1
A study about leadership styles and
managers’ commitment to change in
relation
to
absenteeism
in
organizations.
2012
L.A. Spit – Van Campen Studentnumber: 1384090
University of Groningen
Faculty of Management and Organization Master thesis MscBA
Specialization: Change Management Van Teylingenweg 42 3471 GC, Kamerik tel.: 06-18795831 E-mail: leoniespit@hotmail.com Supervisor/ University:. Dr. J. Rupert
Supervisor/ field of study: Van Campen Consulting
M. Teunis
Acknowledgment:
2
Abstract
In this research the extent to which leadership styles influence decreased absenteeism following the method ‘Oplossingsgericht Verzuimmanagement®’, and whether commitment to this method has a mediating role in this relationship is investigated. Data was gathered through distributing questionnaires. Correlation and regression analyses were performed on a dataset of 96 respondents from several organizations that have implemented and adopted the method in the last two years. These are organizations in various industry sectors; service to business, logistics, production and healthcare. To study various industry sectors makes the research outcomes less specific and more extensive to interpret for different organization sectors.
This study proposed a positive relationship between people-oriented leadership styles and decreased absenteeism, with managers’ commitment to the method as mediating role. This study found that people-oriented leadership styles and continuance and normative commitment to the method are negatively significantly related to decreased absenteeism. Further managers with a people-oriented leadership style have less affective commitment to the method (internally driven), but these managers are more normatively and continuancly committed to the method (externally driven). Particularly, there are no relations found between task-oriented leadership and the other variables. These outcomes do not support the proposition of this study.
3
Index
ABSTRACT ... 2 INDEX ... 3 1. INTRODUCTION ... 5 1.1.CONTEXT ... 5 1.1.1. History ... 5 1.1.2. Mobility model ... 6 1.1.3. Manager behavior ... 7 1.2.RESEARCH QUESTION ... 8 2. THEORY ... 10 2.1.ABSENTEEISM ...10 2.1.1. Four models ... 11 2.2.LEADERSHIP STYLES ...122.2.1. The Competing Values Framework ... 14
2.3.MANAGERS’ COMMITMENT TO CHANGE ...17
2.4.LEADERSHIP STYLES AND MANAGERS’ COMMITMENT TO CHANGE ...19
2.5.MEDIATION RELATION ...20
3. METHODOLOGY ... 21
3.1.DATA COLLECTION AND PROCEDURE ...21
3.2.MEASURES ...22
3.2.1. Factor analysis and reliability ... 22
3.2.2. Dependent variable: decreased absenteeism ... 24
3.2.3. Independent variable: leadership style ... 24
3.2.4. Mediation variable: commitment to change ... 25
3.2.5. Control variables ... 26
3.3.DATA ANALYSIS ...26
3.3.1. Descriptive statistics and correlations ... 26
3.3.2. Regression analysis ... 27
4. RESULTS ... 28
4.1.LINEAIR REGRESSION ANALYSES ...28
4.2.MEDIATION ANALYSIS ...30
5. DISCUSSION AND CONCLUSION ... 33
5.1.SUMMARY OF RESULTS ...33
5.2.THEORETICAL IMPLICATIONS ...34
4
5.4.LIMITATIONS AND FURTHER RESEARCH ...37
5.5.CONCLUSION ...38
REFERENCES ... 39
INTERNET SOURCES ...41
APPENDICES ... 42
APPENDIX 1–THE QUESTIONNAIRE ...42
APPENDIX 2–SAMPLE STATISTICS ...44
5
1. Introduction
1.1. Context
In 1996 the law on health insurance in the Netherlands was changed. Since then this law stipulates that an employer is responsible for support and reintegration of an absent employee. This entails that he has to pay 70% of an absent employee’s wage during the first two years of illness; previously this was a part of the Dutch social security system. Organizations became responsible for the costs of an absent employee. AON, an insurance organization, calculates €160 direct costs for one day sick leave, TNO, a research bureau, calculates also indirect costs like replacement costs, loss of productivity and hidden overcapacity. An employee on sick leave can cost an organization between €275 and €500 a day. This shift in responsibility from government to corporations provides a stronger incentive within companies to minimize absenteeism.
Teunis (2010) developed the method “Oplossingsgericht Verzuimmanagement®”. This method sets out to reduce absenteeism in every organization to a percentage of 3.0% or less. In 2011 the average illness rate in the Netherlands was 4.2% (TNO, 2011). There are several organizations (small business and large enterprises) in varied industry sectors (production, consulting services, healthcare, logistics, education and government) that have implemented this method and made it part of their policy. They have reached an illness rate of 3.0% or less. Looking at absenteeism in other countries with the same sort of organizations, the conclusion can be drawn that the average absenteeism in the Netherlands could be reduced even more. From 2005 till 2008 the Netherlands had the highest absenteeism (4.5%) of four West-European countries (Belgium, France, Germany and the Netherlands). Although, Dutch employees were reported ill less often, the duration of their sickness was longer than that of their colleagues in the neighboring countries (Aon, 2010). So, the result of an average illness-rate of 3.0% turns out to be realistic for all kinds of organizations in all kinds of industry sectors.
1.1.1. History
6 employees first had to recover for 100% before they could start working again. When the costs of absenteeism in organizations became sky high in the late eighties, there were 900.000 disabled employees, the capacity model was introduced by occupational health services and medical officers. This model focussed on sick employees to become active to quicken recovery by only looking at the functional limitations for work. Staying home from work will not always help recovery. These two models focused on illness. Illness was the cause of absenteeism and leads to total disability for work. Since the introduction of the WAO in 1990 (law on disability for work) there is also attention for the capacity for work of the employee during illness, but both models are still passive instead of pro-active. The third model was the behavioral model (Falke en Verbaan, 1992). This model was based on the choice of the employee to be absent during illness. Illness does not necessarily lead to absenteeism. A call centre employee with a broken leg can be 100% productive, a call centre employee with a sore throat cannot be productive for 100%. Absenteeism is no longer a passive process, but is a choice of the employee and employer and can be influenced. But when a manager assesses illness this can be seen as a no-confidence motion by the employee and will not contribute to a good relationship between manager and employee.
1.1.2. Mobility model
The different views on illness and absenteeism from the three models mentioned in the last paragraph provide a few valuable aspects for the model introduced by Marleen Teunis (2010). The medical model makes it clear that it is important to pay attention to the reason of illness. The capacity model adds to this that attention has to be paid to the rest capacity of the employee. The last model, the behavioral model states that absenteeism is a choice of employee and employer, and that choice can be influenced through both. The mobility model brings these three things together, and translates this in a conclusion about mobility of the employee. Not discuss the diagnose but discusses the mobility and the current situation of the employee and their possibilities and constraints for work. In conclusion, this method is based on three components: expertise (expertise to manage absence), process (the prescribed method in cooperation with legislation) and communication (the way of communication).
Figure 1
‘Oplossingsgericht Verzuimmanagement®’
PROCESS
7 The essence of this method is making contact and a reintegration plan at the first day the employee is absent, covered by Dutch legislation. The focus is on the first two weeks of illness. Manager and employee together make a reintegration plan for work in the first week of absenteeism. Analyses show the duration of illness increases when there is no contact or counseling in the first week of absenteeism between manager and employee. This influences the illness-rate negatively. Research shows that employees are astrange from their work when they have not been in contact with their organization for at least two weeks. After three months of absenteeism and no commitment to the organization, 40% of the employees never return to their past situation in the organization (Teunis, 2010). Analyses during the implementation of the method shows that managers often struggle with the aspect of the method to make a reintegration plan in the first week of illness. Teunis’ (2010) explanation for managers’ behavior is that they lack the skills of the method to feel self-confident in a conversation and are not convinced of the benefits of the conversation to minimize absenteeism. These skills include the three components of the method, mentioned in the past paragraph, expertise, process, and communication. The communication part is based on Solution Focused Therapy developed by Insoo Kim Berg and Steve de Shazer (2006). This approach focusses on the future instead of the past. And helps to find a solution for the problem to defining a preferred situation (health) and identify current moments the preferred situation already exist. Characteristic for this method is that the employee tells the manger the best solution. The manager has to trust this will happen using this method of communication and does not have to push or communicate his or her own view on the problem. ‘C’est le ton qui fait la musique’, 93% of a conversation is about the way of communication and will help forming a good relationship between manager and employee. Rather, Vonk (2006) found that notions and moral values of a manager determine their communication and can effect distrust of the employee and destruct the relationship between manager and employee when trying to find a solution to the problem of absenteeism. Once more the conclusion can be that managers’ behavior influences absenteeism.
1.1.3. Manager behavior
8 problems of the employee are central to the manager. The manager with this attitude will forget the interest of the organization. 3. Less acceptation / less willingness to help, these managers only see the interests of the organization. There is no attention for the employee. The employee has to change or leave. 4. Less acceptation / high willingness to help. The manager really wants to help the employee, but also keeps the organization’s interest in mind and is strict. They think the last attitude is the most successful way of acting for a manager to shorten illness duration, because the manager takes care for the employee and the organization, instead of employee or organization. The managers are interested in the employee and the willingness to help him with everything to the best of his ability to solve the problem. On the other hand the manager also tries to stimulate the employee to solve the problem by him- or herself and change the employee’s attitude in illness. The conclusion is that manager has to solve the problem and has to help the employee.
There are organizations in which the implementation of the mobility model is more successful than in other organizations. During the implementation of the model, Van Campen Consulting analyze numbers and statistics of the absenteeism in the organizations every month, during 12 months. One of the conclusions of these analyses is that when managers have not made a valid reintegration plan in the first week of absenteeism, the illness duration gets longer and this influences the performance. The conclusion can be made that managers behavior cause the failure. Teunis’ (2010) experience is that there is a taboo for managers to talk to their absent employees for work in the first days of illness and start a conversation about the problem that has arisen and how to act and solve this problem. This is due to the history of Dutch legislation. Moral values of many organizations, managers and employees are still to the same as the earlier view on absenteeism, the welfare state.
Managers and middle management are key figures during the implementation of the method and the success or failure of the project, because these persons have to put up the method on absenteeism in practice. Managers have to lead absenteeism conversations with their employees and they have to conform to a new vision, policy, protocols and processes concerning counseling absenteeism.
1.2. Research question
The management questions of Van Campen Consulting are: what is the influence of the style and behavior of the manager during implementation and securing the project ‘Oplossingsgericht Verzuimmanagement®’ on absenteeism? And what is the influence of the managers commitment to the change?
9 to the change about the accomplishment of the new organization illness policy must be determined. Does the new vision on absenteeism of the organization fit with their own attitude, moral values about absenteeism?
Subsequently the focus of the thesis will be on leadership styles and on the managers commitment to change and how that is related to absenteeism.
This leads to the following research question:
10
2. Theory
Based on the introduction and the research question the following conceptual model can be made. The goal of this research is to show how these variables are related to absenteeism in organizations. These variables will be explained in the following theory section. Conclusions based on the theory will help to formulate the hypotheses for this research. Figure 2 Conceptual Model
+
+
+
_
_
2.1. AbsenteeismAbsenteeism in organizations simply means employees failing to show up for work when they are scheduled to be there (“Attendance Management”, 2010). The direct effect of absenteeism is a decrease in productivity. An indicator for the level of absenteeism in an organization is the illness rate defined as the total costs, including direct and indirect costs, due to this decrease in productivity expressed as a percentage of total productivity. Direct costs are wages of the employee and the replaced employee. Indirect costs are lower productivity, poor quality of goods and services and excess manager time (Carpenter and Wyman, 2008). The illness-rate is a calculation of the total number of failed calendar days divided by the total number of calendar days available to all employees. This definition is set down by Dutch government in the law ‘Terugdringen Ziekteverzuim/Arbowet’ in 1996 to prevent difference calculations.
11 decrease absenteeism. There is also a lot of literature about absenteeism in organizations and factors that influence it in organizations. In this paragraph absenteeism in organizations will be discussed further by explaining four different models used for or based on research about reducing absenteeism in organizations. These four models show that social support by managers can stimulate health. Managers’ behavior is an important aspect that can influence employees motivation, attitude and absenteeism.
2.1.1. Four models
In this section four models will be described: Job Demand-Control-Support model; Effort-Reward-Imbalance model, Work Ability model and the Job Demand-Resources model. The basic assumptions of the theories underlying these models will be discussed. The Job Demand-Control-Support model (Karasek, 1979) distinguishes between the dimension ‘job demands’, ‘job control’ and ‘social support’. Job demands are defined as the required physical and/or psychological effort and skills for a job. Job control is defined as the degree of authority that employees have in order to make decisions. Social support means that an employee experiences support of the direct and indirect environment. Examples of social support are a good relationship with colleagues and support from managers. This model can be used to improve employee well-being and performance in a wide range of organizations. This model suggests that high job demands, little job control and little social support will cause negative health effects.
Siegrist et al. (2004) explains the Effort-Reward-Imbalance model that distinguishes two characteristics of work: effort (work demands and requirements) and rewards (salary, estimation and promotion). The assumption of this model is that an imbalance between effort and rewards cause (mental) exhaustion and increases health risks. Effects of this model show that perceived organizational support from managers to their employees are positively related to employees turnover intentions and their absence (Kinnunen, Feldt and Mäkikangas, 2008).
The third model that will be discussed is the Job Demands-Resources model (Demerouti, Bakker and Schaufeli, 2001). This model is characterized by the following principles: every industry sector has its own specific risks related to work stress. These can be grouped in task requirements and energy sources. Energy sources are all elements in work that gives the employee more energy than that he or she has to invest (autonomy, performance feedback, social support, and opportunities for learning). Energy sources especially have a positive effect when task requirements are high and will decrease work stress and absenteeism.
12 This model illustrates the factors that determine work strain. Firstly, mental and physical health is the fundament to sustainable employability. Poor health is a consistent risk for decreasing productivity and can be caused by lifestyle. Secondly, adequate competencies are needed to meet the requirements of the work. Thirdly, moral values, the attitude of employees to the development of skills and knowledge influence work strain and sustainable employability. Following the Job Demands-Resource model of Demerouti, Bakker and Schaufeli (2001) motivation is a mediator to energy sources and sustainable employability. Energy sources have a positive influence on motivation and contribute to employees’ health and sustainable employability.
These four models show that balance between job requirements and job control is important for employees’ health. Job control and social support improve employees health. Social support and motivation by managers is an important factor in all these four models to decrease absenteeism. To summarize, besides personal circumstances like physical health and employees competencies, motivation and attitude of employees are important determinants and predictors to reduce absenteeism at the workplace. The four models show that social support by managers can stimulate health. Managers’ behavior is an important aspect that can influence employees motivation and attitude and can be supportive to less absenteeism. So, this study focuses on how managers’ leadership styles influences decreased absenteeism.
2.2. Leadership styles
O’Driscoll and Beehr (1994) suggest that if managers are the most immediate and salient persons in an individuals’ work context, then they are most likely to have a direct influence on employees’ behavior, including absenteeism. However, previous research has seldom looked at the possible association between effective leadership and absence, there are factors which leaders can affect directly, and which in turn affect absence, directly or through a longer causal chain. As noted by Ostroff and Bowen (2000), it would be shortsighted to expect a direct link between particular organizational practices and outcomes without considering a range of other potentially relevant factors. Managers’ behavior may be geared to enhance team spirit, a safe work environment, or other desirable features, and these features may influence other variables, including health and work-family conflict, which in turn cause absence. This study investigates the influence of different leadership styles on absenteeism.
13 the task-oriented leadership that focuses on organizational goals, output and performance and on the other side people-oriented leadership that pays attention to motivating and inspiring people (Bass, 1990). To understand these two aspects and how they are related to each other and the influence of both aspects on effective leadership to decrease absenteeism, an outline of theories about leadership will be made through the last decades.
Leadership in theory and in practice has been a fast-growing part of management knowledge since the 20th century. In most conceptions of management and organization, leadership has a central place in motivating employees (people-oriented aspect) and communicating future goals and visions of the organization (task-oriented aspect). (Crevani, Lindgren and Packendorff, 2010). After that, the determination of what is effective leadership must be made and which balance must be made between people-oriented and task-people-oriented aspects for effective leadership. This question gives rise to a series of different theoretical schools. In the late 1940s, the trait approach dominated in organizations. This approach believes that leaders are born rather than made, so in this period the focus was not on influencing leadership behavior. Leadership style was accepted as a final fact. In the 1960s this point of view changed and the contingency approach was introduced, this model implies that a leader’s effectiveness is based on ‘situational contingency’ which is a result of interaction between two factors: leadership style and situational control. Different situations need different leadership approaches, effective leadership depends on the situation. (Fiedler 1967; Fiedler and Garcia 1987).
14 decision-making, and promote a cooperative and trusting work environment” (Carless, 1998:888). There seem to be plausible reasons to expect transformational leadership to have some influence on absenteeism. The individualized-support component of transformational leadership is likely to be important in reducing absenteeism. Prior research using other concepts of leadership, such as initiating structure and leader consideration, found that attention to individuals’ concerns and supportive behavior were negatively associated with absence from work (Tharenou, 1993; Boumans & Laandeweerd, 1993; Smulders 1984 in Mellor, Arnold and Gelade, 2008).
To summarizing these theories explain the contrast between task-oriented and people-oriented leadership styles. Besides these theories, research shows several different theories to explain different leadership styles and how leadership styles affect organization performance. The different theories, mentioned above, distinguish people-oriented leadership styles and task-people-oriented leadership styles. Therefore this study will distinguish these two types of leadership styles, i.e. people-oriented leadership style and task-oriented leadership styles.
Leadership styles include managers’ behavioural activities, personal characteristics and motives (Stoker and Kolk, 2003). Leadership styles represents how managers act, behave, think and work. Because of the complexity of leadership styles it is difficult to measure. Quinn et al. (1996) define leadership styles as an element of organization culture. They define organizational culture as what is valued, dominant leadership styles, language and symbols, procedures and routines, and the definition of success that make an organization unique. This definition focuses on the understanding of culture, specifically on the practices that are embedded in the organization. To investigate different leadership styles, the Competing Values Framework of Quinn et al. (1996) will be used in this study. This framework includes all the aspects of leadership styles, the behavioural activities, personal characteristics and motives to define leadership styles. They developed an instrument to measure leadership styles, the Organizational Culture Assessment Instrument (OCAI). It has been found that the OCAI is an accurate instrument to assess leadership styles, significant relationships between leadership styles and indicators of organizational effectiveness have been found. For that reason the OCAI will be used to measure leadership styles in this study.
2.2.1. The Competing Values Framework
15 focus, which represents the consent and development of the employee. This dimension fits the people-oriented leadership style defined in this study. The second dimension represents the organizational preference for structure and represents the contrast between stability and control and flexibility and change. This dimension fits the task-oriented leadership style defined in this study. Internal focus and integration are opposites of external focus and differentiation. The framework can be used to put order in the literature of leadership and leadership profiles.
Figure 3
Competing values framework from Quinn et al. (1996)
16 Leadership roles in the clan culture are seen as facilitator or mentor and have people-oriented characteristics. The facilitator manages conflict and seeks consensus. His or her influence is based on getting people involved in the decision-making and problem-solving process. Participation and openness are actively pursued by the facilitator role. On the other hand, the mentor is caring and empathic. This person is aware of others and cheers for the needs of individuals. His or her influence is based on mutual respect and trust. Morale and commitment are actively pursued (Cameron and Quinn 1999:114).
Leadership roles in the adhocracy culture are the innovator or the visionary and are also people-oriented. The innovator is clever and creative. This person envisions change. His or her influence is based on anticipation of a better future and generates hope in others. Innovation and adaptation are actively pursued. The visionary is future oriented in thinking. This person focuses on where the organization is going and emphasizes possibilities as well as probabilities. Strategic direction and continuous improvement for organization and employees is a hallmark of this style pursued (Cameron and Quinn 1999:114).
The roles in the other two quadrants of hierarchy culture and market culture are more task-oriented. Leadership roles in the hierarchy culture are the monitor and the coordinator. The monitor is technically expert and well informed. This person keeps track of all details and contributes expertise. His or her influence is based on information control. Documentation and information management is actively pursued. The coordinator is dependable and reliable. This person maintains the structure and flow of the work. His or her influence is based on situational engineering, managing schedules, giving assignments, physical layout. Stability and control are actively pursued (Cameron and Quinn 1999:114).
17 Zammuto and Krakower (1991) and Jones et al. (2005) focus in their studies on only one axis of the competing values framework to make a distinction between the hierarchy and market cultures, the task-oriented leadership styles and the clan and adhocracy cultures, the people-oriented leadership styles. The reason for clustering the quadrants of the framework is because of the fact that other studies showed the strongest effect with the quadrants along this dimension. Therefore this study will also focus on this dimension, which divides task-oriented leadership styles, the lower side of the framework and people-oriented leadership styles, the upper side of the framework.
Earlier in this theory section it has been mentioned that leadership styles can influence decreased absenteeism. Social support and motivation by managers is important behaviour to decrease absenteeism. These are aspects of the people-oriented leadership style. Another conclusion based on the four models about absenteeism in organizations, defined in an earlier section is that employees job satisfaction is an important factor to decrease absenteeism. Organizations with people-oriented managers tend to have higher job satisfaction among employees, lower turnover rates and absenteeism. On the other hand, although organizations with more task-oriented managers tend to have higher job performance, it is usually accompanied by lower job satisfaction, as well as higher turnover rates and absenteeism (Wee, Wee and Huang, 2006). So, managers with a people-oriented leadership style show more social support and motivation to employees and their employees have higher job satisfaction. These three factors positively influence a decrease in absenteeism. Based on the theory mentioned above, the following hypothesis can be stated:
Hypothesis 1: People-oriented leadership styles (i.e. facilitator, mentor, innovator and broker) are related more positively to a decrease in absenteeism than task-oriented leadership styles.
2.3. Managers’ commitment to change
Implementation of the method to reduce absenteeism in organizations requires that a path of change has to be constructed regarding the counseling of absenteeism by managers. Commitment to change has been identified as an important aspect of behavioral intention to support change (Herscovitch and Meyer, 2002; Fedor et al, 2006). In this study commitment to change means commitment to the method ‘Oplossingsgericht Verzuimmanagement®’.
18 “Commitment to change is a force (mind-set) that binds an individual to a course of action deemed necessary for the successful implementation of a change initiative” (Herscovitch and Meyer 2002: 475). Commitment to change seems to have a positive influence on employees willingness to occupy in change-related behaviors (Jarros, 2010). For this reason managers’ commitment to change is an important variable in relation with the success of the implementation of the method.
Herscovitch and Meyer (2002) developed a model to measure commitment to organizational change. They proposed that commitment could take different forms that have different implications for the nature and level of behavioral support for change. They distinguish three forms of commitment. The first form is ‘affective commitment’. This is defined as support for the change based on personal belief in its inherent benefits. The second form of commitment is ‘continuance commitment’. This implicates that managers have to support the change because of the recognition that there are costs associated with failure. The third form of commitment is ‘normative commitment’. This refers to a sense of obligation to support the change. For this research it is interesting to figure out how these three forms of commitment influence the implementation of the method and decreased absenteeism.
The relationship between commitment to change and individual performance is found by Meyer and Allen (1991). They found that a higher level of affective commitment leads to a higher level of individual performance. Parish et al. (2008) also investigated this relationship. The outcome of their research was that affective and normative commitment to change positively influences individual performance. For instance, high levels of organizational commitment are believed to be associated with low turnover, limited tardiness, low absenteeism, and enhanced job performance (Mowday, Steers and Porter, 1979).
The before mentioned studies suggest that a positive relation should exist between commitment to change and the individual performance level of a manager. Within the context of this study it is expected that managers’ commitment to the method influences the success of its’ implementation which should result in a decrease in the rate of absenteeism. This leads to the following hypothesis:
19
2.4. Leadership styles and managers’ commitment to change
This study describes people-oriented leadership styles and managers’ commitment to change in relation with decreased absenteeism. It is hypothesized that these two factors are related positively to a decrease in absenteeism. In order to elaborate managers’ commitment to the method within different leadership styles, this section will describe how managers’ commitment to the method and people-oriented- and task-oriented leadership styles are related to each other.
As mentioned earlier the implementation of the method implies a path of change in which managers have to change their behavior. This demands the support of managers for change. Commitment to change contributes more than organizational commitment to the prediction of employees’ self-reported behavioral support for change (Herscovitch and Meyer 2002). This means that organizational commitment of the managers does not influence their commitment to the method.
An important finding in literature is that innovative and supportive subcultures have positive associations with commitment to change, while a bureaucratic subculture has a slight negative association with commitment to change (Lok and Crawford, 1999). This finding is consistent with previous findings, which suggested that a bureaucratic environment often resulted in a lower level of employee commitment (Brewer, 1993; Kratina, 1990; Wallach, 1983). Innovative and supportive subcultures can be find in the upper side of the Competing Values Framework, which is defined as the people-oriented leadership styles for this study. The bureaucratic subculture fits with the lower side of the framework, which is defined as the task-oriented leadership style. This implicates that managers with a people-oriented leadership style have more commitment to change than managers with a task-oriented leadership style.
20 Hypothesis 3: People-oriented leadership styles (i.e. facilitator, mentor, innovator and broker) are related more positively to managers’ commitment to the method than task-oriented leadership styles.
2.5. Mediation relation
This theoretical section discussed the three variables of the conceptual model (figure 2). It has been discussed how leadership styles will affect decreased absenteeism and how commitment to the method will affect decreased absenteeism. After that the relationship between leadership styles and managers’ commitment to the method is described. This study also suggests a mediation relation of managers’ commitment to the method, as showed in the conceptual model (figure 2).
Commitment is arguably one of the most important factors involved in employees’ support for change initiatives, searched by a number of researchers (Meyer and Allen, 2002). And according to Conner and Patterson (1982), “The most prevalent factor contributing to failed change projects is a lack of commitment by the people” (p. 18). This implicates that managers’ commitment to the method contributes to the success of the implementation, and thus on decreased absenteeism.
This study discussed that different leadership styles have different outcomes to decreased absenteeism. It is hypothesized that people-oriented leadership styles are related more positively to decreased absenteeism than task-oriented leadership styles. Also managers’ commitment to the method is an important factor to decreased absenteeism. The last paragraph explains the relationship between a people-oriented leadership style and managers’ commitment to the method. Because of these relations this study will investigate whether managers’ commitment to the method has a mediating effect between leadership styles and decreased absenteeism. So, the following hypothesis is proposed:
21
3. Methodology
3.1. Data Collection and procedure
In this study quantitative research was used to measure the influence of the two variables leadership styles and commitment to change on absenteeism. A quantitative study provides verifiable data that could be used for statistic analyses and for drawing conclusions to answer the research question. For the research a questionnaire was used, which tested the respondents on the above mentioned factors; commitment to the method and leadership styles. The questionnaire was composed by combining existing questionnaires translated into Dutch (see Appendix 1 for the complete questionnaire). For this research managers who have participated in the change and have to accomplish and deal with the method every day doing their job, were asked to complete the questionnaire. The questionnaire was sent out by e-mail, respondents could click on a hyperlink in the email message that led them to the questionnaire. Advantages by digital collecting data are easy administration, it is fast to collect the data and perform the analyses (Hair et al., 2007). Respondents were managers from several organizations which have implemented and adopted the method in the last two years. These are organizations in varied industry sectors; service to business, logistics, production and healthcare, in total 156 managers. Because various industry sectors were studied the research outcomes are less specific and they can be interpreted more generally.
Participation was voluntary and anonymity was guaranteed. In this way the respondents were more motivated to give honest answers, because their privacy was protected (Emans, 1986). To maximize the response the board of managers of the organizations were asked to support and promote the research. Because of the partnership between VCC and the participants during the implementation project the participants were willing to cooperate to this research.
The sample was 67.30% (105 responses). 9 questionnaires were not completed and had missing values on one or more items. These data are deleted in the analysis, in total 96 questionnaires were used (61.54%).
22 average number of employees in a team was 55 members. Appendix 2 presents these sample statistics.
3.2. Measures
This section provides inside information about the measures. Besides the measure methods of all variables, the factor analysis and reliability of the survey will be explained.
The questionnaire had structured statements which the respondents had to judge on a rating scale (Hair, Money, Samouel & Page, 2007). The Likert scale is very popular among researchers, because it has a lot of advantages. It is quick and easy to construct and the Likert scale provides more reliable and a greater volume of data than most other scales (Cooper & Schindler, 2006).
The questionnaire was composed by combining two existing and validated questionnaires. The part about commitment to the method was constructed by Herscovitch and Meyer (2002) and measure three components of commitment: affective-, continuance-affective-, and normative commitment. The part about leadership styles was a questionnaire developed by Quinn et al. (1996) and distinguishes eight manager roles mentioned in theory. Originally the questionnaire consisted of 72 items in total.
3.2.1. Factor analysis and reliability
To test if the different factors in this study measured the proposed variables various factor analyses were conducted. In Appendix 3 the various factor analyses can be found. Usually, a factor analysis gives one solution for all items loading on the associated variables. For this research a 5-factor solution has been used.
In order to determine which factor solution was the best for this research, first a factor analysis was forced into a logical structure regarding the theory. The first step was forcing the factor analysis into five components. Three variables of commitment to the method and based on theory the eight managers roles were summarized into two categories people-oriented and task-oriented leadership styles. This analysis showed that three variables of commitment to the method load on two components. Continuance- and normative commitment loaded mainly on one component. People-oriented leadership style loaded on all different components.
23 Because the factor analysis with four components did not show the desired structure either, a factor analysis with eleven components was performed at the level of sub-dimensions, determining three types of commitment to the method and eight leadership styles (Appendix 3). The result was that the items of the people-oriented leadership style were evenly divided between the 5-factor and 4-factor analysis.
Kline (1994) stated that for a good factor analysis there have to be 5 respondents per item. This research distinguishes 72 items, which means there have to be at least 360 respondents. Because of the small amount of respondents (96) in relation with 72 different items, a factor solution was sought in which the variables commitment to the method and leadership style are separated into two factor analyses (Appendix 3). This solution shows desired outcomes for commitment to the method, but again no significant improvement to the structure of leadership styles.
Due to all insufficient performed factor analyses, the choice was made to perform the 5-factor analysis. This choice was made because from all 5-factor analyses, this solution turned out to be the most appropriate, with the least double loadings and the best division of the variables. However this factor solution does not show the desired structure, but for theoretical reasons, the choice was made to stick to five components. The five components refer to the variables: affective commitment, continuance commitment, normative commitment, people oriented leadership styles (facilitator; mentor; innovator and broker) and task oriented leadership styles (producer; director, coordinator and monitor). As opposed to expectations commitment to change loads in this solution on two components, people-oriented leadership on four components and task-oriented leadership on one component. In the headings below the outcomes of the 5-factor analysis are interwoven in the elaboration of the variables. To optimize the factor analysis 18 items were deleted and 36 items were used for this study. Table 1 represents the number of items for each scale and the Cronbach’s alpha’s values.
Table 1 Reliability analysis
Variable Amount of items Cronbach’s alpha
value
Affective commitment 5 0.92
Continuance and Normative commitment 7 0.84
People-oriented leadership styles 12 0.74
24 3.2.2. Dependent variable: decreased absenteeism
Absenteeism will be measured with objective data from the organizations. The average illness-rate is a calculation of the total number of failed calendar days divided by the total number of calendar days available to all employees. It is presented in a percentage. This definition is set down by Dutch government in the law ‘Terugdringen Ziekteverzuim/Arbowet’ in 1996 and must be used in every organization to prevent difference in calculations.
The administrated average illness-rate of the teams of the participating managers are used. The illness-rate at the start of the project and the illness-rate at the end of the project, after implementation were subtracted from each other. The duration of the projects was twelve months. This means that a positive number of decreased absenteeism is a positive result. The range of decreased absenteeism is a minimum of -6.91 and a maximum of 9.60, the average is 1.40 and SD 2.48.
3.2.3. Independent variable: leadership style
The second concept is the leadership style of the manager. For the measurement of these items the questionnaire of the Competing Values Framework developed by Quinn et al. (1996) was used. The questionnaire of the Competing Value Framework given to the managers measured the eight operational roles based on the original work of Denison et al. (1993) and consisted of 36 items, each role was represented by 4 or 5 items. The managers were asked to fill out how many times they showed certain behavior in their role as a manager, like delegating, controlling or mediating. The eight managers’ roles were grouped in two leadership styles for this research; 1. people-oriented leadership styles (i.e. facilitator, mentor, innovator and broker) and 2. task-oriented leadership styles (i.e. producer, director, coordinator and monitor). To optimize validity some items were excluded from this study through the 5-factor analysis, because of double loadings. Deleted items loaded on different components. Removal of these items does not cause a decrease in reliability and an improvement of the factor analysis. Items that were deleted described hands-on activities of managers. The admitted items were more about conceptualization, guidance and support. The contingency theory also distinguishes hands-on activity and supports as two different categories in managers’ tasks (Fiedler, 1967). The theory section mentioned that managers’ influence on decreased absenteeism is about support of the employee, that is why the items about support were added and the items about hands-on activity were deleted for this study.
25 was a two-dimensional model with eight managers roles. The two variables that are distinguished in theory, people-oriented and task-oriented leadership, will be maintained in this study for that reason.
The extreme double loaded items in this scale were removed following the 5-factor analysis added in Appendix 3 (loading on three components or a difference of <.20). The leadership style ‘mentor’, part of the variable people-oriented leadership style, loads on the same component as affective commitment. This is quite likely, the items of ‘mentor’ role formulate the specific tasks to implement the method (concern for employees health, listening to private problems of employees and attention and commitment to employees) that fits with managers’ affective commitment to the method to perform these tasks; ‘I believe in the value of this change’ and ‘This change serves an important purpose’. Removal of all the items of ‘mentor’ role would result in a decreased reliability with 0.10 and give an output of Chronbach’s alpha of 0.64. Social science proved that the ‘mentor’ role is important for managers’ effectivity. The ‘mentor’ role is important to predict managers’ success (Bass, 1990). So for theoretical reasons, three items of leadership style ‘mentor’ were enclosed in the variable people-oriented leadership style.
Task-oriented leadership style loads mostly on one different component (Appendix 3). The items that load on other components were removed, except for leadership style ‘producer’. The items of ‘producer’ loaded inconsistently on three different components in the 5-factor analysis. This can be explained because the producer role is a completion to the director role, and some items overlap. The producer role is about implementing what the manager in the director role planned (Quinn, 1996). The statements of the producer role are about hands-on activities. As discussed earlier all the items about hands-on activity were deleted for this study. For that reason all items of managers’ producer role should be deleted according to the theory. However Hooijberg and Choi (2000; 2001) found high intercorrelations exist among managers’ roles ‘producer’, ‘director’, and ‘coordinator’. They labeled these three roles as Goal Achievement. In this study these three roles and the ‘monitor’ role were put together in the variable task-oriented leadership. Deleting all the items of the ‘producer’ role would damage the scale of task-oriented leadership style and would not contribute to better loadings on the other items of the scale. Excluding managers’ role ‘producer’ influences other items in the 5-factor analysis in a negative way and it has no influence on the reliability. The first three items looked like the three items of the ‘director’ role and gave the best result in the 5-factor analysis. That is why the first three items of producer were included in the study.
3.2.4. Mediation variable: commitment to change
26 consists of 6 items, continuance commitment (cost-based) represents 6 items, and normative commitment (obligation-based) also with 6 items (Appendix 1).
Looking at the variables for commitment it is remarkable that continuance and normative commitment load in the factor analysis on the same component. Both of these variables appeal to extrinsic motives of commitment (cost based and obligation) while affective commitment is based on feelings and intrinsic motivation.
To measure hypothesis 3 the variable commitment to change was split up into two parts, because of the outcomes of the 5-factor analysis (Appendix 3). An analysis will be made that distinguishes two factors of commitment to change: affective commitment (internally motivated) and continuance + normative commitment (externally motivated) and measures the influence of these two factors on decreased absenteeism and leadership styles.
One item of affective commitment is deleted ‘This change is not necessary’ because this item measures a different factor than the other five items. This statement was not relevant for this research, because change was already implemented and respondents had no influence on that. Two items of continuance commitment and three items of normative commitment were excluded following the factor analysis, because of measuring another factor or a double loading. These items appeal to organization interests, without influence of managers by themselves. The items that could be used, mention managers’ own influence and choice.
3.2.5. Control variables
The questionnaire also include some demographical items as control variables: age, gender, nationality, level of education, work experience, work experience in this job, and the amount of employees in managers teams. Given that all respondents were Dutch the control variable nationality was not used in further analysis. The variables age, work experience and work experience in this job intersect with each other. For this study work experience is chosen as control variable in the analysis, because this variable has the strongest correlation on decreased absenteeism (see table 2).
3.3. Data Analysis
For the data analysis of the questionnaire the computer program Statistical Package for the Social Science (SPSS, version 20) was used. The following statistical procedures were conducted.
3.3.1. Descriptive statistics and correlations
27 positively or negatively. Following the descriptive statistics, a correlation analysis was performed using the Pearson correlation coefficient.
3.3.2. Regression analysis
The hypotheses were tested with regression analyses. According to De Heus, Van der Leeden and Gazendam (1995) there are four assumptions that need to be checked before performing regression analyses. First the interval ratio can be verified for this research, because the data used a seven-point Likert-scale. This is ‘quasi-interval’ and can be used as well since the results remain stable and there is no strong variation in the scales of the variables. The second assumption is linearity of the variables. This was tested with residual plots. All plots showed significant linearity and therefore this assumption is also verified. Randomness is the third assumption and means that all independent variables should be fixed and the dependent variable should be random. When using regression analyses it is not necessary to verify this assumption (de Heus, van der Leeden and Gazendam, 1995). The last assumption is about independence, normality and homoscedasticity of residues. Independency of residues was accounted for the procedure of data collection in this research. The other two concepts were tested using residual and q-q plots. These plots did not report deviating signs and thus the final assumption was also verified.
28
4. Results
First, the descriptive statistics and the correlations of the variables in this study, are shown in table 1. After that the outcomes of the analyses are presented. Table 1 shows that there is a negative significant relationship between continuance and normative commitment to change and decreased absenteeism (r = -.23, p<.05). This means that managers with high continuance and normative commitment to the change are associated with a higher level of absenteeism in their teams. These two variables of commitment are external motivated instead of affective commitment which is internal motivated. Both leadership styles are negatively related with affective commitment to the method; people-oriented leadership style (r = -.59, p<.01) and task-oriented leadership style (r = -.49, p<.01). Even people-oriented managers and task-oriented managers have no internal motivation to the method.
There is a negative correlation between people-oriented leadership style and decreased absenteeism (r = -.22, p<.05). Managers with this leadership style have a higher level of absenteeism.
Decreased absenteeism correlates with managers age (r = .32, p<.01), work experience (r = .33, p<.01). The older the manager and the more work experience the lower the absenteeism in their teams.
Task-oriented leadership is positively related to the amount of team members (r = .23, p<.05). Managers with more team members have a more task-oriented leadership style.
4.1. Lineair regression analyses
This section shows lineair regression analyses and gives insight in the relationship between the independent and dependent variable. The control variables age, work experience and team members have been inserted in the correlation matrix. The control variables age and work experience are included in the analyses, because the correlation matrix did show relevant correlations with the other variables for this study.
Table 2
Correlation matrix (N=96)
Mean SD 1 2 3 4 5 6 7 8
1. Decreased absenteeism 1.40 4.48 -
2. Affective Commitment 10.00 5.00 -.01 -
3. Continuance and Normative Commitment 30.59 8.27 -.23* .07 -
4. People-oriented leadership 68.06 5.84 -.22* -.59** .16 -
5. Task-oriented leadership 66.06 9.46 -.08 -.49** .01 .73** -
6. Age 48.94 8.53 .32** -.12 -.21* .18 .16 -
7. Work experience 22.66 12.46 .33** -.08 .01 .10 .06 .81** -
8. Teammembers 54.44 52.52 .20 -.16 .17 .09 .23* .03 -.15 -
29 The first hypothesis proposes that a people-oriented leadership style is related more positively to a decrease in absenteeism than task-oriented leadership style. This hypothesis was tested with a regression analysis which includes the variables leadership styles people-oriented and task-oriented and decreased absenteeism, and is also controlled with the variables age and work experience. The outcomes of this test are presented in the table below.
Table 3
Lineair regression results
Decreased absenteeism
Age .22
Workexperience .18
People-oriented leadership style -.41**
Task-oriented leadership style .18
R2 .21
F 5.86**
*p<.05, **p<.01
There is a significant negative relationship between the two variables people-oriented leadership style and decreased absenteeism (β= -.41, p< .01, R2= .21). The conclusion is that managers with a more people-oriented leadership style have a less decrease in absenteeism. There is not a significant relationship between task-oriented leadership style and decreased absenteeism. Although a relationship between the variables is found, hypothesis 1 is rejected.
The second hypothesis proposes that managers’ commitment to the method is related more positively to decrease in absenteeism. This hypothesis was tested with a lineair regression analysis which includes the variables managers commitment to change: affective commitment (internally driven) and normative + continuance commitment (externally driven) and the variable decreased absenteeism, controlled by age and work experience of the managers. The outcomes of this test are presented in the next table.
Table 4
Lineair regression results
Decreased absenteeism
Age .20
Workexperience .16
Affective commitment (intern) -.02
Normative + Continuance commitment (extern) -.25*
R2 .18
F 4.86**
*p<.05, **p<.01
30 decreased absenteeism. This implicates that managers support the method because of the recognition that there are costs associated with failure. There is no relation between affective commitment to the method (internally driven) and decreased absenteeism. So hypothesis 2 is not confirmed, although there are other significant relations found between the variables.
The third hypothesis proposes that a people-oriented leadership style is related more positively to managers’ commitment to the method than task-oriented leadership style. This hypothesis was tested with a regression analysis that includes the variables leadership styles people-oriented and task-people-oriented and managers commitment to the method and is controlled with age and work experience. Two elements of commitment are formulated through the factor analysis: affective commitment (internally motivated) and normative + continuance commitment (externally motivated). The outcomes of these tests are presented in the table below.
Table 5
Lineair regression results
Affective commitment (intern) Normative + Continuance commitment (extern)
Age .01 .18
Workexperience -.02 -.16
People-oriented leadership style -.49** .31*
Task-oriented leadership style -.14 -.24
R2 .35 .06
F 12.42** 1.39
*p<.05, **p<.01
There is a significant negative relationship between the variables people-oriented leadership style and affective commitment (β= -.49, p< .01, R2= .35). There is a significant positive relation between the variables people-oriented leadership style and normative + continuance commitment to the method (β= .31, p< .05, R2= .06). This means that managers with a people-oriented leadership style have less affective commitment to the method (internally driven), but these managers are more strongly normative and continuance committed to the method (externally driven). There is no significant relationship between managers with a more task-oriented leadership style and commitment to the method. Although, it is remarkable that task-oriented leadership styles relates negatively to affective commitment (internally driven) and normative and continuance commitment (externally driven). Hypothesis 3 is rejected.
4.2. Mediation analysis
31 absenteeism). This is proposed in hypothesis 1. Secondly, a direct relationship needs to exist between the independent variable and the mediator (i.e. commitment to the method). This is proposed in hypothesis 3. Thirdly, the mediator must affect the dependent variable, see hypothesis 2. These three steps were tested with linear regression analyses in the previous paragraphs. An overview of the results is illustrated in figure 4. This figure represents the first three steps of Baron and Kenny (1986).
Figure 4
Significant results of regression analyses
H1:β=-.41, p<.01 H3:β=-.49, p<.01 H2:β=-.02, p>.05 H3:β=.31, p<.05 H2:β=-.25, p<.05 H1:β=.18, p>.05 H3:β=-.14, p>.05 H2:β=-.02, p>.05 H3:β=-.24, p>.05 H2:β=-.25, p<.05
The purpose of steps 1-3 is to establish that significant relationships among these variables exist. If one or more of these relationships are non significant mediation is not possible. Assuming there are significant relationships from steps 1-3, one proceeds a fourth step. In this study there are only significant relationships from step 1-3 for the variables people-oriented leadership style, continuance + normative commitment to the method and decreased absenteeism. So, mediation will only be determined with these variables.
In order to determine a mediation relationship, a fourth step is needed. In the fourth step some form of mediation is supported if the effect of the independent variable (i.e. people-oriented leadership style) on the dependent variable (i.e. decreased absenteeism) remains
Affective commitment to the method People-oriented leadership style Decreased absenteeism Continuance + normative
commitment to the method
Affective commitment to the method Task-oriented leadership style Decreased absenteeism Continuance + normative
32 significant after controlling for the mediator (i.e. commitment to the method). If the dependent variable (i.e. people-oriented leadership style) is no longer significant when the mediator (i.e. commitment to the method) is controlled, the finding supports full mediation. If the dependent variable (i.e. people-oriented leadership style) and the mediator (commitment to the method) are still significant, the finding supports partial mediation. This includes a multiple regression analysis because the dependent variable is regressed on both the independent variable and the mediator. In order to test this fourth step a multiple regression analysis was performed to measure hypothesis 4, which implies a mediation relationship. Hypothesis 4 proposes that commitment to the method has a mediation relationship on leadership styles and decreased absenteeism, with the expectation that this relationship is stronger for people-oriented leadership styles than task-oriented leadership styles.
Table 6 shows the outcome of the regression analysis, which was performed to examine the effect of the independent variable people-oriented leadership style on the dependent variable decreased absenteeism, which is mediated by continuance and normative commitment to the method (externally motivated), controlled by age and work experience.
Table 6
Mediation analysis decreased absenteeism
Model 1 Model 2
B β B β
Age .11 .22 .14 .27
Workexperience .06 .18 .05 .13
People-oriented leadership style -.31 -.41** -.19 -.25*
Continuance + normative commitment -.12 -.31*
R2 .21 .23
F 5.86** 6.97**
*p<.05, **p<.01
33
5. Discussion and conclusion
The objective of this research was to gain insight into how different leadership styles of managers influence absenteeism in organizations that implement the method ‘Oplossingsgericht Verzuimmanagement®’. Subsequently, this research also shows how managers’ commitment to this method influences the effect of leadership styles on absenteeism.
5.1. Summary of results
On the basis of the theory four hypotheses were proposed regarding to what extent leadership styles had an effect on decreased absenteeism, and how commitment to the method influences this relationship. In this research the focus has been on two aspects of leadership styles, the people-oriented leadership style and the task-oriented leadership style along the Competing Values Framework. The reason for clustering the quadrants of the framework was because of the fact that other studies showed the strongest effect with the quadrants along this dimension. Based on the factor analysis, commitment to the method was divided in two variables, affective commitment to the method (internally motivated) and continuance + normative commitment to the method (externally motivated).
It was hypothesized that people-oriented leadership styles are more strongly positively related to a decrease in absenteeism than task-oriented leadership styles. The outcomes do not confirm this proposition. The outcomes present a negative relation between people-oriented leadership styles and decreased absenteeism. This means that managers with a more people-oriented leadership style have higher absenteeism in their teams in the sample. On the other hand, no empirical evidence could be found to support a relation between the degree of a task oriented leadership style and decreased absenteeism.
Regarding externally motivated commitment to the method a negative relationship with decreased absenteeism was established. So, managers who feel external motivation to execute the method are not successful in decreasing absenteeism. On the other side there is no relationship between managers with internally motivated commitment to the method and decreased absenteeism. These outcomes are also in contrast to what was hypothesized. It was expected that managers’ commitment to the method positively relate to decrease in absenteeism. This study finds evidence that suggest that externally motivated commitment to the method by managers has a negative contribution on decreasing absenteeism.
34 confirm this proposition only for externally motivated commitment and not for internally motivated commitment. This implies that managers with more people-oriented leadership style are externally motivated to commit to the method. There is no relationship between managers with task-oriented leadership styles and commitment to the method.
The final hypothesis that was investigated states that the relationship between people-oriented leadership styles and decreased absenteeism was mediated through externally motivated commitment to the method. The results suggest that externally motivated commitment to the method explains the negative relationship between people-oriented leadership styles and decreased absenteeism.
5.2. Theoretical implications
This research focused on the effect of leadership styles, with the use of the competing values framework (Quinn et al. 1996) on decreased absenteeism, with the mediating factor commitment to the method ‘Oplossingsgericht Verzuimmanagement®’, using the model of commitment to change from Herscovitch and Meyer (2002).
This study found that people-oriented leadership styles of managers are negatively related to decreased absenteeism. This is in contrast with what was hypothesized and found in theory. In trying to find a reason for the fact that people-oriented leadership styles have a negative influence on decreased absenteeism, Wee, Wee and Huang (2006) provide some indications related to different leadership styles and organizational performance. They mentioned the importance of balance between both leadership styles people-oriented and task-oriented. According to the Leadership Grid of Blake and Mouton (1975), managers can assess their current level of balance between each leadership style in order to achieve the best results, high levels for people-oriented and task-oriented leadership. This will create an organization with high performance and will reduce absenteeism and also increase job performance. This implicates that through a balance between people-oriented leadership styles and task-oriented leadership styles, effective leadership strategies can be achieved and absenteeism will decrease. So, following Wee, Wee and Huang (2006) only people-oriented leadership styles are not successful; managers need to have both leadership styles people-oriented and task-oriented.
