Sie sind auf Seite 1von 249

A Perception-Based View of the Employee: A Study of Employees’ Reactions to Change

DISSERTATION of the University of St. Gallen, Graduate School of Business Administration, Economics, Law and Social Sciences (HSG) to obtain the title of Doctor of Business Administration

submitted by

Chaiporn Vithessonthi

from

Thailand

Approved on the application of

Prof. Dr. Markus Schwaninger

and

Prof. Dr. Günter Müller-Stewens

Dissertation no. 3040

D-Druck-Spescha, St. Gallen 2005

The University of St Gallen, Graduate School of Business Administration, Economics, Law and Social Sciences (HSG) hereby consents to the printing of the present dissertation, without hereby expressing any opinion on the views herein expressed.

St Gallen, January 20, 2005

The President:

Prof. Dr. Peter Gomez

Abstract

Drawing on several theoretical perspectives (e.g., individual motivation, behavioral decision-making, social exchange theories, organizational justice theories, social cognition, institutional theories and neoclassical economics theories) from different disciplines (e.g., organizational psychology, strategic management, and economics), this dissertation developed a perception-based approach to examine a possibility that employees’ perceptions and/or attitudes will be associated with their decisions in an organizational setting. Specifically, this dissertation examined the effects of employees’ perceptions and/or attitudes on their reactions to organizational change. This dissertation addressed two major research questions relevant to organizational change management, organizational behavior and applied psychology. First, it addressed a question of what perceptions and/or attitudes influence employees’ resistance to change. Second, it addressed a question of what perceptions and/or attitudes influence employees’ support for change? This was done by drawing on several theoretical perspectives and examining relationships between perception and/or attitude variables and resistance to change and support for change. Based on data obtained from two samples of respondents from two different settings (i.e., a downsizing in Study 1 and a privatization in Study 2), this dissertation found significant relationships between perceptions and/or attitudes and resistance to change and/or support for change. The findings provide some empirical support for the perception- based view of the employee. Using multinomial ordered probit modeling, some perceptions and/or attitudes were found to be significantly predictive of employees’ reactions to change. The potential practical value of using perceptions and/or attitudes as predictors of employees’ reactions to change is discussed, as are implications and suggestions for future research.

i

Acknowledgements

This dissertation began with a conversation with Professor Dr. Markus Schwaninger, a professor of management at the University of St. Gallen, in the summer of 2003, when I was about to finalize my master’s degree in international management at the same university. During this conversation we discussed organizational change, and since I had thought that it might make an interesting piece of research, I asked him about the possibility to write the dissertation that lies before you today. As I expected, his response was clear, insightful, interesting, and encouraging. He enthusiastically agreed to supervise my dissertation and told me to proceed with my ideas. So it began. I am reminded as I finalize these notes of my good fortune in being able to do something I enjoy, and to complete my research. It is the rarest of privileges for me, with my limited ability, to do that in a relatively short span of time; this seems tremendously precious to me. But this work could not have been completed without support from many people. I owe a debt of gratitude to the 315 respondents who took time out of their busy schedules to complete and return the questionnaire. I am extremely grateful to Prof. Dr. Markus Schwaninger, who has been not only the referee for this dissertation but also my mentor throughout the past years, for offering his invaluable help, comments, perspectives, and suggestions, and for showing great interest in my research. Undoubtedly, he has pointed me in the direction of a fascinating landscape, not for the first time and, I hope, not for the last.

I also want to express my sincere gratitude to Professor Dr. Günter Müller-Stewens, who has magnanimously taken time out of his busy schedule to become the co-referee, for offering his valuable insights and perspectives on the theoretical, methodological and empirical aspects of my dissertation. I am very grateful to Dr. Klaus Edel as well, not only for offering his valuable suggestions and solutions to statistical issues with enthusiasm, but also for allowing me to use his computer and statistical applications. I am also grateful to Silke Bucher, Bernd Beuthel, and Jasmina Hasanbegovic for their thoughtful and constructive feedback on earlier versions of this dissertation. And, of course, I thank Linda Roberts, my editor and proofreader, at Western Illinois University, who shouldered the editorial and proofreading work on my unpolished lines of English. Last but not least, I would like to thank my parents for their love, incredibly great confidence, and unbounded support throughout the course of this journey and beyond.

Basel, January 2005

ii

Chaiporn Vithessonthi

Table of Contents

Abstract

i

Acknowledgements

ii

List of Tables

vi

List of Diagrams and Figures

 

viii

List of Abbreviations

x

1. Introduction

1

1.1. Research

Issues

 

1

1.2. Research

Questions

3

1.3. The Importance of the Research Questions

6

1.4. The Scope of the Dissertation

 

8

1.5. The Intended Contributions of this Dissertation

10

2. Core Concepts and Relevant Literature

12

2.1. Theories of Change

 

12

2.2. Perceptions

16

2.3. Attitude

19

2.4. Emotion

21

2.5. Individual Decision-Making

 

23

2.6. Reactions to Change

 

27

3. Theoretical Development and Research Model

31

3.1. Perception-Based View of the Employee

32

3.2. Research Model and Hypotheses

35

3.2.1. Perceived

Organizational Support

38

3.2.2. Perceived Procedural Justice

40

3.2.3. Perceived Participation in Decision-making

42

3.2.4. Perceived

Need for Change

45

3.2.5. Attitude towards Organizational Change

48

3.2.6. Fear of Known Consequences of a Change

50

3.2.7. Fear of Unknown Consequences of a Change

52

3.2.8. Perceived

Change

in

Power

54

3.2.9. Perceived

Change

in

Status

56

3.2.10. Perceived

Change

in

Pride

58

3.2.11. Job

Satisfaction

 

60

3.2.12. Job

Security

62

iii

3.2.13.

Job Motivation

64

 

3.2.14. Perceived Employability

66

3.2.15. Self-Confidence for Career-Relevant Learning

69

3.2.16. Affective Commitment

71

3.2.17. Trust in Management

73

3.2.18. Colleagues’ Reactions to Change

75

4.

Research Methodology

77

 

4.1.

Context, Sample and Procedure

78

 

4.1.1. Study 1 – Context, Sample and Procedure

78

4.1.2. Study 2 – Context, Sample and Procedure

79

 

4.2. Alternative Methods of Data Analysis

80

4.3. The Multinomial Ordered Probit Model

81

4.4. Measures of Theoretical Constructs

83

 

4.4.1. Dependent Variables

83

4.4.2. Independent Variables

84

4.4.3. Control Variables

87

 

4.5.

Data Analysis Procedures

87

5.

Results and Discussion

89

 

5.1. Study 1 – Results and Discussion

89

 

5.1.1. Analyses of Correlations among Dependent Variables

89

5.1.2. Analyses of Correlations among Independent Variables

90

5.1.3. Results for Hypotheses – The Multinomial Ordered Probit Models

94

5.1.4. Discussion of Study 1

106

 

5.2. Study 2 – Results and Discussion

108

 

5.2.1. Analyses of Correlations among Dependent Variables

109

5.2.2. Analyses of Correlations among Independent Variables

110

5.2.3. Results for Hypotheses – The Multinomial Ordered Probit Models

114

5.2.4. Discussion of Study 2

127

 

5.3. General Discussion

131

 

5.3.1. Key Contributions of the Dissertation

131

5.3.2. Limitations to this Dissertation

136

5.3.3. Implications and Directions for Future Research

138

5.3.4. Implications and Directions for Practice

139

6.

Conclusions

140

References

142

Appendices

165

iv

Appendix A: Questionnaire Survey Items for Studies 1 and 2

165

Appendix B: Study 1 – Diagrams and Correlations

171

Appendix C: Study 2 – Diagrams and Correlations

195

Appendix D: Additional Regression Analyses for Study 2

219

Curriculum Vitae

236

v

List of Tables

Table 1:

Characteristics of Alternative Regression Models

81

Table 2:

Study 1 – Correlations for All Final Variables

93

Table 3:

Study 1 – Regression Results of Active Resistance to Change

95

Table 4:

Study 1 – Regression Results of Passive Resistance to Change

96

Table 5:

Study 1 – Regression Results of Active Support for Change

97

Table 6:

Study 1 – Regression Results of Passive Support for Change

98

Table 7:

Study 2 – Correlations for All Final Variables

113

Table 8:

Study 2 – Regression Results of Active Resistance to Change

115

Table 9:

Study 2 – Regression Results of Passive Resistance to Change

116

Table 10: Study 2 – Regression Results of Active Support for Change

117

Table 11: Study 2 – Regression Results of Passive Support for Change

118

Table 12: Summary of Results for Hypotheses in Study 1 and Study 2

132

Table 13: Study 1 – Correlations for All Dependent Variables

182

Table 14: Study 1 – Correlations for Active Resistance

183

Table 15: Study 1 – Correlations for Active Resistance (cont.)

184

Table 16: Study 1 – Correlations for Active Resistance (cont.)

185

Table 17: Study 1 – Correlations for Passive Resistance

186

Table 18: Study 1 – Correlations for Passive Resistance (cont.)

187

Table 19: Study 1 – Correlations for Passive Resistance (cont.)

188

Table 20: Study 1 – Correlations for Active Support

189

Table 21: Study 1 – Correlations for Active Support (cont.)

190

Table 22: Study 1 – Correlations for Active Support (cont.)

191

Table 23: Study 1 – Correlations for Passive Support

192

Table 24: Study 1 – Correlations for Passive Support (cont.)

193

Table 25: Study 1 – Correlations for Passive Support (cont.)

194

Table 26: Study 2 – Correlations for Dependent Variables

206

Table 27: Study 2 – Correlations for Active Resistance

207

Table 28: Study 2 – Correlations for Active Resistance (cont.)

208

Table 29: Study 2 – Correlations for Active Resistance (cont.)

209

Table 30: Study 2 – Correlations for Passive Resistance (cont.)

210

Table 31: Study 2 – Correlations for Passive Resistance (cont.)

211

Table 32: Study 2 – Correlations for Passive Resistance (cont.)

212

Table 33: Study 2 – Correlations for Active Support

213

vi

Table 34: Study 2 – Correlations for Active Support (cont.)

214

Table 35: Study 2 – Correlations for Active Support (cont.)

215

Table 36: Study 2 – Correlations for Passive Support

216

Table 37: Study 2 – Correlations for Passive Support (cont.)

217

Table 38: Study 2 – Correlations for Passive Support (cont.)

218

Table 39: Summary of Regression Results of Indicators for Resistance to Change

223

Table 40: Summary of Regression Results of Indicators for Support for Change

224

Table 41: Regression Results of Active Resistance to Change 1

225

Table 42: Regression Results of Active Resistance to Change 2

225

Table 43: Regression Results of Active Resistance to Change 3

226

Table 44: Regression Results of Passive Resistance to Change 1

227

Table 45: Regression Results of Passive Resistance to Change 2

228

Table 46: Regression Results of Passive Resistance to Change 3

229

Table 47: Regression Results of Active Support for Change 1

230

Table 48: Regression Results of Active Support for Change 2

231

Table 49: Regression Results of Active Support for Change 3

232

Table 50: Regression Results of Passive Support for Change 1

233

Table 51: Regression Results of Passive Support for Change 2

234

Table 52: Regression Results of Passive Support for Change 3

235

vii

List of Diagrams and Figures

Figure 1:

Dimensions for Categorization of Reactions to Change

29

Figure 2:

A Categorization of Reactions to Change

30

Figure 3:

Alternative Models Relating Perceptions and Reactions to Change

31

Figure 4:

Conceptual Diagram of the ‘Direct Effects’ Model

37

Figure 5:

Five Stages of Organizational Decline

46

Figure 6:

Summary of Measures of Reactions to Change

84

Figure 7:

Summary of the Sequence of Data Analysis

88

Figure 8:

Study 1 - Indicators for Active Resistance to Change

171

Figure 9:

Study 1 - Indicators for Passive Resistance to Change

171

Figure 10: Study 1 - Indicators for Active Support for Change

172

Figure 11: Study 1 - Indicators for Passive Support for Change

172

Figure 12: Study 1 - Indicators for Perceived Organizational Support

173

Figure 13: Study 1 - Indicators for Perceived Procedural Justice

173

Figure 14: Study 1 - Indicators for Perceived Participation in Decision-Making

174

Figure 15: Study 1 - Indicators for Perceived Need for Change

174

Figure 16: Study 1 - Indicators for Attitude towards Organizational Change

175

Figure 17: Study 1 - Indicators for Fear of Known Consequences of a Change

175

Figure 18: Study 1 - Indicators for Fear of Unknown Consequences of a Change

176

Figure 19: Study 1 - Indicators for Perceived Change in Power

176

Figure 20: Study 1 - Indicators for Perceived Change in Status

177

Figure 21: Study 1 - Indicators for Perceived Change in Pride

177

Figure 22: Study 1 - Indicators for Job Satisfaction

178

Figure 23: Study 1 - Indicators for Job Security

178

Figure 24: Study 1 - Indicators for Job Motivation

179

Figure 25: Study 1 - Indicators for Perceived Employability

179

Figure 26: Study 1 - Indicators for Self-Confidence for Learning

180

Figure 27: Study 1 - Indicators for Affective Commitment

180

Figure 28: Study 1 - Indicators for Trust in Management

181

Figure 29: Study 1 - Indicators for Perceptions of Colleagues’ Resistance to Change

181

Figure 30: Study 2 - Indicators for Active Resistance to Change

195

Figure 31: Study 2 - Indicators for Passive Resistance to Change

195

Figure 32: Study 2 - Active Support for Change Indicators

196

Figure 33: Study 2 - Indicators for Passive Support for Change

196

viii

Figure 34: Study 2 - Indicators for Perceived Organizational Support

197

Figure 35: Study 2 - Indicators for Perceived Procedural Justice

197

Figure 36: Study 2 - Indicators for Perceived Participation in Decision-Making

198

Figure 37: Study 2 - Indicators for Perceived Need for Change

198

Figure 38: Study 2 - Indicators for Attitude towards Organizational Change

199

Figure 39: Study 2 - Indicator for Fear of Known Consequences of a Change

199

Figure 40: Study 2 - Indicators for Fear of Unknown Consequences of a Change

200

Figure 41: Study 2 - Indicator for Perceived Change in Power

200

Figure 42: Study 2 - Indicators for Perceived Change in Status

201

Figure 43: Study 2 - Indicators for Perceived Change in Pride

201

Figure 44: Study 2 - Indicators for Job Satisfaction

202

Figure 45: Study 2 - Indicators for Job Security

202

Figure 46: Study 2 - Indicators for Job Motivation

203

Figure 47: Study 2 - Indicators for Perceived Employability

203

Figure 48: Study 2 - Indicators for Self-Confidence for Learning

204

Figure 49: Study 2 - Indicators for Affective Commitment

204

Figure 50: Study 2 - Indicators for Trust in Management

205

Figure 51: Study 2 - Indicators for Perceptions of Colleagues’ Resistance to Change

205

ix

List of Abbreviations

AR

Active resistance to change

AR1

Indicator 1 for active resistance to change

AR2

Indicator 2 for active resistance to change

AR3

Indicator 3 for active resistance to change

AS

Active support for change

AS1

Indicator 1 for active support for change

AS2

Indicator 2 for active support for change

AS3

Indicator 3 for active support for change

e.g.

Exempli gratia; (for example)

etc.

Et ectera (and so forth)

i.e.

Id est; (that is)

IIA

The independence of irrelevant alternatives

IPO

Initial Public Offering

OLS

Ordinary least square

PBV

Perception-Based View (of the employee)

POS

Perceived organizational support

PR

Passive resistance to change

PR1

Indicator 1 for passive resistance to change

PR2

Indicator 2 for passive resistance to change

PR3

Indicator 3 for passive resistance to change

PS

Passive support for change

PS1

Indicator 1 for passive support for change

PS2

Indicator 2 for passive support for change

PS3

Indicator 3 for passive support for change

S.E.

Standard error

x

1.

Introduction

1.1. Research Issues

The starting point for this research is the challenge of “managing change in organizations.” Managing organizational change is problematic: situations in which changes are undertaken are shifting, it is harder for organizations, and in particular top managers as well as change agents, to prepare for and manage the change in ways that satisfy the demands of both the organization and its employees. 1 How do organizations go about making “structured” and “unstructured” decisions concerning how to cope with resistance to change, so that they achieve the goals of their organizational change efforts? It is not surprising that, over the years, resistance to change has attracted increasing attention from researchers, practitioners, and the general public. A great deal of research has focused on understanding the sources and determinants of resistance to change. The media and the general public are generally interested in various forms of active resistance to change such as strikes or protests. Other forms of resistance such as passive resistance, although less observable, have not gone unnoticed and thus have also warranted extensive research over the years. 2 Not surprisingly, resistance to change is frequently reported as being one of the sources of organizational change failures (Coch and French, 1948; Kotter, 1995; Kotter and Cohen, 2002). Broadly speaking, the concept of organizational change (e.g., Meyer, 1982; Nadler, 1998) refers to an effort or a series of efforts designed to modify certain aspects or configurations of an organization: for example, identity, goals, structure, work processes or human resources. Furthermore, ideas of organizational learning (e.g., Argyris, 1990; Argyris and Schön, 1978; Fiol and Lyles, 1985; Crossan and Berddrow, 2003) or strategic flexibility (e.g., Harrigan, 1985; Sanchez, 1995; Raynor and Bower, 2001) that emphasize the extent to which a firm is capable of learning and adapting itself to changing environments are associated with the antecedents and outcomes of organizational change. We can thus see that the growing attention to these concepts has enhanced both the frequency and scale of organizational change efforts. Hence, one can reason that the likelihood of employees facing some type of organizational change is higher than ever before.

1 According to the institutional school of organizational thought, individuals in organizations have their own interests and generally try to make use of organizations for their own interests. For a more detailed discussion of these problems, see Selznick (1965) or Meyer and Rowan (1977).

2 To me it seems that we should distinguish between active resistance and passive resistance. This view is consistent with those of Hultman (1998) and Judson (1991). For a more extensive discussion of reactions to change, see Section 2.6 of this dissertation.

1

In addition to the greater level of exposure of employees to organizational change, managers within most organizations are also experiencing greater internal and external pressures to initiate change within their organization in order to maintain or improve firm performance. These pressures include, for example, increased competitive pressures (Meyer, Brooks and Goes, 1990), new government regulations (Meyer et al., 1990; Haveman, 1992; Fox-Wolfgramm, Boal and Hunt, 1998), technological change (Haveman, 1992), or declining firm performance (Bibeault, 1982). Given the above-mentioned environments, research on organizational change has been enriched by both empirical and theoretical studies investigating many aspects of organizational change such as change strategies, change processes, or antecedents and outcomes of different forms of change. 3 To search for conditions that promote successful change in organizations, it is crucial to know the implications or organizational change for employees, and more importantly, the reactions employees will have. Much of the past research on employees’ reactions to change seems to have been implicitly based on a rational choice theory about employees’ behaviors 4 , thereby giving little attention to the potential effects of perceptions, attitudes, or social influence on decisions and behaviors. Indeed, rational choice theories 5 have long dominated the research in organization theory, which encompasses research on organizational change and development In their roughest form, rational choice theories would assert that when organizational change efforts are understood to be beneficial to a firm, employees in this firm should support such changes. This raises the question of whether all employees do in fact share the same view on this change. What are the implications for their decisions if they do not share the same view? Within the large body of research on decision-making in the literature on strategic management or management science, several concepts and underlying assumptions—for example, cost-benefit analysis and human rationality—seem to have conditioned both the theoretical and empirical research in organizational change and employees’ reactions to

3 Organizational change can be considered as a class of organization theory.

4 Note that it is important to understand choice theories and underlying assumptions about how people make choices because any kind of reactions to change—for example, resistance to change and support for change—is an outcome of choice-selection process. In its simplest form, the economic model of decision-making assumes that managers have perfect information and thus could make decisions that maximize profits. For a critique of the economic model of decision-making, see Simon (1957) and March and Simon (1958). For a more extensive discussion of the concept of rationality, see Section 2.5 of this dissertation.

5 Simon (1978, 1985, 1986) pointed out that there are at least two main forms of human rationality in social science:

one of them is in an area of cognitive psychology; and the other is in an area of economics. In this dissertation, unless stated otherwise, both rational choice theories and human rationality shall refer to the form of human rationality in the field of neoclassical economics. It is important to note that in economics there are variations in the concepts of rationality.

2

change. This view is consistent with that of Rumelt, Schendel, and Teece (1991), who have suggested that the logic of economics has dominated the field of strategic management. Only recently have researchers become aware of the limitations of decision-making models in economics, and thus have applied a cognitive paradigm in their research on strategic decision-making (e.g., Schwenk, 1984, 1995; Tversky and Kahneman, 1974). 6 Employees who are confronted with changes in their organization face an inevitable choice: whether they should support or resist such changes in order to still (or best) achieve their personal goals and objectives. Despite a large body of normative literature on techniques for managing change, for example, models of implementing change by Judson (1991), Kotter (1995), Galpin (1996), and Kotter and Cohen (2002), empirical studies of their application seem to be too sparse to indicate convincingly and conclusively whether the techniques presented in those models have had significant influences on employees’ reactions to change. Because I do not share the views and assumptions of some prior researchers 7 , this dissertation theoretically deviates from the mainstream research on change management by introducing a perception-based view of the employee as an alternative approach to understand employees’ reactions to change.

1.2. Research Questions

Researchers and practitioners alike posit that employees’ reactions to change have critical implications for change implementation and firm performance (e.g., Kotter, 1995; Kotter and Cohen, 2002). For instance, the issue of intraorganizational conflict as a serious challenge for managers in making strategic asset decisions has been highlighted (Amit and Schoemaker, 1993). The question of how firms, managers or consultants can minimize employees’ resistance to change is a subject of debate and further research. There are a number of theoretical and practical questions, some of which lie more in the area of philosophy than in the area of change management or social science. In this dissertation I am particularly concerned with the role of perceptions and attitudes and how these might constitute determinants of employees’ reactions to change. These perceptions and attitudes about change (e.g., perceived need for change, perceived change in power, and job security) are theorized to be factors leading to subsequent conscious or unconscious decisions and/or behaviors in response to changes in organizations, which may significantly impact the

6 For a more extensive discussion of decision-making, see Section 2.5 of this dissertation.

7 For a more extensive discussion of assumptions in prior studies and assumptions made in this study, see Section 3 of this dissertation.

3

change implementation and firm performance. In particular, this dissertation attempts to answer two research questions:

What perceptions and/or attitudes influence employees’ resistance to change?

What perceptions and/or attitudes influence employees’ support for change?

With the above research questions, I advance and test an argument that perceptions and/or attitudes influence employees’ reactions to change. Rather than hindering or substituting for current change management models where the main focus is on human rationality, well- understood effects of perceptions and/or attitudes may actually promote more comprehensive, effective and pragmatic change management models designed for promoting employees’ support for change and/or for reducing employees’ resistance to change. In response to seemingly limited empirical evidence on the effectiveness of most change management models, well-understood effects of perceptions and attitudes on reactions to change narrow the domain of potentially key factors influencing employees’ reactions to change to which an organization should pay attention. In addition, these research questions are consistent with contemporary research on the role of psychological factors in predicting employees’ behaviors in response to various types of decisions of organizations, but the role of several psychological factors require empirical verification. Thus, this dissertation attempts to fill a gap in current empirical research by empirically examining relationships between several perceptions and/or attitudes on the one hand and resistance to and support for change on the other hand. Despite evidence that certain change management practices during organizational change are related to employees’ resistance to and/or support for change rates (i.e., a percentage of the total number of employees who support or resist a change to the total number of employees) at the organizational level, it would be a fallacy to then assume that such practices are similarly and/or directly related to employees’ resistance to and/or support for change decisions at the individual level. Thus, it is critical to explain the relationship between any type of change management practices and resistance to and/or support for change at the individual level. The results in this dissertation may help scholars explain such relationship by providing a connecting answer. Rather than answering the question of the effect of change management practices on employees’ reactions to change directly, empirical evidence of the role of perceptions and/or attitudes in predicting employees’ reactions to change may promote a better understanding of psychological factors influencing employees’ reactions to change. If certain change management practices were found to influence these perceptions and/or attitudes, then such practices may thereby

4

have the effect on employees’ reactions to change. Thus, the findings in this dissertation add to the change management literature by examining relationships between perceptions and/or attitudes held by employees and their reactions to change. Given these research questions, this dissertation has three major objectives.

In order to provide foundations for developing a theoretical framework in this dissertation, the first objective is to review prior research on perceptions, attitudes, decision-making, theories of change, and employees’ reactions to change. It should be noted that the literature review on decision-making focuses on the normative and cognitive decision theories in the fields of management and economics.

The second objective is to conceptualize a theoretical framework representing the link between various perceptions and attitudes on the one hand, and resistance to change and support for change, on the other hand. Here I propose to bring several theoretical perspectives together, creating a more realistic model of employees’ reactions by combining different conceptions of human rationality. The main aim of the research model is to investigate which perceptions and attitudes are associated with resistance to change and/or support for change. Additionally, it aims to provide theoretical and, perhaps, practical insights to organizations, top managers, as well as change agents to assist them in developing tools that may detect and alter employees’ perceptions and attitudes in order to (minimize resistance to change and) optimize support for change.

The final objective is to empirically test the hypothesized relationships presented in the research model by gathering and analyzing relevant empirical data in a systematic way.

After having identified the main research questions, the next step is to decide the appropriate level of analysis: employee, top manager, or firm level. To answer the research questions above, the employee, not the firm, will be the unit of analysis in this dissertation. Using the employee as the unit of analysis, one can explore a perception or attitude as a predictor of employees’ reactions to change. Further, examination at the decision level of analysis—that is to say, resistance to change and support for change—diminishes at least two concerns. First by relating perceptions and attitudes rather than decision-making process to reactions to change, casual ambiguity is not an issue since (1) the relationships between perceptions and reactions to change are more direct; and (2) such analyses do not have to deal with the extent to which an employee uses rational decision-making processes. Second, to use the decision level of analysis, it is not necessary to assume that employees

5

consistently use specific processes across time or decisions, thereby allowing the notion that different decision-making processes may be at work for each reaction to change. Past research has posited that it is a key interest of a firm to appropriately deal with resistance to change in order to achieve the goals of organizational change efforts (Coch and French, 1948; Kotter, 1995; Kotter and Cohen, 2002). However, it is important to know whether resistance to change always has a negative impact on change efforts and thus firm performance, or whether there might at times be a counter-intuitive implication, i.e., a positive effect, on change efforts and thus firm performance. Despite the claim that resistance has a negative effect on change efforts and therefore should be minimized (e.g., Kotter, 1995; Kotter and Cohen, 2002), the question is, then, whether resistance to change may be strategically valuable or positive when it acts as a means to ensure that the change is indeed designed and implemented to promote an organization’s goals. Thus, one can question the accuracy of the claim made by some researchers (e.g., Coch and French, 1948) that resistance to change is always undesirable. This question seems to have gone unnoticed, providing little recognition of the conditions under which resistance to change may result in superior outcomes of organizational change. Although identifying conditions in which resistance to change has positive or negative outcomes on change processes is not the goal of this dissertation itself, it deserves mention so as to reflect on this issue. In order to state that resistance to change always has negative implications for the firm, one would have to show that such resistance can legitimately be considered negative at any given moment and in any particular circumstance. If this same resistance does not create a negative implication for the firm at another moment and in another similar circumstance, one may not legitimately and precisely conclude that resistance to change is always undesirable and negative. On the other hand, it is probable that resistance to change may at times have a positive effect on the outcome of organizational change, and that it may be strategically valuable to an organization. For example, it is imaginable that resistance to change could be constructive by entailing a high degree of objective evaluation of the change. This should suggest that researcher should not make the critical assumption that resistance to change always has negative effects on the outcomes of organizational change. Instead, they should investigate how to benefit from resistance to change.

1.3. The Importance of the Research Questions

Clearly, improved firm performance is one of the main objectives of organizational change, but intermediate outcomes are more proximal indicators of its success or failure. Employees’ performance can be considered as an immediate outcome or a path through which changes in organizations affect firm performance. Therefore, one can also reason that

6

employees’ resistance to or support for change which are arguably predictive of their performance at the time of the organizational change, can be seen as an indicator predicting the probability of success of the change (Kotter, 1995; Kotter and Cohen, 2002). Thus, the optimization of resistance to change and support for change may enhance the probability of success of organizational change, thereby improving firm performance. In order to optimize these levels of resistance to change or support for change, we need to understand the factors that play an important role in creating or changing them. In this dissertation I focus on the direct relationships between perceptions and attitudes on the one hand, and resistance to change and support for change on the other hand. This perspective assumes that employees’ perceptions and attitudes are likely to influence their reactions to change. I take this a step further by proposing that if we know which perceptions or attitudes affect levels of resistance to change and support for change, we will then have opportunities to develop tools to properly influence reactions to change. Although individual decision-making processes are not programmed, they are programmable (Mintzberg, Raisinghani and Theoret, 1976): the underlying basis of this argument is that strong evidence indicates that a basic logic or structure underlies the actions of a decision-maker and that this structure can be identified by a systematic study of his or her behavior 8 . If individuals have patterns of decision-making processes, the study of employees’ reactions to change may yield some valuable insight. That is, if certain perceptions or attitudes are associated with certain decisions at a given moment and in a particular setting, the same pattern of relationships may persist at other moments and in other settings. One assumption in this dissertation is that employees’ decisions are based on their interpretation and evaluation of the data available to them. As this data is collected and interpreted, different employees may arrive at different perceptions, interpretations and understandings of the same data. Consistent with Simon’s (1957) concept of ‘bounded rationality’ in decision-making processes 9 , I argue that there are potential gaps between the object’s (e.g., organizational change’s) ‘objective’ (what they actually are) characteristics and ‘perceived’ (what people believe or perceive them to be) characteristics, and that the

8 Note that in this dissertation, words like “decision”, “behavior”, or “reaction” are used interchangeably since they all refer to an employee’s resistance to change and/or support for change. In its simplest form, one may find that resistance to change and support for change can be considered as one kind of decision, and that both resistance to change and support for change are expressed in terms of behaviors or reactions.

9 Simon (1957) has discussed the two main problems with the economic model of decision-making; first, managers seldom have perfect information and thus often have to make decisions under uncertainty. Second, managers are not cognitively capable of processing all of the information that they would need to make a profit-maximizing decision. Therefore, Simon (1957) has introduced the concept of “bounded rationality” for the model of decision-making.

7

‘perceived’ characteristics rather than the ‘objective’ characteristics are used as inputs in the decision-making processes determining decisions. Thus, perceptions of the object play an important role in decision-making processes and result in decisions that at least partially reflect these perceptions. As I reflect on economic theories that seem to explain well the utilitarian side of human behavior, but seem to fail to explain the side of human behavior that goes beyond outcome-driven self-interest, I want to explore an alterative approach in understanding employees’ reactions to change, and label this approach the “perception-based view of the employee.” In short, using perception-based logic, this dissertation focuses on the role of perceptions and attitudes as the driving forces leading employees to either support or resist organizational change. Understanding the ways in which employees react to change will certainly provide a potential avenue for developing new change management strategies that may bring employees’ perceptions into alignment with ones desired, thereby eliciting desired reactions to change.

1.4. The Scope of the Dissertation

Obviously, the field of organizational change and its scientific investigation is manifold. For instance, archetypes of a firm’s organizational change can be neatly classified into five groups or dimensions: (1) identity; (2) strategy; (3) business processes; (4) structure; and (5) human resources. Each of these dimensions has different implications for an organization as well as its employees. 10 In view of the fact that organizational change can take on many forms, this dissertation focuses on two aspects of organizational change: “downsizing,” which can be subsumed into the “firm structure” dimension, and “privatization.” which is part of the “firm strategy” dimension. 11 In Study 1, a downsizing effort was chosen because this kind of change typically has direct and significant implications for employees, who directly experience the effects of these changes. For example, employees may have to increase their productivity (Hambrick & Schecter, 1983) or risk losing their jobs. Further, downsizing has often been employed as a means to improve firm performance (Freeman & Cameron, 1993). Additionally, firms in a crisis situation often downsize as part of their turnaround strategies (Robbins & Pearce, 1992; Appelbaum, Everard & Hung, 1999). Similarly, there were

10 Donaldson (1987) pointed out that organizational change can be thought of as an adjustment of strategy, structure, or processes of an organization. For a more extensive discussion of organizational change, see Section 2.1 of this dissertation.

11 It is noteworthy that from different perspectives, any kind of organizational change can be classified into more than one category.

8

numerous reasons for choosing privatization as a context of the study. First, many state- owned enterprises in several countries are, or will eventually be, in a process of privatization, and this privatization has several implications for markets, other firms, and their employees. Second, in addition to being a change in itself, privatization is also a source of other changes within an organization—for example, changes in corporate strategies and corporate structures. Consequently, it is important to note that a study of employees’ reactions to privatization not only has to deal inclusively with reactions to privatization but also with reactions to a set of changes that come along with the privatization initiative in a broader sense. This dissertation aims to develop a research model which suggests relationships between perceptions and attitudes on the one hand and resistance to change and support for change on the other hand, and to empirically test it by using data gathered from employees currently facing organizational change. The number of variables examined in this dissertation is limited predominantly due to two key reasons: theoretical aspect (the greater the number of variables in the model, the less the degree of parsimony of the model), and practical aspect (the greater the number of variables in the model, the lower the response rates in the survey). This dissertation focuses on empirical evidence gathered at a particular point in time from employees in two organizations. In Study 1, a survey was distributed to a random sample of 100 teachers at a large private school in Thailand where the management has recently decided to reduce the number of teachers. Of those sampled, 91 teachers returned the surveys (91% response rate). In Study 2, a survey was distributed to 500 employees at a large state-owned company in Thailand where top managers have attempted to privatize the organization. Of those sampled, 224 employees returned the survey (44.8% response rate). The focus of this dissertation is strictly limited to the examination of the relationships between perceptions and attitudes on the one hand and reactions to change on the other hand at the given moment in time rather than during different points in time. Thus, it is not a longitudinal study. This implies that these studies did not investigate feedback-loops or a so- called dynamic model 12 that addresses: (1) the effects of employees’ resistance to and support for change on the change efforts (e.g., the change goals and processes); and (2) the perceptions of the modifications in organizational change efforts at time t 1 as a consequence of employees’ reactions at time t 0 on their reactions to such modifications at time t 1 . One

12 As March (1955) pointed out problems in determining influence order, one may consider that the influence relationships in this dissertation may represent closed-loop systems. For a detailed discussion of this issue, see March (1955).

9

reason why this dissertation does not include the feedback-loops model concerns the casual link between perceptions or attitudes, and reactions to change. If we were to find such a relationship at one moment in time, then we might expect to find that relationship at another moment in time. Another reason concerns the practical aspect of developing and validating the feedback-loops model using data from the questionnaire surveys: this would require not only comparing results of different surveys but also gathering data from the same respondents at different times, which would be too problematic or beyond the scope of this study. It is important to note that the nature and magnitude of the impact of organizational change on employees depends on, among other things, the type of change and the way in which the change is introduced. For example, changes can be initiated either from top management (a so-called top-down approach) or from employees (a so-called bottom-up approach). Because I assumed that the strength of the impact of the change was inherently expressed in the perceptions of the employees, it was not necessary to separately explore the effects of the change on the employees, or distinguish how the changes were introduced. However, because the changes studied in this dissertation entailed organization-specific, situation-specific, time-specific, and relationship-specific contexts, the extent to which the findings can be generalized to other contexts is limited. It is also useful to note that the implications of cultural differences on reactions to change are not within the scope of this dissertation. Thus, the examination of relationships between predictors and outcomes within one culture (Thailand) is conducted.

1.5. The Intended Contributions of this Dissertation

The principal thesis that emerges from the research model is that employees who are confronted with any form of organizational change tend to develop the initial and subsequent reactions to this change through a variety of decision-making processes. Consistent with the bounded rationality framework (Simon, 1957), this dissertation further argues that certain perceptions and attitudes enhance or prohibit their choices of reactions to change. Specifically, this dissertation focuses on employees’ perceptions and attitudes in a downsizing situation (in Study 1) and a privatization situation (in Study 2). These perceptions and attitudes are theorized to be factors leading to subsequent conscious and/or unconscious decisions and/or behaviors in response to the changes, which may significantly impact the change implementation and firm performance. This dissertation attempts to contribute to the research on organizational change, especially employees’ resistance to change and support for change, in three ways.

10

First, this dissertation examines a variety of actions that employees may choose in response to a change in the organization. Drawing upon prior research, it identifies two primary types of employees’ reactions to change: resistance to change and support for change. These are further divided into: active and passive resistance, and active and passive support. Second, using perception-based logic, this dissertation examines a number of perceptions and attitudes that may influence employees’ choice (conscious or unconscious) to support or resist change, thus shedding light on whether perceptions and attitudes impel, impede or exert no effect on employees’ behavior and decisions. Although the organizational change literature is rich, there seems to be a surprising gap in the literature concerning the role and nature of employees’ perception of change in organizations. In particular, this dissertation aims to contribute in this area by examining the link between various perceptions and attitudes on the one hand and resistance to change and support for change on the other hand. Consensus on these issues will allow theories of employees’ reaction to change to move forward in a systematic fashion. Third, based on the findings in this dissertation, it is probable that we will be able to develop a variety of tools for predicting employees’ reactions to change. More importantly, understanding the ways in which employees establish certain reactions to change will provide a potential avenue for developing a range of change management strategies that may bring employees’ perceptions in alignment with those desired, thereby strengthening the degree to which employees support organizational change. Specifically, the key findings are mainly relevant to the design, implementation, and closing phases of change management strategies in Thailand. Nonetheless, it goes without saying that part of the knowledge derived from the present dissertation will be applicable and transferable to firms in other geographical settings and industries wishing to introduce a change, any change, to their organization. The findings in this dissertation will also be informative for consultants, as a way to improve the current change management practices in dealing with employees’ resistance to change. As discussed earlier, employees’ resistance to change is reported to be a source of problems for organizations and has subsequent negative effects on firm performance. Understanding employees’ perceptions and attitudes before, during, and after the implementation of organizational change may prove to be valuable to firms, managers, and consultants.

11

2. Core Concepts and Relevant Literature

This section discusses the central tenets of theories of change, perception, attitude, emotion, individual decision-making, resistance to change and support for change, focusing on the core theoretical and empirical arguments. It is important to note that I neither seek to provide an exhaustive literature review, nor seek to explicitly review an extended list of the critiques of the core arguments previously made. This narrow focus is deliberate, for my purpose is to concisely outline the main tenets of concepts and theories concerning these topics, to assess how they are conceptualized, to provide a basis for establishing the link between key concepts, and to develop my research model.

2.1. Theories of Change

There are several relevant questions concerning change. What is it? Why do firms need to change? Under which conditions will firms initiate changes in their organization? What kinds of outcome will a change bring to firms? Certainly, these questions already suffice to show that there is need for research on organizational change. 13 The wide range of past research on organizational change has focused on four main categories. One category has to do with content issues, and it mainly focuses on factors related to successful or unsuccessful change attempts (e.g., Hofer 1980; Bibeault, 1982; Hambrick and Schecter, 1983; Barker and Duhaime, 1997). Another category concerns process issues, mainly focusing on steps, phases, or actions undertaken during the implementation of an intended change (e.g., Judson, 1991; Kotter, 1995; Galpin, 1996). An additional category deals with context issues, focusing on internal or environmental forces or conditions affecting a change in an organization (e.g., Schendel and Patton, 1976; Slatter, 1984; Robbins and Peace, 1992). The final category concerns reaction issues, and it focuses on employees’ responses to organizational change (e.g., Weiss and Cropanzano, 1996; DeWitt, Trevino, and Mollica, 1998; Patterson and Cary, 2002). The literature suggests several internal and external factors that lead a firm to commence a change. Examples of these factors include: (1) increased competitive pressure (Meyer, Brooks and Goes, 1990); (2) new government regulation (Meyer et al., 1990; Haveman, 1992; Fox-Wolfgramm, Boal and Hunt, 1998); (3) technological change (Haveman, 1992); and (4) management team change (Castrogiovanni, Baliga and Kidwell,

1992).

13 As mentioned earlier, research on organizational change is one of the areas in organization theory research.

12

Firms that undertake change, any change, in their organization often aim to improve their performance in terms of, for example, higher profits, better responsiveness to the market, and long-term competitive advantage. For example, past studies on corporate turnaround (e.g., Hofer 1980; Bibeault, 1982; Hambrick and Schecter, 1983; Barker and Duhaime, 1997) have found several actions or strategies that can revive the troubled firms through corporate turnaround. We can thus conclude that the real value of organizational change rests on its ability to alter an organization’s identity, strategy, structure, operation or human resources as a means to enhance firm performance. Now let us consider the characteristics of change. Change is defined as a movement away from a current state toward a future state (George and Jones, 1995). In the organizational change literature, at the abstract level, there are two distinct modes of change: first- and second-order change. The phrase “first-order change” is used to describe organizational changes that occur within a relatively stable system that remains mostly unchanged; and for a system to remain stable or unchanged, it requires frequent first-order changes (Weick and Quinn, 1999). 14 On the contrary, second-order change or so-called episodic change modifies or transforms fundamental structures or properties of the system (Weick and Quinn, 1999). The concept of first- and second-order change is very popular and powerful, and its fruits have been many. To give but a brief sample of some of the works that have benefited from this concept, it has advanced several theoretical models such as Argyris and Schön’s (1979, 1996) single- and double-loop learning by individuals, Miller and Friesen’s (1984) adaptation vs. metamorphosis by organizations, and Tushman and Anderson’s (1986) competence-enhancing vs. competence-destroying changes in technology. In summary, there are several patterns or types of change (Miller, 1980; Johnson-Cramer, Cross and Yan, 2003): small or large (Ledford et al., 1989), planned or emergent in nature (Johnson-Cramer et al., 2003), radical or incremental (Weick and Quinn,

1999).

Another aspect of change is that it can occur at differing organizational levels. First, change can occur within a population of organizations. For example, changes occurring at an industry level (e.g., changes in customers’ demands and preferences) have implications for most, if not all, companies within the industry. Similarly, changes occurring at a country level have implications for most, if not all, organizations within the country. In addition, changes can occur in a single organization, having implications for the whole organization

14 As the phrase “continuous change” is used to describe organizational changes that tend to be ongoing, cumulative, and evolving (Weick & Quinn, 1999), the terms “first-order change” and “continuous change” seem to be used interchangeably.

13

or for specific parts of the organization. Last but not least, changes can occur at the level of individuals within an organization—that is, at the level of employees or managers. The important point for us to observe is that changes at differing levels may share some common characteristics but may also possess certain unique characteristics. One of the central issues of organizational change concerns the ability of the organization to enact change. 15 The ability may be partly limited by organizational inertia; that is, the organization may not be inclined to search for new solutions (Meyer and Rowan, 1977; Zucker, 1987). This raises the question of whether organizations can change themselves. 16 That is a difficult question, and no single answer will adequately answer it. My answer is that they cannot, due to the fact that from a legal perspective, an organization is a non-human entity; therefore, we can argue that it is not the organization that changes itself but rather the people in the organization that change themselves and thereby change the organization. But this leads to the question of whether an organization’s capability to adapt is conditioned by its employees’ capability to adapt, which may be determined by the levels of inertia at the individual level. Research has been done on organizational inertia, which examines the role and impact of organizational inertia on organizational structure and design. 17 In the organizational inertia literature, it is argued that various factors generate several forms of inertia in the organization (e.g., strategic, structural, or cultural inertia). Organizational change may be limited by internal factors such as an organization’s investments 18 in plant, equipment, and specialized personnel (Hannan and Freeman, 1989). It is also possible that top managers or decision makers may receive limited or insufficient information to the extent that they may

15 The question concerning the degree to which organizations or managers can change the ways organizations work in response to changes in their operating environments is one of the main questions in the organization theory research. There are two contrasting views: The first perspective is that organizations or individuals in organizations can undertake a surveillance of their environments and change their internal properties to promote their survival. The second view is that organizations or individuals in organizations are constrained by limitations to the possibility for change.

16 It should be noted that this question is different from the question of whether organizations change or not. A key difference between them concerns who is an actor or initiator of organizational change. The former question focuses on whether organizations can be the ones who make organizational change, whereas the latter question focuses on whether organizations, as an object, can be changed. To the latter question, prior research has suggested that organizations do change certain aspects such as strategies or structures (Zajac and Shortell, 1989).

17 For an extensive discussion of organizational inertia, see Arrow (1974), Hannan and Freeman (1984, 1989), or Perrow (1986).

18 In economics theory these investments can be considered sunk costs.

14

fail to make a decision on organizational change or adaptation. 19 If internal politics exist in an organization, they may also contribute to organizational inertia; that is, political disequilibrium in an organization may lead to resistance towards certain proposed changes. Indeed, most organizational changes are designed to benefit the organization as a whole; and these benefits are likely to take time to be realized; however, any political resistance within the organization generates short-run political costs that may either exceed the potential benefits or be high enough that top executives may decide against the intended change (Hannan and Freeman, 1989). 20 Likewise, external factors such as the dynamics of political coalitions, costly or limited information with regard to relevant environments, and legal and other barriers to entry or exit from the market may also restrict the nature and degree of organizational change or adaptation in organizations (Hannan and Freeman,

1984).

Research on organizational change has led to various views and perspectives. However, there are at least three most prominent views on organizational change. The first view, based on population ecology theory, 21 argues that most of the variations in organizational structures occur through the creation of new organizations and organizational forms, and the demise of old ones (Hannan and Freeman, 1977, 1989; Freeman and Hannan, 1983). According to Hannan and Freeman (1989), this perspective, which may be called “selection theory,” argues that existing organizations, particularly the largest and most powerful ones, seldom change their strategy and structure quickly enough to keep up with the demands of uncertain and changing environments. The second view, based on random transformation theory, proposes that endogenous processes induce structural changes in the organizations, but the changes are loosely associated with the goals of the organization and the demands of the uncertain and changing environments (March and Olsen, 1976; Weick, 1976). The third view, based on the rational adaptation theory developed by March and Simon (1958), argues that organizational variability generates changes in strategy and structure of organizations in response to threats, opportunities, and environmental changes.

19 This proposition is consistent with Simon’s concept of bounded rationality. For more complete details, see Simon

(1957).

20 From the political and institutional perspectives, it is difficult to enact changes in organizations without a major organizational crisis. For a detailed discussion on this issue, see Fligstein (1996) or Myer and Rowan (1977).

21 Population ecology theory, which is based on biological theory, especially Darwinian natural selection theory, was developed in response to a growing dissatisfaction with and critiques on the adaptation and strategic contingency models.

15

There are some variations in this view. For example, strategic contingency theories 22 focus on structural changes that match organizational structures (Thompson 1967), whereas resource dependence theories 23 focus on structural changes that neutralize sources of environmental uncertainty (Pfeffer and Salancik, 1978). In summary, as the review of the literature has shown, organizational change, regardless of its form, will have implications for the organization as well as its employees. Simon (1991: 32) noted that “employees, especially but not exclusively at managerial and executive levels, are responsible not only for evaluating alternatives and choosing among them but also for recognizing the need for decisions.” Accordingly, it is useful to understand how employees view and react to organizational change. In support of this view, the main focus of this dissertation is on the implications of organizational change for employees rather than for organizations and, specifically, on how employees respond to organizational change.

2.2.

Perceptions

As this dissertation will deal to a high degree with perceptions and reactions to change, it is especially important to elaborate on the nature and implications of perceptions on decisions. Although research on perceptions is rich and comprehensive, the intent of this literature review is not to present an exhaustive list of extant definitions of perception. Instead, my intent is to establish two key points. First, perception, as a psychological construct, is associated with other constructs such as attitude or emotion. Despite the differences among these constructs, most, if not all, of them seem to share common properties that shall be seen later. Second, perceptions influence the ways in which humans understand the world around them and how they make decisions. With deeper insights into how people understand the world, we can better comprehend the ways in which humans make decisions and why they behave in certain ways. First of all, what is perception? Perception can be defined as a “complex process by which people select, organize, and interpret sensory stimulation into a meaningful and coherent picture of the world” (Berelson and Steiner, 1964: 88). In the same vein, perception is “about receiving, selecting, acquiring, transforming and organizing the

22 The strategic contingency theory assumes that owners and managers of organizations establish organizations that allow them to monitor the goals and procedures in the organization so that they will be able to respond to external problems. See Thompson (1967) for a theoretical approach along these lines.

23 Slightly different from the logic of the strategic contingency theory, the resource dependency theory argues that managers strategically create organizational structures and procedures that help organizations mitigate the effects of external environments on the organization. See Pfeffer and Salancik (1978) for more complete references.

16

information supplied by our senses” (Barber and Legge, 1976: 7). The research on perceptions can be traced back to Bartlett’s (1932) influential works on the constructive nature of cognition, which argues that schematic thinking dominates human perception in ways that human generic beliefs about the world influence and shape information processes. Several researchers (e.g., Allport, 1954) have extended Bartlett’s (1932) work and have advanced our understanding of perception, attitude, judgment, and several other concepts. 24 The preceding discussion has suggested that from a psychological perspective, individuals’ perceptions have a directive influence upon their decision-making and the outcome of their decisions; thus, it is not surprising that organization theorists are now interested in relationships between perceptions and various aspects of organizations. For example, a work by Anderson and Paine (1975) has posited the influences of the perception of uncertainty in the environment on the perception of the need for change in a firm’s strategies. The research on the roles and effects of perceptions on people’s decisions and behaviors is yet to be completed, and the search for a better understanding of various perceptions on employees’ behaviors such as turnover or commitment in the field of human resource management continues its momentum. However, empirical research has begun to show that in organizational settings, certain perceptions such as the perception of uncertainty are associated with people’s behaviors. An empirical study by Ashford and colleagues (1989), for example, has shown evidence for a positive relationship between perceived job insecurity and intention to quit. Another empirical study by Eisenberger, Fasolo and Davis-LeMastro (1990) has demonstrated that employees’ perceived organizational support is related to various attitudes and behaviors. In a more recent study, Gopinath and Becker (2000) found that perceived procedural justice concerning the divestment activities of the firm is positively related to post-divestment commitment to the firm.

Thus far, I have dealt with a holistic review of perceptions. However, the discussion of the general concept of perceptions would be incomplete without mentioning two other related concepts – recognition and action. The concept of recognition deals with the ability to discriminate among familiar classes of objects, 25 and it is related to the concept of categorization. Thus, at an abstract level, recognition is one’s ability to place objects in a

24 Note that the dominant assumption in much of human perception is that one’s schematic preconceptions drive his or her evaluations of, and reactions towards, an object. We will come back to this issue in later chapters.

25 For a detailed discussion of the concept of recognition, see Langley and Simon (1981).

17

category. 26 To understand the relationship between recognition and categorization, it is necessary to consider how humans make sense of reality in a complex world. Perhaps the key answer to this question is the assertion in psychology that in an attempt to make sense of a complex world, humans often construct and use categorical representations to simplify and streamline the perception process (Fiske and Taylor, 1984, 1991; Gilbert and Hixon, 1991). In psychology, the term “categorization” is typically regarded as a process in which people group together objects and/or things (Zentall, Galizio, and Critchfield, 2002). Within psychology literature, there are several theories of categorization, for example, exemplar models (Brooks, 1978) and decision bound theory (Ashby and Gott, 1988). The other relevant concept in connection with perception is “action.” Action refers to one’s activities such as moving the body in response to the perceptual process. As pointed out by Argyris (1999), humans possess certain kinds of mental programs on how to act effectively in different types of interaction; and there are two theories of action that humans hold. The first one is normally expressed in the form of stated beliefs and values. The second one is actually used and can thus only be inferred from observing their behaviors. Up to now, most people studied have a theory-in-use, which is called Model I (Argyris, 1999). 27 Model I theory-in-use requires defensive reasoning (Argyris, 1999). In his view, individuals tend to keep their premises and inferences tacit for fear that they may lose control, and the use of defensive reasoning prevents questioning the defensive reasoning. 28 The consequences of the model of the theory-in-use strategies are that defensiveness, misunderstanding, and self-fulfilling and self-sealing processes are more likely (Argyris,

1999).

If perceptions are derived from or based on incomplete information and limited observation, perceptual biases will occur, and thus affect a person’s decisions and actions. 29 But what is the point of getting to know the concept of perception? Here, it is the contention that several perceptions of change are acting as determinants of employees’ reactions to

26 The term “category” typically refers to the totality of information that one has in mind about various groups of objects (Smith, 1998).

27 In short, Model I theory-in-use (Argyris, 1982, 1990, 1993, 1999; Argyris & Schön, 1996) has four governing values: achieve your intended purpose; maximize winning and minimize losing; suppress negative feelings; and behave according to what you consider rational.

28 This is because most people who follow Model I theory-in-use employ the following prevalent action strategies:

advocate your opinion; evaluate the thoughts and actions of others (and your own thoughts and actions); and attribute causes for whatever you are trying to understand (Argyris, 1999).

29 Individuals do not see or receive everything that happens in a particular situation. More importantly, they tend to be selective in what they attend to and what they perceive. This selectivity in the perceptual process leads to the tendency or bias to perceive one thing and not another. This is called “perceptual bias.”

18

change. That is, humans usually try to make sense of what has happened, what is happening, and what will happen. A number of researchers have noted a link between the perceptual process and the interpretation of information; they have argued that the interpretation of information is based on the perceptual process (e.g., Anderson and Pained, 1975). Further, during organizational change processes, employees create their own perspectives and interpretations of what is going to happen, what others are thinking, and how they themselves are perceived. Additionally, if there is a lack of information about the change, then evidence of employees’ own perspectives and interpretation of the change is more likely to be observed (Coghlan, 1993).

2.3.

Attitude

Like the preceding discussion of human perceptions that has given an overview of how humans perceive and make sense of the world, this section discusses how research on perceptions has advanced our understanding of attitude(s). In psychology, attitude has been examined extensively for a long period of time. The main focus of research on attitudes concerns the nature and function of attitudes and how individuals construct them. The application of current knowledge on attitude to business settings and the implications of attitude for individuals’ decisions and behaviors are of interest in this dissertation. This dissertation asserts that employees’ attitudes can influence their predisposition to formulate a pre-determined response to a change. The definitions of attitudes have been many. In social psychology, the term “attitude” refers to an individual’s preference for or disinclination toward an idea, issue, item or object; it is subjective in nature, and can be positive or negative. There are three other definitions that have influenced subsequent studies on attitude. 30 One definition is that attitude is “the affect for or against a psychological object” (Thurstone, 1931: 261). Another definition that seems to be more comprehensive is that attitude is “a mental and neural state of readiness, organized through experience, exerting a directive or dynamic influence upon the individual’s response to all objects and situations with which it is related” (Allport, 1935: 8). A final definition that is slightly different from Thurstone’s (1931) is that attitude is “a disposition to react favorably or unfavorably to a class of objects” (Sarnoff, 1960:

261). There are two important aspects of attitude: one of them is a belief aspect that uses cognitive processes to describe an object and its relation to other objects, the other is an affective aspect that leads to liking or disliking an object (Katz, 1960).

30 For a detailed review of attitudes, see Greenwald and Banaji (1995).

19

Another fundamental question in the research on attitudes is about how individuals acquire attitude. In psychology, attitudes arise from concepts, which are constructed through experience; and concepts become attitudes though a process in which an evaluative aspect is added on to them (Rhein, 1958). To understand the role of attitude in human behavior, a model by Fishbein and Ajzen (1975) has suggested that: first, an individual’s positive or negative beliefs about an object form an attitude towards that object; second, this attitude determines the individual’s intention to behave with respect to the object; and finally, this intention to behave is related to the actual behaviors acted. Given this observation, we may assume that attitudes towards organizational change tend to result in pre-determined intentions to behave and then subsequent behaviors. In this sense, managers and/or employees who have a negative attitude towards organizational change are more likely to resist efforts to change. In the same way, it is probable that managers or employees who have a positive attitude toward organizational change are more likely to support efforts to change. There are two basic assumptions guiding the directions of research on attitude. The so- called traditional view assumes that attitudes are dispositional in nature. According to the dispositional approach, the so-called traditional view, attitudes are seen as stable dispositions that have developed within the individual. In its roughest form, this view emphasizes the role of an individual’s disposition to the development of attitudes (Salancik and Pfeffer, 1978). A more recent theory, developed in the late 1970s, can be seen as a breakaway from the traditional view on how attitudes are formed is based on the assumption that attitudes are situational in nature. According to this approach, attitudes are viewed as reactions to social situations that change when social context changes (Salancik and Pfeffer, 1978). For instance, the social information processing perspective (e.g., Salancik and Pfeffer, 1977, 1978) asserts the role of social information on the behavioral reactions of individuals to situations. Both predictions may help us gain insights into how humans acquire and change their attitudes. By considering the two different views, whether attitudes are dispositional or situational in nature, my main conclusion is that the two views complement one another. Attitudes may be stable dispositions, but may be influenced by social situations. There is also the question of whether attitudes are conscious or unconscious in nature. Recent research, which has suggested that attitudes are conscious in nature, has been implicitly embedded in much of the prior research on attitudes (Greenwald and Banaji, 1995); and most of the previous studies have focused on conscious cognitive involvement in debate judgments and decisions. On the contrary, another group of researchers has begun to recognize the unconscious aspect of attitude. The key proposition of this stream of research

20

is that there is an implicit operation of attitudes. 31 By acknowledging both implicit and explicit operations of attitudes, we can assert that humans’ attitudes can influence thinking, decision-making and behaviors in situations in which people recognize the existence of their attitudes as well as in situations in which they do not recognize the existence of their attitudes.

2.4.

Emotion

Because the human rationality approach to decision-making in economics has dominated much of the research on organization theory for years (Ashkanasy, Härtel and Daus, 2002), empirical research on the effects of emotion on decision-making and behavior in the field of management has been limited. Therefore, we can reason that because of the predominance of the rational choice models, the concept of emotion has been largely unnoticed or ignored for some time in the mainstream research in management science. Whereas there is a relatively sparse body of management literature dealing with emotion, research on emotion in the field of psychology has been voluminous. Research on emotion has a long history, perhaps starting with Charles Darwin’s book The Expression of the Emotions in Man and Animal (1872). Since then, the development of the literature on emotion has lead to at least four prominent perspectives: the Darwinian, the Jamesian, the Cognitive, and the Social-Constructivist. 32 All of these approaches have attempted to provide answers to the same fundamental questions, for example, what is emotion, and how does it evolve? This literature review will not discuss each of the key questions in light of the different perspectives. In fact, it will only touch on certain aspects of the last two. Since my objective is to give only a brief overview of the research on emotion and underpin arguments for my research model, this literature review will take us through the research on emotion in the simplest form. The term “emotion” has been defined as “a relative short-term positive or negative evaluative state that involves neurophysicological, neuromuscular, and cognitive components” (Kemper, 1978). Although in the psychology and sociology literature, there seems to be little consensus concerning the meaning of emotion and related terms such as mood, feeling, and sentiments (Kemper, 1987), it is beyond scope of this paper to offer a new definition. For that reason I shall simply review the role of and influence of emotion on

31 There has recently been some debate over the views on conscious or unconscious aspects of attitudes. The traditional view does not make an explicit distinction as to whether attitudes operate in conscious mode or in unconscious mode, while a recent view explicitly acknowledges that attitudes also operate in unconscious mode. For a detailed discussion of this issue, see Greenwald and Banaji (1995).

32 For an overview of all four perspectives on emotion, see Cornelius (1996).

21

people’s decisions and behavior. Because emotions exert effects on people’s decisions and behavior, it is important for this dissertation to recognize and understand how and why emotions can exert such effects on employees’ decisions and behavior in the context of organizational change. Previous research on emotion has suggested that humans’ affective states, decisions and behavior are influenced by how they process information. For example, a central theme of the Affective Infusion Model (AIM) proposed by Forgas and George (2001) is that the influence of affective states on individuals’ judgments and behaviors depends on the type of information-processing strategies the individuals adopt in a particular situation. In its roughest form, the basic idea is that affect (referring to both moods and emotions) impacts organizational behavior because it influences both what people think (the content of thinking) and how people think (the process of thinking); and most social thinking and action occurs in naturally complex and ambiguous situations, and requires the use of active, constructive information process strategies (Forgas and George, 2001). According to Forgas (1995), affect infusion can be defined as the process whereby affectively loaded information exerts influence upon and becomes incorporated into an individual’s cognitive and behavioral processes, entering into their constructive deliberations and eventually coloring the outcome in a mood-congruent direction. Basically, the AIM model (Forgas and George, 2001) asserts that there are four information- processing strategies based on different affect infusion potentials: (1) a direct access; (2) a motivated processing; (3) a heuristic processing; and (4) a substantive processing. Both the direct access and motivated processing strategies require little constructive processing, limiting the extent of mood infusion. On the contrary, both the heuristic and substantive processing strategies require a high degree of open and constructive thinking, allowing greater mood infusion to occur, and resulting in the creation of new knowledge from the combination of new information and stored information. Moreover, task characteristics, personal variables, and situational features determine processing choices (Forgas and George, 2001), implying that the influence of affect is context-dependent. Central among issues of emotion is whether there is only one direction, either positive or negative, for each relationship between emotions and the other variables. This is crucial because it complicates and shapes how researchers conduct their research on emotions. Based on numerous studies on emotions, it is obvious that there can potentially be inverse relationships between moods and emotions on the one hand and behaviors and attitudes on the other hand. For example, George and Zhou (2001) theorized that under certain conditions, positive moods might hinder, and negative moods might enhance creative performance. This is because when creativity is an objective for people and that they are

22

high on clarity of feeling, they may use their mood as input to determine the sufficiency of their efforts. In this sense, negative moods may signal that things are not going well and that additional effort is needed. Along similar lines, positive moods may signal that things are going well and that an additional effort is not required. George and Zhou’s (2001) findings were consistent with their theory, that is, positive mood was negatively associated with creative performance; and negative mood was positively associated with creative performance. In other conditions, there may be a positive relationship between positive mood and creative performance (see, e.g., Isen, Daubman and Nowicki, 1987). According to the literature on mood, its effects on work motivation are manifold. First, positive moods may enhance spontaneity and helpfulness toward coworkers (George, 1991). Second, positive moods may facilitate a flexible and open cognitive style in social situations (Forgas, 1999a, 1999b). Last but not least, positive moods may influence the performance of leaders (George 1995, 2000). In contrast, a negative mood has been linked to several negative decisions and behaviors such as absenteeism and turnover. For example, empirical research has shown an inverse relationship between employees’ positive moods and levels of absenteeism (e.g., George, 1989). Another example is that the interaction between value attainment, job satisfaction, and positive mood is likely to predict turnover intentions among employees (George and Jones, 1996).

2.5. Individual Decision-Making

In economics, research on decision-making and judgment seems to have begun in 1950’s, focusing on a rational approach. 33 Since then, research on decision-making (e.g., strategic decision-making in organizations) has been growing. Another stream of research on decision-making in the field of psychology has also advanced our understanding of how individuals make judgments and decisions. It is important to note that with regard to human rationality, the forms of human rationality in the area of psychology, which differ from those of theories of human rationality in neoclassical economics, have begun to receive greater attention in strategic management research. Examples of works that have been influential in strategic management or management science (e.g., Dean and Sharfman,

33 This by no means suggests that the concept of rationality is not being applied in other social sciences. Simon (1978, 1979, 1985, 1986) pointed out neatly that most social sciences implicitly or explicitly assume human rationality; however, the forms of human rationality that they adopt may differ. Thus there is a point of agreement concerning human rationality: that is, humans have reasons for what they do or for how they behave. The differences have to do with the question of what constitutes rationality. For instance, economic theories take a special form of human rationality – the rationality of the utility maximizer who will objectively aim for the best possible choice in terms of the given utility function (Simon, 1978).

23

1996) are those of Simon (1957), March and Simon (1958), and Tversky and Kahneman (1974). Central to these works are the arguments that there are several forms of human rationality, and that human rationality is bounded to external and/or internal constraints. A review of past research on strategic decision-making has shown that there are several models of the strategic decision-making process. 34 One example is Hofer and Schendel’s (1976) model that outlines seven steps of the strategic decision-making process:

(1) strategy identification; (2) environmental analysis; (3) resource analysis; (4) gap analysis; (5) strategic alternatives; (6) strategy evaluation; and (7) strategic choice. Another is the model of Mintzberg et al. (1976), which suggests three phases and seven steps of the strategic decision-making process: (1) identification phase consisting of decision recognition and diagnosis steps; (2) development phase consisting of search and design steps; and (3) selection phase consisting of screening, evaluation, and authorization steps. Likewise, Fredrickson (1984) suggested that from the perspective of a managerial decision maker, the rational decision-making process involves five interrelated cognitive stages: (1) pay attention to a problem or opportunity; (2) gather information; (3) develop a series of options; (4) value the options using expected costs and benefits; and (5) select the option with the greatest utility. Another key aspect in decision-making is learning, which involves developing new understandings. The learning process involves the acquisition and interpretation of knowledge (Linsay and Norman, 1977). Learning is the process of modifying one’s cognitive map or understandings (Friedlander, 1983: 194), thereby altering the range of one’s potential behaviors (Huber, 1991). So we may speculate that since learning capability refers to individuals’ ability to develop a new understanding of the world around them, it may promote or limit their understanding of a proposed change. Past research has led to several concepts and theories to explain certain aspects of decision-making with the goal of explaining decision-outcome deviations from normative expectations of the rational decision-making approach. One such theory is Beach’s (1990) image theory that incorporates Einhorm and Hogarth’s (1981) idea that humans make use of mental simulation to evaluate options by applying strategies from known situations to new situations. 35 Another example is the model called framing effects that has suggested how apparently irrelevant variables can influence decision-making. According to Kahneman and Tversky (1979), framing or editing phases occurring during a process of choice concerns

34 For a detailed review of the literature on strategic decision-making, see Eisenhardt and Zbaracki (1992).

35 For a detailed discussion of image theory, see Einhorn and Hogarth (1981), Klein and Crandall (1995), and Beach

(1990).

24

with the preliminary analysis of alternatives, their outcomes and contingencies. 36 Note that the concept of framing effects is theoretically related to the concept of categorization discussed in the previous chapter. Most strategic decision-making models that have been influenced by economic theories assert implicitly or explicitly that a manager as well as an employee, as an agent of a firm, should arrive at a decision that will achieve the firm’s goals, one of which is the maximization of the firm’s value. This observation suggests a key difference between strategic decision-making models for firms and decision-making models for employees. That is, decisions (e.g. reactions to change) of employees may be oriented towards the individual-level maximization of certain objectives such as career advancement or social status rather than towards the firm’s goals such as maximizing the value of the firm. However, we may argue here that the ways in which different individuals arrive at decisions (e.g., as a manager making a choice that achieves the firm’s goals or as an employee making a choice that achieves his or her personal goals) may not be fundamentally different. That is, as employees react to change, they are likely to carry out: (1) objective identification; (2) decision/outcome alternatives; and (3) evaluation and selection. In this sense, employees are assumed to be rational; however, their form of rationality does not necessarily correspond to the form of rationality in economics or the form of rationality that the firm may wish its employees to hold. Let me now turn to a study leading to a model of reaction to change proposed by Isabella (1990). In this empirical work, 40 executives from a medium-sized, urban, financial services institution were asked to describe and discuss five events that had occurred in the organization over the previous five years. The results showed that members of the organization construe key events linked to the process of change and that there are four stages that individuals go through as changes unfold. The four stages are anticipation, confirmation, culmination, and aftermath. In the anticipation stage, people gather rumors, scattered pieces of concrete data, to construct a construed reality. In the confirmation stage, following the standardization of events into a conventional frame of reference, people reflect or refer their frames of reference which have worked in the past. In the culmination stage, people compare the conditions before and after an event, at which time they amend their frame of reference to either include new information or omit invalid information. In the aftermath stage, people review and evaluate the consequences of a change. From this

36 See Tversky and Kahneman (1981), John et al. (1993), and Paese, Bieser, and Tubbs (1993) for empirical research on framing effects.

25

example we begin to better understand the process which individuals undergo when they are confronted with a change in their organization. At this point, I shall discuss how individuals interpret data and information. The term “interpretation” has been defined as the process through which people give meaning to information (Daft and Weick, 1994). The process of interpretation is important because it is used to understand information. Accordingly, employees use the process of interpretation to give meaning to, and to understand the information concerning a change. It is logical to argue that different people may give differing meanings to the same information, 37 and that the differing meanings prompt differing decisions. Moreover, an individual’s emotions and behaviors depend upon the way they structure their thoughts (Ellis and Harper, 1975). Thus, one can reason that the processes of interpretation and decision-making may be related. Indeed, the evaluation phase of a decision-making process requires interpreting information. Past research on interpretation processes has suggested several models of the decision- making processes. For example, Jaffe, Scott, and Tobe (1994) have proposed a four-stage model of how employees interpret events as an organizational change unfolds. The four stages are denial, resistance, exploration, and commitment. However, one may argue that this general explanation is an incomplete view of real decision-making processes (Beach, 1993). For instance, personal biases, failures of memory, and misunderstood probabilities have been found to cause decision mistakes (see, e.g., Kahneman, Slovic, and Tversky, 1982). In addition, several researchers have emphasized the existence of intuitive and irrational decision-making (Isenberg, 1986; Fiske, 1992). That is, decision-making processes sometimes involve experience-based mental routines, creating quick decisions without rational thought. Now let us have another look at mood theory. Forgas and colleagues have conducted studies on mood theory concerning the manner whereby moods determine behaviors in social (see, e.g., Forgas, 1995) and organizational (see, e.g., Forgas and George, 2001) settings. In short, past empirical research on emotions such as positive or negative moods has suggested that emotions may affect people’s attitudes, values, and behaviors toward other objects and their world. This observation suggests that the effects of emotion on judgments, thought processes, decision-making, and behaviors should not be neglected when one wishes to study people’s decisions and behaviors.

37 Let us consider whether we always give the same meaning to the same information—that is to say, whether we sometimes assign differing meanings to the same information in other circumstances. If this is the case, then we might reason that it ought to be possible that different people may give a different meaning to the same information as well.

26

2.6.

Reactions to Change

Resistance to change has been identified as a negative and undesired response for organizations because it can lead to failures of change efforts (Martin, 1975; Regar, Mullance, and Gustafson, 1994; Spiker and Lesser, 1995). Indeed, studies of organizational change often attribute outcomes of change efforts to behaviors of employees, especially acceptance of change and resistance to change (e.g., Kotter, 1995; Galpin, 1996). Given the frequent occurrence and persistence of resistance to change in most change initiatives, it is not surprising that much research has been devoted to examining the problems of resistance to change, especially the ways in which resistance to change can be minimized. It is understandable that research on organizational change management has a pessimistic view on resistance to change. After all, resistance to change may disrupt or suppress efforts to change. However, little work has directly addressed the possibility of gaining a positive effect from resistance to change. As discussed earlier, the question is whether resistance is always negative. It might be that resistance to change can become strategically valuable. If we are to understand why resistance to change has been considered the source of organizational change failures, we need to examine closely the characteristics and role of resistance to change itself. It appears that Kurt Lewin (1945, 1947, 1951) was the first author who used the notion of resistance to change. According to his field theory 38 , the status quo represents the equilibrium between the forces supporting change and the barriers to change. Some difference between these forces is therefore required to generate the “unfreezing” that initiates change. To make the change permanent, “refreezing” at the new level is required. In this sense, resistance is a system phenomenon. It is part of the change process and is not necessarily a negative factor. Many studies have posited that resistance to change is negative and should be removed or minimized. For example, Coch and French’s (1948: 521) view on resistance to change is that it is a combination of an individual reaction to frustration with strong group-induced forces. Similarly, Zander has defined resistance to change as “a behavior which is intended to protect an individual from the effects of real or imaged change” (Zander, 1950: 9). In the same view, Agócs (1997) has defined resistance as a process of refusal by decision-makers to be influenced or affected by the views, concerns or evidence presented to them by those who propose change. In summary, resistance to change generally refers to the behaviors of individuals or groups of individuals who are opposed to or unsupportive of changes that top executives want or decide to implement in the organizations.

38 For Lewin (1947), a change process consists of three phases: (1) unfreezing, (2) moving, and (3) refreezing.

27

Ford, Ford, and McNamara (2002) noted that, from a constructivist perspective, resistance to change is a function of the socially constructed reality in which a person lives, and that depending on the nature of that constructed reality, the form of that resistance will vary. On the contrary, from a modernist perspective, with the assumption that the same objective and homogeneous reality is shared by everyone, all people involved in a change are believed to confront the same change within the same context. An important conclusion to be drawn from these extremely different perspectives is that we need to develop a better understanding of how individuals really construct reality or see the world. However, for the purpose of this dissertation, I suggest that any difference between both perspectives is not of a great concern since in any circumstance the reality that a person holds will ultimately be expressed in terms of perceptions and/or attitudes. According to Agócs (1997), a typology of forms of resistance consists of: (1) denial of the legitimacy of the case for change; (2) refusal to recognize the responsibility to address the change issue; (3) refusal to implement a change initiative that has been adopted by the organization; and (4) the reversal or dismantling of a change initiative once implementation has begun. Recently, some researchers (e.g., Dent and Goldberg, 1999) have argued that people do not resist change, but rather losses of status, pay or comfort, and that this is not the same as resisting change. In the literature on organizational change, several factors are thought to be determinants of resistance to change; they include fear of real or imagined consequences (Morris and Raben, 1995), fear of unknown consequences (Mabin, Forgeson, and Green, 2001), a threat to the ways in which people make sense of the world (Ledford et al., 1989), a threat to the status quo (Beer, 1980; Hannan and Freeman, 1988; Spector, 1989), a threat to social relations (O’Toole, 1995), distrust toward those leading change (Bridges, 1980; O’Toole, 1995), and different understandings or assessments of the situation (Morris and Raben, 1995). Thus, it can be reasoned that a person does not resist organizational change but rather the consequences of organizational change. However, it can also be reasoned that the consequences of organizational change are part of change efforts and thus cannot be clearly separated. As discussed above, a central issue raised by previous research in change management is the role and implications of resistance to change, that is, how resistance to change evolves. At least one issue emerges from previous studies. Despite the seemingly extensive research on resistance to change, with the exceptions of the aforementioned definitions, seldom has previous research provided a definition of resistance to change. It seems that the term ‘resistance to change’ is used as a given. Thus, it is useful, if not critical, to examine

28

the dimensions of “reaction to change” to better understand and conceptualize the term and the concept of “resistance to change” as well as “support for change.”

Figure 1: Dimensions for Categorization of Reactions to Change

Discontent

Discontent

Discontent

Active

Active

Active

Discontent Discontent Active Active Active Contentment Contentment Contentment Passive Passive
Discontent Discontent Active Active Active Contentment Contentment Contentment Passive Passive
Discontent Discontent Active Active Active Contentment Contentment Contentment Passive Passive

Contentment

Contentment

Contentment

Passive

Passive

Passive

As illustrated in Figure 1, one dimension of reactions to change is whether a reaction represents one’s contentment or discontent with the change. Resistance to change can be described as a reaction that represents a high degree of one’s discontent with change. In contrast, support for change can be described as a reaction that represents a high degree of one’s contentment with change. Another dimension of reactions to change is whether it is active or passive in nature. An active action refers to any action that is active in nature and, thus, can be easily recognized, whereas a passive action refers to any action that is passive in nature and, thus, cannot be easily observed. I treat each dimension as a continuum along which any reaction can be located. The position of the reaction along each continuous dimension affects the categorization and nature of information required to detect it. That is, it is possible that one can easily observe most active actions with relatively little effort. On the contrary, a relatively large amount of efforts is required to detect passive actions. Likewise, the position of the reaction along each continuous dimension affects the relative ease of interpretation of the reaction. To illustrate:

By seeing employees listening quietly to the announcement of organizational change, any observers will find it relatively difficult to know whether they will support or resist the change; therefore, addition information will be needed to develop a better understanding of their position with regard to the change. On the contrary, by seeing employees acting in ways that they oppose change efforts, any observer can easily determine that they resist the change. Several researchers (e.g., Hultman, 1998; Judson, 1991) have already distinguished between active and passive resistance. Thus, it can be reasoned that these dimensions of reactions to change are capable of creating the parameters in which the definitions of resistance to change and support for change could be conceptually established. However, each dimension alone is insufficient to constitute a reaction that can be considered as resistance or support for change. Therefore, I suggest that a combination of the two dimensions can form the basis for defining and

29

categorizing various reactions to change. As illustrated in Figure 2, any reaction to change can be neatly classified along two dimensions into one of four categories: (1) active resistance; (2) passive resistance; (3) active support; and (4) passive support.

Figure 2: A Categorization of Reactions to Change Active Active Active Resistance Active Resistance Active
Figure 2: A Categorization of Reactions to Change
Active
Active
Active Resistance
Active Resistance
Active Support
Active Support
Discontent
Discontent
Passive Passive Resistance Resistance
Passive Passive Support Support
Passive
Passive

Contentment

Contentment

According to this classification, active resistance to change can be defined as a reaction that is active in nature and represents discontent with the change. Consider, for example, expressing opposition to a proposed change. It is an active action and reveals discomfort with or disagreement with the change; therefore, it can be considered active resistance to change. In contrast, passive resistance to change can be defined as a reaction that is passive in nature and reveals discontent with the change. For example, the act of ignoring is passive in nature and implicitly indicates one’s discomfort with change; thus, it can be called passive resistance to change. Active support for change can be defined as a reaction that is active in nature and reveals contentment with the change. Consider, for example, praise of the change. This is an active action and reveals one’s comfort with or agreement with the change; therefore, it can be considered active support for the change. In contrast, passive support for change can defined as a reaction that is passive in nature and represents contentment with the change. For example, expressing agreement is a passive action and reveals one’s contentment with the change. In summary, to explore the concept of resistance to change and support for change, this dissertation proposes two dimensions that specify the properties that should be considered the criteria for defining both resistance to change and support for change. It is expected that these definitions will increase the level of clarity of the definitions of resistance to change and support for change.

30

3. Theoretical Development and Research Model

Recall that the research questions of this dissertation are: What perceptions and/or attitudes influence employees’ resistance to change? And, what perceptions and/or attitudes influence employees’ support for change? A review of the literature suggests not only direct relationships between perceptions/attitudes and reactions to change but also paths (indirect relationships between them). Reflecting on the research questions, I first examine whether there is a direct relationship between each perception and each reaction to change without considering the potential moderating effects of other variables. This model is called a ‘direct effects’ model in Figure 3.

Figure 3: Alternative Models Relating Perceptions and Reactions to Change

1. 1.

Direct Effects Model

Direct Effects Model

Perceptions / Attitudes Perceptions / Attitudes Reactions to Change Reactions to Change 2. 2. Partially
Perceptions / Attitudes
Perceptions / Attitudes
Reactions to Change
Reactions to Change
2.
2.
Partially Moderated Model
Partially Moderated Model
Perceptions / Attitudes
Perceptions / Attitudes
Perceptions / Attitudes
Perceptions / Attitudes
Reactions to Change
Reactions to Change

3. 3.

Fully Moderated Model

Fully Moderated Model

Perceptions / Attitudes

Perceptions / Attitudes

Perceptions / Attitudes Perceptions / Attitudes
Perceptions / Attitudes Perceptions / Attitudes
Model Perceptions / Attitudes Perceptions / Attitudes Perceptions / Attitudes Perceptions / Attitudes Reactions

Perceptions / Attitudes

Perceptions / Attitudes

Perceptions / Attitudes Perceptions / Attitudes
Perceptions / Attitudes Perceptions / Attitudes
Attitudes Perceptions / Attitudes Perceptions / Attitudes Reactions to Change Reactions to Change It is important

Reactions to Change

Reactions to Change

It is important to note that the main focus of this dissertation is on the direct effects model. This narrow focus is deliberate, for my objective is to concisely develop and empirically test the model of perception-based factors affecting employees’ reactions to change. If a majority of hypothesized relationships in the direct effects model were to receive significant support, I will also examine a set of intervening models to assess if inclusions of some perceptions and attitudes improve the explanation of reactions to change, and, if so, which intervening model is most appropriate to the data. As suggested earlier, one reason to develop the degree of employees’ resistance to change and support for change as the dependent variables in this research is that organizational change may strongly affect work performance in some change efforts, affect only average in others, and affect weakly in still others. To this point, one can reason that resistance to change tends to have a negative relationship with work performance (Kotter, 1995). One can also reason that employees’ overall performance during organizational

31

change processes depends, among other things, on the net effect of organizational change on employee’s reactions to change; that is to say, organizational change has a direct effect on employee’s reactions to change, which in turn have a direct effect on work performance, which subsequently affects firm performance. Thus, it is sensible to study what factors enable resistance to change and support for change to emerge. Indeed, employees’ resistance to change and support for change often depend on, among other things, how employees construct numerous perceptions, and assign the degree of importance to each perception. Each of these perceptions can lead to higher or lower levels of resistance to change and/or support for change. Aggregating the outputs of these numerous perceptions can impose a difficulty at examining whether a particular set of perceptions and attitudes actually generates resistance to change and/or support for change. In this setting, it is also useful to recognize the importance of employees’ ranking of level of importance attached to each perception with regard to a decision choice.

3.1. Perception-Based View of the Employee

What is a perception-based view (PBV) of the employee? In contrast to the rational decision-making approach commonly used in the mainstream research in management science, an alternative approach which is labeled as ‘a perception-based view of the employee’ in decision-making focuses on the use of perception, attitude or emotion for a purpose of selecting a sensible alternative in pursuit of one’s goals. The main purpose of the perception-based view of the employee is to explain variations in decision and/or behavior among employees in the same context. That is, it attempts to answer two primary questions:

(1) why do individuals in the similar setting and facing the same object have differing decisions; and (2) why do individuals make decisions that might seem irrational, and be contradictory to those predicted by rational choice theories? Perceptions are multi-dimensional and have behavioral implications for humans’ decision-making. What, then, are the implications of perceptions on decisions and behaviors? Why is this related to reactions to change? First, let us consider a question of whether perceptions are a source of input for making a decision. If we are to agree that humans use perceptions as inputs for arriving at decisions, we face a question of whether we perceive the world around us as others do. The central idea here is that a person may obtain a different perception of a stimulus than others do, 39 and each reacts to this stimulus according to his or her interpretation process and is thereby motivated to make a decision

39 There is also the question of whether we assume that humans have an objective description of the world as it really is; that is, whether their perception of the world corresponds to the world as it really is.

32

that is different to one another. This point emphasizes at least two implications. First, if different people have different perceptions of a similar object at a given moment and in a given circumstance, and if one perception, as a certain input for a process of choosing, contributes to one decision and/or behavior rather than another, then one may conclude that variations in decisions and behaviors can be explained by variations in perceptions. Second, if, on the other hand, a perception varies across space and time—that is to say, a person’s perception of a similar object does not hold true at another moment and in another circumstance, then we may acknowledge consequences of time and space on humans’ perceptions. Here, it seems to me, if the concept of perceptions is to be applied in this dissertation, it is important to recognize that one shall have to understand the basic elements of this construct. The PBV perspective deals with how perceptions, attitudes, and emotions are used by individuals to solve a problem or to make a decision. At the abstract level, it deals with the effects and implications of psychological factors on individuals’ decision-making in a choice situation. In this sense, the central question concerns the extent to which psychological and emotional factors exert an influence on humans’ decisions and behaviors. By approaching humans’ decision-making with the PBV logic rather than with the rational approach, one will observe at least three key differences between the two approaches. A first difference is that the PBV perspective does not assume that decision-makers focus on a choice that maximizes expected utility, which is at the heart of the rational choice perspective, influenced by the neoclassical economic theory. 40 Viewed from an economic perspective, an individual is outcome-driven self-interested, thereby attempting to maximize expected utility. More importantly, we should make a distinction between a rational choice of an individual and a rational choice of the firm. This is important because one choice may seem irrational to one person, but it may seem rational to another. For illustration, I can use the logics of economics—the utility function. If two persons arrive at a rational choice that maximizes expected utility of their utility function respectively, and if both do not have a similar utility function, then both will not arrive at the same decision. If one of them presumes the other to hold the similar utility function, it is possible that this person may view the other’s choice as an irrational choice. 41

40 In its simplest form, the economic model of decision-making assumes that managers have perfect information and thus could make decisions that maximize profits. For a critique of the economic model of decision-making, see Simon (1957) and March and Simon (1958). The logic of the economic model of decision-making may be extended to employees’ decision-making.

41 Assume that he thinks such choice that the other has made gives a lower level of expected utility (according to his own utility function).

33

A second difference is that although I tend to agree with prior research on decision-

making asserting that humans make rational decisions, 42 I do not believe that humans always arrive at a rational decision. Now consider a simple example of an angry man. Should we assume that all decisions he makes at the moment of angriness are rational? If your answer is “no”, then you are getting close to agreeing with my argument. As this

example illustrates, several psychological constructs such as emotion or attitude exert a directive or influence on individuals’ decision-making process and outcomes, thereby diminishing the use of rationality in the process of making a decision.

A third difference is that the PBV perspective does not assume that decision-makers

consistently use a rational decision-making process that focuses on analytical comprehensiveness. According to Miller (1987), analytical comprehensiveness is a concept that focuses on individuals’ systematic scanning and analysis of environments in decision- making processes. Consider a situation in which people have to make quick responses; that is, there is a time constraint for arriving at decisions. In these circumstances, they may not go through all steps in decision-making process for making their decisions. Most important for the PBV perspective is that the greater the existence of the psychological factors for an individual in a choice selection situation, the stronger the predictive power of the PBV logic; that is, the PBV posits that the nature and magnitude of psychological factors existing at the time of decision-making will condition the effects of these psychological factors on a decision choice, and moderate the extent to which individuals make use of the rational-decision making process. In this dissertation, perception-based predictions of employees’ behavior are tested using the degrees of resistance to change and support for change of employees as dependent variables. For the moment, to illustrate a potential limitation to the PBV logic, consider a case that resistance to change of an employee only depends on two perceptions: perception A and perception B. Suppose that each of these perceptions has a positive effect on an employee’s resistance to change: that is, the greater the degree of perception A, the greater the degree of resistance to change becomes. Similarly, the greater the degree of perception B, the greater the degree of resistance to change becomes. Suppose that the firm has a change management strategy that is effective at minimizing perception A, but is ineffective at minimizing perception B. Suppose that this employee has a low level of perception A but a high level of perception B. The net effect of these two perceptions on resistance to change may be that

42 Rational decisions in this sense are not limited to rational choices within a boundary of the rational choice theory in economics. The notion of rational decisions here is in a broader sense and includes both a rational decision and a decision that is derived from a rational decision-making process.

34

this employee only produces average levels of resistance to change, and subsequently

average levels of work performance. If researchers were to measure the degree of perception

A that influences this employee to resist a change, and correlate this perception with this

employee’s resistance to change, they might conclude that perception-based predictions were not supported; i.e., resistance to change is higher than predicted by the perception- based logic. On the other hand, if researchers were to measure the degree of perception B that influences this employee to resist a change, and correlate this perception with this employee’s resistance to change, they might conclude that perception-based predictions were not supported, but in a different way; i.e., resistance to change is lower than predicted by perception-based logic. Both of these findings apparently contradict the perception-based logic which will be discussed in greater detail in the following section, although, at this employee’s overall perceptions, perception-based logic may be strongly supported.

3.2. Research Model and Hypotheses

In view of research issues and research questions discussed earlier, the model of perceptual

and attitudinal factors exerting an influence upon employees’ resistance to change and support for change presented here represents an attempt to develop a conception of the

directive influence of perceptions and attitudes on decisions of the employees in the context

of organizational change upon which to support the perception-based view of the employee.

As such, the research model links various perceptions and attitudes on the one hand and

resistance to change and support for change on the other hand. It is theorizing that, through

a decision-making process, a range of perceptions and attitudes may exert the effects upon the nature and magnitude of resistance to change and support for change. While prior empirical research has examined the role of some of these factors, much of prior research has focused on a smaller set of variables than the one this dissertation examines. Additionally, empirical research pertaining to the examination of factors

influencing employees’ support for change is relatively sparse. Here, it seems to me, much

of past empirical research has emphasized more on the examination of resistance to change

and less on the investigation of acceptance of change or support for change. An example of

empirical works on acceptance to change is an empirical study by Iverson (1996) examining the relationships between various factors such as organizational commitment, job satisfactory and job motivation on the one hand and employees’ acceptance of organizational change on the other hand. Another example is an empirical study by Iverson

and Pullman (2000) investigating the determinants of voluntary and involuntary turnover in

a repeated downsizing environment. The final example is an empirical study by Erez,

35

Earley, and Hulan (1985) examining the impact of participation on goal acceptance and performance. The review of the literature on employees’ reactions to various organizational decisions (e.g., layoff, turnaround strategy, and employee compensation plan) has identified several perception variables that are expected to exert an influence upon their resistance to change and support for change. Based on the results of the initial literature review, the research model, as illustrated in Figure 4, was developed. The explanatory variables proposed in this research model using the perception-based predictions can be neatly categorized into four groups or dimensions: (1) factors concerning change processes; (2) factors concerning real and expected consequences of change; (3) factors concerning employees’ ability; and (4) factors concerning employees’ relationship with the firm and their colleagues. To summarize, this dissertation focuses on (1) the degree to which employees support or resist organizational change and (2) whether direct relationships exist between various perceptions and/or attitudes and reactions to change. In the following section, I will discuss each of hypothesized relationships in greater detail.

36

Figure 4: Conceptual Diagram of the ‘Direct Effects’ Model

Perceived Organizational Support Perceived Organizational Support Perceived Procedural Justice Perceived Procedural
Perceived Organizational Support
Perceived Organizational Support
Perceived Procedural Justice
Perceived Procedural Justice
Factors concerning
Factors concerning
Perceived Participation in Decision-Making
Perceived Participation in Decision-Making
Change Processes
Change Processes
Perceived Need for Change
Perceived Need for Change
Attitude toward Organizational Change
Attitude toward Organizational Change
Fear of Known Consequence of Change
Fear of Known Consequence of Change
Employees’ Reactions to Change
Employees’ Reactions to Change
Employees’ Reactions to Change
Fear of Unknown Consequence of Change
Fear of Unknown Consequence of Change
Active Resistance to Change
Active Resistance to Change
Active Resistance to Change
Perceived Change in Power
Perceived Change in Power
Passive Resistance to Change
Passive Resistance to Change
Passive Resistance to Change
Factors concerning
Factors concerning
Perceived Change in Status
Perceived Change in Status
Actual & Expected
Actual & Expected
Active Support for Change
Active Support for Change
Active Support for Change
Consequences
Consequences
Perceived Change in Pride
Perceived Change in Pride
of the Change
of the Change
Passive Support for Change
Passive Support for Change
Passive Support for Change
Job Satisfaction
Job Satisfaction
Job Security
Job Security
Job Motivation
Job Motivation
Perceived Employability
Perceived Employability
Factors concerning
Factors concerning
Employees’ Ability
Employees’ Ability
Self-Confidence for Learning/Development
Self-Confidence for Learning/Development
Factors concerning
Factors concerning
Affective Commitment
Affective Commitment
Employees’
Employees’
Relationships
Relationships
Trust in Management
Trust in Management
with the Firm and
with the Firm and
Colleagues
Colleagues
Colleagues’ Reaction to Change
Colleagues’ Reaction to Change

37

3.2.1.

Perceived Organizational Support

The variation in employees’ reactions to change – resistance to change and support for change – may be systematically attributable to certain perceptions, attitudes and emotions that employees form about a change. Previous studies have suggested that perceived organizational support is related to a wide array of work-related attitudes and outcomes (Eisenberger, Fasolo and Davis-LaMastro, 1990). Accordingly, if perceived organization support is correlated with certain decisions of employees, we would then expect that employees are more likely to take into account perceived organizational change when making decisions regarding whether to resist or support organizational change. Consider employees who search for a response to organizational change by considering the extent to which their organization has supported them. The value of subjective perceptions of organizational support comes from the possibility of avoiding negative consequences of organizational change. Intuitively they would be willing to support organizational change efforts to the value of perceived organizational support they receive. On the contrary, they will be more likely to resist organizational change if they hold a perception that their organization does not support them. Based on the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960) 43 , which have been widely used for research on the relationship between organizations and employees, it can be reasoned that employees’ perceived organizational support affects their feeling of obligation to their organization. Based on the norm of reciprocity, employees tend to respond positively to favorable treatments from their employer, or they feel obliged to help those who helped them, implying a positive norm of reciprocity (Gouldner, 1960). On the contrary, employees tend to respond negatively to unfavorable treatments from their employer or feel obliged to retaliate against those who injured them, implying a negative norm of reciprocity in which individuals retaliate to the injuries or the negative benefits enacted by others (Gouldner, 1960). Prior empirical studies (e.g., Eisenberger et al., 1986) have found a positive correlation between perceived organizational support and feelings of obligation. The key to this idea is that based on the organization’s procedures and policies employees form perceptions about the organization’s intentions and attitudes towards them. Then, they assign certain human- like attributes/characteristics to the organization on the basis of the treatment they receive.

43 Gouldner (1960) noted that, in a view that exchanges of benefits or favors among individuals induce the imbedded obligation, there are three different components of the process that form the governance of the norm of reciprocity. A first component is equivalence, which is the extent to which the amount of return almost equates to what was received. A second component is immediacy, which is the time period between repayment and receipt of favors. A final component is interest, which is the motive of the partner in making the exchange of benefits.

38

In this view, employees who perceive greater support from their organization feel obliged to repay their organization (Eisenberger et al., 1986; Eisenberger et al., 1990). In one empirical study, Allen, Shore and Griffeth (2003) have illustrated that perceived supportive human resource practices contributed to the development of perceived organizational support, of which mediated relationships with organizational commitment and job satisfaction. Likewise, the concept of the inducements-contributions framework of voluntary turnover (March and Simon, 1958) has suggested that a balance between the inducements offered by the organization and the contributions expected of employees is more likely to determine employees’ decision to continue their participation in the organization. In this view, employees who have a perception of greater inducements would be less likely to desire to leave the organization than those who have a perception of fewer inducements. Research on psychological contracts 44 examining implications of psychological contract breach 45 on employees’ attitudes and behaviors (e.g., Robinson and Morrison, 1994; Robinson and Morrison, 1995; Robinson, 1996; Turnley, Bolina, Lester and Bloodgood, 2003) has suggested that the breach of psychological contracts often results in a wide range of negative consequences such as declined job satisfaction or declined trust in the organization. Thus, it is possible that employees may consider organizational change as the breach of psychological contracts committed by their organization, and this promotes their negative reaction to change. In summary, an organization that offers support to its employees, particularly during change processes, may be seen as offering greater inducements to employees, thereby promoting a sense of obligation for the employees to repay to the organization. In this sense, we might expect that perceived organizational support would be correlated with reactions to change. Specifically, it is positively correlated with support for change and is negatively correlated with resistance to change. Accordingly, the following hypotheses are formulated:

H1a:

Employees’ perceived organizational support will be negatively related to their

H1b:

resistance to change. Employees’ perceived organizational support will be positively related to their support for change.

44 Psychological contract refers to an employee’s beliefs about the terms and conditions of the exchange agreement between himself or herself and the organization (Robinson, Kraatz & Rousseau, 1994)

45 An employee feels that the organization breaches the psychological contract when the organization has failed to fulfill one or more of the perceived obligations comprising the psychological contract (Morrison & Robinson, 1997).

39

3.2.2.

Perceived Procedural Justice

An important question arises regarding the likelihood of interaction between perceived procedural justice and reactions to change. Before considering whether, or in what sense, perceived procedural justice can explain employees’ reactions to change, one must first ask what fairness or justice can do. Several researchers in management science have indicated that fairness of organizational policies and procedures exerts an impact on people in organizations (e.g., Adams 1965; Thibaut and Walker, 1975, Gopinath and Becker, 2000). By recognizing the role of fairness of organizational policies and procedures on employees’ decisions, it is possible to investigate if one facet of justice – perceived procedural justice – has an effect on employees’ reactions to change. The literature dealing with (1) how employees react to inequitable processes and outcomes (e.g., Greenberg, 1982; Folger and Greenberg, 1985) and (2) how they try to establish equitable conditions (e.g., Greenberg, 1982; Folger and Greenberg, 1985) has suggested that perceptions of fairness of organizational decision-making processes have significant effects on employees’ attitudes and behaviors. 46 Moreover, empirical research has shown that perceived justice of the organization’s decision-making process has been found to have effects on employees’ reactions to strategy implementation (Kim and Mauborgne, 1993) and pay raise decisions (Folger and Konovsky, 1989). In addition, perceptions of fairness have been found to be positively associated with employees’ organizational commitment (McFarlin and Sweeney, 1992) and job satisfaction (Conlon and Fasolo, 1990). It should be noted that there are several forms of justice in the social psychology literature: one of them is procedural justice; the other is distributive justice. What is procedural justice? For Thibaut and Walker (1975), procedural justice refers to decision control and process control in determining fairness. In a same vein, Tyler (1994) noted that procedural justice deals with the fairness of a procedural or set of procedures. There are

several aspects of procedural justice: for example, consistency and bias suppression (Leventhal, 1980). What is distributive justice? Distributive justice refers to the perceived fairness of resource allocations or outcomes (Moorman, 1991). Rooted in the field of social psychology, distributive justice theory has been used to study various organization phenomena such as conflict resolution process (Karambayya and Brett, 1989) and work group incentive pay plans (Dulebohn and Martocchio, 1998). In addition to its root in social psychology literature, distributive justice is also grounded in equity theory, which states that when outcomes are consistent with individuals’ inputs; they believe outcomes are fair

46 For a detailed review, see Lind and Tyler (1988)

40

(Folger and Crapanzano, 1998). Thus, one may reason that employees evaluate the fairness of their own outcome by comparing it to a reference point (e.g., the outcome of their colleagues). Now let us cast an eye on empirical research on procedural justice in an organizational setting in greater detail. A first example of empirical work on procedural justice is a study by Brockner et al. (1994) which has found that perceptions of procedural justice significantly influenced the reactions of employees who survive layoffs; that is, when survivors of layoffs felt that procedural justice was high, the perception of negative outcomes had no effect on their reactions, whereas survivors who felt that procedural justice was low reacted more adversely to perceived negative outcomes. Another example is an empirical study by Gopinath and Becker (2000) which has found that high levels of perceived procedural justice are correlated with high levels of trust in new ownership and high levels of post-divestiture commitment to the organization. This study has also found that communications from management explaining the events helped increase perceptions of the procedural justice of the divestiture and layoffs. My final example of empirical study on procedural justice is a work by Korsgaard, Sapienza and Schweiger (2002) which has found that procedural justice would moderate the impact of planning change on employees’ obligations, trust in management, and intention to remain in the organization. In light of previous theoretical and empirical research aforementioned, extending these results to the context of organizational change by linking perceived procedural justice to employees’ resistance to change and support for change seems plausible. If employees perceive a high level of procedural justice in organizational change efforts, then consequences of organizational change may seem justifiable. Intuitively that will promote the level of legitimacy of the change. Consequently, it may enhance employees’ support for change, and reduce employee’s resistance to change. On the contrary, if employees hold a perception that procedural justice regarding organizational change is low, then consequences of organizational change may seem unjustifiable. Intuitively that will reduce the degree of legitimacy of the change. Consequently, it may reduce employees’ support for change, and increase employees’ resistance to change. In sum, the following hypotheses are proposed:

H2a: Employees’ perceived procedural justice will be negatively related to their resistance to change. H2b: Employees’ perceived procedural justice will be positively related to their support for change.

41

3.2.3.

Perceived Participation in Decision-making

So far the argument that perceived procedural justice is likely to influence employees’ reactions to change has been made. The argument raises the issue of whether the extent to which employees participate in decision-making might have an effect on their reactions to change. It is clear that while perceived procedural justice and perceived participation in decision-making may be parallel, they are not the same. Therefore, we should distinguish between the two. Consider employees who hold a perception that procedural justice is high, who seek to hedge or minimize the downside consequence of organizational change by influencing decisions regarding the change. The perception of their participation in decision-making is then not contingent upon their perception of procedural justice, implying that one cannot assume that high levels of perceived procedural justice implies or equates high levels of perceived participation in decision-making since they are two different constructs. Several researchers have emphasized the role of participation in decision-making for promoting employees’ acceptance of change (e.g., Blumberg, 1976; Coch and French, 1948, Lewin, 1951). Extending the logic of previous research, one might expect that employees who perceive a low level of their participation in an organization’s decision-making concerning organizational change tend to react more negatively to change than those who perceive a high level of their participation in decision-making. Using the same logic, one might expect that employees who perceive a high level of their participation in decision- making process tend to react more positively to change than those who perceive a low level of their participation in decision-making. The key to these propositions is that if employees who face organizational change efforts evaluate their level of participation in decision- making or the extent to which their organization allows them to participate in the decision- making concerning organizational change or to express their opinions, concerns, or suggestions, and if they correlate this perception of participation in decision-making to their reactions to change, we may expect that a degree of perceived participation in decision- making tends to be associated with employees’ resistance to change and support for change. Empirical research on participation in decision-making has suggested that participation in decision-making is related to a variety of work-related attitudes and decisions. For example, a study by Ruh, Kenneth, and Wood (1975) has found a correlation between participation in decision-making and job involvement. A more resent study by Allen, Shore and Griffeth (2003) has found a significant negative and significant correlation between participation in decision-making and turnover intentions and a significant positive relationship between participation in decision-making and perceived organizational support. However, prior empirical studies on participation in decision-making have shown mixed

42

results. For example, a study by Locke, Frederick, Lee, and Bobko (1984) has suggested that participation in decision-making was less effective in production setting. Thus, findings of this dissertation may help clarify these issues to some extent, and allow a theory to move forward in a systematic fashion. There are at least three reasons to support a proposition linking perceived participation to change on the one hand and resistance to change and support for change on the other hand. One reason is that according to the self-interest model (Thibaut and Walker, 1978), which is based on social exchange theory (Blau, 1964), an organization’s decision has implications for employees; therefore, employees will want to gain control (to some extent) over the decision-making, and exert an influence on any major decision. This is because it is possible that employees consider a control over decision-making as a means to mitigate the magnitude of their own exposure to a potential negative effect. Consistent with the logic of the self-interest model, I argue that if low levels of participation in a decision-making are to promote uncertain feelings concerning a preferred outcome of organizational change, and if employees perceive themselves as a decision-taker rather than a decision-maker because they have a low level of participation in this decision-making process, and correlate this perception with their subsequent decision, we might expect that perceived participation in a decision-making is correlated with their reactions to change. Empirical evidence has suggested the relationship between perceived participation in decision-making and reactions to decisions of organizations. For example, a study by Erez, Earley and Hulin (1985) has found that participative goal setting results in higher goal acceptance than goal assignment and that when a variance in individual goal acceptance exists, the performance of participative goal setting groups are higher than assigned goal setting group. A second reason is that if perceived participation in decision-making is related to a feeling of uncertainties in the organization due to the uncertainties surrounding organizational change and its consequences, then employees who feel uncomfortable about organizational change or a situation with which it is related may attempt to participate in decision-making so that they may be able to obtain information about organizational change (early enough and/or sufficient enough), and to have an opportunity to change it if needed. By having a strong sense of control 47 over a situation, especially decision-making, employees may feel more secure about the situation regarding organizational change. Empirical evidence has suggested that perceived control is negatively associated with

47 A sense of control has been defined as “a generalized belief on the part of the individual about the extent to which important outcomes are controllable” (Parks, 1989: 21).

43

emotional distress (Spector, 1986) and negatively associated with burnout during a downsizing process (Westmann, Etzion and Danon, 2001). A final reason is that employees may view the merit of organizational change not only on the content of organizational change (e.g., reasons for change and the expected outcomes) but also on the process of organizational change (e.g., how a change is designed, evaluated, selected and implemented). Specifically, a perception that an organization does not allow employees to participate in a decision-making or express their opinion, concerns, or suggestions may give a signal that an organization does not care about employees’ feelings. 48 Based on the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960), employees may return in kind the favors of the organization; that is, they may react negatively to the change. Following the above discussion, two hypotheses regarding the effects of perceived participation in decision-making on employees’ reactions to change are developed. In developing the hypotheses, it is assumed that employees have a perception that a higher level of participation in the decision-making they have, the higher a probability of receiving a positive consequence of organizational change (at least for themselves) becomes or the lower a probability of receiving a negative consequence of organizational change (at least for themselves) becomes. It is also assumed that employees have the same preference of a level of participation in decision-making as a baseline. 49 In addition, it is assumed that employees take on the same initial degree of exposure to organizational change. 50 In summary, the effects of the perception of participation in a decision-making process on employees’ reactions to change can be expressed in the following hypotheses:

H3a: Employees’ perceived

participation

in

decision-making

processes

regarding

H3b:

organizational change will be negatively related to their resistance to change.

Employees’

organizational change will be positively related to their support for change.

perceived

participation

in

decision-making

processes

regarding

48 In my view, despite the fact that participation in decision-making is regarded as one facet of procedural justice or fairness, we should not assume that low perceived participation in decision-making implies low perceived procedural justice.

49 The alternative would be to assume that employees have a different preference of a level of participation in decision- making.

50 The alternative would be to assume that employees take on the different initial degree of exposure to a change.

44

3.2.4.

Perceived Need for Change

For organizational change to emerge there must be some underlying causes or antecedents of a change in an organization. For organizational and/or strategy theorists, a combination of internal and external pressures influences an organization to undertake a certain archetype of organizational changes. For any change to occur, one can expect that most organizations that want to undertake organizational change will communicate their compelling reasons for change to employees at one point in time. Research on organizational change has suggested that a proper communication from management tends to help employees understand a situation and a need for organizational change, thereby facilitating change processes and reducing employees’ resistance to change (see, e.g., Kotter 1995; Kotter and Cohen, 2002). One of purposes of communication between organizations and their employees is to legitimize organizations’ decision to enact organizational change, subsequently promoting employees’ acceptance of the change. Empirical research on organizational change management seems to support the role of communication during change processes. For example, an empirical study by Gopinath and Becker (2000) has found that in contexts of divestiture, communications from management that help employees understand the events relating to the sales of business units are positively correlated with the perceived procedural justice of the divestiture and the absolute justice of layoffs. Up to this point, it raises a question of whether employees will share the similar view of a situation with that of their organization. 51 Starbuck and Milliken (1988) suggested three reasons concerning why perceptual filters of managers might affect noticing changes in the environment. First, individual habits and beliefs will exert an effect on what they notice. Second, some stimuli that are not actively noticed must change significantly in order to get noticed. Third, the organizational institutionalization may induce problems of noticing because resources are assigned to track those stimuli that top managers have perceived as important, while other stimuli that top mangers have not identified as important may not be traced, and go unnoticed. Additionally, a study by Barr, Stimpert and Huff (1992) attempting to investigate a relationship between changes in mental models of top managers and changes in organizational action has suggested that top managers may not notice changes in the environment because these changes are not central to existing mental models

51 Or, to be more precise, that of the organization’s top management. Note that, unless stated otherwise, the terms “top management” and “organization or organizations” are used interchangeably, i.e. top management’s view refers to the organization’s view, and vice versa.

45

of top managers, and it has also found that the organization that renewed its strategy and still survives shows a definite succession in mental models. Based on the preceding discussion, it raises a question of whether the concept of the perceptual filters of managers may be applicable to employees. If employees have such filters, we might reason that they may not notice changes in the environment because their existing mental models do not focus on those changes. One possible implication of this is that employees who have not noticed changes in their environment tend to have a differing view on organizational change than those (e.g., top managers or other employees) who have noticed these changes in the environment.

Figure 5: Five Stages of Organizational Decline

environment. Figure 5: Five Stages of Organizational Decline Source: Weitzel and Jonsson (1989), p. 102 But

Source: Weitzel and Jonsson (1989), p. 102

But what is a perceived need for change? Anderson and Paine (1975: 815), from a strategic formulation standpoint, defined a perceived need for change as “the perception by the

46

strategic formulator (boundary spanner) of a distinct lack of competence, capabilities, or internal resource to carry out a planned program of action”. If organizations fail to perceive a need for change and to carry out a corrective action, they may experience a decline. As illustrated in Figure 5, Weitzel and Johnsson (1989) suggested a model that portrays five stages of organizational decline from the early stage of decline to the demise of organization as follows: the blinded stage, the inaction stage, the faulty action stage, the crisis stage, and the dissolution stage. One plausible interpretation on Anderson and Paine’s (1975) definition of a perceived need for change and on Weitzel and Johnsson’s (1989) model of organizational decline is that for organizational change to be perceived as needed, one condition must be fulfilled:

that is, a currently planned program of actions is less likely to be fulfilled, subsequently resulting in lower performance. In this sense, it may suggest that people (e.g., top managers and employees) may perceive that their organization needs to change when viewing that a considerable threat to certain aspects of the organization exists. Hence, this definition of perceived need for change does not take another important aspect of organizational change into account. That is, organizational change may be initiated in order to capitalize on an opportunity to enhance firm performance or to preempt competitors from taking opportunities that may present a great threat to the firm. One may argue that for organizations with good performance, there may not be a need for change to survive, but there may be a need for change to take an advantage of available opportunities to perform better. Similarly, one may argue that although organizations may perform well, they should try to enact organizational change that may improve their performance. In this sense, we clearly observe an ambiguity of the term “perceived need for change”, suggesting a subjective nature of how one defines and quantifies a need for change. It is useful to note that from employees’ point of view, the context in which organizational change takes place tends to exert an effect on employees’ perceptions of need for change. For example, organizational change in a turnaround situation tends to be perceived of greater need than organizational change in a normal situation, demonstrating the implication of the context in which organizational change occurs on employees’ perceived need for change. In summary, two hypothesize are proposed:

H4a: Employees’ perceived need for change will be negatively related to their resistance to change.

Employees’ perceived need for a change will be positively related to their support for change.

H4a:

47

3.2.5.

Attitude towards Organizational Change

Several researchers (e.g., Kirton and Mulligan, 1973) have suggested that there are effects of employees’ attitudes towards organizational change on their reactions to change. As discussed earlier, any attitude may register directive effects on a person’s predisposition to formulate a certain type of pre-determined reaction to a wide range of objects, it ought to be possible that employees’ positive or negative attitudes towards organizational change may exert an effect on their perceptions of organizational change and a situation with which it is related. Much of the attitude literature has suggested that there is considerable consistency in the underlying belief that an individual with attitudes about an idea, item or object is likely to approach such idea, item or object with a set of pre-determined behaviors, essentially demonstrating a distorted information processing and interpreting process and a form of preference. As Allport (1935: 8) has defined an attitude as “a mental and neural state of readiness, organized through experience, exerting a directive or dynamic influence upon the individual’s response to all objects and situations with which it is related”, one can reason that attitude induces not only pre-determined responses but also tendencies towards performing of such responses. Prior research has shown that attitudes are related to a variety of decisions of people in organizational setting. For example, Robey (1979) argued that attitudes could affect the use of a computer based information system, and found that users’ attitudes are significantly related to management information system use. Past research has suggested that employees’ experiences with other decisions of an organization have an impact on their evaluation of the current decision of the organization. For example, Davy, Kinicki, and Scheck (1991) found that the ways in which employees evaluate an organization’s decision-making process (e.g., decision-making process concerning a layoff) are undertaken in the context of their experiences with other decisions. Therefore, it is logical to argue that prior evaluation of organizational change is likely to exert a directive effect on the evaluation of a present organizational change effort. If a series of the evaluations results in similar outcomes, one might reason that these outcomes, through the attitude construction mechanism, may form or promote a positive or negative attitude about organizational change. What is attitude towards organizational change? Generally speaking, there is no general agreement on the definition of attitude towards organizational change in the literature. Grounding on Sarnoff’s (1960) definition of attitude, attitude towards organizational change is defined here as a disposition to react favorably or unfavorably to a change within an organization. It must be noted that there is no restriction on this definition with regard to whether attitudes toward organizational change are conscious or unconscious

48

in nature. My position is that a question of whether employees recognize the existence of their attitudes is of irrelevance to this dissertation so long as an influence of attitudes towards organizational change on shaping employees’ reactions to change does in fact exist. Parallel to the logic of hypotheses 4a and 4b, it is expected that there is a direct relationship between employees’ attitude towards organizational change and their reactions to change. Specifically, it is expected that a positive attitude towards organizational change may result in greater positive perceptions of a current change and/or greater feelings of comfort with the change and, through positive perceptions of a current change and/or feelings of comfort with the change, lead to lower levels of resistance to change and/or higher levels of support for change. Likewise, it is possible that a positive attitude towards organizational change may result in a distortion of perceptual process. Thus, a positive attitude towards organizational change may promote a positive perception of the change, weaken any feeling of uneasiness with the presence of the change, and thus facilitate a decision to accept or support a change. On the other hand, a negative attitude towards change may create negative perceptions of a current change and/or feelings of uncomfortable with a current change, which can result in a distortion of perceptual process. In this sense, employees who have a negative attitude towards organizational change may have a negative perception of organizational change that promotes feelings of uneasiness with the presence of organizational change and, thus, leads to higher levels of resistance to change and/or lower levels of support for change. In sum, the effects of attitudes towards organizational on employees’ reactions to change can be expressed in the following hypotheses:

H5:

Employees’ (positive) attitude towards organizational change will be negatively

H5:

related to their resistance to change. Employees’ (positive) attitude towards organizational change will be positively related to their support for change.

49

3.2.6.

Fear of Known Consequences of a Change

One can question whether fear serves as a determinant of humans’ reactions. Or for that matter interferes or hinders the process of arriving at any reaction. In either case, it is a valuable moment to examine if there is an effect of fear on humans’ decision. For purposes of discussing fear of known consequences of change, one must first note that fear is one kind of emotional states (Ortony et al., 1988) that are capable of exerting influences on humans’ decision. How can fear be conceptualized so as to make this concept useful for management theory? Interestingly, it is relatively difficult to find a clear and precise conceptualization of fear in the management literature. But according to Merriam-Webster’s Collegiate Dictionary, fear refers to “an unpleasant often strong emotion caused by anticipation or awareness of danger”. Due to a lack of empirical or theoretical knowledge of fear in the management literature, this dissertation follows much of theories and frameworks from other social sciences (e.g., psychology) to understand effects of fear in the organizational setting, particularly in an organizational change situation. But what does this fear of known consequences of a change effort have to do with employees’ reactions to change? Fear is often considered as a factor that triggers employees’ resistance to change (Agócs, 1997, Kotter and Cohen, 2002). But why is this so? In the case before us, where the organization is trying to undertake organizational change, which is intended to improve organizational performance, fear of known consequences of a change becomes a barrier to employees’ acceptance of change, because it exerts a negative effect on any person’s rational thinking. If we accept the notion that fear can affect our thinking and reasoning, we might expect that fear can also affect our decision-making in general and our decision- making concerning a reaction to change in particular. Thus it is not hard to observe that researchers have associated fear with resistance to change. For example, Dubrin and Ireland (1993) noted that resistance to change is attributed to employees’ fear of poor outcomes, the unknown, and realization of pitfalls with the change. In the same vein, Kotter and Cohen (2002) posited that fear or panic drives self-protection or immobilization. There are numerous explanations for a question of why we react negatively to an object or idea while we fear such objective or idea. First, fear can suppress our rational thinking and decision-making—that is to say, fear may shorten a person’ sequences of thinking process or duration of such processes because it reduces a set of information needed for making a decision and encourages impulsive decisions. Second, in circumstances where fear exists, we may arrive at decisions that we may not have made in other circumstances. In the extreme case, fear can force anyone to kill someone. Although we do not expect that fear will force employees to kill managers, we must recognize that fear is

50

powerful and capable of invoking a person to do things unimaginable. Third, the known consequences may be psychologically unbearable—that is to say, employees who view the know consequences as psychologically unacceptable or as a threat to their psychologically well-being may react negatively to ideas of experiencing this change. Finally, the known consequences may be viewed as practically unattainable at the level of individual or firm. For example, if employees who realize new demands of them and limitations of their resources and/or abilities do not hold a belief that they could meet these new demands, then they will be more likely to react negative to this change. Through the fear-generating mechanisms and the effects of fear of known consequences, employees tend to react negatively to change. The argument for employees making resistance to change is relatively straightforward. To the extent that employees are afraid of or fear the known consequences, they will feel that this change is not good, at least not good for them. Or is it the other way around? In either case, fear will drive them into protection mode. In order to get rid of feelings of fear, they either (1) accept the situation as it is and live with it or (2) remove a source of fear—that is to alter the change. Using the logic of social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960), it is argued that feelings of fear are likely to result in a more negative evaluation of the organization making organizational change and of the change because this change is viewed as a negative treatment a person receives from the organization. Therefore, if they are afraid of or fear the known consequences, and they feel that the organization treats them badly with this change, they return the organization a negative treatment—they resist the change. Although much of prior research on resistance to change has suggested that, under an underlying assumption that employees’ fear tends to recede when they are well aware of what is going on (e.g., Kanter, 1995), the positive effects of communication in various phases of the change process may minimize their resistance to change. However, the organization’s communication of organizational change alone may not be sufficient to reduce employees’ resistance to change because fear induced by knowing the content (e.g., the consequences of the change) may be responsible for their resistance to change. In summary, to test the potential effect of fear of known consequences of a change on employees’ reactions to change, the following hypotheses are formulated:

H6a:

Employees’ fear of known consequences of a change will be positively related to

H6b:

their resistance to change. Employees’ fear of known consequences of a change will be negatively related to their support for change.

51

3.2.7.

Fear of Unknown Consequences of a Change

As the preceding discussion of fear of known consequences of a change concerns primarily the effects of fear induced by knowing consequences of a change on employees’ reactions to change, I now turn my attention to a related-but-distinct factor—fear of unknown consequence of a change. Specifically, I examine whether there is any relationship between employees’ fear of unknown consequences of a change and their reactions to change. From a practical perspective, shifting attention from explaining employees’ reactions to fear of known consequences to explaining their reactions to fear of unknown consequences allows an organization to decide on the extent to which the information on organizational change should be communicated to its employees. In the organizational change literature, there are suggestions that change communications are more likely to reduce employees’ resistance to change (Kanter, 1995; Cobb, Wooten and Folger, 1995) because these communications will reduce anxiety about change (Kanter, 1995), and increase a perception of fairness of change (Cobb et al., 1995). Because past research has found that organizational change is related to anxiety and stress (Jick, 1985; Leana and Feldman, 1992), most change management models and frameworks (Judson, 1991; Kotter, 1995; Galpin, 1996) have emphasized the roles and effects of communication during the change process. Anticipating the effect of fear of unknown consequences of a change on employees’ resistance to change is supported by two reasons. One reason is that the majority of people tend to be afraid of moving from the known to the unknown—that is to say, if organizational change entails moving employees from the known sphere to the unknown sphere, and if employees are afraid of the unknown, then we might expect that employees who do not know the consequences of a change tend to have a negative response to that change (Steinburg, 1992; Coghlan, 1993). Another reason is that if the majority of people tend to form a negative perception about the unknown, we might expect that employees may form a negative perception about the unknown consequence, leading to fear of not knowing consequences of change. This fear of unknown consequences of a change then contributes to resistance to change. As Mabin, Forgeson and Green (2001) noted that fear of unknown may be defined as being uncertain about the nature of change, feeling that he or she does not know what is going on, and what the future is likely to hold, it is possible that employees may have a rather dark image of the unknown consequences of a change. As discussed earlier, most people do not want to experience a feeling of fear. If employees decide to direct their fear of unknown consequences of a change towards the organization because they feel that such change is a bad treatment of the organization for

52

them, using the logic of the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960), we might expect that they will react negatively to organizational change that has induced fear to them. However, past research has not discussed the likelihood that without fear employees will support the change. Therefore I also examine whether there is a relationship between fear of unknown consequence of a change and support for change— that is to say, if employees have a low level fear of unknown consequence of change, then they may support such change. Although this prediction does not directly address a question of whether communication during the change process will enhance employee’s support for change, it will indirectly address a question of whether communication that minimizes employees’ fear of unknown consequence of a change will promote their support for change. 52 To summarize, the above discussion suggests two hypothesized relationships. On the one hand, I predict a negative relationship between fear of unknown consequence of a change and resistance to change. Using the same theoretical lens, I examine whether there is a negative relationship between fear of unknown consequences of a change and support for change. Accordingly, the following hypotheses are formulated:

H7a:

Employees’ fear of unknown consequences of a change will be positively related to

H7b:

their resistance to change. Employees’ fear of unknown consequences of a change will be negatively related to their support for change.

52 Many researchers (e.g., Kotter, 1995; Galpin, 1996; and Kotter and Cohen, 2002) have argued that organizations’ communication to employees during organizational change processes will lower employees’ fear of unknown consequences, and increase employees’ support for change.

53

3.2.8.

Perceived Change in Power

Research in a variety of setting has revealed a consistent view of power. In particular, several studies have noted that loss of power is associated with resistance to change (e.g., Agócs, 1997; Trader-Leigh, 2002). This dissertation therefore examines if there is a relationship between perceived change in power and reactions to change. What is power? The term “power” can be seen in a wide range of disciplines: for example, political science, organization theory, sociology and psychology. Several definitions of power have been offered and adopted in research over year. One of influential definitions of power is French and Raven’s definition. That is, power has been defined as “the perception by P that O has a legitimate right to prescribe behavior for him” and that P accepts “that O has a legitimate right to influence P and P has an obligation to accept that influence” (French and Raven, 1959: 151). Another example of definitions is the one offered by Pierce and Newstrom (1995: 21), defining power as “the ability to change one’s environment”. A review of literature on power has suggested that the concepts of power, influence, and control have been used almost interchangeably. For example, legitimated power is one form of influence, and individuals develop power to have a control in order to meet organizational and personal objectives (Pfeffer, 1981). Similarly, influence refers to “the process whereby one party changes the views or preferences of another so that they now conform to their own” (Dawson, 1996: 170). In addition, a concept of power distance, which refers to the extent to which less powerful members of an organization or a society accept an unequal or hierarchical distribution of power within an organization or in a society (Hofstede, 1980), has been as a key explanation to various studies. Here, it seems to me, concepts of power, influence and control are, at least, overlapping each other. In organizational setting, power is generally associated with a position in an organization; that is, a higher level of a hierarchical position in an organization, a higher level of power is associated with it. In this view, any change in an organization that affects a power or hierarchical structure in an organization may result in creating a combination of three distinct groups of employees. A first group consists of any employee who will receive or gain greater power in this organization as a consequence of this change. A second group consists of any employee who will have lesser power in this organization as a consequence of this change. A final group consists of any employee who will experience no change of their power in this organization. It is not surprising that employees who will have less power in an organization as a result of this change may react to this change differently than those who will have more power as a result of this change or than those who will maintain their current level of power. One important aspect requires attention here. That is, there is a question of whether employees’ goals and an organization’s goals are always aligned. To

54

this question, we may reason that both managers and employees regarded as an agent of an organization may have a set of personal goals (e.g., career goals) that is not necessarily coherent or aligned with that of an organization. 53 The key here is that an effort to achieve the organization’s objectives by inducing organizational change may affect the attainment of employees’ personal goals. Thus, employees whose personal goals have been negatively affected by organizational change may attempt to defend their personal goals through various means. A perceived negative change in power as a consequence of a change can lead to aversive psychological outcomes. Using logics of the social exchange theory (Blau, 1964) and the norm of reciprocity (see, e.g., Blau, 1964; Eisenberger et al, 1986; Fasolo and Davis-LaMastro, 1990), one can reason that employees’ strong feelings of deterioration in power may result in their resistance to change because they may feel that their organization has mistreated them. If a perceived change in power is considered as an injury or mistreatment committed by the organization, then employees may have a negative response to the change. On the other hand, employees who have a perception that they have, or will have, more power as a consequence of a change may have a more positive response to the change because they may view this perceived gain in power as a benefit provided by the organization, and feel obliged to repay the organization by providing support for change. Based on the social exchange theory and the norm of reciprocity, it can be argued that employees in a change situation will have a normative expectation about the likelihood of various outcomes regarding their power position in an organization that might occur in that setting. The actual outcomes that occur in a change setting do not necessarily conform to these expectations for a variety of reasons. Suppose that the least employees expect is to maintain the current level of their power. Any such deviation of actual outcomes from expectation has an effect on their reactions to change. Employees may reduce negative deviations (loss in power) by resisting such change (with an attempt to alter the change so that their power will not be reduced). On the contrary, employees may not reduce positive deviations (gains of power), and thus support the change. Thus, the following hypotheses are proposed:

H8a:

An employee’s perceived rise (fall) in power resulting from a change will be

H8b:

negatively (positively) related to their resistance to change. An employee’s perceived rise (fall) in power resulting from a change will be positively (negatively) related to their support for change.

53 In economics, this kind of problem is known as a principal-agent issue, see Fama (1980) for a detailed discussion.

55

3.2.9.

Perceived Change in Status

Does it bother you when you get a lower status in your social group? Or when your friend or colleague rejects, neglects, or ignores your request that was once easily accepted? Or for that matter disagrees or argues with you? Of course, you are left feeling hurt, upset, and angry, but the question is why? Why does that bother us? Why do you care if you get a lower status in a society? 54 So often we throw around the words like status, ranking, and position. But what do they all mean and how are they connected to our decision? In general, status indicates the relative standing of a person in his or her social group. But what does this have to do with being upset, hurt, or angry? Researchers have argued that a person’s status is based on the prestige, honor and deference accorded him or her by other members of the group (Lovaglia and Houser, 1996). 55 When we talk about the concept of status, we refer to two interrelated concepts: (1) the creation of status ranking, and (2) the attempt to achieve high status ranking (Waldron, 1998). That would be great if we could get it ourselves directly, but status is a by-product of how we behave or live our life. It cannot be gained directly. Notice that I say “by other members of the group”. In order to gain the prestige, honor and deference accorded by other members, we must do what is right, and more importantly be perceived to do something right, meaning that we must be able to behave for a period of time that can be weeks, months, or even years. This is why any situation that takes away our status in effect harm our self-esteem, because when we feel being demoted in status, we think we do something wrong. Therefore we find that our self-esteem is intertwined. And that is the key. A change in anyone’s status implies a change in a set of relationship between this person and other people in the social group. As we will see, it is the loss of status that exerts effects on our decision. When we see our status in a society is declining, we will often see ourselves fighting back against whatever seems to cause that change. We become defensive and that is understandable. For instance, when we are ordered to take up a position that is considered to be lower than the one we are currently occupying, we do not feel good about ourselves, and we may even feel guilty and fight back. Any situation that robs us of our status forces us consciously or unconsciously to react negatively. Prior research on organizational change management has suggested that employees’ status quo in

54 Research on status has been undertaken in at least three disciplines: (1) sociology, (2) anthropology, and (3) psychology. Arguments made in this dissertation are mainly from theories and frameworks from sociology and psychology.

55 Note that we should distinguish between status in organizational contexts and the formalized positions in the organizations. As Waldron (1998) points out, formalized positions and status rankings may often be parallel, but they are not identical.

56

an organization induces resistance to change, particularly when a change negatively affects their status quo (e.g., Smith, 1982; Spreitzer and Quinn, 1996). Additionally, Spreitzer and Quinn (1996) discovered that middle managers claimed that higher-ranked executives resisted change efforts. Similarly, Smith (1982) found that people in power often try to maintain the status quo rather than change it and that when a proposed change is perceived of having a negative impact on their power, such perceived loss of power enacts a defense mechanism, leading to resistance to change. Research has also shown that the achievement of status is one of the important concerns of employees (e.g., Kovach, 1987). Many people tend to associate power with social status, maintaining that the greater the power, the higher the rank of social status becomes. When organizational change affects our status in a negative way, it causes us to question ourselves as a person, and we have a tendency, conscious or unconscious, to behave in ways that may maintain our status. Do you feel that words like power and status sound similar to us? Although power and status may often be parallel, they are not the same. Therefore it is important to distinguish between the two, and that will allow us to study the effects of perceived change in status. There is also another aspect to consider. Based on the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960), it is argued that employees will view a decline in their status in an organization as a negative treatment they receive from an organization, and feel obliged to return a negative treatment to the organization (Gouldner, 1960). This is why people are easily annoyed and upset when something demotes their status. And this is exactly why, when a change affects employees’ status, they tend to have a negative response to such change. On the other hand, one can argue that employees facing organizational change will react positively when they feel that such change will provide them a higher status in an organization. This is because employees may perceive a higher status as a positive treatment they receive from an organization, and feel obliged to return a positive favor or benefit to the organization, leading them to support the change. In line with this discussion, it can be reasoned that there may exist the relationship between employees’ perceptions of change in status resulting from organizational change and their reactions to change. Thus, the following hypotheses are proposed:

H9a:

An employee’s perceived rise (fall) in status resulting from a change will be

H9b:

negatively (positively) related to their resistance to change. An employee’s perceived rise (fall) in status resulting from a change will be positively (negatively) related their support for change.

57

3.2.10. Perceived Change in Pride An idea that employees’ perceived change in pride affects their reactions to change has yet become well studied. A question in this context is: Does a perceived change in pride resulting from organizational change exert an effect on a choice of reactions to change and the magnitude of any chosen reaction? Based on the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960), a general hypothesis is that any modification to a person’s pride will be the major driver of his or her subsequence reactions to a cause of such modification. It is argued that employees facing organizational change will have a normative expectation about the likelihood of various outcomes of a change that might occur in that setting. Any outcome of a change that robs or decreases employees’ pride forces them to react negatively to such change—that is to say, they will resist such change. On the other hand, any outcome of a change that promotes or increases employees’ pride encourages them to react positively to such change—that is to say, they will support such change. What is pride? In general, pride is “the quality or state of being proud”. In the organizational setting, pride can be defined here as the quality or state of being proud about oneself in one’s social group (e.g., in one’s workplace). A question then is why any change in pride has an effect on employees’ reactions to change. Initial answers to this question focus on the psychological motivations for reacting to any change in pride. It is logical to expect that any person who feels being robbed of his or her pride will feel hurt and upset, and try to regain his or her pride. This is because we, as a human being, socialize with other member of our social group. Interactions with other members and the group’s acceptance of us in a social group are the key to our psychological well-being. When a person is left feeling a loss of pride, these emotions or feelings tear away at his or her self-esteem. 56 Loss of pride can lead to embarrassment and disruptive modifications of ties or linkages between people and their social group. Thus, they will try to regain their pride, and remove their embarrassment. One of possible cures for this loss of pride is to remove its cause—the change. Thus, if employees perceive organizational change accountable for their loss of pride, and feel that removing such change will bring back their pride, they will try to remove such change, meaning that they will resist the change. As mentioned earlier, the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960) can be used to explain the possible relationship between perceived change in pride on the one hand and resistance to change and support for change on the other hand. Consider organizational change that decreases employees’ pride. Suppose

56 See Maslow (1954, 1970) for a detailed discussion of the hierarchy of need’s theory.

58

that employees’ pride is based on how they are proud of themselves in their social group. If employees have a feeling that organizational change causes a negative change in their pride (that they have in their workplace), and perceive this negative shift in pride as a negative treatment or injury they receive from an organization, we might expect that they will want to return in kind this injury to the organization—that is to say, they will resist the change. On the contrary, if employees have a feeling that organizational change causes a positive change in their pride (that they have in their workplace), and perceive that this positive change in pride as a positive treatment or favor they receive from an organization, we might expect that they will feel obliged to return in kind this benefit to the organization—that is to say, they will support the change. This view is consistent with past research that has suggested that pride is related with humans’ decision. For example, Kotter and Cohen (2002) found that false pride or arrogance drives complacency, and that prevents the implementation of planned change. This means that when employees perceive that their pride may be obstructed or negatively affected by organizational change, they tend to behave in such a way that may maintain the current state of pride, meaning that they will resist the change. It is important to note that pride is often associated with power; that is, the greater the power, the greater the pride becomes. But it is crucial that we distinguish between the two because even if power and pride may often be parallel, they are not the same. To illustrate this point, consider whether employees who are proud of their contribution to the firm’s success will always have more power in the organization. Of course their contribution may lead them to be promoted to a higher position within the firm. A question is then whether we can firmly conclude that pride and power are parallel in a situation where the contribution does not lead to any promotion or more power 57 , and they are still proud of themselves for what they have achieved for the organization. Thus, we should not assume that pride and power are the same in this context. In sum, the following hypotheses are proposed:

H10a: An employee’s perceived rise (fall) in pride resulting from a change will be negatively (positively) related to their resistance to change. H10b: An employee’s perceived rise (fall) in pride resulting from a change will be positively (negatively) related to their support for change.

57 Note that I focus only the power associated with a position in an organization. For simplicity, other kinds of power or how others’ perceptions of his power are not taken into account.

59

3.2.11. Job Satisfaction Now I should like to discuss another factor in my model—job satisfaction. The question is whether there is a direct relationship between job satisfaction and reactions to change. Past studies have tried to explain relationships between job satisfaction and a variety of decisions and behaviors. For example, Mobley (1977) noted that job dissatisfaction often leads employees to quit the organization due to the effect of the expectation that leaving an organization will result in getting a more satisfying job. Therefore, it seems logical to argue that job satisfaction may place considerable condition on an exchange relationship between employees on the one hand and the organization and their reactions to change on the other hand.

To fully understand what is going on, we have to answer the question: what is job satisfaction? Research on job satisfaction has generated several definitions of job satisfaction. I can only mention a few definitions here. First, Smith, Kendall and Hulin’s (1969: 6) defined job satisfaction as “the extent to which an employee has a positive affective orientation or attitude toward their job, either in general or towards particular facets of it”. Second, Locke (1976: 1300) defined job satisfaction as “a pleasurable or positive emotional state resulting from the appraisal of one’s job or job experiences”. Finally, Dorman and Zapf (2001: 486) defined job satisfaction as “a pleasurable emotional state resulting from the appraisal of one’s job”. In a nutshell job satisfaction tells us about how a person feels about his job—for example, whether he is happy or unhappy with his job.

When employees are satisfied with their job, they may be happy because it satisfies how they want to see themselves. And this helps them establish their ability to feel in control. On the other hand, when they are not satisfied with their job, they may not be happy because it hurts how they see themselves. It robs them of their ability to feel in control. When they are not satisfied with their job, they feel uncomfortable. When they are told doing wrong things at work, they go into protection mode—they try to defend themselves of what they did, and more importantly, of a person they are. This is why organizational change that changes the degree to which they are satisfied with their job may bother them. A person with a perception that this change means he has performed not well enough will fell uncomfortable, hurt, and upset. Because these feelings eat away at his self-respect and self-esteem, he feels less good about himself. So he becomes defensive and wants to protect his self-respect and self-esteem. He will react negatively to this change. On the other hand, when a person thinks that this change indicates that he has worked well, he will feel good, happy, and joyful. Because these feelings enhance his self-respect and self-esteem, he feels better about himself. So he becomes more

60

open and wants to gain even more self-respect and self-esteem. Therefore he will react positively to this change. Prior research on job satisfaction has suggested that job satisfaction is positively correlated with attitude towards change (Gardner et al., 1987). Moreover, it has been empirically confirmed that low job satisfaction is related to high turnover (e.g., Mobley, Horner and Hollingsworth, 1978; Porter and Steers, 1973). Another empirical study by Farrell (1983) has found that low job satisfaction is related to four categories of behaviors— exit, voice, loyalty and neglect. Past studies have also suggested that job satisfaction is sigificnatly correlated with organizational commitment (e.g., Brooke, Russell and Price, 1988, Davy et al., 1991). A more recent study by Iverson (1996) has found that employees with high levels of job satisfaction will have higher levels of acceptance of organizational change. In addition, it should be noted that despite a general tendency to measure job satisfaction at the overall job, several researchers have started to measure satisfaction with job facets such as pay or promotion opportunities (e.g., Smith et al., 1979; Taber and Alliger, 1995). In addition, some searchers (e.g., Weiss et al., 1967; Davy, Kinicki, and Scheck, 1997) have started to distinguish between intrinsic job satisfaction and extrinsic job satisfaction. Intrinsic job satisfaction denotes a pleasurable emotional state resulting from the appraisal of one’s intrinsic aspects of the job, whereas extrinsic job satisfaction denotes a pleasurable emotional state resulting from the appraisal of one’s extrinsic aspects of the job.

As the above discussion suggests, employees who perceive lower levels of job satisfaction as a direct or indirect consequence of a change are more likely to react negatively to change. Likewise, employees who have high levels of job satisfaction as a direct or indirect consequence of a change are more likely to react positively to change. In summary, the following hypotheses are formulated:

H11a: Employees’ job satisfaction will be negatively related to their resistance to change. H11b: Employees’ job satisfaction will be positively related to their support for change.

61

3.2.12. Job Security Is it possible that job security or insecurity might exert an influence on employees’ reactions to change? The big question here is whether employees can make their decision properly and correctly when they do feel insecure about their job situation. A variety of studies have shown some consistency in the strength or even the direction of the relationships between job security and several work-related decisions observed. For example, an empirical study by Westman et al. (2001) has found that there is a relationship between job security and burnout, whereas an empirical study by King (2000) has found that job security has a significant impact on work efforts, organizational loyalty as well as on citizenship behavior. Another empirical study by De Witte (1999) has also found that job security has effects on psychological well-being. As a result, it ought to be possible that perceptions of job security may have effects on employees’ reactions to change. What is job security? Unlike job satisfaction, job security reflects the perceived continuity in a job a person receives for his or her contribution to the organization. Job security has been defined as “one’s expectations about continuity in a job situation” (Davy et al., 1997: 323). What is job insecurity? Job insecurity has been referred to the degree of uncertainty an individual has about his or her job continuity (see e.g., Greenhalgh, 1982). Job insecurity has been defined as “the lack of control to maintain desired continuity in a threatened job situation” (Hui and Lee, 2000: 216). But how does job security or insecurity have to do with reactions to change? We all know that for a change, any change, in an organization, we might sense its impact in one way or another. If we do not feel anything about this change with regard to job security, then it is fine. But if we feel that this change affects our job security, it causes us to question our own self-worth and react on it. This is why employees with a perception of job insecurity are highly sensitive—because they question themselves of their ability to work and remain with the organization. When we feel insecure about our job situation, we feel less good about who we are and what we do. The greater our feeling of job insecurity, the greater hurt we feel. Therefore we go into protection mode—we react negatively. Resistance to change is the impulsive response to this hurt feeling because we direct our resistance toward the source that we feel responsible for eating away at our self-respect and self- esteem. In sociology and psychology, the literature on trust suggests that trust lowers transaction costs by ‘replacing contracts with handshakes’ (Adler, 2001). Therefore, we may expect that both parties—employees and employers—would try to honor both formal and informal contracts because withdrawing from the contacts is usually costly, and parties are willing to accept such costs which can be in monetary and/or non-monetary forms only if positive consequences of a contractual cancellation are considerable. Has there ever been a

62

time in your life when you had implicit expectations about your job security? Prior studies have insisted that we do have implicit expectation about our job security (Rousseau, 1989). Researchers have suggested that any perceived threat to the job security represents a possible violation of the informal or psychological contract, and that leads to withdrawal cognitions (Davy et al., 1997). Feelings or perceptions of job insecurity during change processes can lead to aversive psychological outcomes. This occurs through a complex process (e.g., interpretation), which tends to weaken a person’s rational thinking mechanism. Strong feelings of job insecurity that stem from organizational change can result in resistance to change because employees may view resistance to change as a means to reverse the change. Why is this so? By focusing on the negative aspects of change (e.g., job insecurity), employees will feel less good about the change. When their feeling of job insecurity becomes so strong, they become very emotional and tend to forget about other potentially good aspects of the change. In addition, perceptions of uncertainty regarding the existence of one’s current job may be as damaging as actual job losses (Latack and Dozier, 1986). Another question is whether employees who feel secure at their job during change processes are likely to be less emotionally affected by the change, thereby decreasing a tendency to engage in such behaviors that opposes the change. Intuitively the magnitude of perceptions of job security or insecurity is likely to influence how employees will react. It is noteworthy that it may occur that a negative shift along a continuum—from job security to job insecurity—tends to have a far greater effect on employees’ response than a positive shift along the same continuum—from job insecurity to job security. But do you feel that traditional notions of job security remain in many workplaces nowadays? Rousseau (1995) posited that in some workplace settings employees tend to focus on employability rather than job security. Thus, the extent to which employees value job security may mediate the effect of job security on reactions to change. Additionally, a notion of work-life balance 58 was found positively correlated with organizational commitment (Scandura and Lankau, 1997). This may also condition the effects of feelings of job security on reactions to change. In sum, the following hypotheses are proposed:

H12a: Employees’ job security will be negatively related to their resistance to change. H12b: Employees’ job security will be positively related to their support for change.

58

The concept of work-life balance concerns with career advancement and family responsibilities (Wolfe & Kolb,

1980)

63

3.2.13. Job Motivation Recent developments in the management literature, and particularly in the field of organizational behavior, have raised great interests as to how organizations can raise the extent to which employees are motivated to perform their job (e.g., Herzberg, 1968). However, it is not the purpose of this dissertation to discuss these interests, or to determine whether organizations can motivate employees, or to examine what they should do to motivate their employees. 59 Rather, I shall assume that most, if not all, actions of an organization affect employees’ job motivation, and shall put forth some proposition as to the effect of job motivation on employee’s reactions to change might be. Broadly speaking, an organization’s actions are expected to affect employees’ job motivation because these actions may lead to significant changes at the employee level such as new job characteristics, new hiring, or even dismissal (Korsgaard et al., 2002). Because past research has suggested that job motivation tends to affect work performance (e.g., Herzberg, 1968), it can be reasoned that organizational changes may have an effect on employees’ job motivation, and that job motivation may in turn determine not only the choice of reactions to change but also the strength of such reaction. In this respect, this dissertation focuses on examining if there is a direct relationship between job motivation on the one hand and resistance to change and support for change on the other hand. What is job motivation? Or for that matter, what is motivation? In psychology, motivation has long been defined in relation to need strength. For example, McClelland’s need theory has posited that humans are motivated by need for power, achievement and affiliation (McClelland and Boyatizis, 1984). Prior research on job motivation in the management literature seems to follow the frameworks of research on motivation in psychology. For instance, the motivation-hygiene theory of job attitude (Herzberg, 1968) has suggested that there are two sets of different needs of humans: one of which relates to animal nature; another relates to unique characteristic of human—the ability to achieve and to experiment with psychological growth. Research on job motivation has distinguished between intrinsic motivation and extrinsic motivation (Herzberg, 1968; Hui and Lee, 2000; Sansone and Harackiewicz, 2000). Intrinsic motivation represents “the relationship between a person and his/her job itself” (Hui and Lee, 2000: 216) and is derived from within the person or form the activity related to the job itself (Sansone and Harackiewicz, 2000). Research on intrinsic motivation has suggested that task variety, task significance, task identity, and feedback from the task

59 In the research on job motivation, there are several theories: for example, Kanfer’s taxonomy of motivation theories (1990) and Vroom’s expectancy theory (1964).

64

are the key characteristics that induce intrinsic motivation (Hackman and Oldham, 1976). On the other hand, extrinsic motivation represents the relationship between a person and externally administered rewards such as pay or prestige from others (Komaki, 1982). An empirical study by Stumpf and Hartman (1984) has found that there is a positive relationship between work motivation and perceived work performance and that work motivation has a moderate correlation with intention to quit. Thus, it can be reasoned that if there is a considerable relationship between job motivation and intention to quit, then, it ought to be possible that there may be a relationship between job motivation and resistance to change or support for change. It can be argued that employees who have a low level of job motivation 60 are less likely to provide support for organizational change than those who have a high level of job motivation. In the same vein, employees who have a high level of job motivation may be less likely to provide resistance to change than those who have a low level of job motivation. Why is this so? Drawing upon research on job motivation (e.g., Herzberg, 1968) suggesting that employees seek to align motivation levels with efforts and that alignment processes have certain implications for their decision, one may expect that the nature and magnitude of employees’ reactions to change may also be influenced by their job motivation. In addition, one may also reason that if past actions of organizations influence employees’ job motivation, then organizational change may also influence their job motivation. Suppose that organizational change has negative consequences for employees: for example, greater levels of workloads or lower compensation. In this situation, employees will be less motivated to work and, thus, more likely to provide resistance to change. On the other hand, suppose that organizational change has positive consequences for employees:

for example, higher compensation or better job conditions. Then, they will be greater motivated to work and, thus, more likely to render support for change. It must be noted that in this dissertation, it is assumed that any effect of consequences of organizational change on employees will be subsumed to a level of job motivation. In summary, it is hypothesized that:

H13a: Employees’ job motivation will be negatively related to their resistance to change. H13b: Employees’ job motivation will be positively related to their support for change.

60 For simplicity, I assume that the degree of job motivation of employees during organizational change represents the net degree of job motivation after any effect of organizational change that may have on him.

65

3.2.14. Perceived Employability As discussed earlier, a general attitude towards an idea, issue or object formed by an individual is more likely to promote a positive or negative inclination on the perceptual process and decision-making process. 61 If employees form a perception about their employability, which refers to how employable they are for another identical or better job in the market, perceived employability may play a role in determining a kind of reactions to change. In contrast to a relative large body literature in the area of human resource management dealing with a wide array of job characteristics such as job motivation (e.g., Herzberg, 1968), job security (e.g., Westman et al., 2001), or job satisfaction (e.g., Mobley et al., 1978; Porter and Steers, 1973), research on employability has been extremely scant. Indeed, I have found no theoretical or empirical work focusing on employability. The path I propose to follow in theorizing the effect of perceived employability on reactions to change is by using arguments of the concept of sense of control (e.g., Parks, 1989; Westman and Etzion, 1995; Westman et al., 2001). Sense of control is “a generalized belief on the part of the individual about the extent to which important outcomes are controllable” (Parks, 1989:21). Following from this, I propose that employees may consider “employability” as an important outcome and that they may have a certain degree of sense of control over their employability—that is to say, they may hold a certain belief about the extent to which acquiring new jobs are controllable. In this view, it can be reasoned that perceived employability, at its core, has at least one similar property as that of sense of control: that is to say, both constructs concern the extent to which important outcomes are perceived to be controllable. Thus, I maintain that one may extent the concept of sense of control to develop a better understanding of perceived employability. Empirical evidence has shown that sense of control exerts effects on several work- related outcomes of employees. For example, an empirical study by Westman and Etzion (1995) has found a negative relationship between sense of control and burnout of employees: that is, sense of control helps reduce degrees of burnout. In a more recent study, Westman et al. (2001) also found that employees’ burnout is associated with their feelings of job insecurity as well as self-control. Thus, if sense of control has the effects on one’s attitudinal and work outcomes, so has sense of perceived employability, and the prediction that perceived employability may influence employees’ reactions to change will seem plausible.

61 For a detailed discussion of the concept of attitudes, see Section 2.4 of this dissertation.

66

The nature and magnitude of perceived employability may arise from several sources such as self-confidence (e.g., self-efficacy for performing), career aspiration, and current characteristics of job market. Specifically, factors such as a high level of self-confidence, a booming job market, good economic conditions or a high degree of knowledge possessed and demanded by the market may promote a positive perception of one’s employability which can be defined as a perception of having no difficulty for getting an identical or better job if needed. On the other hand, factors such as a low level of self-confidence, a poor job market, poor economic conditions or a low degree of knowledge possessed and demanded by the market may promote a negative perception of one’s employability which can be defined as a perception of having some difficulty in getting an identical or better job than the current one if needed. The important for us to observe is that the perceived employability construct also has one property of the concept of self-confidence: that is, to say, for employees to form their perceived employability, they need to form a level of confidence that they will achieve a certain goal, of which is getting a new job. Thus, one may also extend the logic of the concept of self-confidence to explain potential effects of perceived employability for employees’ resistance to change and support for change. 62 Here, it is argued that employees’ perceived employability might affect their perceptual process and decision-making process when dealing with how to react to organizational change. Suppose that employees have already established a perception of their employability, it is possible that their perception may moderate a relationship between them and their organization. That is, perceived employability may determine the extent to which employees depend on the organization in terms of employment. Is it possible to expect that employees who feel capable of having a new job easily will feel less dependent on their current employment than those who feel less capable of getting a new job? Up to this point, a key assumption in this discussion is that the degree to which employees value the employment contract with their current employer can be explained by the degree of their employability in the job market they hold. One can also reason that employees’ perception of their dependency on the organization may be determined by perceived employability. Then, it is probable that employees who have a low level of perceived employability may be motivated to behave in ways to remain with their organization than those who have a high level of perceived employability. One key question remains: how does perceived employability shape or affect employees’ reactions to change? It remains unclear whether a positive perception of

62 For a detailed discussion of self-confidence, see Section 3.2.15 of this dissertation.

67

employability will have a positive or negative relationship with resistance to change and/or support for change. Similarly, it remains unclear whether a negative perception of employability will have a positive or negative relationship with resistance to change and/or support for change. To illustrate this point, let us consider, for example, a simple case where employees are afraid of losing their job, and they do not expect to get a new job easily. One possible scenario is that they will strongly resist the change because keeping their current job is very important. On the other hand, it is also possible that they will accept the change because they are afraid that their resistance to change may cause their job. Similar to the above case, let us consider whether employees will support or resist organizational change when they have a high degree of perceived employability. Certainly we cannot answer this question outright because this perception seems to mediate relationships between other variables and reactions to change rather than to have a direct effect on reactions to change. In addition, because employees who have a high degree of perceived employability may not be dependent on their current job, they may have more options regarding reactions to change, and feel relatively free to make a reaction. 63 The important point for us to observe is that both high and low levels of perceived employability can lead to different reactions to change. However, because there is no prior study examining the effect of perceived employability on reactions to change, I shall first focus on examining direct relationships between them. Thus, I propose the following hypotheses:

H14a: Employees’ perceived employability will be positively related to their resistance to change, H14b: Employees’ perceived employability will be negatively related to their support for change.

63 Note that I assume that an employee with a high degree of perceived employability is not afraid of choosing a decision that he feels is right (for him). In contrast, an employee with a low degree of perceived employability tends to be afraid of choosing a reaction that may eventually cause his job even if that choice may be his preferred reaction.

68

3.2.15. Self-Confidence for Career-Relevant Learning Individuals have different beliefs or perceptions about the factors responsible for what happens to them. Individuals with an internal locus of control consider what happens to them as determined by factors under their control, whereas those with an external locus of control consider what happens to them as determined by factors outside their control (Rotter, 1966; Elangovan and Lin Xie, 1999). A question is then whether it is possible that one’s self-confidence exerts an influence on his or her decision? Specifically, does employees’ self-confidence for career-relevant learning and competent development exert effects on their reactions to change? In exploring a direct effect of self-confidence for career-relevant learning and competent development (or in short ‘self confidence for learning and development’) on employees’ reactions to change, this dissertation focuses on resistance to change and support for employee of employees as (dependent) variables of interest. There has been a growing awareness in the organizational psychology literature (e.g., Bandura, 1977) that self-efficacy is a key determinant of individuals’ intention and choice to pursue an activity. Self-efficacy has been defined as the belief that one possesses the ability to complete a certain task (Foley, Kidder and Powell, 2002). Moreover, Bandura (1997) suggested three levels of self-efficacy: (1) task specific self-efficacy; (2) domain self- efficacy; and (3) general self-efficacy. Consistent with the definition of self-efficacy, self- confidence for career-relevant learning and competence development can be categorized at the task specific self-efficacy, which is self-confidence for dealing with performance of a specific task, and at the domain self-efficacy level, which is self-confidence for dealing with performance within a certain domain of tasks. In addition, competence development should encompass learning of new skills or new level of existing skills. Consequently, self- confidence for learning and development, at its core, refers to one’s confidence in learning new things and developing new skills. It has been argued that self-efficacy for development and learning and self-efficacy for performance should be distinguished (Maurer, 2001). According to Maurer (2001), self- efficacy for development and learning refers to one’s confidence in developing skills and learning new things, whereas self-efficacy for performance refers to one’s confidence in performing a task that one already possesses the skills required to perform that task. It is useful to note that research has suggested that a decline in older employees’ self-confidence in their ability to learn and develop may contribute to the age effect on participation in learning and development (see e.g., Fossum, Arvey, Paradise and Robbins, 1986). Another question is whether it is plausible to expect that employees who have a high level of self-confidence for learning and development tend to feel more comfortable with

69

organizational change than those who have a low level of self-confidence for learning and development. One may speculate that employees who have a high level of self-confidence for learning and development will feel more comfortable than those who have a low level of self-confidence for learning and development. One plausible reason to support this proposition is that the degree to which a knowledge, skill or competence is obtainable may be influenced by self-confidence for learning and development. Another possible reason is that a work performance may be influenced by a result of learning new knowledge and skills; therefore, employees who have a low level of self-confidence for learning and development may be afraid of having a poorer work performance because they fail to learn new knowledge or skills. Likewise, those who have a high level of self-confidence for learning and development may consider learning new skills as an interesting challenge sufficient so as to enjoy doing it, whereas those who have a low level of self-confidence for learning and development may have a negative view on learning new skills. Last but not least, it is also possible that employees who have a high level of self-confidence for learning and development may consider learning new skills (as part of organizational change) as an opportunity to boost their career prospect rather than as a threat to their work performance or career advancement, whereas those who have a low level of self-confidence for learning and development may have another idea. In sum, this dissertation examines whether there is a negative relationship between self-confidence for learning and development and resistance to change and whether there is a positive relationship between self-confidence for learning and development and support for change. To test these hypothesized relationships, the following hypotheses are proposed:

H15a: Employees’

self-confidence

for

career-relevant

learning

and

competence

development will be negatively related to their resistance to change.

 

H15b: Employees’ self-confidence

for

career-relevant

learning

and

competence

development will be positively related to their support for change.

70

3.2.16. Affective Commitment Despite the general agreement on the effect of affective commitment—that is to say, the relationship between affective commitment and various work-related variables such as willingness to turnover (Parker et al., 2003) or trust in management (Kim and Mauborgne, 1993; Pearce, 1993) have been found, there seems to be no prior empirical studies examining effects of affective commitment on reactions to change, especially resistance to change and support for change. Following from this, one might expect that employees who are committed to their organization are not only less likely to behave in ways that resist organizational change but also more likely to behave in ways that support organizational change because they want their organization to be successful. Therefore, this dissertation examines whether there is a relationship between affective commitment and resistance to change and support for change. A review of the literature dealing with organizational commitment has revealed several terms such as affective commitment (Porter, Steers, Mowday and Boulian, 1974; Allen and Meyer, 1990), normative commitment (Weiner, 1982), organizational commitment (Mowday, Steers and Porter, 1979), and affective organizational commitment (Mowday, Porter and Steers, 1982). Affective commitment refers to emotional attachment to, identification with, and involvement in organization (Meyer and Allen, 1997), whereas organizational commitment refers to identification with organizational goals, willingness to exercise effort on behalf of the organization, and interest in remaining with the organization (Mowday et al., 1979). There is one issue here to which I would like to draw attention. Although these terms seem to have been used interchangeably across much of research, there is still some disagreement concerning the definitions of organizational commitment (Iverson and Roy, 1994). The point of departure amongst researchers over the definitions of organizational commitment centers on an issue about whether commitment is attitudinal or behavioral in nature. However, this is not of a great concern to this dissertation because one can take either perspective or both of them for the purpose of studying the effect of employees’ commitment to their organization on their decisions. In spite of their differences, however, I tend to follow the “attitudinal commitment” perspective because it can be argued that humans tend to follow what their heart tells to do. 64 While researchers seem to disagree on the definitions, they have provided empirical evidence supporting the notion that commitment, regardless of whether it is attitudinal (or

64 Note that to be more specific and correct on this point, one has to assume that actions that the mind or heart tells one to do are actionable. For example, if someone intends to quit his job for whatever reason, he might not actually do so when he does not think he can get another comparable job in a reasonable time period.

71

affective) commitment or behavioral commitment, has the effects on employees’ decisions (e.g., Kim and Mauborgne, 1993; Parker et al., 2003). Additionally, research on organizational commitment has pointed out that a degree of the organization’s commitment to its employees is one of predictors of organizational commitment. This argument is supported by an argument that employees’ perception of their organization’s actions will induce the reciprocity in their attitudes or behaviors (Shore and Tetrick, 1991). Moreover, empirical evidence has also suggested that perceived organizational support is significantly correlated with organizational commitment (e.g., Shore and Tetrick, 1991), and organizational commitment is significantly related to trust in management (e.g., Kim and Mauborgne, 1993; Pearce, 1993; Gopinath and Becker, 2000; Whitener, 2001). Based on past theoretical underpinning and empirical evidence, linking employees’ affective commitment to their reactions to change (e.g., resistance to change and/or support for change) is supported by at least two reasons. A first reason is that based on the norm of reciprocity (Gouldner, 1960), employees who have a low level of affective commitment may hold a perception that they are under no obligation to return certain benefits to their organization because a low level of affective commitment indirectly implies a low level of perceived organizational support (Shore and Tetrick, 1991). On the other hand, employees who have a high level of affective commitment may hold a perception that they have an obligation to return benefits to their organization. Following from this, organizational change that shifts levels of affective commitment tends to have altered perceptions of an obligation to the organization. Employees who have a perception that they are under an

obligation to return a favor to their organization are more willing to repay their organization

in one or more ways than those who have a perception that they are under no obligation to

their organization. On the other hand, employees who have a perception that they are under no obligation to their organization are unlikely to refrain themselves from resisting the change, if it deems appropriate. It can be reasoned that if employees were to consider

support for change as a means to repaying his organization, those who have a perception that they are under an obligation to repay their organization will provide higher levels of support for change. A second reason to support the potential relationship between affective commitment and resistance to change and support for change is that if levels of affective commitment play a role in determining levels of work performance (Parker et al., 2003), employees who have a high level of affective commitment tend to perform their work better than those who have a low level of affective commitment. If employees have a low level of affective commitment during organizational change, then we might expect that they will tend to have

a lower level of work performance. Intuitively, changes in work performance may be

72

influenced by resistance to change. Similarly, if employees have a high level of affective work performance, then we might expect that they will tend to have a higher level of work performance. Changes in work performance may be influenced by support for change. This position is then supported by an empirical study by Parker and colleagues (Parker et al., 2003), which has found that effects of work attitudes (job involvement and commitment) on work performance are partially mediated by motivation. If support for change is positively correlated with work effort devoted to an organization, and if resistance to change is negatively correlated with work efforts, then affective commitment differences among employees will result in a variation in the nature and the degree of reactions to change. Specifically, it is expected that there is a positive relationship between affective commitment and support for change, and there is a negative relationship between affective commitment and resistance to change. Thus, the following hypotheses are formulated:

H16a: Employees’ affective commitment will be negatively related to their resistance to change. H16b: Employees’ affective commitment will be positively related to their support for change.

3.2.17. Trust in Management Parallel to the logic of H16a and H16b, this dissertation predicts that there is a relationship between trust in management and reactions to change. An important question is concerned with whether employees will make a different decision between when having a high degree of trust in management and when having a low degree of trust in management. Past research has recognized the significance of trust in management or leadership (e.g., Argyris, 1962; McGregor, 1967). Not only has the concept of trust in management lead to several leadership theories—for example, trust is a key element of leader-member exchange theory (Schriesheim, Castro and Cogliser, 1999) and employees’ trust in management as an important factor for leader effectiveness (Bass, 1990)—but it has been emphasized in other concepts such as psychological contracts, organizational relationships, or change management. It has been suggested that trust in management would inevitably become one of key concepts in organizational theory (Kramer, 1999). Trust, which is a social construct, has been defined as a “psychological state comprising the intention to accept vulnerability based upon positive expectations of the

73

intention or behavior of another (Rousseau, Sitkin, Burt and Camerer, 1998: 395). 65 Trust in management refers to employees’ attitude towards top management that indicates their willingness to be vulnerable to top management (Whitener, Brodt, Korsgaad and Werner, 1998). Similarly, the trustworthiness has been defined as “an individual belief that top management can be trusted” (Spretizer and Mishra, 2002: 710). Prior research on organizational change management has noted that when top managers attempt to undertake changes within their organization, they should build trust among their employees in order to facilitate and sustain effective change (Webb, 1996). Indeed, top managers should build trust among employees so as to increase organizational effectiveness (Argyris, 1962). In addition, perceptions of legitimacy of organizational change can be enhanced by trust in management (Rousseau and Tijoriwala, 1999). Several researchers (e.g., Bridges, 1980; O’Toole, 1995) have noted that distrust towards those leading changes is one of factors that lead to employees’ resistance to change. Empirical research examining the role of trust in management has shown that there are certain relationships between employees’ trust in management and their attitudinal outcomes (Dirks and Ferrin, 2001). For example, Kim and Mauborgne (1993) and Pearce (1993) found a relationship between trust in management and affective commitment. More recently, Spreitzer and Mishra (2002) also found a positive correlation between the trustworthiness of management and affective commitment. In addition, empirical evidence has suggested that trust in management has a positive effect on group performance (Klimoski and Karol, 1976) and business unit performance (Davis, Shoorman, Mayer and Tan, 2000). Although prior research on trust has suggested that employees’ trust in management develops through a social exchange process over time; that is, a process of social exchange generates trust (Blau, 1965), there are different explanations about the process through which trust forms as well as the process through which trust affects work- related outcomes (Dirks and Ferrin, 2002). 66 There is another issue here deserves attention. The issue is that there is a great diversity in constructs of trust across much of past studies. Not only have researchers provided many definitions of trust in management but also they have specified the construct with different leadership referents (Dirks and Ferrin, 2002). For example, some researchers have given attention to trust in a direct leader, whereas others have given attention to trust in top management. Clarification on this issue is very important not only for theoretical

65 For Kim and Mauborgne (1993), this definition suggests that employees’ trust in management reflects two important aspects: (1) an employee’s faith in organizational goal attainment and an organizational leader; and (2) an employee’s belief that he or she will receive benefits from an organizational action.

66 For a more detailed discussion of this issue, see Dirks and Ferrin (2002).

74

reasons, but also for practical reasons, as it may provide guidance on whether this dissertation should focus on trust in a direct leader or trust in leadership. Considering that the forms of organizational change studied in this dissertation are a downsizing and a privatization that have been originated by top managers, 67 I therefore propose that the theoretical construct of trust in management in this dissertation rests on trust in top management rather than on a direct leader. In light of this discussion, it can be reasoned that employees who have high levels of trust in management are likely to provide higher levels of support for change and lower levels of resistance to change. Accordingly, the following hypotheses are proposed:

H17a: Employees’ trust in management will be negatively related to their resistance to change. H17b: Employees’ trust in management will be positively related to their support for change.

3.2.18. Colleagues’ Reactions to Change Research on the effect of group and organizational context on attitudes and behaviors of employees has attracted growing attention in the field of management (Kidwell, Mossholder and Bennett, 1997). The social context has been described by relational phenomena that cannot be understood in terms of individual separately but rather in terms of characteristics of the environment, organization, and work group together, of which have an impact on attitudes or behaviors of individual members of the group (Kidwell et al., 1997). Both the social exchange theory (Blau, 1964) and the norm of reciprocity (Gouldner, 1960) can explain the interactions and interpersonal relationships between members of the group and the influence of the group’s norms on individual members of the group. In addition, social network theory (Alderfer, 1986), which suggests that social relationships are embedded within a larger structural context that encourages or precludes various kinds of social contact, can also be used to explain the effect of the social group’s actions on members of the group. In the organizational behavior literature, the influence of social factors on individuals’ behavior has been demonstrated in several studies. Social identity theory and social comparison research have demonstrated how other people exert an influence on individuals’

67 Note that middle-level managers were not the one who initiated the change (both for a case of downsizing at the private school and for a case of privatization at EGAT); therefore, trust in middle-level manager (or in a direct leader) is of less relevance to employees.

75

behavior (Festinger, 1954; Terry and Hogg, 2000). According to a social identity theory (Capazza and Brown, 2000; Tajfel and Turner, 1986) people tend to classify others and themselves into social categories/groups and identify more with members of their own categories (in-group) than with members of other categories (out-group). People may have multiple social identities along several dimensions: for example, gender, ethnicity, and sexual orientation (Frabble, 1997). In addition, the social learning theory (Bandura, 1977) postulates that role models for behavior have the effects on individuals’ behavior; that is, individuals obtain a collection of certain behaviors by observing others’ behaviors and the consequences in their social environment. According to the social information processing theory (Salancik and Pfeffer, 1978), individuals use information about norms, values, expectations, and behavior outcome contingencies derived from others in their social environment to steer behavior. Moreover, the attraction-selection-attrition perspective (Schneider, 1975) suggests that a group attracts, selects, and retains individuals with characteristics or traits similar to those of other members of the group. For example, individuals with easy-going tendencies would be more likely to be attracted to, selected by, and stay with groups with similarly easy-going members, resulting in the establishment of a group with relatively easy-going characteristics. According to Hackman (1992), characteristics of a work environment may be conceived of as either (1) discretionary stimuli that are transferred to individual members of the group differently or (2) ambient stimuli that are diffused throughout the group setting and are possibly available to all members of the group. In a conflicting social identities situation, people tend to identify more with those who are similar along the dimension of social identity that is most salient to them. Brewer (1991) has also noted that the value of group identification comes from distinctiveness and shared identity. In light of the above discussion, the social group’s decision is expected to exert an effect on individual members’ decision. That is, colleagues’ resistance to change may promote employees’ resistance to change, and weaken employees’ support for change because employees, as a member of the group, tend to feel obliged to follow their group’s decision so that they are not excluded from the group. On the other hand, colleagues’ support for change may weaken employees’ resistance to change, and promote employees’ support for change. In line with this reasoning, this dissertation hypothesizes that:

H18a: Employees’ perceptions of colleagues’ resistance to change will be positively related to their resistance to change, and be negatively related to their support for change H18b: Employees’ perceptions of colleagues’ support for change will be negatively related to their resistance to change, and be positively related to their support for change.

76

4.

Research Methodology

The overall objective of this dissertation is to examine if perceptions and/or attitudes effect employees’ reactions to change. To achieve this objective, this dissertation seeks to test the hypothesized relationships between perceptions and/or attitudes on the one hand and employees’ resistance to change and support for change on the other hand. If one wishes to understand the implications of employees’ perceptions and attitudes for their reactions to change, and if one postulates a direct relationship between perceptions and reactions to change, then two important consequences follow. First, the understanding of different research methodologies that have been used in social science in general and management science in particular was required. Second, a research strategy that was capable of providing answers to the research questions in this dissertation had to be determined. It is understandable that empirical research has many variants. Three of which are the most prominent: (1) empirical experimental research; (2) empirical statistical research; and (3) empirical case study (Wacker, 1998). According to Wacker (1998), each type of research strategies seeks to address different research objectives: for the empirical experimental research, the focus is on examining the relationships between variables by manipulating controlled treatments to determine the exact effect on specific dependent variables; the purpose of the empirical statistic research is to empirically validate the assumed relationships between variables in a large sample from actual environments; and the purpose of the empirical case study is to develop insightful relationships within smaller or limited samples. For the moment, one can consider this dissertation to be the empirical statistic research as it was to empirically test the hypothesized relationships between predictors and outcomes. Reflect on my research questions: what perceptions and/or attitudes influence employee resistance to change? And what perceptions and/or attitudes influence employees’ support for change? Quantitative research seems to better address these research questions than qualitative research. One plausible reason to lend strong support for the use of quantitative research in this dissertation is that quantitative research is better suited than qualitative research where the purpose of the study is to investigate relationships as pairs of variables. In particular, quantitative research provides an opportunity to generalize the results statistically to the population. To address the research questions, this dissertation follows a positivistic tradition that, as Schwaninger (2004) has described, focuses on facts, adopts an objectivist worldview, and relies on quantification. Another issue to be considered is concerned with research design

77

choices which imply trade-offs between three dimensions: (1) generalization to the population that supports the external validity; (2) precision in measurement and control of the behavioral variables that affect internal and external validity; and (3) realism of context. It is also useful to note that this dissertation was primarily concerned with statistical generalization and precision in measurement. The following sections address sampling criteria with the emphasis on the method used for selecting the original population for studies 1 and 2, for selecting a sample size, and for collecting data. Different methods of data analysis and model specification for the main research model will be discussed. This dissertation proceeds by discussing measures of study variables: dependent variables, independent variables and control variables. Finally, the overall sequence of data analysis will be discussed in detail.

4.1. Context, Sample and Procedure

4.1.1. Study 1 – Context, Sample and Procedure

The setting for Study 1 was a private school in Thailand employing 120 employees at the time of the survey. Declining numbers of enrolled (both new and retained) students over past few years (e.g., from approximately 200 students per class in 1990 to 100 students per class in 2004) had caused the management team to make multiple efforts to improve the school’s efficiency and profitability. However, the numbers of enrolled students still continue to decline every year, pressurizing the management team to engage in workforce reductions. The downsizing program was initially aimed to reduce approximately 10 teachers by the start of the next academic year (i.e., 2005-2006) so as to improve the student/teacher ratio and to remain profitable. Teachers were informed about the downsizing decision in August 2004. It is useful to note that at the time of the survey (during the first two weeks of September 2004), teachers did not have the full details of the downsizing program (e.g., the details of the involuntary workforce reductions program). A multiple-item survey in Thai was administered during working hours to a random sample of 100 teachers at the school. In order to determine the clarity and the readability of the original questionnaire written in the English language, the referee and the co-referee conducted a review of the questionnaire items. Then, I whose first language is Thai translated the questionnaire into the Thai language. Then, a professional Thai-English translator back-translated the questionnaire into the English language and I examined each item for translation error. The inspection did not find any instances where an item’s meaning had significantly changed due to the translation. Survey instructions stressed that participation in this survey was voluntary and confidential and that this survey was undertaken for the purpose of writing a doctoral dissertation. Ninety-one surveys were

78

returned, presenting a response rate of 91%. Of these, 3 surveys were removed from the analysis due to missing values in the main part of the returned questionnaire. Thus, the sample comprised 88 cases. Respondents were representative of the larger population (25% male, 80 % of respondents had a bachelor degree or equivalent, mean age = 44.1 years, mean position tenure = 14.8 years, mean organizational tenure = 17.8 years). Listwise deletion of missing values of the remaining sample reduced the sample to 86 cases (for further data analysis with three control variables: age, education, and gender).

4.1.2. Study 2 – Context, Sample and Procedure

The setting for study 2 was a stated-own company in Thailand employing around 63,700 employees. Since 1992 this organization has started to form subsidiaries with the government’s privatization policy to increase private sector participation in this industry and reduce investment burden of both the organization and the government. Since 2003 the organization has been in the final phrase of the privatization process; that is, the management team has planned to list the company on the Stock Exchange of Thailand. Although the initial public offering (IPO) was scheduled in mid-2004, this has yet been materialized due to the fact that a final decision on the IPO has yet been made as a result of disagreement between four main stakeholders (i.e., employees, top management, the government and the public) on the details of the IPO. It should be noted that at the time of the survey (during the first two weeks of September 2004), the final decision on the IPO was still pending. A multiple-item survey in Thai was administered during working hours to a random sample of 500 employees at four offices located in Bangkok. It is useful to note that the four offices had about 900 employees in total. For the purpose of comparability, the questionnaire items in Study 2 were similar to those of in Study 1. As with Study 1, survey instructions stressed that participation in this survey was voluntary and confidential and that this survey was undertaken for the purpose of writing a doctoral dissertation. Two hundred and twenty-four questionnaires were returned, presenting a response rate of 44.8%. Of these, 17 surveys were removed from the analysis due to those respondents not completing the main part of the questionnaire survey. The final sample comprised 207 cases. Respondents were representative of the larger population (64.7% male, 40.5 % of respondents had a bachelor degree or equivalent, mean age = 39.1 years, mean position tenure = 9.0 years, mean organizational tenure = 14.9 years). Listwise deletion of missing values of the remaining sample reduced the sample size to 197 respondents (for further data analysis with three control variables: age, education, and gender).

79

4.2.

Alternative Methods of Data Analysis

In this section, several methods of data analysis for the research model are discussed in great details. Let us first consider an ordinary regression analysis since it is often used in many, if not most, studies in management science. Given that dependent variables were measured with six items on a five-point scale, which indicate the ordinal nature of the dependent variable, using the ordinary regression analysis (ordinary least square or OLS) as an analytical tool for estimation may be inappropriate. In the tests of employees’ resistance to change (which was divided into two subsets:

active and passive resistance to change) and support for change (which was divided into two subsets: active and passive support for change) as dependent variables, an ordinary regression analysis (OLS) would be inadequate because it would fail to account for the ordinal nature of the dependent variable. The ordinary regression analysis would treat the difference between a level of scales (e.g., resistance to change) 4 and 5 as the same as the difference between a level of scales (e.g., resistance to change) 1 and 2 (McKelvey and Zavoina, 1975; Goodstein, 1994). Given a view that the differences between the levels of resistance to change need not be similar, the parallel regression assumption may not be true, implying that with the ordinary regression model, the estimates of the effects of independent variables would be wrong, i.e., biased and inconsistent. Let us then consider a multinomial logit or probit model. From the outset, both logit and probit models seem to be more appropriate than the ordinary regression model. However, the examination of these models suggests a key issue concerning the use of these models in this dissertation. For both multinomial logit and probit regression models, there is no order in the different categories that dependent variable can take; these models would not take the existence of a ranking into account. Therefore, both models are also inefficient in this dissertation. It is also important to note that the multinomial logit regression model makes an implicit assumption, which is known as the independence of irrelevant alternatives (IIA) property. The independence of irrelevant alternative property implies that the ratio of the probability of choosing one choice to the probability of choosing a second choice unchanged for individual decision-makers if a third choice enters the race (for review, see Alvarez and Nagler, 1998). On the contrary, as the multinomial probit regression model allows for correlations between the disturbances for difference choices, it does not assume the independence of irrelevant alternative property (see Alvarez and Nagler, 1998). Based on the limitations to several regression models discussed here, the multinomial ordered probit regression model seems to be the most appropriate tool for estimation in this

80

context because (1) it takes the existence of a ranking into account, (2) it assumes that the difference between any two adjacent levels is not known, and (3) it does not assume the independence of irrelevant alternative property. To test hypotheses, I therefore conducted the multinomial ordered probit regression model, which does not violate the independence of irrelevant alternative property. But it must be noted that due to a computation difficulty, the estimation of large choice multinomial probit models is almost impossible using the numerical integration of bivariate normal distributions (Alvarez and Nagler, 1998). A comparison of the key properties of several regression models, which have been discussed here, is illustrated in Table 1.

Table 1:

Characteristics of Alternative Regression Models

 

Ordinary

Multinomial

Conditional

Multinomial

Multinomial

Regression

Logit

Logit

Probit

Ordered

(OLS)

Probit

Dependent Variable

Interval or

Ordinal or

Ordinal or

Ordinal or

Ordinal or

Ratio

Nominal

Nominal

Nominal

Nominal

Independent Variable

A linear function of interval/ratio of binary variables

 

Method

e.g.,

e.g., Maximum Likelihood Method

 

Moment-

 

based

Method

Type of Distribution

Normal

Binomial

Binomial

Cumulative

Cumulative

Distribution

Distribution

Distribution

Normal

Normal

Distribution

Distribution

Correlated Disturbances

No

No

No

Yes

Yes

Includes Position of Choice

No

No

Yes

Yes

Yes

Assumes the IIA Property

No

Yes

Yes

No

No

Can Correctly Measure Omission of a Choice

No

No

No

Yes

Yes

Source: Adopted from Alvarez and Nagle (1998: 59).

4.3. The Multinomial Ordered Probit Model

Let us define two categorical variables y 1 ∈ {1…m} and y 2 ∈ {1…m} indicating the observed degree of resistance to change and support for change, respectively. As discussed earlier, the explanatory variables include (1) five that are related to change process; (2) eight that are related to actual and expected consequences of change; (3) two that are related to employees’ ability; and (4) three that are related to the relationships between employees and their organization and colleagues. The discrete probability function of y conditioned on all

81

explanatory variables is commonly known as an ordered probit model. It is based on a latent variable intended to present the employees’ reactions to change. Let us assume that the unobserved continuous measures, resistance to change (Y 1 *), is a linear function of a set of explanatory variable X, with the corresponding parameter vector β, and a normally distributed error term ε:

Y 1 *

=

β’X + ε,

Although Y 1 * is unobserved, it is related to an observed discrete variable y 1 , whose value is dependent upon the value that Y 1 * takes. It is assumed that employees’ resistance to change is measured in one of three alternatives (y 1 = 1, 2 or 3, where 1 = low, 2 = medium, and 3 = high) corresponding to the range of observed ratings assigned by employees.

y 1

=

= 2

=

3

1

if 0 < Y 1 * ≤ µ 11 ,

if µ 11 < Y 1 * ≤ µ 12 ,

if µ 12 ≤ Y 1 *,

Where µ 11 and µ 12 (0 < µ 11 < µ 12 ) are unknown threshold parameters of Y 1 * that define the range of resistance to resistance; these parameters will be estimated in conjunction with the β vector by maximizing the joint probability or likelihood function. Let us assume that ε is normally distributed across observations, and the mean and variance of ε are normalized to zero and one. With the normal distribution, the multinomial ordered probit model suggests the following probabilities, where Ф denotes the cumulative normal distribution function.

Pr (Y 1 = 1|X 1 , X 2 , …, X k )

=

Ф(µ 11 - β’X) ,

Pr (Y 1 = 2|X 1 , X 2 , …, X k )

=

Ф(µ 12 - β’X) - Ф(µ 11 - β’X) ,

Pr (Y 1 = 3|X 1 , X 2 , …, X k )

=

1 - Ф(µ 12 - β’X), and

Pr (Y 1 = 1) + Pr (Y 1 = 2) + Pr (Y 1 = 3) = 1

Similarly, let us assume that the unobserved continuous measures, support for change (Y 2 *), is a linear function of a set of explanatory variable X, with the corresponding parameter vector β, and a normally distributed error term ε:

Y 2 *

=

β’X + ε,

82

Although Y 2 * is unobserved, it is related to an observed discrete variable y 2 , whose value is dependent upon the value that Y 2 * takes. It is assumed that employees’ support for change is measured in one of three alternatives (y 2 = 1, 2 or 3, where 1 = low, 2 = medium, and 3 = high) corresponding to the range of observed ratings assigned by employees.

y 2

=

= 2

3

=

1

if 0 < Y 2 * ≤ µ 21 ,

if µ 21 < Y 2 * ≤ µ 22 ,

if µ 22 ≤ Y 2 *,

Where µ 21 and µ 22 (0 < µ 21 < µ 22 ) are unknown threshold parameters of Y 2 * that define the range of support for change; these parameters will be estimated in conjunction with the β vector by maximizing the joint probability or likelihood function. Let us assume that ε is normally distributed across observations, and the mean and variance of ε are normalized to zero and one. With the normal distribution, the multinomial ordered probit model suggests the following probabilities, where Ф denotes the cumulative normal distribution function.

Pr (Y 2 = 1|X 1 , X 2 , …, X k )

=

Ф(µ 11 - β’X) ,

Pr (Y 2 = 2|X 1 , X 2 , …, X k )

=

Ф(µ 12 - β’X) - Ф(µ 11 - β’X) ,

Pr (Y 2 = 3|X 1 , X 2 , …, X k )

=

1 - Ф(µ 12 - β’X),

Pr (Y 2 = 1) + Pr (Y 2 = 2) + Pr (Y 2 = 3) = 1

Thus, we can calculate the probability of levels of both resistance to change and support for change conditioned on all explanatory variables. For parameter estimation, I used the SPSS package version 12. Specifically, parameters were estimated by maximum likelihood function using the command PLUM (PoLytomous Universal Model procedure) with a probit link function in the SPSS package (see e.g., Borooah, 2002). The size of sample and number of parameters being estimated can impose a potential estimation problem for both models: that is, the greater the number of parameters, the larger is the sample size needed.

4.4. Measures of Theoretical Constructs

4.4.1. Dependent Variables

Resistance to change was measured using a newly developed six-item scale. Three newly developed items were used to measure degrees of behaviors that were believed to represent employees’ active resistance to change. Specifically, three types of behaviors, i.e., opposing, arguing, and objecting, were measured. In the same vein, three newly developed items were used to measure degrees of behaviors that were believed to represent employees’ passive

83

resistance to change. Specifically, two types of reaction, i.e., withdrawing and ignoring, were measured. Respondents were asked to report the degree to which they agree with each of items using a five-point scale ranging from 1 (strongly disagree) to 5 (strongly agree). Support for change was measured using a six-item scale. Three newly developed items were used to measure degrees of behaviors that were believed to represent employees’ active support for change. Specifically, three types of reactions, i.e., embracing, cooperating, and supporting, were measured. Similarly, three newly developed items were used to measure degrees of behaviors that were believed to represent employees’ passive support for change. That is, three types of reactions, i.e., agreeing, accepting, and complying, were measured. Respondents were asked to report the degree to which they agree with each of three items using a five-point scales ranging from 1 (strongly disagree) to 5 (strongly agree). Figure 6 presents an overview of measures of behaviors classified into four categories of reactions to change in this dissertation.

Figure 6: Summary of Measures of Reactions to Change

Discontent

Discontent

Active

Active

Active Resistance to Change

Active Resistance to Change

Active Support for Change

Active Support for Change

Opposing a change

Opposing a change

Arguing against a change

Arguing against a change

Objecting a change

Objecting a change

Embracing a change

Embracing a change

Cooperating with the firm

Cooperating with the firm

Giving support for a change

Giving support for a change

Passive Resistance to Change

Passive Resistance to Change

Passive Support for Change

Passive Support for Change

Withdrawing support

Withdrawing support

Ignoring a change

Ignoring a change

Agreeing to a change

Agreeing to a change

Accepting a change

Accepting a change

Complying with a change

Complying with a change

Passive

Passive

Contentment

Contentment

4.4.2. Independent Variables

Perceived organizational support was measured using a three-item measure reflecting perceived organizational support. These items were adapted from Eisenberger et al.’s (1986,

84

1990). Respondents were asked to indicate the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived procedural justice was measured using a three-item perceived procedural justice scale adapted from the 31-item measuring perceptions of the procedure justice of the sale and layoffs developed by Gopinath and Becker (2000). Respondents indicated the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived participation in decision-making process was measured using a three-item measure reflecting perceived participation in decision-making. Respondents are asked to indicate the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived need for change was measured using a newly developed three-item measure reflecting employee’s perception of need for a change in the organization. Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Attitude towards organizational change was measured using a newly developed three- item scale reflecting employees’ attitude towards organizational change. Respondents indicated the degree to which they agreed with the items using a five-point scales ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Fear of known consequence of change was measured using a newly developed three- item scale. Respondents indicated the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Fear of unknown consequence of change was measured using three newly developed items. Respondents were asked to report the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived change in power was measured using a newly developed three-item measure reflecting perceived change in power. Respondents indicated the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived change in status was measured using a newly developed three-item scale reflecting perceived change in status. Respondents were asked to indicate the degree to which they agreed with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived change in pride was measuring using a newly developed three-item scale reflecting perceived change in pride. Respondents indicated the degree to which they agreed

85

with these items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Job satisfaction was measured using a three-item scale adopted from those used by Weiss et al. (1967) and Davy et al. (1997). As job satisfaction items were found to indicate two sub-dimensions: intrinsic and extrinsic satisfaction (Weiss et al, 1967), I used three items to measure the indicators of extrinsic job satisfaction. These items include: (1) pay and amount of work; (2) chances for advancement; and (3) working conditions. Respondents indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Job security was measured using a three-item scale adopted from those used by Caplan et al.’s (1975) and by Hui and Lee (2000). These items consisted of: (1) ‘I am certain about what my future career picture looks like in this company’, (2) ‘I am certain about what my responsibilities will be six months from now’, and (3) ‘I am certain about my job security in this company’. Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Job motivation was measured using a three-item measure adopted from those used by Hui and Lee (2000). These items measured extrinsic job motivation. Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Perceived employability was measured using a newly developed three-item measure reflecting employees’ perceptions their ability to find a new job. Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Self-confidence for learning and development was measured using two newly developed items and one item adopted from a study by Maurer et al (2003) reflecting employees’ perceptions of their ability to learn new knowledge and develop new skills. Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Affective commitment, a psychological attachment to the organizational, was measured using a three-item sale adopted from a 10-item scale used by Gopinath and Becker (2000). Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”. Trust in management was measured with a three-item scale adopted from those used by Gapinath and Becker (2000) and Robinson and Rousseau (1994). Respondents indicated the degree to which they agreed with the items using a five-point scale ranging from “1 = strongly disagree” to “5 = strong agree”.

86

Perceptions of colleagues’ reactions to change were measured using a newly developed three-item scale reflecting employees’ perceptions about colleagues’ resistance to change and/or support for change. Respondents were asked to indicate the degree to which they agreed with the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)”.

4.4.3. Control Variables

In addition to the endogenous and predetermined variables discussed above, this dissertation aimed to control for an influence of age (in years), gender (male = 0, female = 1), education (below a bachelor level = 0, a bachelor level = 1, a master level = 2, higher than a master level = 3), position tenure (in years), organizational tenure (in years), family status (single = 0, married/co-habiting = 1, divorced = 2), a number of children when conducting analyses to separate individual differences in age, gender, position tenure, organization tenure, family status and a number of children. Data for all control variables were obtained from the questionnaires completed by the respondents.

4.5. Data Analysis Procedures

In this section, the sequence of data analysis conducted in both Studies 1 and 2 that were designed to test all hypotheses presented in this dissertation is presented. The overview of the sequence of data analysis is illustrated in Figure 7. First, descriptive statistics (e.g., value range, minimum and maximum values) for all variable indicators were examined. Second, histograms illustrating distributions of all variable indicators were examined. In a third step, for the purpose of reducing the number of indicators, the patterns of directions of indicators for each variable were examined. To see this, I first averaged all indicators for each variable, plotted this aggregate indicator along with original indicators in a graph, and examined the pattern of directions. If the results did not reveal a satisfactory level of internal consistency, I removed an indicator that appeared to have a different pattern of directions from the model and conducted the same sequence of analyses until the aggregate indicator that seemed to provide the highest level of internal consistency was found. As the Spearman correlations also indicate the pattern of directions between indicators, I also examined the correlation coefficients among indicators for each variable for the purpose of reducing the number of indicators. Upon the completion of steps 3 and 4, I reduced the number of indicators for both dependent and independent variables.

87

Figure 7: Summary of the Sequence of Data Analysis

Examine descriptive statistics for all indicators

Examine descriptive statistics for all indicators

Examine descriptive statistics for all indicators Examine histograms for all indicators Examine histograms

Examine histograms for all indicators

Examine histograms for all indicators

for all indicators Examine histograms for all indicators Examine patterns of directions of indicators for each

Examine patterns of directions of indicators for each variable

Examine patterns of directions of indicators for each variable

patterns of directions of indicators for each variable Examine correlations for indicators for each variable

Examine correlations for indicators for each variable

Examine correlations for indicators for each variable

Examine correlations for indicators for each variable Reduce the number of indicators for each variable Reduce

Reduce the number of indicators for each variable

Reduce the number of indicators for each variable

variable Reduce the number of indicators for each variable Examine correlations among final variables Examine

Examine correlations among final variables

Examine correlations among final variables

final variables Examine correlations among final variables Recode scales of variables measurement Recode scales of

Recode scales of variables measurement

Recode scales of variables measurement

measurement Recode scales of variables measurement Conduct restricted ordered probit models Conduct

Conduct restricted ordered probit models

Conduct restricted ordered probit models

probit models Conduct restricted ordered probit models Conduct unrestricted (multinomial) ordered probit models

Conduct unrestricted (multinomial) ordered probit models

Conduct unrestricted (multinomial) ordered probit models

Upon this point, there was only one indicator (the aggregate indicator) for each variable for further analyses. Then, correlation analyses for the aggregate indicators were conducted. Due to the sample size and number of parameters being estimated (for the multinomial ordered probit models), I recoded levels of measurement for both dependent and independent variables. That is, the items using a five-point scale ranging from “1 (strongly disagree)” to “5 (strongly agree)” were recoded to a three-point scale ranging from “1 (low)” to “3 (high)” as follows: “1” and “2” = “1 = low”; “3” = “2 = medium”; and “4” and “5” = “3 = high”. Next, 76 separate restricted models that included only one of the independent variables and only one of the dependent variables at a time were conducted. Then, I ran four separate full multinomial ordered probit models that included all independent variables, one of the dependent variables, and three control variables (age, education, and gender). However, the results indicate that the size of sample in Studies 1 (n = 86) and 2 (n = 197) did not allow for conducting models that had more than one independent variable. That is, multiple-variables models with the sample size of 86 and 197 cases (for Studies 1 and 2, respectively) inevitably produced a large number of empty cells. Thus, I did not include control variables in the models and, indeed, ran only two-variable models to test Hypotheses 1a-18b for both studies.

88

5.

Results and Discussion

5.1. Study 1 – Results and Discussion

In Study 1, the first step was to examine whether the indicators for each variable were internally consistent, i.e., whether it was possible to reduce the number of indicators. To examine this possibility, I first averaged the indicators to form an aggregate indicator, plotted the aggregate indicator along with the original indicators in a graph, and examined the pattern of directions (see the Appendix B for further information). Using this procedure, it could be seen whether the indicators for each variable followed the same pattern of directions. It is important to note that the method in this study differs from other studies (e.g., Mathieu and Farr, 1991; Hui and Lee, 2000) due predominantly to the ordinal nature of the data in this study. Thus, this dissertation did not follow the procedure used in prior studies that treated the ordinal data as the matrix data. Therefore, a factor analysis was not conducted. 68 The results of the procedure conducted in the present study indicate that it was possible to reduce the number of indicators for each construct by either averaging all three indicators to an aggregate level or averaging a pair of indicators to an aggregate level; thus, I reduced the number of indicators for each variable, and conducted the testing of hypotheses using the aggregate indicators in the multinomial ordered probit regression. In addition to the use of graphs to examine the possibility of reducing the number of indicators, analyses of the Spearman correlation coefficients were also conducted.

5.1.1. Analyses of Correlations among Dependent Variables

Tables in the Appendix B provide the Spearman correlation coefficients among all indicators for dependent variables: namely, active