units. Next to, this they wish to know what influenced the change process in such a way that they can explain the difference in outcome. The management question is:

How can the differences after the reorganisation between the B unit and the A unit be explained and what lessons could be learned for future reorganisations within the municipality of XYZ ?

Answers to this management question can be found in the field of change management. To measure the differences after the reorganisation, the dependent variable, it seems obvious to use resistance to change. The feelings of resistance among employees, as described earlier, are a useful indicator to measure the differences in results after both reorganisations has been implemented.

Based on the case, two independent variables arise. These variables are relevant from a theoretical perspective since they influence the level of resistance and from a practical perspective because they are relevant to the case described earlier in this study. Firstly, the different approaches the change agents used to reorganise the units. In the field of change management, different change approaches exist to implement change. Further on in this paper, these change approaches are further elaborated.

(9)

9 1.3. Research question

To answer the management question, the following research question has been developed: How do change approaches and team identification influence the level of resistance to change

of employees during a reorganisation?

The variables that are investigated in this research are change approach as applied by the change managers, team identification and the role of uncertainty as mediator between team identification and resistance to change.

1.4. Contributions of this paper

The practical implications of this paper are mainly for the management of the municipality of XYZ. This paper provides insights about their change processes and what influences these processes. The knowledge developed in this paper can help them to plan and implement the other five units more successfully.

Whilst organisational change appears to be happening with increasing frequency and magnitude in both the public and private sectors, most of the major studies of and approaches to change focus on the private sector and tend to derive their approaches to change from that sector (Coram & Burnes, 2001). This paper will focus on two change approaches, namely top-down and bottom-up approaches towards change in a public sector organisation.

As Hogg and Terry (2000: p135) describe, “ the challenge for the future is to integrate new social identity mechanisms centrally into theories of organisational behaviour”. They continue mentioning that the important role that identification with the workteam may play in organisational behaviour has yet to be fully articulated. This article tries to show how team identification using the social identity perspective influences resistance to change during organisational change.

1.5. Scope limitations

As explained earlier, the management of the municipality of XYZ wishes to learn from the reorganisation of the units A and B to succesfully reorganise the other five units. This study hopes to provide this information to support the management. An important limitation of the scope of this study is that it does not explore the effects of the outcomes of the reorganisation of the first two units have on the other five units. One could imagine that employees of the units A or B who are confronted with negative outcomes of the reorganisation in some way effect their collegeaus of other units. In particular where employees of different units work on the same floor or even in the same room. For the management it could be important to take note of this effect although it is not included in this study.

(10)

10

function obligated to actively involve and cooperate with different parties, both internal as well as external to the organization. This requires a certain openess to the environment and flexibility to work with many different parties. This difference in openess and flexibility can influence the ability to change and implement a reorganisation. Again, for the management it could be important to take note of this effect although it is not included in this study.

1.6. Outline

(11)

11

2. Literature on change management

In the 21st century, organisations and managements are faced with continuous demands for

change (Buono & Kerber, 2010). Influenced by its fast pacing surrounding, organisations are constantly seeking to stay in touch with their environment and adapt to new situations. Because of this, there has been an increasing amount of literature on change management.

2.1. Resistance to change

An important topic within the field of change management, and the dependent variable in this paper, is resistance to change. Kurt Lewin constructed the concept of resistance to change is in 1946 by defining resistance as a restraining force moving in the direction of maintaining the status quo. Resistance is a natural part of the change process and is to be expected. It occurs because change involves going from the known to the unknown (Coghlan, 1993) Change can generate deep resistance in people and in organisations (Cummings & Worley, 2008:10). The majority of the qualitative and quantitative research on employees’ emotional reaction on organisational change suggests that the emotions triggered by change are mainly negative (Piderit, 2000; Seo, Taylor, Hill, Zhang, Tesluk & Lorinkova, 2012).

In this paper, we will follow Piderit (2000) who constructs resistance to change as a multifaceted construct. In her article, resistance to change is conceptualised in a multidimensional view of attitudes toward organisational change consisting of cognitive, affective and behavioural components.

The cognitive component refers to an individual’s beliefs about the organisational change. These beliefs may range from strongly positive to strongly negative beliefs as response. For example, an individual could think that an organisational change is necessary for expanding the existence of the company. For example, negative beliefs focus on possible devastating effect of organisational changes. The second component is the affective component. This component refers to individual’s feelings in response to the organisational change and ranges from strongly positive emotions to strongly negative emotions. The third component, behavioural component, refers to the positive intentions of the individual to support or negative intentions to oppose the change.

Now the dependent variable is described, the independent variables are presented. For this paper, based on the situation of the organisation described above and the literature, the independent variables change approach and team identification are applied. As described earlier the difference in approach of the two change agents could have resulted in the different levels of resistance and therefore they are subject of study within this paper. Following, different change situations require different approaches (Burnes, 2004). Therefore different levels of resistance can be expected which make it interesting in this case to see what these differences are.

(12)

12

to protect the prestige relative to other teams because of team evaluation is self-evaluation (Hogg, Abrams, Otten & Hinkle, 2004). During a reorganisation as described earlier this protection is one example how team identification can lead to changes in the level of resistance to change.

Mediating the relationship between team identification and resistance to change is uncertainty. Uncertainty is one of the most commonly reported psychological states in the context of organisational change (Bordia, Hobman, Jones, Gallois & Callan, 2004). Therefore, it also could have played a role in the change program within the organisation focused on here. Next to this, uncertainty reduction is the primary motivation for team identification (Hogg & Mullin, 1999). This shows the link between team identification and the level of uncertainty.

2.2. Change approaches

Leading organisational change has been a research topic over a long period of time. In the change management literature there is considerable disagreement regarding the most appropriate approach to changing organisations (Bamford & Forrester, 2003). According to B urnes (2009:) “the argument is in essence (Beer & Nohria, 2000; Burnes, 2004; Kanter et al., 1992; Stace & Dunphy, 2001) that there are a variety of approaches to change, and that an im portant element in achieving successful change is to choose the most appropriate approach for the type of change being undertaken and the circumstances in which it is being undertaken”. For example, in a longitudinal case study about an organisation that transformed its culture, management practices and structure, Burnes (2004) shows how different change situations require different approaches. The organisations in Burnes’ article successfully apply both a planned and an emergent change approach. By doing so Burnes shows that using multiple approaches is possible.

Planned approach

Different approaches have been developed with the planned approach, originated by Lewin (1947), and the emergent approach being the one most often mentioned (Burnes, 2004). Planned change is primarily aimed at improving the operation and effectiveness of the human side of the organisation (Burnes 2004). Lewin (1947) proposed that before change and new behaviour can be adopted successfully, the previous behaviour has to be discarded. Lewin describes three steps to do this. First step is unfreezing from the present level of behaviour, second step is changing to the preferred level of behaviour and third step is refreezing on the new level of behaviour.

Emergent approach

(13)

13

the key decisions about matching the organisation’s resources with opportunities, constraints and demands in the environment evolve over time (Hayes, 2002).

Both planned and emergent approaches to change are seen as the one best way by their advocates. This means that only one approach is most suited for all change events according to their advocates. Opposed to this is the contingency approach that supports a “one best way for each” organisation approach instead of “one best way for all” (By, 2005). The contingency approach is based on the idea that the structure and the performance of an organisation are dependent on the situational variables that it faces (Dunphy & Stace, 1992). They advocate that there is a range of approaches to achieve a fit with the organisation as well as the constantly changing environment. Based on the environment and organisation which together lead to a specific situation, there is one optimal approach to change.

Bold stroke & Long march

Following the planned and emergent approaches are the approaches of Kanter (1985) and Beer and Nohria (2000). In their book Kanter, Stein and Jick (1992) go into more detail about the organisational change and especially on transformational change. They describe two approaches, namely bold stroke approach and a long march approach. The first, bold stroke can be used for rapid overall change, for example structural change. The second, the long march can be used for incremental change leading to transformation over an extended period of time, for example cultural change in an organisation.

Theorie E & O

Beer and Nohria (2000) present two theories: Theory E and Theory O. Theory E, which is related to the bold stroke approach of Kanter (Burnes, 2004) is based on economic value. It is described as the “hard” approach focusing on maximizing shareholder value through top-down leadership and emphasized on structure and systems. Theory E usually involves heavy use of economic incentives, downsizing and restructuring (Beer & Nohria, 2000). One of the underlying dimensions of Theory E is the planned approach (Beer & Nohria, 2000)

Theory O, which relates to the long march approach of Kanter (Burnes, 2004), is described as the “soft” approach. The goal of this approach is to develop corporate culture and organisational capabilities through individual and organisational learning. The leadership style is known to be bottom-up and the focus is on building up the culture, behaviour and attitudes of employees within the organisation (Beer & Nohria, 2000). One of the underlying dimensions of Theory O is the emergent approach (Beer & Nohria, 2000).

2.2.5. Change approach and level of resistance to change

Based on the literature presented above and the situation within the organisation as described earlier in this paper, we focus on top-down approach versus bottom-up approach. These approaches describe how the change agents implemented the change in the municipality of XYZ.

(14)

14

management of change, while non-managerial employees generally have little say in the decision-making process (Ryan, William, Charles & Waterhouse, 2008).

The bottom-up approach is the opposite of the top-down approach. The bottom-up approach creates conditions for direct employee participation that top-down change generally does not provide. The term is used to describe widely different organisational arrangements that vary, especially with respect to the amount of decision-making authority delegated to staff (Ryan, William, Charles & Waterhouse, 2008). This lack of participation of the lower level employees is a reason why resistance is more likely with a top-down approach than with a bottom-up approach (Giangreco & Peccei, 2005)

Top-down and bottom-up approaches to change are besides approaches to change themselves, also included in several other approaches to change. The change approaches described above have included either top-down or bottom-up approach to change. For instance, Planned change of Lewin has top-down characteristics.

Emergent change has bottom-up characteristics (Burnes, 1996, 2004; By, 2005). The rationale behind this is that the pace of change is so rapid and complex, that when change occurs, it is impossible for senior management to identify, plan and implement every action required (Bamford & Forrester, 2003)

Also Beer and Nohria (2000) make a clear distinction in the way change should be led in their Theory E and Theory O approaches. Theory E uses the top-down approach. In this theory, leaders set goals with little involvement from their management teams and certainly without input from lower levels (Beer & Nohria, 2000). On the other hand, with Theory O, leadership is directed from the bottom-up to “encourage participation”. This means that lower levels of the organisation are actively involved in the change process and are encouraged to participate in the process.

Although several authors used these approaches in their articles (see for example: Coram & Burnes, 2001; Ashburner, Ferlie & Fitzgerald, 1996; Conway & Monks, 2011; Sminia & van Nistelrooij, 2006; Ryan, Williams, Charles & Waterhouse, 2008) the authors of this paper never found a tool to measure these approaches within an organisational setting in the research available on this topic. Therefore questions were developed based on the interviews held in the orientation phase.

In this study, we argue that a top-down led approach to change will result in higher levels of resistance to change as opposed to bottom-up change.

(15)

15

the lower levels of the organisation, lower levels feel forced to change, which could counteract the successful implementation of the change.

Several other authors relating the level of participation of the employee also describe this. Lack of participation in decisions negatively influences employee commitment and thereby undermines the successfullness of a top-down approach (Kanter et al., 1992; Doyle et al., 2000).

Participation can both lead to designing high-quality changes and to overcoming resistance to implement the change (Nutt, 1986). Cummings and Worley (2005) continue, mentioning that participants will be more committed to implement changes when involvement in planning the changes increases. This is because the likelihood that members’ interests and needs will be accounted for during the intervention increases when they are actively involved.

Another advantage for applying a bottom-up approach is that the top-down approach has been criticized for failing to recognize the complexity and uncertainty of organisational change processes (Dawson, 1994:5). Change is too complex and too subtle to be controlled completely from one place or by one small team of executive "leaders" (Mohrman & Cummings, 1989). Again, involvement of all levels within the organisation seems important for the successfulness of the implementation of the change.

Based on literature, the following hypothesis is constructed

H1: A top-down approach will result in higher levels of resistance to change than a bottom-up approach

2.3. Team identification

To describe the effect of the change on team identification the social identity theory of Tajfel (1974) is utilized. Social identity theory has been applied often in organisational settings to gain insights about work-based identity. As Hogg, Abrams, Otten and Hinkle (2004) describe in their review article focused on the social identity perspective: “Because organisations are teams, the social identity perspective is perfectly suited for analysing organisations”.

(16)

16

To show that social identity theory can successfully be applied to team identification the article of Janssen and Huang (2008) is relevant. They investigated team identification and individual differentiation as complementary drivers of team members’ citizenship and creative behaviour. To test the hypothesis in that research, results of a survey among 157 middle management team members were used. All of these team members worked for a Dutch company in the banking industry.

The social identity theory as described by Tajfel (1974) shows how individuals have two types of identities. The first one is the personal identity. This is a combination of idiosyncratic personality attributes that are not shared with other people or close personal relationships that are tied entirely to the specific other person in the dyadic relationship. This type is focused more on the “I” and “Me and You”.

The second type of identity and the focus of this paper is social identity. This type of identity describes the relationship between two or more people (team) and how they identify themselves in the same way and have the same definition of who they are, what attributes they have and how they relate to and differ from specific out teams. This is described as team membership and expressed in terms of “we” and “them” (Hogg, Abrams, Otten & Hinkle, 2004.). These teams may be formal working teams but also informal teams within a new organisational setting or teams based on sex or gender.

2.3.1. Team identification and level resistance to change

Team identification could influence the single member’s opinion and behaviour concerning the change due to the strong relation and influence between the members within the team. In teams with high levels of team identification, individuals are strongly committed to its team. A key assumption of the theory is that individuals derive their social identity from the teams to which they belong (e.g. student, team member). The more one conceives of oneself in terms of the membership of a specific team, the more importance this social identity carries for the self (Jetten, o’Brien & Trindall, (2002). This is supported by Van der Vegt and Bunderson (2005) who state that the team member is more committed to the goals of the team rather than their personal goals and the esteem of the team is strongly linked to the individuals’ self-esteem. When the team continuity comes under pressure because of organisational change, for instance due to reorganisation, team members are most likely to resist the change since losing the team means lowering of the individual self-esteem and changing goals.

(17)

17

organisational change (van Dijk & van Dick, 2009). As a response to this threat the members may resist the change.

Following, van Knippenberg and van Schie (2000) suggest that team identification is likely to be a stronger predictor of behaviours and actions than organisational identification because teams are typically smaller and individuals have more in common with members of their team. If members within a team resist the organisational changes, through high team identification these members may influence other team members to resist the organisational changes. Asforth and Mael (1992) confirm this by mentioning that stronger identification also increases the individual’s susceptibility to influence by other team members (1989).

We argue that higher levels of team identification will lead team members to resist the change strongly. Based on the above, the following hypothesis is developed:

H2a: Higher levels of team identification will lead to higher levels of resistance to change among organisational members during organisational change

2.4. Uncertainty

The mediating variable is the level of uncertainty. Milliken (1987) defines uncertainty as an individual’s perceived inability to predict something accurately. During organisational change, employees are likely to experience uncertainty in relation to a range of different organisational issues (Allen, Jimmieson, Bordia & Irmer, 2007). For instance, when due to changes the team composition changes but individuals are unknown about the underlying reason for the change, this may lead to uncertainty. Uncertainty has been identified as one of the major consequences of organisational change for employees (Allen, et al. 2007). Uncertainty arising from a changing and dynamic work environment is a major contributor to feelings of change related psychological exhaustion (Bordia et al., 2004).

2.4.1. The mediating role of uncertainty

Research on mergers and acquisitions (Schweiger & DeNisi, 1991), which is a special type of organisational change, shows how lack of clearness about the members future leads to increases of the uncertainty level of members of an organisation. This uncertainty is often more stressful than the organisational changes themselves (Schweiger & DeNisi, 1991). During these change the role of the team is even more important. This has mainly to do with the safety the team provides.

The safety the team provides is challenged by the uncertainty about the changes. Given that organisational teams do not exist in isolation without any changes to their membership, role structure, task, or external environment examining how change affects team work is critical (Summers, Humphreay & Ferris, 2010). Bordia et al. (2004) confirm this showing that employees may experience uncertainty regarding the security of their position and their future roles and responsibilities.

(18)

18

have of each other. Hormuth (1990) continues stating that during times of transition, the relationship between the individual and his or her environment changes, and the person must adapt to those changes in some way. Because of this, a complete set of new norms have to be developed to successfully adapt to the new situation.

Research has indicated that employees may experience uncertainty regarding the security of their position, and their future roles and responsibilities (Bordia et al., 2004). Uncertainty about their future role, position and self-esteem derived from their membership will increase if team identification is high since members have more to loose and place higher value on their team membership. Uncertainty about negative results will result in higher levels of resistance to change.

Higher levels of uncertainty will result in higher levels of resistance to the change. Schweiger and DeNisi (1991) show in their survey studying two merging Fortune 500 companies that increases in uncertainty during an merger led to increases in stress and a decrease in satisfaction, commitment and intentions to remain with and organisation. All of these may cause resistance. For this study, the relationship between uncertainty and resistance to change is important. Kanter (1985) describes in her article how too much uncertainty leads people to resist change. “If people don’t know where the next step is going to take them (…), change seems dangerous”. This leads to resistance to change.

Based on the above, we argue that when team members have high levels of team identification, uncertainty about this team will increase the level of resistance to change.

H2b: The level of uncertainty mediates the relationship between team identification and level of resistance to change

(19)

19

3. Conceptual Model

Figure 2: Conceptual model

(20)

20

4. Methodology

There are several different research types. Choosing a research type is important since it has consequences for the design and implementation of the research (Baarda & De Goede, 2006:3). For this study, an explanatory case research is performed (Yin, 2009:1). In order to test the hypothesis, an explanatory case research is conducted to explain the presumed causal links in real-life interventions (Yin, 2009:1). The case studied in this study is a single embedded case study, with one case focusing on more than one unit of analysis (Yin, 2009:2).

Based on the pre-study exploratory interviews described below, it was concluded that higher levels of resistance to change could be expected in the unit B and lower levels of resistance to change in the unit A. To avoid respondents that are only moderately influenced or interested in the change, this study focuses on the extreme cases. These are the respondents with either high levels of resistance to change or low levels of change.

Following this, the researcher and the programme manager organisational development selected respondents from the units B and A. These were respondents with respectively the highest levels from the unit B and the lowest levels from the unit A of resistance to change. This process is known as purposive sampling (Baarda & De Goede, 2006:5) where researchers choose participants arbitrarily for their unique characteristics or their experiences, attitudes or perceptions. In this study their ascribed level of resistance to change made them interesting for including them in this study. When asked, the interviewees were told that they were chosen through random sampling to avoid interviewees answering according their ascribed level of resistance and refusal since being described of having high levels of

resistance to change could be negatively interpreted. The research population of this study is all employees of both units; in total 130 employees.

The sample size of this study included 40 respondents, evenly distributed over both units. Baarda and De Goede (2006:5) state that 30 respondents is the absolute minimum amount of respondents needed to statistically test causal relation and find significance. With 40

respondents we included around one third ofthe total population in this study.

Table 1: Description of the interviewees

Sample Percentage

Gender Male n=22 55%

Female n=18 45%

Age Highest 63 yr.

Lowest 27 yr.

Average 45 yr.

Educational level Primary school 0 0%

(21)

21 Organisational

tenure

Average 12,2 yr.

Job tenure Average 4,6 yr.

4.1. Data collection

For the data collection and derived from the research question, three tools were used in this study namely: semi structured interviews, questionnaires and document analysis. First the interviews will be described below; followed by the questionnaire and at last a document analysis.

For this study six interviews were conducted. The change agents from both units were interviewed, as well as two team leaders from the unit A and two team leaders from the unit B. These interviewees were selected based on their position during the change process. These positions, leading positions both on operational level as well as on the change process level, provided them access to contextual information about the two units. This information contained information about the situation of the two units before, during and after the change process. Using only a questionnaire could not provide this data because the interviewees can be seen as experts on the topic (and the people that are involved) of the change process. Interviews were chosen because they are an effective method for collecting data (Myers, 2009:10). Semi structured interviews were chosen because they enable the interviewer to be adaptive during the interviews and enabled the interviewer to probe emergent issues during the interview process (Cummings & Worley, 2005:7) while the prepared questions provide some focus as well (Myers, 2009:10). The interview guide is included in Appendix B.

These interviews were mainly used for orientation purposes. The aim of the interviews was to obtain more insights in the change process and the roles played by different people within that process. This information was also used to construct the questionnaire and also for the selection of the sample size. This is described in the following section. During the interviews, essential information was obtained and therefore some results of the interviews were included in the results section of this study.

The interviewees were contacted either in person (team leaders) or by phone and e- mail. The four interviews with the team leaders were conducted in meeting rooms in the city hall. On request of the change agents, their interviews were conducted at their homes. Meeting interviewees’ requests about where to hold the interviewees increased the quality of this study since the interviewees were more comfortable (Cooper & Schindler, 2008:7)

(22)

22

The second data collection tool of this study is the questionnaire. The questionnaire was largely based on existing measures found in literature. These measures are described below. The questionnaire was included in Appendix A.

All respondents were contacted in person by the researcher and received a hard copy questionnaire. A short oral introduction about the study, anonymity and the questionnaire followed. The questionnaires were collected three days following the hand out. To minimize the amount of missing values the questionnaires were checked directly after collecting. Because of anonymity issues, the questionnaires were checked for any missing value. For the researcher it would be impossible to trace back the respondent. The anonymity issue was mentioned again in the introduction section of the questionnaire.

The third data collection tool is document analysis. This analysis mainly uses information that is internally available, by analysing existing documents. For example, the developmental plans regarding the two units were used to gain more insights about the start-up phase of the different change approaches and the processes. Also, intranet turned out to be a rich source of information for documents concerning the change process. This information can supplement the data provided by interviews and provide much useful background information (Myers, 2009:12).

In this study, both qualitative and quantitative methods are combined using different data collection tools. This is called triangulation. Triangulation of multiple collection tools allows the researcher to gain a broader picture of what is happening (Myers, 2009:2). Applying triangulation will increase the quality of this study.

4.2. Quality control

Four tests have been commonly used to establish the quality of an empirical social research (Yin, 2009:2). An explanatory case study is one form of such research. The quality criteria are: construct validity, internal validity, external validity and reliability. How these are dealt with in this study is described below.

The first quality criterion is the extent to which a measuring instrument measures what it is intended to measure. This is called construct validity (Aken, Berends & van der Bij, 2007). Construct validity can increase by identifying correct operational measures for the concepts being studied (Yin, 2007:2). Most measures used in this study are derived from articles, which successfully applied and tested the measures in the past. The Cronbach’s alpha was used as an estimate for the applicability of the measures for this study.

(23)

23

& van der Bij, 2007). In this study, triangulation of multiple collection tools is used to increase internal validity

External validity is the third quality criterion and deals with the question whether the present study’s findings are generalizable beyond this case study (Yin, 2009:2). Studies investigating multiple cases increase the external validity. Since this study only focuses on one case, this limits the external validity and the generalizability of this study beyond the case focused on in this study.

Reliability is concerned with estimates of the degree to which a measurement is free of random or unstable error (Cooper & Schindler, 2008:11). The results of a study are reliable when they are independent of particular characteristics of that study and can therefore be replicated in other studies (Yin, 1994, Aken, Berends & van der Bij, 2007). Aken, Berends and van der Bij (2007) describe in their book on research methodology four types of reliability. These four types of reliability are dealt with below.

4.2.1. Researchers and reliability

The reliability of the results of a study is higher when they are independent of the specific researcher conducting the study. To increase this type of reliability, different types of data collection tools are used. Using a questionnaire decreases the possible influence on the results, and therefore makes the results more reliable. Although interviews leaves more room for the researcher to have influence on the results, applying a semi-structured interview with the same questions for every interview limits the amount of influence largely.

4.2.2. Instruments and reliability

Research results should be independent of the specific instruments used (Aken, Berends & van der Bij, 2007). As mentioned earlier, triangulation also helps to increase this type of reliability. Triangulation can remedy the specific shortcomings and biases of the instruments used in this study by complementing and correcting each other (Aken, Berends & van der Bij, 2007). By doing so, triangulation increases the reliability of this study.

4.2.3. Respondents and reliability

Research results should be independent of the respondents included in the study (Aken, Berends & van der Bij, 2007). In this study, the researcher and the programme manager organisational development chose the sample arbitrarily. This decreases the reliability since including different respondents in the sample could result in different answers. A random sample with a larger number of respondents would increase this type of reliability.

4.2.4. Circumstances and reliability

(24)

24

prevent disturbances. Besides this, the meeting rooms also prevent others from listening during the interview. In this way, the interviewees could speak freely.

4.3. Measures

Most measures were derived from existing questionnaires. Some were customized to fit the current study. The measures described below are all in English. The target organisation is located in The Netherlands and to avoid errors related to language, all measures were translated into Dutch by the researcher. Back-translation (Brislin, 1970) was used to increase the accuracy of the translations. Unless mentioned differently, all measures were answered based on a likert-scale ranging from 1= strongly disagree until 7= strongly agree. The questionnaire can be found in Appendix A. Questions marked with an asterisk (*) were reversely coded.

4.3.1. Resistance to change

For the operationalization of the dependent variable, level of resistance to change, Piderit’s (2003) multidimensional view of attitudes toward organisational change was used. This consists out of cognitive, affective and behavioural components. Questions have been drawn from the article of Oreg (2003) where a multifaceted construct based on the ideas of Piderit (2000) was applied to measure resistance to change. The reliability scores for the behavioural, affective and cognitive components were .77, .78 and .86 respectively.

4.3.2. Change approach

The first independent variable is change approach. As described earlier in this study we will focus on a top-down approach to change and a bottom-up approach. Although several authors used these approaches in their articles (see for example: Coram & Burnes, 2001; Ashburner, Ferlie and Fitzgerald, 1996; Conway & Monks, 2011; Sminia & van Nistelrooij, 2006; Ryan, Williams, Charles and Waterhouse, 2008) the researcher of this paper did not find a tool to measure these approaches within an organisational setting in the literature available on this topic. Therefore questions were developed based on the interviews held in the orientation phase. These were the six semi structured interviews described earlier with the two change agents and two team leaders per unit.

4.3.3. Team identification

(25)

25 4.3.4. Uncertainty

Uncertainty was measured by a nine-item scale developed by Bordia et al. (2004) based on the questions of Schwieger and DeNisi (1991). These questions were transformed to fit the specific situation of the specific case used in this study. All responses were made on a Likert scale ranging from 1=very uncertain until 7=very certain. Question focus on uncertainty about the organisational change, own job and new situation.

4.4. Analytical approach

(26)

26

5. Results

In the following section we present the results of the quantitative section of this study. The results are drawn from the analysis conducted on the data set. This data set is based on the results of the questionnaire. As was mentioned earlier in the method section, some variables are measured by multidimensional views. For example, Resistance to change has a Cognitive, Affective and Behaviour component. In this section of this study, the following abbreviations are applied: RTC stands for Resistance To Change, TI stands for Team Identification and CA stands for Change Approach. The results are shown for every individual component.

In this section we will start with a descriptive analyses. Following are the results of the independent sample T-test where the two units are compared. This section ends with the testing of the hypotheses as composed in chapter two of this paper.

5.1. Descriptive analysis

Table 1 shows the descriptive analysis of the data set drawn from the questionnaire. The table shows the mean and the standard error mean for every variable and for every unit. The mean and standard mean error (SE) of the total population are shown in the last column.

Table 2: Descriptive analyses of enquettes DESCRIPTIVES

Unit B Mean (SE) A Unit Mean (SE) Total Mean (SE)

RTC_cognitive 16.06 (1.36) 13.85 (1.16) 14.96 (0.90) RTC_affective 15.95 (1.18) 12.05 (0.86) 14.00 (0.78) RTC_ behaviour 16.15 (1.31) 13.55 (0.94) 14.85 (0.82) TI_self categorization 11.60 (0.87) 11.50 (0.93) 11.55 (0.63) TI_self esteem 22.40 (0.95) 24.15 (0.85) 23.28 (0.65) TI_team commitment 15.45 (1.02) 15.90 (0.82) 15.68 (0.65) CA_Bottom-‐up 14.10 (1.44) 18.15 (0.97) 16.13 (0.92) CA_Top-‐down 24.80 (1.36) 15.70 (1.04) 20.25 (1.12) Uncertainty 36.15 (2.34) 34.75 (1.56) 35.45 (1.41)

5.2. Comparison of the two units

An independent sample T-test was conducted to compare the unit B and the A unit in terms of the level of Resistance to change, Team identification, Change approach and Uncertainty. In the following section, the results of this analysis are shown.

As for RTC_Cognitive, it was found that there was no significant difference between unit B and the A unit, t(38)=1.229, p=0.227. Regarding RTC_Affective, there was a significant difference between Unit B and Unit A at confidence level 95%, t(38)=2.675, p=0.011.

For RTC_Behaviour, it was found that there was no significant difference between the two units, with t (38)=1.610, p=0.117.

(27)

27

As for CA_Bottom-up, there was a significant difference between unit B and the A unit. At confidence level 95%, t(38)=-2.334, p=0.025. Regarding CA_Top-down, it was found that there is a significant difference between Unit B and the A unit at confidence level of 95%, t(38)=5.312, p=0.000.

Regarding Uncertainty, it was found that there was no significant difference between unit B and the A unit, t(38)=0.490, p=0.627.

Analysis: A significant difference was found between the two units for the Affective component of resistance to change. This result show that there was a difference in how people reacted to change and how they felt about the change. Next to this, a significant difference was found between the two units and the change approach used. One plausible explanation is that there are different approaches to change applied by the two change agents. This supports the assumption made in chapter one of this paper.

5.3. Test of Hypotheses

In the following section, the hypotheses are tested. The results of the interviews are presented first followed by the statistical results. First, we will test the hypothesis on the entire data set, while no difference is made between the units. Second, we will split the data set on the basis of the two units. In this part, the hypotheses are tested for the individual units.

5.3.1. Complete data set

The data concerning the complete data set are presented in table 3 and 4 below.

Hypothesis 1. As for our first hypothesis, we hypothesized that change approach concerning bottom-up was negatively correlated with resistance to change, while there is a positive correlation between top-down change approach and the resistance to change.

According to the one-tailed correlation analysis, as for the relationship between bottom-up and RTC-Cognitive, Affective, Behaviour, the null hypothesis was rejected. This means that there was a significant correlation between the CA_Bottom up and the three RTC items. As for the correlation between CA_Bottom-up and RTC_Cognitive, we’ve found a significant, negative and moderate correlation, r=-0.388 (p=0.007). As for the correlation between CA_Bottom-up and RTC_Affective, we’ve found a significant, negative and quite weak correlation, r=-0.299 (p=0.030). As for the correlation between CA_Bottom-up and RTC_Behaviour, we’ve found a significant, negative and quite weak correlation, r=-0.291 (p=0.034).

(28)

28

(p=0.000). As for the correlation between CA_Top down and RTC_Behaviour, we’ve found a significant, positive and moderate correlation, r=0.413 (p=0.004).

Analysis. Based on the above, Hypothesis 1 was supported. The correlation between applying a bottom-up down change approach and Resistance to change indicates that higher levels of bottom-up approach will result in lower levels of Resistance, whereas the positive correlation between applying a top-down approach and Resistance to change indicates that higher levels of a top-down approach will result in higher levels of Resistance to change.

Hypothesis 2a. For the second hypothesis, we hypothesized that higher levels of team identification will lead to higher levels of resistance to change.

According to the one-tailed correlation analysis, as for the relationship between TI_Self esteem and RTC-Cognitive, Affective, Behaviour, the null hypothesis was rejected. This means that there was a significant correlation between TI_ Self esteem and the three RTC items. As for the correlation between TI_Self esteem and RTC_ Cognitive, we’ve found a significant, negative and quite weak correlation, r=-0.292 (p=0.034). As for the correlation between TI_Self esteem RTC_ Affective, we’ve found a significant, negative and moderate correlation, r=-0.458 (p=0.001). As for the correlation between TI_Self esteem RTC_ Behaviour, we’ve found a significant, negative and quite weak correlation, r=-0.290 (p=0.035).

According to the one-tailed correlation analysis, no significant correlation were found for TI_Self categorization, TI_Team commitment and the three RTC items.

Analysis. We failed to support this hypothesis. Contrary to the theory described in chapter one, TI_Self esteem is negatively correlated with Resistance to change. This means that higher levels of TI_Self esteem will result in lower levels of Resistance to the change.

Hypothesis 2b. For this hypothesis, the mediating rol of uncertainty on the relationship between Team identification and Resistance to change is tested. To test this hypothesis, a one-tailed partial correlation was conducted controlling for uncertainty.

According to the one-tailed partial correlation analysis, as for the relationship between TI_Self esteem and RTC_Cognitive, Affective and Behaviour, the null hypothesis was rejected. There was a significant correlation between TI_Self esteem and the three RTC items. As for the correlation between TI_Self esteem and RTC_Cognitive, we’ve found a significant, negative and moderate correlation, r=-.441(0.002). As for the correlation between TI_Self esteem and RTC_ Affective, we’ve found a significant, negative and moderate correlation, r=-0.499 (0.001). As for the correlation between TI_Self esteem and RTC_Behaviour, we’ve found a significant, negative and moderate correlation, r=-0.370 (0.010).

(29)

29

Analysis. Following hypothesis 2a, TI_Self esteem was found negatively correlated with Resistance to change when controlling for Uncertainty. Moreover, the correlation between TI_Self esteem and Resistance to change proved to be stronger when controlling for uncertainty.

Table 3: Correlations for total data set

RTC_Cognitive RTC_Affective RTC_Behaviour

CA_Bottom up -‐0.388 (0.007)** -‐0.299 (0.030)* -‐0.291 (0.034)* CA_Top Down 0.390 (0.006)** 0.501 (0.000)** 0.413 (0.004)* TI_Self categorization 0.080 (0.311) 0.010 (0.497) -‐0.028 (0.432) TI_Self esteem -‐0.292 (0.034)* -‐0.458 (0.001)** -‐0.290 (0.035)* TI_Team commitment -‐0.090 (0.291) -‐0.093 (0.284) 0.018 (0.456)

Table 4: Correlations for total data set after controlling for uncertainty

5.3.2. Splitting the data set

In the following section the data set is split up based on the two units to gain more insights

and obtain more details about the hypotheses. The data are represented in table 5 and 6 below. In the table, the number one represents the results for the B unit and the number two the results for the A unit. The results of the interviews are included in Appendix C

Some remarks concerning the p-values in table 5 and 6. Some correlation were found unsigninficant with values just over p=0.05. For instance, the correlation between TI_Team commitment and RTC_Cognitive for Unit B with p=0.051 (Table 5). This could indicate a Type II error where a correlation shows unsignificant results, although a larger sample size could provide significant results. The same is applicable for p-values just under p=0.05. For instance, the correlation between CA_Bottom-up and RTC Behaviour for UNIT B, p=0.047 (Table 4). Here, a Type I error could be applicable. A larger sample size could result in an insignificant result, p<0.05. We will further adress this in the Discussion Section of this paper.

RTC_Cognitive RTC_Affective RTC_Behaviour

TI_Self categorization 0.079 (0.316) -‐0.002 (0.497) -‐0.036 (0.414) TI_Self esteem -‐0.441 (0.002)** -‐0.449 (0.001)** -‐0.370 (0.010)* TI_Team commitment -‐0.210 (0.100) -‐0.124 (0.226) -‐0.043 (0.397)

**. Correlation is significant at the 0.01 level (1-tailed) *. Correlation is significant at the 0.05 level (1-tailed)

(30)

30

Hypothesis 1. As for the first hypothesis, it was hypothesized that change approach concerning bottom-up was negatively correlated with resistance to change, while there is a positive correlation between top-down change approach and the resistance to change.

Results of the interviews

Based on the interviews it is shown that there was a difference between the two units and the change approach used. All three interviewees of the unit A stated how important the input of the employees of the new unit was during the planning and implementation of the change. “We’ve done this with intensive consultation of all relevant colleagues and employees of the unit”. The employees of the A unit were actively involved and their input was used to write the development plan. During central meetings, but also by informal contact, the employees were informed, asked for input for the next phases and were able to discuss the new plans with the change agent and the team leaders. As one interviewee stated: “we let the employees participate by listening to them about what ideas they had, and what they thought of the plans”.

When there was resistance, for instance because some employees were to be re-placed outside the unit, the change agent addressed this in a personal way. All three interviewees mentioned how the change agent coached the individuals to be able to deal with the change. But the opposite is also true; the lump sum of the employees supported the change leading to positive reactions from the employees to the team leaders and the change agent.

Turning to the second unit, unit B, different result were found. All three interviewees stated that the Change Agent was less interested in the input of the employees. “the change agent did not want that every employee had their say in the change process”. The change agent mainly consulted with her two team leaders to develop the development plan and plan the change process. One interviewee mentioned that one reason for the resistance was the fact that input provided by the employees did not result in changes in the development plan. This lack of identification with the plan resulted in low amounts of support for the plan, leading in the end to resistance. The intention at the start of the process was different. Two interviewees note that if the management team (two team leaders and the change agent) were to prepare the change process, they wanted to do that with the help of the employees.

Statistical results

Unit B. According to the one-tailed correlation analysis, as for the relationship between Bottom-up Change approach and the RTC items the null hypothesis accepted. No significant correlations were found.

(31)

31

Unit B. According tot the one tailed correlation analysis, as for the relationship between Top down approach and RTC_Affective, we’ve found a significant, negative and moderate correlation, r=-0.403 (p=0.039).

A Unit. For the A Unit, no significant correlations were found.

Analysis. This hypothesis was partly supported. The results show that only for the bottom-up approach and for the A unit a significant result was found. This is supported by the results of the interviews. The employees were actively involved during the change process; their input was valued and used to develop the development plan. This resulted in positive reactions and low levels of resistance. Moreover, we also found a significant correlation for the A unit between a top-down approach and RTC_Cognitive. This shows that applying a top-down change approach in the A unit leads to a higher levels of resistance to change. More specific, it shows us that employees of the A unit will think more negatively if a top-down approach was used.

A significant relationship was also found between the B unit and a top-down approach, where a significant relation was found with RTC_Affective. Contrary to the theory and the interviews, a negative relationship was found. This means that a top-down approach resulted in lower levels of resistance to change. This is also contradictory with the interviewees, which stated that because of a lack of influence on the change process and development plan employees started to resist the change.

For the part that higher levels of a bottom-up approach will result to lower levels of resistance to change is fully supported, both by the statistics and the interviewees. For the part that higher levels of a Top-down approach will result in higher levels of resistance is only partly supported. Only the results of the interviews support this hypothesis, whereas the statistics show that it is the other way around; it leads to lower levels of resistance.

Hypothesis 2a. This hypothesis states that higher levels of team identification will lead to higher levels of resistance to change.

Results of the interviews

For both units the level of team identification was noticeable. All six interviewees mentioned how in the old situation before the change, the teams were completely separated. This was the result of three divisions “which had different cultures, directors and work processes”. For the unit A, this resulted in two separated teams, which “hardly ever cooperated with each other”. One reason for this was the physical distance; the teams did not work in the same building. For each unit, all unit employees were placed in the same building and if possible on the same floor. Four out of six interviewees stated that this has not resulted in problems. After an adjustment period, most employees saw the advantages of working together. So although there were large differences between some teams within the unit, none of the interviewees mentioned that this would lead to higher levels of resistance to change on the long term. Both change agents put effort to increase cooperation.

How change approaches and team identification influence the level of resistance to change of employees during a reorganisation.

Table of contents

1. Introduction

2. Literature on change management

3. Conceptual Model

4. Methodology

5. Results